Citation
Breastfeeding Practices in the Neonatal Intensive Care Unit before and after an Intervention

Material Information

Title:
Breastfeeding Practices in the Neonatal Intensive Care Unit before and after an Intervention
Copyright Date:
2008

Subjects

Subjects / Keywords:
Breast milk ( jstor )
Breastfeeding ( jstor )
Breasts ( jstor )
Health care industry ( jstor )
Hospitals ( jstor )
Infants ( jstor )
Medical personnel ( jstor )
Milk ( jstor )
Neonatal intensive care units ( jstor )
Professional education ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
8/31/2006

Downloads

This item has the following downloads:


Full Text












BREASTFEEDING PRACTICES IN THE NEONATAL INTENSIVE CARE UNIT
BEFORE AND AFTER AN INTERVENTION PLAN















By

ROBERTA GITTENS PINEDA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Roberta Gittens Pineda




























This dissertation is dedicated to all the mothers who have premature and medically
fragile infants in the neonatal intensive care unit (NICU). It is hoped that this and other
works with infants and mothers in the NICU will give you the hope, courage, and
information needed to "mother" in the complex environment of the NICU, during your
infant's first precious days. This dissertation is also dedicated to my husband, Jose, for
his endless source of love and inspiration. It is also dedicated to my children, Alan and
Marissa, whose early birth made me realize the importance of education and support for
mothers, as well as to my daughter Abigail, whose premature birth in the middle of this
research gave me the extra steam to see it through. This dissertation is especially
dedicated to my mom, Barbara Gittens Valentine, whose expertise with mothers and
babies was critical during my own son's hospitalization.















ACKNOWLEDGMENTS

This research would not have been possible without the support and guidance of

Dr. Lorie Richards, who has been my advisor, mentor, and friend. I want to thank Lorie

for having faith in me and inspiring me to do my best. I would like to thank the nurses

and health care professionals at Shands Hospital who participated in this research

endeavor. In particular, I give special thanks to Annmarie Brennan, who enabled this

research project to occur in the neonatal intensive care unit (NICU) at Shands and

supported the project every step of the way. I also would like to thank Cammy Pane, the

co-author of the Educational Module; Stephanie Meeks for your hours of work on "A

Mother's Gift", and other members of the Lactation Committee at Shands who helped

with my research: Elayne McNamara, Sandra Sullivan, Brenda Owens, Sheila Walker

and Jeannette Sexton. I want to give special thanks to Susan Frazier from Medela for

your support at the inservices. I also want to thank all those who provided donations as

incentives for participation: Sonny's Barbeque, Atlanta Bread Company, Scholotsky's

Deli, and TGIF. I want to thank Dr. David Burchfield, the medical director of the NICU

at Shands, for assisting with this project. I thank Sarah Boslaugh for guiding me through

the statistics and for all your patience from the many questions that came up along the

way. I would also like to thank my committee for sticking with me through the years,

and the move to St. Louis and the addition of the new baby. I appreciate your endless

patience, high expectations and sincere enthusiasm for my interests and work. I want to

thank Drs. Richards, Foss, Krueger, Seung, and Rosenbek!









I would finally like to thank my parents who always showed unconditional love and

always motivated me to strive to do better. I extend special thanks to my husband, Jose,

for always being there when I needed you most and giving me patience and love every

step of the way. You enabled me to go back to school and were there when it came to

crunch time. You have made this all possible and I am eternally grateful for your love

and support.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ........ ..... ... .... .......... ........... .......... ..... viii

LIST OF FIGURES ......... ........................................... ............ ix

A B STR A C T ................................................. ..................................... .. x

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

The Importance of Breast Milk and Breastfeeding....................................................2
Health Benefits of Breastfeeding for the Full Term Infant ................................
Health Benefits of Breastfeeding for Premature Infants ....................................
Long Term Benefits of Breastfeeding .....................................................5
Developmental Benefits of Breastfeeding............... .......... .................5
Benefits of Breastfeeding for the M other............... ...................... .............6
Current Breastfeeding Recom m endations ........................ ......... .................... ... 7
Why More Women Are Not Giving Their Infants the Benefits of Breast Milk...........8
G general B reastfeeding B barriers ........................................ ......... ............... 9
Barriers to Breastfeeding Prem ature Infants .......................................................9
Health Care Professionals Can Hinder the Breastfeeding Process in the
Neonatal Intensive Care Unit........... ........... ... .................. .. 14
Treatments to Foster Improved Breastfeeding Rates.............................................16
Need for an Educational Package for Health Care Professionals and Mothers of
Infants in the Neonatal Intensive Care Unit................................. ..... ............. ...19
Synactive Theory and Breastfeeding Interventions in the Neonatal Intensive Care
U nit................................... ........... ....... ....... ............ 20
Theory Governing the Behavior of Health Care Professionals ............................. 25
Transtheoretical Model of Behavior Change and Methods of Education ..................31
Sum m ary and R research Questions ........................................ ........................ 33

2 M E T H O D O L O G Y ............................................................................ ................... 35

P a rtic ip a n ts ........................................................................................................... 3 5
R research Interventions................ .. .. .................. .......... ............ ........ 36
Intervention 1: Breast Pump Loaner Closet ....................................... .......... 36









Intervention 2: Health Professional Education Initiative .................................37
Intervention 3: Breastfeeding Guideline ........................ .....................38
Intervention 4: Educational Pamphlet for New Mothers of Neonatal
Intensive Care Unit Infants .......................... ........................ 38
Intervention Plan Modification..................................................... 39
Design .............. ........ .................................39
P ro c e d u re s ..................................................................................................... 4 0
Program Evaluation ....... ... .. ...... .... ... ........... .. .... .... .......41
D ata C o lle ctio n ..................................................................................................... 4 2
D ata A nalysis................................................... 42
H y p oth eses ..............................................................4 3
Adjusting the Alpha Level ............. ... ......... ............. 44

3 R E S U L T S .............................................................................4 6

Intervention Im plem entation .............................................. ............... 46
The Sample ......................... .............................48
Inter-Rater Agreement ............. ......... .......... ........ 50
Demographics ...... .................. ........................ 50
Investigation for Selection Differences ..................................................... 52
Results Per Research Question................... ......................52

4 D IS C U S S IO N ......... ....................................................... .......................................59

The Effect of the Interventions on Breastfeeding Practices in the Neonatal
Inten siv e C are U nit ..................................................................... .....................59
Lim stations ................. ............. ... ......... ..................... .................64
Recommendations for Further Research ............................................... ......69
C conclusions ....................................... ........ .... .. ........ .......... 71

APPENDIX

A OUTLINE OF THE EDUCATION MODULE.......................................................73

B OUTLINE OF ITEMS ADDED TO THE INDIVIDUALIZED CARE PLAN.........75

C OUTLINE OF THE EDUCATIONAL BOOKLET FOR MOTHERS...................76

L IST O F R E F E R E N C E S ......... .. ............... ................. ................................................78

B IO G R A PH IC A L SK E TCH ..................................................................... ..................87
















LIST OF TABLES

Table pge

1-1 Stages of change in which particular processes of change are emphasized.............28

2-1 Hypothesis testing according to dependent variable............................. .............43

3-1 Demographics of the pre-intervention and post-intervention groups with test
statistics for selection differences ........................................ ......... ............... 51

3-2 B reast m ilk feeding initiation rates ........................................ ....................... 53

3-3 Comparison of rates of breastfeeding after 30 weeks gestation.............................55

3-4 Rates of ever breastfed in the neonatal intensive care unit ......................................56

3-5 Rates of breast milk feedings at discharge........................... ...............57
















LIST OF FIGURES


Figure pge

3-1 Number of times breastfed per day after 30 weeks gestation .............................54

3-2 Proportion of the stay that breast milk was provided.............................................58















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

BREASTFEEDING PRACTICES IN THE NEONATAL INTENSIVE CARE UNIT
BEFORE AND AFTER AN INTERVENTION PLAN

By

Roberta Gittens Pineda

August 2006

Chair: Lorie Richards
Major Department: Rehabilitation Science

The benefits of breastfeeding for both mother and infant are cited extensively in the

literature. Premature infants hospitalized in the neonatal intensive care unit (NICU) have

a great need for the benefits that breast milk offers, due to their fragile health states.

However, mothers of very low birth weight infants hospitalized in the NICU have

magnified barriers to the breastfeeding process due to the complexity of medical

conditions that warrant admission to the NICU and the separation of the infant from the

mother to enable medical care. Studies have cited lack of education about lactation

among health care professionals and discrepancies in education dissemination to mothers

as a major barrier to the breastfeeding process. A three-part intervention within the

NICU was implemented that consisted of an educational initiative for health care

professionals who instruct and support mothers, modifications to the individualized care

plan that included a new breastfeeding pathway, and an educational booklet for mothers

with infants hospitalized in the NICU. Change in breast milk feeding initiation rates,









breastfeeding rates, breast milk at discharge rates, and proportion of the hospital stay that

breast milk was provided was investigated between pre-intervention and post intervention

groups. Results indicated general positive trends in all variables, but only one variable

achieved statistical significance. The percentage of infants who were ever breastfed

while in the hospital increased from 25.9% before the intervention to 44.4% after the

intervention, and this reached statistical significance with ap value of .025. Full

implementation of strategies learned in the interventions was questionable. This study

provides partial support of the three-part intervention in facilitating breastfeeding in the

NICU. Possible reasons for lack of change across all variables, as well as other possible

interventions that could affect change, are explored.














CHAPTER 1
INTRODUCTION

Breastfeeding is an important part of the occupation of mothering. However,

mothers of infants admitted into the neonatal intensive care unit (NICU) are not able to

function in the traditional role of mother. They are usually separated from their infants,

and the role of caregiver shifts to health care professionals. In addition, many infants

may be attached to life-saving or monitoring equipment, which can be intimidating for

new parents. Even more intimidating is that many of these infants are fragile or lack

neurological maturity, which affects how the mother will interact with and care for her

infant. This environment presents significant barriers to the provision of breast milk,

including the fragility of the infant, the separation of the infant from the mother, and the

behavior of the health care professionals who are focused on the medical interventions

necessary for these infants. This is unfortunate, because the established benefits of breast

milk may be even greater in these medically fragile and maturationally immature infants.

The rate of breastfeeding in the United States, despite repeated advertisement of

its benefits, is only 71.4% (Li, Darling, Maurice, Barker, & Grummer-Strawn, 2005).

Unfortunately, due to the many barriers to breastfeeding in the NICU environment, this

percentage is significantly lower for infants discharged from the NICU, with breast milk

feedings in premature infants reaching only approximately 50% (Espy & Senn, 2003).

However, despite the medical complexities of the NICU and the shift of care to health

care professionals, with adequate circumvention of barriers, mothers can be supported in

the occupation of mothering through support of breastfeeding.









Studies have identified that health care professional support is predictive of success

with breastfeeding (Swanson & Power, 2005). However, health care professionals must

be given the tools to foster breastfeeding in the complex NICU environment. Therefore,

the aim of this study was to test the efficacy of an intervention to support breastfeeding

practices in the NICU. The intervention centered on health care professional behavior

change through an educational initiative for health care professionals, modifications to

the individualized care plan (ICP) with a breastfeeding protocol, and educational

materials for mothers with infants in the NICU. It was hypothesized that the intervention

plan would foster change in health care professionals, which would then enable positive

changes in breastfeeding practices in the NICU.

The Importance of Breast Milk and Breastfeeding

Breast milk can be provided to the infant either directly through infant suckling at

the breast (breastfeeding) or by having the mother express the breast milk with a pump

and providing the milk via enteral feedings or bottle (breast milk feedings). The health

benefits of breastfeeding for the infant are cited in the literature extensively (Wolf, 2003).

Breast milk has a protective effect against many childhood health problems. Breast milk

differs from formula in that it has unique ingredients that are difficult, if not impossible,

to duplicate. Important components of breast milk are IgA antibodies, which aid in

preventing infection by creating a non-inflammatory response in body cells. This enables

a more active immune system, which demonstrates better defense against infection.

Other factors in breast milk, such as lactoferrin and oligosaccharides have also been

isolated and are believed to prevent mucous attachment, the origin of most infections

(Hanson, 1998; Hanson et al., 2002).









Infant formulas continue to strive to be similar to breast milk and have become

nutritionally advanced in the last decade, however, research continues to illustrate that

breast milk is far superior to formula (Agostoni & Haschke, 2003; Baker, 2003; Wold &

Adlerberth, 2000). Thus far, formula companies have been unable to replicate the exact

ingredients of breast milk. Perhaps predominantly due to the IgA antibodies found in

human milk, breastfed infants have superior protection from many ailments that

compromise health and prevent optimal functioning.

Health Benefits of Breastfeeding for the Full Term Infant

When comparing babies who are fed breast milk to those who are formula fed,

there is a significant reduction in respiratory infections, diarrhea, necrotizing

enterocolitis, meningitis, sepsis, urinary tract infections, atopic dermatitis, celiac disease,

and inflammatory bowel disease in the breastfed babies (Dai & Walker, 1998; Hanson,

1998; Hylander, Strobino, & Dhanireddy, 1998; Laubereau et al., 2004; Marild, Hansson,

Jodal, Oden, & Svedberg, 2004; Wold & Adlerberth, 2000). Although preliminary

studies have not been conclusive, it is also suggested that allergies and asthma are also

diminished among breastfed babies (Kemp & Kakakios, 2004; Oddy et al., 2004).

Breastfed babies have a diminished risk of sudden infant death syndrome (Alm et al.,

2002; McVea, Turner, & Peppler, 2000), as well as a significantly lower risk of mortality

after the neonatal period (Chen & Rogan, 2004).

Because it is associated with less infant illness, breast feeding may cut medical

expenses for the infant. Ball and Wright (1999) addressed excess medical costs for 3

common childhood illnesses: gastrointestinal infection, respiratory tract infection and

otitis media among breast fed versus formula fed infants in the first year of life. There

was evidence that children who were never breast fed incurred significantly more office









visits, hospitalizations, prescriptions and subsequently had higher health care costs (Ball

& Wright, 1999). Thus, the health advantages associated with breastfeeding create less

financial burden as health care costs diminish (Ball & Wright, 1999) and, more

importantly, they improve the quality of life and health status among mother-infant

dyads.

There have been a multitude of studies that have also investigated health and

developmental benefits of breast milk for premature and high risk neonates (Callen &

Pinelli, 2005). The fragile health states of these infants make them more susceptible to

infection, gastrointestinal problems, and life threatening illnesses than full term infants

(Lanari et al., 2001; Lugo-Vicente, 2003). Therefore, breast milk is perhaps more

important in this fragile population, because it diminishes the risk of multiple medical

problems, which can complicate the medical course and put them at a higher risk of

developmental sequelae.

Health Benefits of Breastfeeding for Premature Infants

Breast milk fed infants from the NICU differ significantly from formula fed infants

in incidence of infection and diagnosis of sepsis/meningitis (Hylander et al., 1998),

necrotizing enterocolitis, and retinopathy of prematurity (Hylander et al., 1998; Hylander,

Strobino, Pezzullo, & Dhanireddy, 2001; Schanler, Hurst, & Lau, 1999). Breast fed

premature babies have been noted to experience less stress than bottle fed infants as

evidenced by fewer episodes of oxygen desaturation and temperature instability (C. H.

Chen, Wang, Chang, & Chi, 2000). Breastfeeding has been cited as an intervention that

has lasting, long term benefit beyond discharge from the hospital (Harrold & Schmidt,

2002), and studies have detected significant reductions in length of stay among breastfed

premature infants (Gomez, Acosta, Sevillano, Curbelo, & Alvarez, 1997).









Long Term Benefits of Breastfeeding

More recent studies are suggesting that the effects of breast milk extend beyond the

period of infancy and early childhood, and promote long term immunity and protection

from chronic diseases. Lower risk of developing childhood cancers, obesity, type I

diabetes, and cardiovascular disease have been cited as long term benefits of

breastfeeding (Davis, 2001; Hanson, 1998; Schack-Nielsen & Michaelsen, 2006; Singhal,

Cole, Fewtrell, & Lucas, 2004). Studies of long term effects of breast milk on premature

infants have also found benefits with lower blood pressure readings in adolescence

(Owen, Whincup, Odoki, Gilg, & Cook, 2002). Despite concerns that breastfeeding

results in suboptimal growth in infancy, studies have demonstrated an increased growth

velocity in late childhood in breastfed groups (Schack-Nielsen & Michaelsen, 2006).

Additionally, improved parental attachment in the teenage years has been linked to

breastfeeding (Fergusson & Woodward, 1999).

Developmental Benefits of Breastfeeding

Research also points to the importance of breastfeeding on infant development.

Breastfeeding results in improved oral motor development and orthodontics (Page, 2001),

with early weaning increasing the risk of malocclusion, mouth breathing, dysfunctional

oral motor development and subsequent suboptimal speech development (Neiva, Cattoni,

Ramos, & Issler, 2003; Viggiano, Fasano, Monaco, & Strohmenger, 2004). Breastfed

infants have improved visual motor skills (Birch et al., 1993), have better responses to

pain (Gray, Miller, Philipp, & Blass, 2002) with improved neurobehavioral organization

(Hart, Boylan, Carroll, Musick, & Lampe, 2003) and have demonstrated improved scores

on mental functioning (Gomez-Sanchiz, Canete, Rodero, Baeza, & Avila, 2003).









Premature infants have a greater risk of poor neurological outcome, which suggests

that breast milk may be critical to enable optimal developmental functioning. Research

has demonstrated improved cognitive and motor functioning scores among premature

infants who had breast milk feedings (Lanari et al., 2001; Schanler et al., 1999). Studies

demonstrate improved cognitive scores and intelligence quotients that continued to be

evident through middle childhood among breastfed infants (Lucas, Morley, Cole, &

Gore, 1994; Lucas, Morley, Cole, Lister, & Leeson-Payne, 1992; Morley, Cole, Powell,

& Lucas, 1988; Smith, Durkin, Hinton, Bellinger, & Kuhn, 2003).

Benefits of Breastfeeding for the Mother

In addition to the benefits given to the baby through breastfeeding, there are also

benefits for the mother. Women who succeed with breastfeeding comment on the special

bonding experience (Torgus, Gotsch, & La Leche League International., 1997). Women

who breastfeed have less postpartum bleeding (Chua, Arulkumaran, Lim, Selamat, &

Ratnam, 1994) and have a faster rate of pregnancy related weight loss (Dewey, Heinig, &

Nommsen, 1993). Women who breastfeed also postpone ovulation (Rea, 2004), and

breastfeeding has been demonstrated to serve as a natural and effective birth control

method in the postpartum period ("How breast-feeding postpones ovulation," 1985;

"What is best birth control to use after having a baby?," 1989). In addition, women who

succeed with breastfeeding lower their risk of osteoporosis (Chantry, Auinger, & Byrd,

2004), obesity (Rooney & Schauberger, 2002), ovarian cancer, breast cancer (Mikiel-

Kostyra, 2000), diabetes and rheumatoid arthritis (Rea, 2004).

Mothers of preterm infants have additionally reported an improved sense of well

being, as they feel that they are actively contributing to the health of their babies

(Schanler et al., 1999). Having an infant in the NICU is a difficult challenge. Mothers









may feel shut off from their infant as the nurses take on the role of primary caregiver.

Being able to provide the best source of nutrition can be one task that embraces the

mother in her role and fosters parental involvement, as it is something only she can do for

her baby.

Current Breastfeeding Recommendations

The health benefits of breastfeeding for mother, baby and health care systems are

evident and extensive. Therefore, the American Academy of Pediatrics (AAP), as well as

the American Dietetic Association, have responded to the benefits of breastfeeding for

mother and baby by recommending exclusive breastfeeding for the first 6 months with

breastfeeding and supplemental solids until the infant is 1 year old ("Breastfeeding and

the use of human milk. American Academy of Pediatrics. Work Group on

Breastfeeding," 1997). The World Health Organization recommends breastfeeding for at

least 2 years ("The optimal duration of exclusive breastfeeding: results of a WHO

systematic review," 2001). However, breastfeeding statistics continue to demonstrate a

gap between these recommendations and how the general population of mothers in the

United States chooses to feed their infants (Li et al., 2005).

Some women never breastfeed, some breastfeed exclusively, some supplement

breastfeeding with bottle feeds of human milk, some supplement breastfeeding with

bottle feeds of formula, some bottle feed formula only, some bottle feed breast milk only,

and some women start out breastfeeding and completely wean once formula is

introduced. Statistics from the year 2003 indicated that 71.4. % of women in the general

population initiated breastfeeding while in the hospital, and 35.1% of mothers were still

breastfeeding when their babies turned 6 months of age. At one year of age, 16.1%

continued to provide some breast milk for their infants (Li et al., 2005). Variable rates of









breastfeeding have been reported for infants in the NICU. The rates for infants receiving

some breast milk at some point range from 50% to 83% (Byrne & Hull, 1996; Espy &

Senn, 2003; Meier, Engstrom, Mingolelli, Miracle, & Kiesling, 2004; Smithers, McPhee,

Gibson, & Makrides, 2003; Yip, Lee, & Sheehy, 1996). However, studies have found that

the rates of breast milk feeds at discharge are 64%, with the rate of breastfeeding being

38% (Yip et al., 1996). One study found that at 4 months of age, only 24% of infants born

at less than 33 weeks gestation continue to receive some breast milk feedings (Smithers et

al., 2003). Subsequently, breastfeeding rates at hospital discharge for infants born

prematurely are significantly lower than those of full term, healthy infants (Yip et al.,

1996).

To understand the suboptimal breastfeeding rates for premature and high risk

neonates, it is beneficial to investigate the barriers to breastfeeding. By understanding

the barriers to breastfeeding, appropriate interventions can be developed and

implemented to facilitate improved breastfeeding practices.

Why More Women Are Not Giving Their Infants the Benefits of Breast Milk

Maternal demographics are strong predictors of breastfeeding. Women with higher

socioeconomic status, more education, previous children but smaller family size,

Caucasian race, and women who are married are more likely to succeed with

breastfeeding (Bueno et al., 2003; Kronborg & Vaeth, 2004; Mitra, Khoury, Hinton, &

Carothers, 2004). However, perinatal medical condition is also an important predictor of

successful breastfeeding (Espy & Senn, 2003; Powers, Bloom, Peabody, & Clark, 2003).

Scott (2006) discovered that the infant being admitted to the intensive care unit was the

strongest predictor of not being exclusively breastfed at discharge (Scott, Binns, Graham,

& Oddy, 2006). Other studies have concluded that having a cesarean section, as well as









having a low birth weight infant, makes a woman less likely to breastfeed (Hwang,

Chung, Kang, & Suh, 2006). Demographic factors as well as medical condition and type

of delivery have been shown to be strong influences on the decision to breastfeed and the

success of breastfeeding.

General Breastfeeding Barriers

The barriers to breastfeeding full term, healthy infants include lack of family and

spouse support and perceptions of lack of support; (Arora, McJunkin, Wehrer, & Kuhn,

2000; Matthews, Webber, McKim, Banoub-Baddour, & Laryea, 1998; Scott et al., 2006),

social withdrawal and isolation (Stewart-Knox, Gardiner, & Wright, 2003), perceived

inconvenience (Zimmerman & Guttman, 2001), perceived inadequacy to provide

adequate nutrition (Arora et al., 2000; Matthews et al., 1998), early supplementation or

first feeding of formula (Wheeler, Chapman, Johnson, & Langdon, 2000), lack of

appropriate education (Arora et al., 2000), functional problems with the process of

breastfeeding; (Bick, MacArthur, & Lancashire, 1998), intent to return to work (Arora et

al., 2000; Matthews et al., 1998; Piper & Parks, 1996; Ryan, Wysong, Martinez, &

Simon, 1990), and maternal illness (Black & Hylander, 2000; Riskin & Bader, 2003).

Barriers to Breastfeeding Premature Infants

Breastfeeding challenges are stronger and even more numerous for the high risk

neonate, despite these babies having an even greater need for human milk. Mothers of

infants who are born prematurely have unique challenges to successful breastfeeding.

One barrier to breastfeeding the premature infant is that when an infant is born

prematurely and warrants admission into the NICU, the mother is separated from her

baby (Black & Hylander, 2000). The time after birth is very different for these mothers

compared to those with full term infants. There is usually not a period of being able to put









the baby to breast immediately after birth, and breastfeeding may not be possible for

several weeks or months, depending on the infant's level of prematurity and medical

instability.

When visitation is possible, mothers may visit their baby in the intensive care unit.

Here, they may have difficulty with the transition to motherhood as the doctors and

nurses make decisions related to the care of the baby, including whether or not the mother

may hold her new baby (Holditch-Davis & Miles, 2000; Lupton & Fenwick, 2001). The

machines and equipment present and being utilized by the baby in the NICU can be

overwhelming for many parents (Wheeler et al., 2000), and this environment is very

different from the quiet, home-like environment one would typically envision during the

first days of the baby's life. An additional barrier is that the ability to achieve let down, in

which breast milk begins to flow during infant feeding and pumping, is hindered by the

inability to relax in this stressful environment (Beresford, 1984; Nyqvist, Ewald, &

Sjoden, 1996; Wheeler, Johnson, Collie, Sutherland, & Chapman, 1999).

Many low birth weight infants are unable to breast feed for several weeks or

months following birth (Hill, Andersen, & Ledbetter, 1995). Their gastrointestinal

systems are immature and feedings can be dangerous or life threatening. During the first

days, a baby may be fed intravenously or through an orogastric or nasogastric tube, in

which feeds may be slowly introduced and advanced. When the gastrointestinal system is

ready for bolus feeds directly into the stomach, the baby's immature central nervous

system may not enable consistent presentation of sucking and swallowing responses to

enable safe oral feeding (Nyqvist, Sjoden, & Ewald, 1999; Ziemer & George, 1990).

Although breastfeeding may not be possible initially, breast milk can be expressed by the









mother with a breast pump, and the infant can be advanced on gastric feeds with breast

milk.

Just as the first feeding by breast is a good predictor of sustained breastfeeding in

full term infants, timely pumping for those mothers who are unable to put the baby to

breast is an important predictor of sustained breastfeeding in the premature baby (Jaeger,

Lawson, & Filteau, 1997). Women of premature babies may express their breast milk and

supply it to hospital staff so that the baby may be tube fed with human milk instead of

infant formula (Meier & Brown, 1996). Additionally, this process establishes and

maintains a milk supply so that the mother will not have diminished or absent milk

supply, when the baby is stable enough to engage in the breastfeeding process. Barriers to

breastfeeding related to this early process include increased amounts of stress (Docherty,

Miles, & Holditch-Davis, 2002; Miles, Funk, & Kasper, 1992) and time constraints

placed on these new mothers, difficulty in acquiring hospital grade breast pumps for milk

expression, lack of special bonding and emotional feedback received from using a pump,

delayed initiation of milk expression, separation from the infant, reliance on medical

technology to feed the baby, and psychological adjustment to the idea of not being able to

breastfeed for weeks or even months (Byrne & Hull, 1996). With the mother

experiencing stress associated with coping with her sick baby (Miles et al., 1992) and a

shift of care from the mother to the baby after the birth, there may be delayed initiation of

pumping and lack of accessibility of hospital grade pumps to promote milk supply in an

efficient manner.

New studies are highlighting the importance of investigating barriers at different

time periods during an infant's hospitalization (Callen & Pinelli, 2005). If a mother









successfully overcomes the challenge of maintaining her milk supply, there are additional

challenges as an infant approaches discharge from the hospital. Poor central nervous

system maturity may initially prevent complete success with breastfeeding, and

dysphagia is common in this population (Hill, Hanson, & Mefford, 1994). The literature

cites problems with the mechanics of breastfeeding a premature baby as a barrier to

breastfeeding (Kavanaugh, Mead, Meier, & Mangurten, 1995). Once discharge is

approaching, there frequently is little time to enable a mom and baby to achieve

successful breastfeeding (Meier & Brown, 1996). Bottle feeding is often preferred as it

allows nurses to orally feed the baby when the mother is not present and the exact amount

ingested can be accurately measured (McGrath & Braescu, 2004). Additionally, infants

can be fed more passively with bottle feeding compared to the active process of

breastfeeding. Although this can have negative side effects of desaturations and

bradycardic events as well as increased risk of gastroesophageal reflux, it is frequently

preferred because of the efficiency of oral feeding. However, breastfeeding can be

achieved in this population despite the preferences of health care staff and the challenges

that must be overcome.

Infants in the NICU are fed according to a schedule, typically every 3 to 4 hours,

and may be fed via bottle, tube or breast to optimize the nutritional status. The inability

of a preterm baby to breastfeed on demand in an environment with scheduled feedings

via different modes is a significant barrier to breastfeeding (Black & Hylander, 2000). It

undermines the typical procedures associated with breastfeeding a full term infant which

involves feeding a baby when he/she shows hunger signs and not supplementing until

breastfeeding is well established. This allows for the infant to ingest a smaller feeding









and thus become hungrier and to have a more rigorous, larger feeding for the next one.

However, scheduled feedings of specific amounts can affect the transition to active

breastfeeding in the NICU. If the infant typically receives a prescribed amount of breast

milk by bottle or nasogastric tube and the mother attempts to breastfeed, not knowing the

exact amount of breast milk ingested by breastfeeding may result in the health care

professional doubting if there was adequate intake. Thus supplementation frequently

occurs, which inhibits the next breastfeeding session, decreases the demand for breast

milk produced by the mother and diminishes milk supply, and thus becomes a cyclic

problem.

Diminished milk supply is cited extensively in the literature as one of the

significant barriers to breastfeeding in the NICU (Callen & Pinelli, 2005). Among the

earliest of premature infants, the average duration of providing breast milk is 4-5 weeks

(Byrne & Hull, 1996). Lack of ability to engage in active breastfeeding due to the health

status of the infant, lack of presentation of sucking and swallowing capabilities of the

infant and the need for the mother to demonstrate consistent milk expression via a breast

pump to establish and maintain a milk supply all contribute to diminished milk supply in

mothers of infants hospitalized in the NICU.

Infants who are born prematurely have different nutritional needs than full term

infants. Thus, there are premature infant formulas that are utilized in the neonatal period

and many are used until one year of life. When gastric feeds are being established,

physicians are concerned with establishing a good weight gain trend. If an infant is not

gaining weight as desired, the physician may increase caloric density or add lipids to

infant formula or expressed breast milk to foster weight gain. Frequently nutrients and









calories are added to human milk by the way of human milk fortifiers (Chan, 2001),

which promote establishment of a good weight gain curve. However, this is a barrier to

breastfeeding as the mother perceives that the composition of her breast milk is not

adequate to promote the health of her child. She may perceive that formula or fortified

breast milk by bottle is essential to enable the appropriate milk composition (Kavanaugh

et al., 1995). However, studies suggest that mother's milk of premature babies differs

from that of full term infants with the most notable differences evident between 4-6

weeks after delivery (Gross, David, Bauman, & Tomarelli, 1980). Additionally, the use

of hind milk, the milk at the end of a breastfeeding session that is very high in fat content,

has been shown to facilitate weight gain in premature babies (Slusher et al., 2003).

Research is identifying that there are factors in the hospital setting that influence

breastfeeding decisions. The site of care is a strong predictor of choice and success with

breastfeeding (Powers et al., 2003). In addition, literature is highlighting the important

role of health care professionals on the decision to initiate and continue breastfeeding

(Nyqvist, Sjoden, & Ewald, 1994; Swanson & Power, 2005).

Health Care Professionals Can Hinder the Breastfeeding Process in the Neonatal
Intensive Care Unit

There are many inconsistencies in what parents are educated about and many

disparities in what parents are instructed to do by health care professionals, and this can

be confusing and frustrating for new mothers (Byrne & Hull, 1996; Nyqvist et al., 1994).

One study identified that 48% of mothers reported receiving conflicting advice about

breastfeeding in the NICU (Jaeger et al., 1997).

Mothers rely on health care professionals in the NICU to provide accurate,

complete, and consistent information about breastfeeding their high risk infant. Many of









the mothers of preterm infants have not had an opportunity to fully prepare for

motherhood before the birth of their baby. Some may have planned on taking a

breastfeeding or parenting class, but the early arrival dampened these plans. In addition,

the NICU is a medically complex environment, and parents need guidance on how to

function in their role as mother with the environmental constraints. The literature

suggests that there is a lack of health care professionals who are trained in lactation and

breastfeeding with premature babies, and that many health care professionals have

incorrect knowledge and negative beliefs about lactation (Berens, 2001; Pantazi, Jaeger,

& Lawson, 1998; Register, Eren, Lowdermilk, Hammond, & Tully, 2000; Spicer, 2001).

Yet, it is the health care professionals in the NICU, despite lack of education, who are

teaching and instructing these new mothers on breast milk feedings and breastfeeding.

Health care professionals can influence breastfeeding behaviors, and their own

values and beliefs concerning breastfeeding can have supporting or damaging results on

the breastfeeding process (Ekstrom, Matthiesen, Widstrom, & Nissen, 2005). Studies

have identified that education and training can affect attitudes and knowledge about

breastfeeding (Bernaix, 2000; Siddell, Marinelli, Froman, & Burke, 2003; Swanson &

Power, 2005). However, to date, there are no studies that have investigated the indirect

impact of health care professionals' behavior change with acquisition of knowledge and

attitude change on breastfeeding outcomes in mothers and infants in the NICU.

The American Academy of Pediatrics issued a statement in February 2005 that

stated that breastfeeding or human milk feedings are recommended for all healthy,

premature and high-risk infants for whom breastfeeding is not specifically

contraindicated. It further states that health care professionals should provide complete,









current and accurate information to parents on the benefits and techniques involved with

breastfeeding (Gartner et al., 2005). However, to date there has been no specific,

standard set of tools developed and utilized to achieve education of health care

professionals to enable consistent information dissemination to parents.

Treatments to Foster Improved Breastfeeding Rates

The United States Surgeon General, David Satcher, identified breastfeeding as a

national health priority and released the "Health and Human Services, Blueprint for

Action on Breastfeeding" in October 2000. In response to this, The World Health

Organization in conjunction with UNICEF is promoting breastfeeding through the baby

friendly hospital initiative. To be designated as "baby friendly," the hospital must follow

the ten steps to successful breastfeeding:

Every facility providing maternity services and care for newborn infants should:

* Have a written breastfeeding policy that is routinely communicated to all health
care staff.

* Train all health care staff in skills necessary to implement this policy.

* Inform all pregnant women about the benefits and management of breastfeeding.

* Help mothers initiate breastfeeding within half an hour of birth.

* Show mothers how to breastfeed, and how to maintain lactation even if they should
be separated from their infants.

* Give newborn infants no food or drink other than breast milk, unless medically
indicated.

* Practice rooming-in (i.e., allowing mothers and infants to remain together) 24 hours
a day.

* Encourage breastfeeding on demand.

* Give no artificial teats or pacifiers (also called dummies or soothers) to
breastfeeding infants.









* Foster the establishment of breastfeeding support groups and refer mothers to them
on discharge from the hospital or clinic.

Source: Protecting, Promoting and Supporting Breastfeeding: The Special Role of
Maternity Services, a joint WHO/UNICEF statement published by the World Health
Organization.

These ten steps specifically involve how health care professionals and the hospital

system will deal with mother-infant dyads on regular maternity floors within a hospital.

Once these ten steps are put into practice, the hospital may apply for designation as a

baby friendly hospital. Hospitals that have been through the process of baby friendly

designation have demonstrated improvement with breastfeeding rates (Philipp, Malone,

Cimo, & Merewood, 2003). For example, Boston Medical Center was designated as

baby friendly in 1999, with increased breastfeeding rates of 58% in 1995 to 86.5% in

1999. Breastfeeding rates were maintained at this high rate from 1999 to 2001.

Although the baby friendly designation is specifically for the maternity floors of a

hospital system and not designed for the unique needs of the high risk population, there

have been positive effects on breastfeeding practices in the NICU following designation

(Vannuchi, Monteiro, Rea, Andrade, & Matsuo, 2004). However, a program specifically

designed for the high risk population with its significant barriers to breastfeeding, could

have the potential for greater enhancement in breastfeeding rates in the NICU Premature

infants are a unique population and warrant individualized breastfeeding strategies and

interventions (Kavanaugh et al., 1995; Meier, 2001)

Many hospitals have implemented practices that will educate and promote

breastfeeding practices for infants within the NICU. However, only a few have evaluated

their programs for effectiveness. The Rush Mother's Milk Club has proven to be

effective in increasing breastfeeding rates (Meier et al., 2004) by enabling free access to









hospital grade breast pumps, by offering lactation support 24 hours a day, by use of cue

based feeding when an infant consumes at least 50% of feeds orally, and by providing of

breastfeeding peer support. Other studies have found positive increases in breastfeeding

initiation with the introduction of counseling as well as contact with lactation consultants

among mothers with low birth weight infants (Pinelli, Atkinson, & Saigal, 2001; Sisk,

Lovelady, Dillard, & Gruber, 2006). A workbook program introduced at 2 different time

periods during the hospital stay was also shown to have positive effects on breastfeeding

with premature infants in one hospital setting (Jang, 2005).

Many papers have documented specific protocols to instruct mothers and promote

breastfeeding (Isaacson, 2006; Premji, Paes, Jacobson, & Chessell, 2002; Spicer, 2001),

but no research has been conducted to determine the effectiveness of such

recommendations. Although studies have shown increased knowledge acquisition by

health care professionals in the NICU following an education plan (Siddell et al., 2003),

there are no studies that have investigated the effect of education of health care

professionals coupled with protocols and standard written information for parents on

changes in breastfeeding practices in the NICU.

Of all interventions for breastfeeding with the high risk neonate, the Rush Mother's

Milk Club is probably the most well known. The health care providers in the NICU at

Rush University have an increased level of knowledge regarding breastfeeding in the

NICU. With this knowledge, they are able to implement advanced strategies, such as

putting breast milk in a centrifuge to modify the fat content to promote weight gain

(Meier, 1998). Many studies have been conducted to evaluate the effectiveness of the

Rush Mother's Milk Club with positive results (Meier et al., 2004). However, the high









level of education about breastfeeding among health care professionals at Rush

University likely underlies the capability to implement the more advanced interventions.

For many hospitals, basic education on lactation with high risk infants is lacking. In

addition, many interventions that have proven to be effective, including the use of free

access to pumps and accessibility to lactation counseling, have associated costs, which

many hospitals do not have budgets to support.

Need for an Educational Package for Health Care Professionals and Mothers of
Infants in the Neonatal Intensive Care Unit

Breastfeeding and human milk feedings are possible and beneficial in the NICU,

however, there is significant support and education that must occur to enable success

among mothers in the NICU (do Nascimento & Issler, 2004). Education and treatments

should be based on research with premature and high risk infants, as they have unique

needs in the breastfeeding process (Meier, 2001). The use of developmental care

practices can drive the understanding of appropriate breastfeeding interventions based on

infant readiness cues (Karl, 2004).

Although there have been many articles and books written on the subject of

breastfeeding with the high risk neonate, there was no up to date, comprehensive,

evidence based education packet with complementary information for both parents and

health care professionals available on the market. By understanding each of the

challenges to breastfeeding the medically fragile infant and the specific developmental

and nutritional needs of the high risk infant, an education initiative can be developed and

then evaluated for efficacy.









Synactive Theory and Breastfeeding Interventions in the Neonatal Intensive Care
Unit

While investigating, developing and implementing appropriate interventions

targeted to improve breastfeeding rates, it is important to consider the vulnerability of the

special population in the NICU and the impact of environmental stressors on this

population. One theory that can be used to guide appropriate interventions in the NICU

is the synactive theory, which identifies the process of neurobehavioral maturation of the

infant. Breastfeeding interventions should be individualized, based on infant readiness

cues and tailored to the responses of the infant (Blackburn, 1998). Review of the

synactive theory and its application to breastfeeding should be part of any education

initiative for health care professionals who serve the vulnerable infants in the NICU.

The synactive theory was developed by Heidelese Als in the early 1980's (Als,

1982). The process of developmental care, related to the synactive theory, is intended to

facilitate a well organized, stable infant who may optimally grow and develop.

Developmental care has been instituted in many neonatal intensive care units around the

country as a developmental care initiative. It provides a framework for interacting with

these fragile infants without jeopardizing health. The synactive theory of development

describes the process of neurobehavioral maturation related to an infant's internal and

external environment. As the infant attempts to interact with the external environment, a

dynamic process occurs internally among 5 different subsystems. The dynamic process

among the 5 subsystems can explain the behaviors and responses exhibited by the

premature infant and can guide appropriate interventions.

The 5 distinct, yet interdependent subsystems are physiological or autonomic,

motor, state organization, attention and interaction, and the state regulation subsystems.









These subsystems are believed to impact the functional organization of the infant's

system in an ordered fashion. The subsystems are not hierarchical, but they are believed

to be ordered and interdependent (Als, 1982; Als, 1994).

The physiological subsystem is considered the core of the system. It is the

foundation for which all of the other systems gain stability. This physiological subsystem

allows the infant to have control over autonomic functions such as voiding, breathing,

maintaining steady vital signs, and processing nutrition. The motor system provides

control over movement, muscle tone, and posture. The state subsystem gives the infant

control over his/her level of consciousness. It enables the infant to move through

identifiable states and move smoothly from one state to another. The

attentional/interactive subsystem enables control over functional responses to stimulation

in the environment and governs the ability to interact. The state regulation subsystem

gives the infant the ability to balance environmental stressors and recover by modulating

all the other systems (Als, et al, 1982). While the autonomic subsystem serves as the

foundation of the system, the state regulation subsystem serves as the gate keeper and is

achieved with increasing maturity.

A cone shaped diagram is used to represent the complex development of the infant

as it relates to the five subsystems (Als, 1982). The cone has its tip at the bottom with the

funnel going upward. The five subsystems can be viewed at the top of the open cone. At

the smallest center is the physiological subsystem with the remaining (motor, state,

attentional/interaction) forming layers outside the center, much like an onion. The

youngest fetus is represented at the bottom of the cone and has with it only components

(not yet a fully developed system) from the autonomic subsystem. This indicates that









infants who are born early are unable to integrate the higher order systems. Stressors

within the system interfere with the physiologic capabilities of the infant. Subsequently,

early premature infants are incapable of any interaction and need all their energy to

maintain homeostasis of the system to sustain life. There is also instability in the

physiologic system, which is why premature infants frequently have medical or

physiological problems when born early and have to contend with the stressful

environment.

With the earliest fetus at the bottom of the cone, increasing gestational age is

associated with increasing maturity spreading out to the other layers of the system. With

increasing gestational age and thus maturity, the infant may extend its control out to the

next level, the motor subsystem. The infant may demonstrate improved muscle tone and

postural control. This concept parallels the literature, which demonstrates improvement

in muscle tone and reflex development with increasing gestational age (Allen & Capute,

1990). Further maturity may extend the infant's control out to the state subsystem

enabling the infant to demonstrate some awake periods and to smoothly transition from

one state to another. As maturity continues, the infant may be able to achieve some

attention and interaction with caregivers and the environment. Lastly, as the infant

approaches term and achieves more maturity, he/she will be able to tolerate stressors,

cope with them, reorganize and continue interaction without being knocked down to

functioning at the lower subsystems.

The term synaction refers to the relationship between all the subsystems and how

instability in one system has the potential to affect all the other subsystems and thus the

integrity of the child's health and well-being (Als, 1982). On the right side of the cone









are gestational ages that reflect the increasing maturity of the system (dependent on the

subsystems) with increasing gestational age. On the left side of the cone are influences of

the environment on the maturation of the system, with a break in the intrauterine and

extrauterine environment before term to indicate the premature birth, thus representing

the role of environmental stresses before full maturity occurs. With earlier birth and

more stressors from the environment, there will be a resultant decrease in

neurobehavioral maturity.

When an infant is stressed from the environment, he/she may initially demonstrate

stress reactions based on the predominant level of neurobehavioral maturation. If he/she

is primarily functioning in the physiological state, he/she may demonstrate bradycardic

events, oxygen desaturations, hiccups, stooling, or spitting up. If he/she is primarily in

the motor state, he/she may demonstrate grimacing, arching, saluting, finger splaying, or

sitting on air. If he/she is in the state subsystem he may shut down or move to a light

sleep state. If he/she is in the attentional/interactive subsystem, he/she may avoid

interaction by turning away. The infant has the capability to re-achieve organization with

time outs or specific strategies designed to help him cope. Interventions designed to help

infants cope with stressors include providing boundaries, swaddling, positioning in

flexion, bringing hands to mouth, minimizing environmental stimulation, non nutritive

sucking, and enabling grasping. Infants additionally will demonstrate approach signals

such as smiling, mouthing, ooh face, cooing, quiet and alert state, and soft and relaxed

facial expressions when they are ready for interaction (Hussey-Gardner, 1996). Once

reorganized and demonstrating approach signals, the stressor may be reintroduced slowly.









The synactive theory defines the subsystem along with stress and coping signs

consistent with each of the subsystems to enable caregivers to identify and respond to

behaviors appropriately. When stress signs are recognized, the caregiver can then

withdraw the stressor that contributed or help facilitate the infant to cope. Once a time

out is given and the infant reorganizes, the treatment or interaction can continue. This

"give and take" enables the infant to function optimally within the environment and

allows him/her to continue to benefit from interaction and stimulation, including

breastfeeding, as he/she tolerates. The synactive theory proposes an approach for each

individual child that is adapted to fit the needs of that infant. It promotes infant

development to occur as normally as possible, despite medical complications and

immaturity brought on by an early birth.

Infants born at earlier gestational ages and with decreased neurobehavioral

maturation are not capable of handling environmental stressors typically experienced by

newborns. When interventions for these neurobehaviorally immature infants are done

without respect for readiness cues, the infant is at risk for regressing to one of the more

primitive states, putting them at risk of developmental and medical sequelae.

Breastfeeding is an environmental stressor. Without observing infant readiness cues and

introducing breastfeeding at the appropriate time individualized for each infant, optimal

responses to the environment as well as optimal neurological maturation are delayed.

Therefore, breastfeeding cannot be introduced at a prescribed time or introduced in the

same way that it would be for a full term infant, but it must be based on the

neurobehavioral maturation of the infant and advanced according to stress and readiness

cues. These concepts need to be in any educational program for both health care









professionals and for mothers of infants in the NICU to help them implement the best

breast milk feeding program for these infants.

While the synactive theory defines the appropriate time and way to introduce

interventions, it also assists with understanding that many extremely low birth weight

infants and low birth weight infants are too neurologically immature and fragile to

engage in any breastfeeding. Subsequently, mothers of infants in the NICU need

equipment that will enable them to achieve and maintain a milk supply, in the absence of

infant suckling at the breast, until the infant is appropriate for nutritional breastfeeding.

Hospital grade breast pumps that will enable long term milk expression are necessary for

mothers with infants in the NICU to maintain adequate milk supplies while they are

waiting for their babies to become medically and developmentally stable enough to

engage in feeding at the breast.

Understanding the synactive theory and implementing developmentally supportive

care can instruct and guide interventions and NICU practices as they relate to

breastfeeding the premature infant. One case study in the literature highlighted the

significant benefits of a developmentally supportive plan on the breastfeeding process in

a premature infant (Nyqvist et al., 1996). The synactive theory should guide the

development of any educational module and inservice for health care professionals who

serve infants in the NICU.

Theory Governing the Behavior of Health Care Professionals

Studies have identified that there is a lack of education about lactation and lack of

consistent support and instruction about breastfeeding among health care professionals in

the NICU (Ekstrom, Widstrom, & Nissen, 2005; Pantazi et al., 1998). Having educated

health care professionals will not necessarily impact breastfeeding practices. It is how









those health care professionals respond and utilize that education to execute new

interventions that will foster change and subsequent improvement in human milk

feedings. The behavior of health care professionals in the NICU needs to change to

support the breastfeeding process. Studies have shown that behavior change is much

more successful when interventions are matched to the stage of readiness to change

(Prochaska, Prochaska, & Levesque, 2001). The transtheoretical model (TTM) provides

a description of how the individual's state of readiness to change translates into

behavioral change.

The premise of the TTM is that there are several stages associated with behavior

change. Individuals go through these stages on their way to making a change. They may

start anywhere along the continuum of the 5 stages and they may move forward or

backward or skip stages, but there is some progression through the stages on their way to

behavior change. The TTM has been used to describe many health behavior changes,

such as use of sunscreen, use of condoms, self examination breast checks, smoking

cessation and initiating an exercise plan. Appropriate interventions can be tailored to the

stage of readiness to change.

The five stages of the TTM are precontemplation, contemplation, preparation,

action and maintenance. Each stage identifies if the individual has an intention of

changing behavior and identifies how significant the intention to change behavior is

(Prochaska & DiClemente, 1983). Someone in the precontemplation stage does not

intend to take action within 6 months. A person in the contemplation stage intends to

take action within the next 6 months. Someone in the preparation phase intends to take

action in the next 30 days. The action stage refers to persons who have made obvious









changes less than 6 months ago. The maintenance stage refers to individuals who have

made significant changes more than 6 months ago.

Interventions that are implemented to enable behavior change should be conducted

in such a way that the intervention matches the stage an individual is in, or should be

tailored to how ready the person is for change. Ten fundamental processes that can affect

change have been identified along with interventions that can be matched to the stage of

readiness to change (Table 1-1) (Prochaska et al., 2001). Certain interventions will only

be effective if they are appropriate for the stage of readiness for behavioral change of the

individual. For example if someone is in the precontemplation stage, interventions should

be focused on educational initiatives and strategies to promote reflection about how the

change may impact the person's situation and how it will benefit others, while

interventions for the person in the contemplative stage should be about supporting and

motivating the person to actually initiate the intended behavioral change. These latter

interventions provided to the precomtemplater would not be effective because the person

has not yet formed the conviction that change is needed or desired and could actually

create significant resistance and prevent behavioral change from succeeding (Prochaska

et al., 2001). By enabling stage matched interventions, education and strategies can be

implemented to foster movement across stages to enable successful change.

The TTM has been used to address behavior change within organizations by

targeting employees. By providing stage matched interventions, all employees can be

given opportunities to participate in the change process. Although stage matched

interventions have been shown to facilitate movement toward action, not all employees

may achieve action. Change at the organizational level should include interventions that









are stage matched to each stage of change to give all employees the opportunity for

participation in the change initiative (Prochaska et al., 2001).

Table 1-1. Stages of change in which particular processes of change are emphasized
Stages
Precontemplation Contemplation Preparation Action
Maintenance
Process Conscious
Raising,
Dramatic Relief,
Environmental -
Reevaluation
Self
Reevaluation
Self -
Liberation
Contingency-
Management,
Helping
Relationship,
Counter
Conditioning,
Stimulus Control

To foster behavior change regarding support and information dissemination about

breastfeeding within the NICU, the TTM can be utilized to structure an intervention

program. By matching breastfeeding interventions to stages of readiness to change, all

health care professionals can have the opportunity to participate in the change process.

By introducing interventions that can target health care professionals in each stage of

readiness to change, a meaningful intervention plan can be implemented to foster change

at the organizational level. A breastfeeding intervention for the NICU with 4 parts could

theoretically target individuals in all of the stages of readiness to change.

For those health care professionals in the pre-contemplation stage, there is no

intention of making a behavior change. Health care professionals in the contemplation

stage intend to make a change within the next 6 months. Both of these stages describe









individuals who are not yet ready for action, and interventions for those in either of these

stages would be the same. Interventions for those in these stages should be two-fold.

One intervention, education, would be aimed directly at the health care professional.

With education, conscious raising can be fostered. With it, it is hoped that health care

professionals will have the resources needed to become aware of the need for

breastfeeding support in premature infants and will start to see solutions to the problem.

However, because those in the pre-contemplation stage have no intention of making a

behavior change, there would be no motivation to participate in an educational initiative.

Likewise, those in the contemplation stage also may need incentives to push them to

participate. Therefore, incentives on annual review, food, prizes and continuing

education units could serve as motivation for participation among those who lack

motivation to attend without some perceived personal benefit. With the participation in

the educational initiative, they would be exposed to content of the education that

highlights the great benefits of breast milk and the need for support and education among

mothers. With this increased awareness of the problem and possible solutions, behavior

change could be fostered.

The other intervention should be aimed at achieving some level of support and

education for the mothers. Materials that provide consistent and thorough information

could be issued to all new parents in the NICU to ensure that all mothers receive

information about initiating and sustaining breast milk feedings in the NICU. Although

this intervention would serve to enable education of mothers, it also may serve as a

conscious raising effort for the health care professional, who may be asked for guidance

and support by the mother on information contained in the educational materials. The









health care professional may then better understand the problem and the need for

behavior change to facilitate success with the breastfeeding process.

Individuals in the preparation phase intend to make a change in the next 30 days.

These individuals are ready for action oriented interventions. Therefore, clinical

pathways or protocols could foster change in how they deal with breastfeeding mothers.

Protocols or pathways, which become a part of required paperwork, could theoretically

facilitate professionals to make a commitment to change by giving protocols that

necessitate action.

Those in the action and maintenance stages have already made changes. The

mother's positive experiences could serve as motivation from the environment, and there

could be other motivators for continued compliance from within the organizational

structure, such as acknowledgement on the annual review and identification as one who

has expertise in breastfeeding with high risk infants. Those in the action and preparation

phase also may take an active part in motivating others and facilitating more positive

change related to breastfeeding interventions in the NICU.

Thus based on the TTM, an educational initiative that includes educational

materials to parents, opportunities for education with incentives for health care

professionals, and protocols or pathways of care could be effective in facilitating change

in breastfeeding practices in the NICU, and each is theoretically matched to all of the

stages of readiness to change.

Although the primary focus of the intervention for this research is health care

professional behavior change, interventions structured to enable change in breastfeeding

practices also must target behavior change in the mothers. Theoretically, developed









interventions for the health care professionals can also be matched to mothers in each

stage of readiness to change. An educational booklet for new mothers could enable

conscious raising for mothers in the precontemplation and contemplation stages. A

breast pump loaner closet as well as milk expression guidelines and a breast milk log

(that could be included in the educational booklet) could be appropriate interventions for

those in the preparation and action stages. Appropriate interventions for those mothers in

the action and maintenance stages would include concepts such as the first feeding being

at the breast and enabling breastfeeding while in the hospital to support continued breast

milk feedings in the presence of the decision to initiate breast milk feedings.

Interventions targeted at health care professional change can be structured to

move mothers to decide to breastfeed and help them maintain that behavior once they

start. Theoretically, interventions including a breast pump loaner closet, an educational

booklet for new mothers, a breastfeeding pathway, and an educational initiative for health

care professionals can support behavior change in two different groups, the health care

professionals as well as the mothers. The health care professional group functions as a

primary support for the mothers to initiate and sustain breastfeeding.

Transtheoretical Model of Behavior Change and Methods of Education

Equally important with providing stage matched interventions for health care

professionals is consideration of what mode of learning to utilize for the conscious

raising strategy. Self learning modules can be considered easy to implement and enables

staff to participate in the learning initiatives at their own pace, enables them to take

modules home if work responsibilities prevent participation during working hours, and is

rather inexpensive when compared to other modes. The literature reflects good success

with self-directed learning modules, with good performance on post tests and learning









retention (Coleman, Dracup, & Moser, 1991). When compared with lecture-discussion

formats, self learning methods were comparable in achieving the educational objectives

being targeted within the nursing field (Nikolaj ski, 1992).

Computer based training is another method of education gaining increased

acceptance and use in the last several years with the increasing capabilities of technology.

Harrington and Walker (2004) discovered that, although both groups significantly

improved their post test scores, a group of individuals who engaged in computer based

training did significantly better than individuals in an instructor led course on fire safety

(Harrington & Walker, 2004). Research and experience are beginning to define computer

based learning as a viable option for educational purposes. However, the access to

technology and to the people who format and design the computer systems is a significant

barrier to widespread use of such learning practices today.

Not all individuals have the same learning style. While some may prefer self paced

methods, others may be more motivated and embraced in a face to face lecture and

discussion with peers. Goldrick, Gruendemann, and Larson (1993) found that 64% of

nurses in a pediatric intensive care unit had an abstract learning style and preferred self

learning modules. However, there remained 36% who preferred more traditional methods

(Goldrick, Gruendemann, & Larson, 1993).

Self directed learning, through an educational module, is an effective form of

educating health care professionals. However, not all individuals possess the learning

style necessary for successful completion of self learning modules and prefer lecture-

discussion formats. By providing both forms of educational opportunities, more health









care professionals could be encouraged to participate in an educational initiative.

Computer based training could also be effective if access to technology can be achieved.

Summary and Research Questions

The provision of breast milk has important benefits to infants, especially those in

the NICU who are less healthy and less mature at birth than full term infants. Yet

significant, but not insurmountable, barriers to breast milk feedings and breastfeeding

exist in the NICU. Health care professionals are a powerful influence (Swanson &

Power, 2005). It is proposed that change from the health care professionals can enable

mothers and infants to overcome many of the barriers to breastfeeding, and subsequently,

positive changes in breastfeeding practices can occur in the NICU. Therefore, in this

study, the effect of a 4-pronged education and support intervention to promote

breastfeeding in the NICU, based on the synactive theory and transtheoretical model of

behavior change, will be explored. The four parts of the proposed intervention are a

breast pump loaner closet, a breastfeeding pathway on the individualized care plan, an

educational booklet for mothers, and an educational initiative for health care

professionals who work with infants and mothers in the NICU. The research questions

are

Is there a significant difference in breast milk feeding initiation in very low birth
weight (VLBW) infants admitted to the NICU before and after implementation of
the intervention plan?

Is there a significant difference in the rate of breastfeeding in the hospital among
women with VLBW infants hospitalized in the NICU before and after the
implementation of the intervention plan?

Is there a significant difference in breast milk feedings at discharge in VLBW
infants admitted to the NICU before and after the implementation of the
intervention plan?






34


* Is there a significant difference in the proportion of the hospital stay that breast
milk is provided in VLBW infants admitted to the NICU before and after the
implementation of the intervention plan?














CHAPTER 2
METHODOLOGY

The purpose of this study was to test the efficacy of a 4-part intervention on

improving breastfeeding practices in the neonatal intensive care unit (NICU). The

overall goal was to attempt to develop an effective intervention to assist mother-infant

dyads in the complex NICU environment.

Participants

Very low birth weight (VLBW) infants (<1500 grams) were included in the study if

they were 1) admitted to the Level II or III nursery at Shands Hospital during the study

periods and 2) had a length of stay greater than or equal to 7 days, 3) were admitted to the

NICU within the first 3 days of life, 4) were hospitalized less than 4 months, 5) achieved

full gastric feeds during their stay, and 6) had a hospital stay that did not cross over from

the pre-intervention group time period into the education initiative time period. Very low

birth weight infants were excluded from the study if they 1) had a length of stay less than

7 days, 2) were transferred to Shands Hospital after the third day of life, 3) were

hospitalized greater than 4 months, 4) did not achieve full gastric feeds during the

hospital stay, 5) had a hospital stay that crossed over from the pre-intervention group

time period to the education plan time frame, or 6) had conditions that would make

breastfeeding contraindicated as established by the physician.

Power indicates the probability of rejecting the null hypothesis, if a condition

exists. With a power of 80%, which is frequently used in the literature, there is a 20%

chance of failing to reject the null hypothesis when it should be rejected.









Prior to conducting the study, a power analysis was conducted to determine the

appropriate sample size. The mean and standard deviations of breast milk feeding

initiation were unavailable from other studies to compute an effect size and subsequently

a sample size. Therefore Cohen's Criteria was utilized to make sample size estimations

(Cohen, 1988). According to Cohen, a .2 standard deviation change is a small effect, a .5

is a medium effect, and a .8 is a large effect. For the purposes of this study, a medium

size effect was selected. By using Cohen's criteria and determining the sample size

necessary with a power of 80%, alpha of .05, and looking for a medium size effect of .5,

Cohen's Criteria indicated a needed sample size of 82 per group. Therefore the research

plan consisted of intent to conduct quota sampling with participants enrolled from the

beginning study dates for both the pre-intervention and post-intervention groups until 82

were achieved in each group.

The planned pre-intervention group consisted of all very VLBW infants admitted to

Shands Hospital NICU from April 15, 2004 forward until 82 participants were enrolled in

the study. The intervention started on March 1, 2005 with conclusion of the educational

initiative on April 15, 2005. The planned post-intervention group consisted of all very

low birth weight infants admitted to Shands Hospital NICU after implementation of the

intervention plan, from April 15, 2005 until 82 were admitted into the study. Data from

participants were collected from the same time of year to account for seasonal confounds.

Research Interventions

Intervention 1: Breast Pump Loaner Closet

Intervention 1 consisted of the development of a breast pump loaner closet for use

by mothers with infants hospitalized in the NICU. Hospital grade breast pumps could be

checked out by mothers who had infants in the NICU to enable them to express their milk









the recommended 8 to 12 times per 24 hour period. This would enable a supply of

expressed breast milk for initiation of breast milk feedings in the infant and would enable

the mothers to establish and maintain a milk supply until the infant was able to go

directly to breast.

Intervention 2: Health Professional Education Initiative

Although there are many different recommendations and published articles about

breast milk feedings and breastfeeding in the NICU, there was no up to date, available

educational plan that could be utilized for staff education. Therefore, an education

initiative encompassing key areas of education on breastfeeding special care babies was

developed to educate as many of existing staff in the NICU as possible. The initiative

consisted of education to staff on breastfeeding to enable health care providers to have

the education and tools to support mothers in the breastfeeding process. The education

was offered through completion of a self study educational module on breastfeeding in

the NICU or through attendance at an inservice on breastfeeding in the NICU.

Education topics contained in the self study module and discussed in the inservice

included the benefits of breastfeeding, the barriers to breastfeeding, the physiology of

lactation, use of breast pumps, pre feeding interventions based on the synactive theory

and breastfeeding interventions that acknowledge the readiness of the infant. All the

information contained in the module was based on an extensive literature review to

represent evidence based practice and was designed to foster success with breastfeeding

in the high risk neonate population while acknowledging their unique needs. The

educational module was reviewed by two individuals considered to be experts in the area

of breastfeeding for establishing validity of information provided. Minor adjustments

were made to the education plan based on the expert feedback. Refer to the outline of the









educational module, appendix A, or contact the author for further details. The successful

completion of the health professional education was defined as completion of the module

or attendance at one of the inservices and a passing score of at least 80% on a post test

that was identical for either form of the education.

Intervention 3: Breastfeeding Guideline

Each medical chart contains an individualized care plan (ICP) for documentation

by nurses. This ICP was modified to also have a pathway of care for providing

breastfeeding support to new mothers (appendix B). This ICP necessitated

documentation of education and support by nurses at critical times in the breastfeeding

process. The guideline called attention to and necessitated documentation on specific key

points that were identified in the literature to be predictive of success: achieving and

maintaining a milk supply, timely pumping, skin to skin contact, and first feeding being

at the breast. It also included areas to check off, date, and sign at the following critical

times in the breastfeeding process: within 6 hours of delivery, issue and instruct in proper

pumping and breast milk storage techniques; within 24 hours, ensure proper pumping and

storage technique; on day 3 to 5, ensure that the milk has come in and trouble shoot any

problems; weekly, foster continued pumping and skin to skin care; first oral feeding,

ensure that it is a breastfeeding session; 10 days, monitor milk supply and make referrals

as appropriate.

Intervention 4: Educational Pamphlet for New Mothers of Neonatal Intensive Care
Unit Infants

An educational pamphlet, "A Mother's Gift", for mothers who had an infant

admitted to the NICU was developed. The outline of the educational booklet (see

appendix C) addressed the following key points: benefits of breastfeeding, how to









express and store human milk, pre-breastfeeding strategies, and cue based breastfeeding

interventions. The back of this pamphlet also included a place for mothers to document

breast milk production to facilitate communication with nurses about their milk supply.

This educational pamphlet was developed to ensure that all mothers received a standard

set of educational points during their infant's hospitalization, and that the information

contained in it was consistent with the education that the health care professionals

received.

Intervention Plan Modification

The original intervention plan consisted of 4 parts: a breast pump loaner closet, an

education module and inservicing, changes to the individualized care plan and an

educational booklet. Prior to the initiation date of March 1, 2005, it was learned that

external funding for the breast pump loaner closet could not be obtained. Therefore, this

prong of the intervention had to be deleted from the intervention program. The study was

then conducted with the following being the intervention/education plan: the education

initiative, the mother's educational booklet, and the breastfeeding pathway addition to the

individualized care plan.

Design

This study was a quasi experimental, matched through cohort controls, design

(Shadish, Cook, & Campbell, 2001), investigating indirect changes in breastfeeding

practices following a 3-part breastfeeding intervention in the NICU. Through this design

the pre-intervention group consisted of a group of VLBW infants hospitalized in the

NICU before the implementation of the intervention plan. This group was then compared

to the post-intervention group, which consisted of a group of VLBW infants who were

hospitalized in the NICU after the implementation of the intervention plan.









The independent variable was the implementation of the intervention as described

above. Dependent variables included 1) breast milk feeding initiation rate (was breast

milk ever consumed/breast milk feeds initiated? (yes/no)), 2) breastfeeding rate (number

of times the infant was put to the mother's breast after 30 weeks gestation divided by the

number of days hospitalized after 30 weeks gestation), 3) breast milk feeding at discharge

rate (did the infant continue to have breast milk feedings at discharge? (yes/no)), and 4)

the proportion of the hospital stay that breast milk was provided (total number of days

into the hospitalization that breast milk was provided divided by the length of stay).

Procedures

The educational intervention was implemented March 1, 2005 to April 15, 2005

with opportunities for health care professionals to complete the self study educational

module or participate in an inservice. "A Mother's Gift", the educational booklet for

mothers was issued to all new mothers with infants admitted to the NICU on or after

March 1, 2005. Last, the modified individualized care plan was used in the medical chart

on all new admissions after March 1, 2005.

To promote completion of the educational initiative, incentives were given to those

who participated in the breastfeeding education initiative by way of food, prizes,

continuing education credits and documentation on the annual review of their

performance. Following the six week educational initiative, completion of the self study

educational module on breastfeeding in the NICU became part of the orientation process

to enable the same education for those staff who were not employed at Shands Hospital

during the six week educational initiative.

The educational opportunities during the initiative dates included a self study

module and/or inservices. A breastfeeding module was available for health professionals









to check out and complete at home or work. Food and prize incentives as well as 2

continuing education credits were awarded for those who completed the educational

module. For those who preferred lecture-discussion formats for learning, 1 hour

inservices were offered at least one time per week throughout the education initiative

period. Those who attended the inservices were educated on the same information

contained in the education module, however in a condensed amount of time. Therefore

one continuing education credit, in addition to food and prizes, were awarded to those

who attended an inservice during the initiative dates.

The booklet entitled "A Mother's Gift" was issued to mothers with infants admitted

into the NICU after March 1, 2005. There was a central location at the reception desk

where nurses who had new admissions could access and issue them to mothers. Nurses

were instructed to issue these booklets during staff meetings, through the monthly

bulletin and in the breastfeeding inservice that occurred over the six week period.

The modified individualized care plan with the breastfeeding pathway replaced the

old ICPs and were placed in the chart as routine paperwork as of March 1, 2005. Nursing

staff were instructed to use it by way of a monthly written bulletin. It was also discussed

in staff meetings and further reminders were given to document on it during the

breastfeeding inservices that occurred over the six week period.

Program Evaluation

The desired impact of this program was increased breastfeeding in the NICU.

However, the intervention strategies used in the current study can only be effective if

they are implemented. The full implementation of the 3-pronged intervention was

evaluated in four ways. All the educational tools (the educational module, the inservice,

the educational booklet for he mothers and the modified ICP) stressed that the first oral









feeding should be at the breast. Therefore the primary outcome measure to determine

implementation of the intervention was whether the first oral feeding was at the breast.

Second, attempts were made to track the percentage of mothers of infants newly admitted

to the NICU to whom educational booklets were issued to determine if, in fact, most

mothers were being issued this educational booklet. Last, weekly communications with

the nursing administrator indicated the degree of compliance with educational key points

based on her monthly experiences as a bedside nurse, in which she worked directly with

mothers and their babies in the NICU.

Data Collection

Participants were recruited by way of a data base containing all admissions and

discharges from the NICU during the two different time periods. For each infant

admitted to the hospital during the applicable time periods, an extensive retrospective

chart review was conducted. Each identified chart was first investigated to ensure that the

infant did not have any exclusion criteria. Given that inclusion criteria were met, the

dependent variables as well as demographics were collected and recorded on a laptop

computer.

Inter-rater reliability was determined in 3 different participants to ensure that

accurate variables were collected from the charts. This occurred by having another

researcher collect data on the same participants following data collection by the principal

investigator and comparing if the variables collected by the two different researchers

were in agreement.

Data Analysis

Retrospectively, charts were reviewed and data was analyzed for significant

differences in the proportion of mothers who initiated breast milk feedings, the number of









times per day the mother breastfed after 30 weeks gestation, the proportion of mothers

who provided breast milk at discharge, and the proportion of the hospital stay that breast

milk feedings occurred. Table 2-1 summarizes the dependent variables and null

hypotheses.

Table 2-1. Hypothesis testing according to dependent variable
Group Breast milk The Number Breast milk Proportion of the hospital
feedings of Time provided at stay that breast milk was
initiated s the Mother discharge provided
(yes, no) Breastfed Per (yes, no)
Day After 30
EGA

Pre- Al B1 Cl Dl
Intervention
Group

Post A2 B2 C2 D2
Intervention
Group

Hypotheses

The following hypotheses and data analysis plan guided this study.

The rate of breast milk feeding initiation will be higher in the post

intervention group (A2>A1).

Data Analysis Plan: A Pearson's Chi Square was used to test two proportions for

significant differences between the two groups.

The number of times per day that an infant is breastfed after 30 weeks

gestation will be higher in the post-intervention group (B2>B 1).

Data Analysis Plan: A one-way analysis of variance (ANOVA) was not possible

secondary to a violation of the assumption of normality. Therefore, the nonparametric

Mann Whitney was used to test for differences between the 2 groups.









The rate of breast milk feedings at discharge will be higher in the post

intervention group (C2>C1).

Data Analysis Plan: A Pearson's Chi Square was used to test two proportions for

significant differences between the two groups.

The proportion of the hospital stay that breast milk was provided will be

higher in the post intervention group (D2>D1).

Data Analysis Plan: A one-way analysis of variance (ANOVA) was not possible

secondary to a violation of the assumption of normality. Therefore, the nonparametric

Mann Whitney was used to test for differences between the 2 groups.

In this study, the pre-intervention group and post-intervention group were

compared for significant differences in four different variables. For the purposes of this

study, an alpha level of .05 was chosen, which is standard throughout the literature.

Adjusting the Alpha Level

There are no statistical procedures that can simultaneously test multiple outcomes,

some of which are continuous and some of which are dichotomous. Therefore, the

significance levels of the individual tests were adjusted by the ranked Bonferroni

adjustment. There has been criticism of the standard Bonferroni adjustment being too

conservative and that, in theory, if many tests were run, the level of significance would be

so low that no differences could be detected. The ranked Bonferroni adjustment was

preferred over a standard Bonferroni adjustment to enable maximum power in initial

comparisons, by adjusting the alpha level with each additional comparison to prevent

inflation of the type I error rate. This would help to prevent the researcher from rejecting

the null hypothesis inappropriately while minimizing inappropriate stringentp value









constraints (Benjamini & Hochberg, 1995). For this study, the questions were ranked in

order of importance. The first question, whether or not there was a difference in breast

milk feeding initiation, was tested at an alpha of .05. The second question, whether or

not there was a difference in number of times breastfed after 30 weeks gestation, was

tested at an alpha of .025 (.05/2). The third question was tested at an alpha of .017

(.05/3). The fourth question was tested at an alpha of .013.

Each statistical analysis was conducted as a one sided test as it was assumed that

trends would be toward increased rates of breastfeeding with the interventions that were

implemented.














CHAPTER 3
RESULTS

Intervention Implementation

One hour inservices were conducted 1 to 3 times per week for a total of 10

inservices during the intervention period of March 1, 2005 through April 15, 2005.

General attendance at each inservice was low with approximately 2 to 5 participants at

each one. Self-study modules were also available for check out during this time. Overall

response to complete the self-study modules was also low in the month of March.

Therefore, in April, the researcher started directly asking health care professionals to

complete the modules and offered food prizes for those who did. It appeared that

directly requesting participation was beneficial in promoting participation by the health

care professionals. There were 11 health care professionals who completed the self study

education modules from March 1 through March 31, 2006, and there were 45 health care

professionals who completed the educational module from April 1 through April 15,

2006.

The total number of health care professionals who participated in the educational

initiative was 88, which was 63% of health care professionals working in the neonatal

intensive care unit (NICU). The total number of nurses who participated in at least one

of the methods of education was 75, which was 77% of all nurses who care for infants in

the NICU. There were 3 rehabilitation therapists (100%), 1 nurse practitioner (9%), 2

neonatologists (20%), 2 social workers (100%), 1 respiratory therapist (10%), and 5 other









health professionals (83%). All those who participated in the education achieved a

passing score of 80% on a post test.

Nursing managers reported variable levels of compliance with the new strategies

presented in the educational initiative, contained in the educational booklet for mothers,

and on the modified individualized care plan (ICP). Starting on March 1, 2005 the nurses

initiated use of a new, revised individualized care plan (ICP) for documentation. The

revised ICP was supposed to replace the old one. However, in mid April, it was realized

that some old stores of the previous ICP, that did not include the breastfeeding pathway,

had been pulled from the shelf and were being utilized. According to the nursing

manager, this problem was resolved with full use of the new ICP by May 1, 2005.

Although all nurses were expected to follow the established guidelines on the

breastfeeding pathway, during data collection it was observed that the new ICP was not

utilized fully. One example of the lack of full implementation of the new pathway

concerned whether the first oral feeding was at the breast. On the breastfeeding pathway,

all mothers should have been encouraged to have the first feeding at the breast with

documentation accordingly or documentation stating why care deviated from the

pathway. However, the first feeding being at the breast occurred in only 25% of mothers

in the post intervention group, and with full implementation it should have approached

100%. Although it is possible that mothers were encouraged, but declined to participate

in the first feeding at breast, it is more likely that there was lack of full compliance with

the educational key points and the modified individualized care plan.

Starting on March 1, 2005 "A Mother's Gift", the educational booklet for mothers,

was available to be issued to new mothers with infants in the NICU. Initial "Mother's









Gift" educational booklets were tracked to be able to determine if the number of booklets

that were issued matched the number of admissions. Not all mothers were given the

pamphlet over the first few weeks of the intervention. There were reports of running out

of the booklets and not being able to find them. Multiple copies of these were distributed

during and after the educational initiative, but they became impossible to track as they

were frequently misplaced, redirected to the maternity floor rather than remaining in a

central location in the NICU, and others outside of the research initiative made copies of

the booklet for distribution.

Nursing managers reported variable levels of compliance with the new strategies

presented in the educational initiative, contained in the educational booklet for mothers

and on the modified ICP. One nursing administrator, who would function in the role of

bedside nurse approximately once a month and would work directly with mothers and

their babies during this time, reported certain personnel to be implementing strategies

while others, even those who participated in the educational initiative, to be consistently

ignoring the pathway of care contained in the medical chart. The nursing administrator's

occasional role of bedside nurse revealed that there were mothers who never received the

educational materials and that ICPs in the medical chart had inadequate documentation.

The Sample

The pre-intervention sample data was obtained before the education plan

implementation using quota sampling from the beginning study date of April 15, 2004.

The post-intervention group was obtained after the intervention period implementation

from April 15, 2005 onward. Eighty one participants were obtained for the pre-

intervention group from April 15, 2004 through discharges on December 7, 2004. Data

collection in the pre-intervention group was stopped at 81, because the subsequent 2









admissions crossed into the treatment period. Data from only fifty four participants in the

post-intervention group was collected from April 15, 2005 through discharges on

November 29, 2005. There were no discharges from the NICU of participants who met

inclusion criteria from November 29, 2005 to December 7, 2005.

This sample included all admissions of VLBW infants admitted during the pre-

intervention study dates except for 17 infants who did not meet inclusion criteria.

Among the 17 infants who were excluded, 13 of them were extremely low birth weight

and expired shortly after birth, thus never achieved full gastric feeds. Two of them did

not achieve full gastric feeds before being transferred to another hospital, and 2 of them

had a length of stay that extended into the treatment period. The pre-intervention group

consisted of 83% of all admissions of VLBW infants admitted to the NICU at Shands

during the study dates. The sample included all admissions of VLBW infants admitted

during the post-intervention study dates except for 11 infants. Among those 11 infants

who were excluded were 9 infants who never achieved full gastric feeds and expired

shortly after birth and 2 who had genetic disorders that made eventual oral feeding

contraindicated. The post-intervention group also consisted of 83% of all admissions of

VLBW infants admitted to the NICU at Shands during the study dates.

The data collection period was not extended in order to capture the remaining 17

participants for two reasons: the first is that a long period of time had passed since the

intervention plan, and new interventions were scheduled to be implemented in the NICU.

These would have introduced significant additional confounds into the study. Secondly, a

new power analysis based on actual effect sizes of this partial sample indicated a need for

data from an additional 124 participants in the post-intervention group and 95 in the pre-









intervention group to achieve 80% power because of the already high breast milk

initiation rate (74.1%). Continuing data collection to enroll 82 in each group based on

the original research plan would have increased power from 38.6% to 45.6%, an increase

that was considered to not be feasible given the potential confounders listed above, or

likely to change the statistical outcomes.

Inter-Rater Agreement

To ensure accurate documentation of the research variables, inter-rater agreement

was tested on the chart review procedures. Another researcher conducted data collection

on 3 charts that the principal investigator had already collected data from. There was

100% agreement in 2 out of 3 of the charts. However, one chart revealed agreement of

92%, for a total inter-rater agreement of 97% for this study. The principal investigator

reviewed the chart that did not have complete agreement to find 100% agreement with

her initial findings.

Demographics

Table 3-1 includes sample demographics andp values for statistical tests to rule out

selection differences. All demographics were collected as continuous or dichotomous

variables, with the exception of race. Race in the medical chart was classified as White,

Black, Asian, Hispanic or Other. Therefore, race is documented with the same

classifications. The pre-intervention group was 4% Hispanic, 42% Black, and 54%

White. The post-intervention group was 2% Asian, 3% Hispanic, 49% Black, 42% White

and 4% with undocumented race in the medical chart. Due to the majority of participants

being Black or White, with minimal representation of other racial backgrounds, and due

to Black being a known predictive factor in the literature, race was dichotomized into

Black and not Black for statistical purposes to rule out selection differences.










Table 3-1. Demographics of the pre-intervention and post-intervention groups with test
statistics for selection differences
Low Race Maternal Marital Transferred Length Birth EGA No.
SES (Black) Age Status Instead of of Weight of
(not DC Home Stay Sibs
married)
Pre- .775 .42 25.46 .56 .432 50 1074 28.57 1.01
Intervention
Group
Post- .70 .49 25.62 .57 .327 54 1114 28.7 .86
Intervention
Group
p Value to .339 .256 .899 .860 .225 .534 .368 .762 .297
Investigate
Selection
Differences


There was a large percentage of participants of low socioeconomic status (77.5% in

pre-intervention group and 70% in the post-intervention group), Black race (42% in the

pre-intervention group and 48% in the post-intervention group), and unmarried mothers

(56% in the pre-intervention group and 57% in the post-intervention group). Average

maternal age in the pre-intervention group was 25.46 and in the post-intervention group

was 25.62 years. The average birth weight in the pre-intervention group was 1074 grams,

and the average birth weight in the post-intervention group was 1114 grams. The average

gestational age (abbreviated EGA) at birth was 28.57 weeks gestation in the pre-

intervention group and 28.7 weeks in the post-intervention group. The average number of

siblings (abbreviated No. of Sibs) in the pre-intervention group was 1.01 and in the post-

intervention group was .86. Eighty four percent of the pre-intervention group consisted

of single births, and 83.3% of the post-intervention group consisted of single births. In

the pre-intervention group there were 43.2% of participants who were transferred to

another hospital instead of discharge home, and in the post-intervention group there were

32.7% who were transferred to another hospital. Average length of stay in the pre-

intervention group was 50 days and in the post-intervention group was 54 days.










Investigation for Selection Differences

Due to the matching through cohort controls research design, it was important to

first determine if there were selection differences in the two groups being compared.

Socioeconomic status was categorized into Women, Infants and Children (WIC) or

Medicaid eligibility or not WIC/Medicaid eligible. Difference in this variable between

the two groups was investigated by use of a z test for 2 proportions. Hypothesis testing

of two proportions with a z test was used to test for group differences in maternal race,

which was dichotomized as Black or not Black. Group dissimilarity based on maternal

age was investigated through an independent samples t test, while differences in marital

status (married, not married) and sex of the infant were investigated by use of a z test for

two proportions. Gestational age at birth, birth weight and number of siblings was

investigated by use of an independent samples t test. Discharge status was investigated

with a z test of 2 proportions and length of stay with an independent samples t test. By

testing each of the demographic variables at an alpha of .05, none of the demographic

variables were significantly different between the two groups (see table 3-1).

Subsequently, having no selection differences supports the ability to use the matching

through cohort controls design.

Results Per Research Question

The primary aim of this study was to implement a breastfeeding intervention that

would improve breastfeeding practices in the NICU. The results of this study are

provided per research question.

Is there a significant difference in breast milk feeding initiation in very low birth
weight (VLBW) infants admitted to the NICU before and after implementation of
the intervention plan?









Table 3-2 summarizes breast milk feeding initiation results. The breast milk

feeding initiation rate in the pre-intervention group was 74.1%. The breast milk feeding

initiation rate in the post-intervention group was 85.2%. This represents an increase of

11.1%. However, through a Pearson's Chi Square Test of 2 proportions, the p value is

.124, indicating no significant difference between groups when tested at an alpha of .05.

The odds ratio of breast milk feeding initiation is 2.013 with a confidence interval of .818

to 4.952.

Table 3-2. Breast milk feeding initiation rates
Was Brest Milk Ever Total Test
Provided? Statistic
No
Yes

Group Pre-Intervention Count 21 60 81
% within 25.9% 74.1% 100.0%
subject
Post- Count 8 46 54
Intervention % within 14.8% 85.2% 100.0%
subject
Total Count 29 106 135
% within 21.5% 78.5% 100.0%
subject
Pearson's Chi-Square .124
Significance
Odds Ratio 2.013
Odds Ratio Confidence .818 to
Interval 4.952

Is there a significant difference in the rate of breastfeeding in the hospital among
women with VLBW infants hospitalized in the NICU before and after the
implementation of the intervention plan?

For the continuous variable of number of times breastfed per day after 30 weeks

estimated gestational age (EGA), see figure 3-1. The graph is clearly skewed toward 0.

Due to the violation of normality, a Mann Whitney nonparametric test was used to test

significance of this variable. Interpretation of this graph and variable is difficult as the








54



rate of breastfeeding in the NICU is significantly low at .059 in the pre-intervention


group, which is once every 17 days, and .139 in the post-intervention group, which is


once every 7 days (see Table 3-3). This variable proved to be significantly different


between the two groups with ap value of .011.


70-

60-



40-

30-

20-

10-



70-

so-

so50-

40-

30-

20-

10-

0-- F=


7-
0

0
0-
a
a)
+-i




0
0
0


n
03




a-

a,



0a)


0.00000 0.20000 0.400O 0 0.60000 0.80000 1.00000 1.20000

Number of Times Breastfed Per Day After 30 Weeks Gestation

Figure 3-1. Number of times breastfed per day after 30 weeks gestation










Table 3-3. Comparison of rates of breastfeeding after 30 weeks gestation
Test Statistic
Number of Times
subject Breastfed Per Day
After 30 EGA

Pre-Intervention Men .
Mean .0593937
Group
N 81
Std. Deviation .18818812

Post-Intervention M
Mean .1389242
Group

N 54
Std. Deviation .24433376
Total Mean .0912059
N 135

.21513898
Std. Deviation .21513898
Mann Whitney .011
Significance

To enable easier interpretation of this variable, it was dichotomized into whether a

mother ever participated in breastfeeding while in the hospital. In the pre-intervention

group, there were 25.9% of mothers who ever breastfed their infant in the hospital. In the

post-intervention group, there were 44.4% of mothers who ever breastfed their infants in

the hospital (see Table 3-4). This represented an increase of 18.5%, which achieved ap

value of .025 through a chi-square test of 2 proportions. Therefore, there were significant

differences in proportion of women who ever breastfed in the two groups, using an alpha

of .025. The odds ratio of ever breastfed in the hospital was 2.286 with a confidence

interval of 1.1 to 4.750.









Table 3-4. Rates of ever breastfed in the neonatal intensive care unit
Frequency Was the Infant Ever Total Test
Breastfed While in the Statistic
Hospital?
No Yes
Group Pre- Count 60 21 81
Intervention % within 74.1% 25.9% 100.0%
subject
Post- Count 30 24 54
Intervention % within 55.6% 44.4% 100.0%
subject
Total Count 90 45 135
% within 66.7% 33.3% 100.0%
subject
Pearson's Chi
Square Significance .025

Odds Ratio 2.286
Odds Ratio
Confidence Interval 1.1 to
4.750


Is there a significant difference in breast milk feedings at discharge in VLBW
infants admitted to the NICU before and after the implementation of the
intervention plan?

There were 35.8% of infants who were provided with breast milk at the time of

discharge in the pre-intervention group. There were 40.7% of infants in the post-

intervention group who were provided with breast milk at discharge. This 4.9% increase

resulted in a p value of .562 through a chi-square test of 2 proportions, indicating no

statistically significant difference among groups. The odds ratio was 1.233 with a

confidence interval of .607 to 2.502.









Table 3-5. Rates of breast milk feedings at discharge
Was Breast Milk Total Test
Provided at Discharge Statistic
No Yes
Group Pre- Count 52 29 81
Intervention % within 64.2% 35.8% 100.0%
subject
Post- Count 32 22 54
Intervention % within 59.3% 40.7% 100.0%
subject
Total Count 84 51 135
% within 62.2% 37.8% 100.0%
subject
Pearson Chi-Square .344
Significance
Odds Ratio 1.233

Odds Ratio Confidence .607 to
Interval 2.502



Is there a significant difference in the proportion of the hospital stay that breast
milk is provided in VLBW infants admitted to the NICU before and after the
implementation of the intervention plan?

Looking at figure 3-2, both groups have peaks at 0 and 1. However, there is a

larger peak at 0 in the pre-intervention group. The variable, proportion of the hospital

stay that breast milk was provided, did not achieve the assumption of normality as the

graphs are u-shaped. Therefore, an ANOVA could not be run on this variable without

violating assumptions. The nonparametric Mann-Whitney test was used to test for

significant differences. This test indicated that thep value was .108, therefore there were

not significant differences between the two groups in proportion of the hospital stay that

breast milk was provided.







58





0
4-1
4a
--
a)
a


0
01


- 25-
hi- .I

C


10- .0
aI

1- 1-



O.0000 0.2000 0.40000 0.60000 0.8000 1.00000
Proportion of Hospital Stay That Breast Milk is Provided


Figure 3-2. Proportion of the stay that breast milk was provided

To look at effect size in a variable that does not have a normal distribution, the

proportion of the hospital stay was dichotomized into breast milk provided for most of

the hospitalization or not. This variable represented whether breast milk feeds were

supplied to the infant more than 50% of the length of stay not. There were 51% of the

pre-intervention group who provided breast milk for most of the hospital stay, and there

were 57% of the post-intervention group who provided breast milk for most of the

hospital stay. This gives an odds ratio 1.219 with a confidence interval of .608 to 2.444

for breast milk feeds being provided for most of the hospital stay.














CHAPTER 4
DISCUSSION

The Effect of the Interventions on Breastfeeding Practices in the Neonatal Intensive
Care Unit

This research investigated if a three-part intervention plan designed to promote

increased breastfeeding in premature infants would have an effect on breastfeeding

practices in the neonatal intensive care unit (NICU). The three-part intervention

consisted of opportunities for education of health care professionals over a 6 week

educational initiative, an educational booklet for new mothers in the NICU, and

modifications to the individualized care plan (ICP) with a pathway of care for

breastfeeding. This intervention did have an effect on breastfeeding practices in the

NICU, but it did not result in changes across all breastfeeding variables as hypothesized.

Evidence of improved breastfeeding practices was that rates of breastfeeding

(mothers putting their infants directly at the breast) in the NICU improved following

intervention. The number of times infants were breastfed per day after 30 weeks gestation

was significantly greater in the post-intervention group than in the pre-intervention group.

Mothers in the pre-intervention group breastfed their infants after 30 weeks gestation

.059 times per day, and in the post intervention group they breastfed .139 times per day.

This works out to an average of once every 17 days in the pre-intervention group and

once every 7 days in the post intervention group. Although a statistically significant

increase was observed following the intervention, the resultant rate of breastfeeding in

the NICU remained low. After dichotomizing this variable, it was noted that there were









25.9% of mothers who ever breastfed their infant in the hospital in the pre-intervention

group. In contrast, there were 44.4% of mothers who ever breast fed their infants in the

post intervention group. This represented a significant increase of 18.5%. This provides

some support that the program had a positive effect with increased participation in

breastfeeding among mothers in the NICU. However, the overall rate of breastfeeding

participation remains low in the NICU.

The variable of ever breastfed while in the hospital is an important one, as it

requires active participation by the mother and infant. Diminished milk supply is cited as

one of the most significant barriers to breastfeeding in the NICU population. Maternal

stress has been linked to inhibition of oxytocin, which is responsible for the let down

response during pumping and breastfeeding (Lang, 1996). However, physiologically,

oxytocin is facilitated with increased mother-infant contact and environments that foster

breastfeeding ("How breast-feeding postpones ovulation," 1985). Mothers with infants in

the NICU typically experience high levels of stress and anxiety and many are

overwhelmed by the NICU environment (Nyqvist et al., 1994). Many also comment on

the loss of control of their infant to others during stays in the NICU (Lupton & Fenwick,

2001). Close contact, as in skin to skin, as well as breastfeeding, are important in

maintaining bonding in a difficult environment, promoting the milk supply and providing

some control over care for mothers (Kirsten, Bergman, & Hann, 2001). This direct

interaction of the mother and infant can be assumed to be critical to the other

breastfeeding outcomes.

Although positive trends in favor of the post-intervention group were observed

across all variables, the intervention did not have a strong enough impact to result in









significant changes in breast milk feeding initiation rates, breast milk feeding at discharge

rates and proportion of the hospital stay that breast milk was provided. This could be

due to lack of a strong enough impact of the intervention with a need for a more

extensive list of interventions or adjustment of the interventions proposed in this study,

need for consistent compliance with the research interventions to promote change, or a

different time frame for the study to detect changes.

The breast milk feeding initiation rate (was breast milk ever provided) in the pre-

intervention and post-intervention groups appears to be fairly comparable to other

research findings, which have documented rates of breast milk feeding initiation in the

NICU at 64% (Byrne & Hull, 1996), 72.9% (Meier et al., 2004), and 83% (Yip et al.,

1996). Although breast milk feeding initiation failed to reach statistical significance,

there was an increase in breast milk feeding initiation of 11.1%, which was a positive

change in the right direction as more infants received some breast milk following the

intervention. Scientists are beginning to refer to breast milk as medicine and have

initiated discussing breast milk in terms of a dose (Meier et al., 2004). Thus following

the intervention, 11.1% more infants in the post-intervention group received breast milk

at the most critical stage of their recovery. However, this variable did not achieve

statistical significance, which could indicate that the impact of the interventions was not

strong enough or that there was not enough compliance with the interventions to promote

change. It is important to note that breast milk feeding initiation reflected whether breast

milk was ever provided and not if breastfeeding ever occurred or if breast milk continued

to be supplied after the 2nd day of life or at discharge.









Although the rate of breast milk feedings at discharge (was breast milk provided to

the infant at discharge) increased slightly by 4.9%, this difference failed to reach

statistical significance. Succeeding with breast milk feedings until discharge in the NICU

population is a significant challenge for mothers with infants in the NICU. Discharge for

many of these infants did not occur until they were 1 to 4 months old. This supports

other research findings that state that some of the strongest predictors of not

breastfeeding by discharge is being low birth weight, having decreased gestational age

and being admitted into the NICU (Hwang et al., 2006; Li et al., 2005; Powers et al.,

2003; Scott et al., 2006). The findings of this study in both the pre-intervention (35.8%)

and post-intervention (40.7%) groups are comparable to other research findings, which

document breast milk feedings at discharge to be 38% (Yip et al., 1996). The low rate of

breast milk feeding in the NICU at discharge gives some insight into the significant

challenges that women encounter with succeeding with breast milk feedings through a

NICU hospitalization. However, to enable mothers with this process, it is critical that

strategies to promote long term success are implemented. Although lack of full

implementation and compliance with intervention strategies may have factored into this

research study, these results suggest that this 3-pronged intervention was not sufficient to

enable mothers to overcome the barriers to maintaining breastfeeding until the time of

hospital discharge in this complex and challenging environment.

There was no difference between the groups in the proportion of the hospital stay

that breast milk was provided. These results again highlight the importance of

determining strategies for long term success to enable mothers to succeed with the









breastfeeding process for a larger proportion of the hospitalization. Such strategies could

enable breast milk feedings until discharge and beyond.

This study complements conclusions from other studies that attention to and

education about lactation affects health care professional knowledge and support of the

breastfeeding process (Siddell et al., 2003). Many studies cite lack of health care

professional education as a significant barrier to the breastfeeding process in the NICU

(Register et al., 2000). Although improvement in breastfeeding rates in the NICU was

observed in the post education group, there is no way to determine which part of the

intervention plan may have had an effect on this variable. Other studies have concluded

that health care professionals play a significant role in breastfeeding practices in the

NICU (Swanson & Power, 2005) as does increased knowledge about breastfeeding

among mothers in the NICU (Bernaix, 2000). However, because all the interventions

were implemented together as a 3-pronged approach, it is unclear if the educational

materials for mothers, the educational initiative, or the modifications to the

ICP/breastfeeding pathway resulted in the effect on breastfeeding in the NICU.

According to the transtheoretical model of behavior change, change is a process

and sometimes requires multiple approaches that are stage matched in addition to the

passage of time. The real impact of the education perhaps could not be fully observed in

assessing the indirect impact on breastfeeding practices among mothers and infants. By

assessing the indirect impact of the education intervention on breastfeeding strategies

without measuring the knowledge and behavior change of the health care professionals, it

cannot be determined if there was a change in the health care professionals and what the

magnitude of such a change was. This calls into question whether the educational









initiative did what it was intended, which was to change health care professional

behavior. With knowledge of a change in health care professional behavior, the true

impact of education and intervention strategies on breastfeeding outcomes in the NICU

could be assessed. In addition, there were 2 behavior changes that could have been

assessed, the health care professional as well as the mother. Further measuring the

mothers' behavior change, in the presence of education from the health care professional,

would have provided useful information on whether the intervention was strong enough

to elicit positive changes in the mothers.

Although positive changes are evident, the reported compliance with strategies and

follow through of education key points was called into question by the nursing

administration. Other studies have determined that education about breastfeeding has had

an effect on breastfeeding knowledge and supportive behaviors (Ekstrom, Widstrom et

al., 2005). This research study did not measure the behaviors of health care professionals

following education. Therefore, it could be that the 3 part intervention plan was not

strong enough to elicit changes in health care professional behavior, that not enough time

passed post intervention to enable successful change or that the health care professional

change elicited following the intervention was not enough to result in positive outcomes

across all proposed breastfeeding variables. Subsequently, this study provides only

partial support for positive changes in breastfeeding practices in the NICU following the

3 part intervention.

Limitations

This study is not without limitations. Limitations included inadequate

implementation of all the intervention strategies by the health care professionals, lack of

methods to determine behavior change and implementation by health care professionals,









lack of ability to give ample time for changes to be implemented without introducing

other confounds, lack of participation by key decision makers in the NICU, the inability

to control for other changes in the NICU environment, lack of a more comprehensive

breastfeeding intervention plan, and lack of a randomized sample.

An important limitation of this study was the questionable full implementation of

the strategies learned in the educational initiative as well as inconsistent use of the

modified ICP and inconsistent distribution of"A Mother's Gift". Although there was a

high health care professional participation rate in the education initiative of 63%, health

care professional behavior change and attitudes were not measured. Therefore, there is

no way to know the direct effect of the education and placement of the modified ICP in

the medical chart on health care behavior and attitudes. It can be assumed that although

there was good participation in the educational initiative by bedside nurses and the

breastfeeding pathway on the ICP was added to the medical record, there was a lack of

movement to action among many health care professionals based on the observations by

nursing administration as well as by the first feeding at the breast variable remaining low

in the post-intervention group.

Despite education about promoting breastfeeding in the NICU and how to

introduce such practices, there remained a large percentage of women (56%) who never

breastfed while in the hospital. This demonstrates that although the education may have

occurred, change was not fully embraced and implemented. Health care professionals

care for infants, including caring for nutritional feeds, in the NICU 24 hours a day, and

parents are visitors to the NICU when they come to see their infant. Breastfeeding is

typically something that health care professionals would enable, and thus its low rate









could be attributed to low health care professional support of the breastfeeding process.

This study investigated breastfeeding outcomes in the natural NICU environment in

which health care professionals may have had other priorities, may have had negative

beliefs about breastfeeding in the NICU, may have had control issues that interfered with

promoting a feeding method that they eventually cannot participate in, or may have been

suffering from bum out. However, without the intermediate measure of health care

behavior, it is difficult to determine if the intervention could be effective if health care

professionals demonstrated an appropriate behavior change and were called to action.

According to the trantheoretial model of behavior change (TTM), change is a

process with people moving through different stages of readiness to change (Prochaska et

al., 2001). The intervention plan was designed to provide interventions that are

appropriate for individuals in each stage of readiness to change. However, the

educational initiative and interventions were introduced with data collection occurring in

the post intervention group 6 short weeks later. This may not have been a long enough

period of time to enable individuals to respond to the intervention and move them to

action. According to the TTM, those in the pre-action stages are the precontemplators

(no intention of making a change), the contemplators (intention of making a change in the

next 6 months) and those in the preparation stage (intend on making a change in the next

30 days). Theoretically a six week period may not have been enough to enable

movement through the stages of change to promote action toward increasing

breastfeeding support and implementing strategies learned in the educational initiative.

In contrast, other confounds were being introduced in the NICU in the months following

the conclusion of this study, as more interventions were set to be implemented.









Therefore, this study could not account for the possible inadequate time frame due to the

potential introduction of confounds that could bias the results.

Another limitation was the weak participation in the education modules by

physicians and nurse practitioners, who are key decision makers in the NICU. Prior to

the educational initiative dates, they participated in a short, general inservice about

breastfeeding to facilitate discussion and direction. Due to this recent meeting, it was

difficult to get them to participate in the research educational initiative. Therefore,

participation was low with only 20% of neonatologists and 9% of nurse practitioners

attending an inservice or completing the self study educational module. Although the

researcher did meet with the physician medical director of the unit to discuss key points

of the educational initiative, the initiative may have been much more successful if the

neonatologists and nurse practitioners had higher levels of participation. Nursing staff

and parents look to physicians and neonatal nurse practitioners for direction, and their

lack of participation was a significant limitation to this study.

The NICU is a constantly changing environment that can not be fully controlled in

a study like this, where all infants during a specific time frame are being enrolled. This

study does not account for other changes that may have occurred, such as staff

changeover, other education that health care professionals may be receiving and

implementing, and changes made to the physical NICU environment. Cosmetic changes

were made to the NICU environment during the course of this study, and it was

impossible to determine if this may have had an effect on the results.

The lack of optimal resources that may facilitate breastfeeding may also pose a

limitation to this study. During the early stages of planning this research project, funds









were applied for to provide a breast pump loaner closet that would provide a needed

resource for long term maintenance of the milk supply (Meier et al., 2004). These funds

were not achieved, making it impossible to provide this resource to enable long term

success with maintaining the milk supply, and subsequently with addressing the idea of

providing breast milk all the way until the infant was discharged from the NICU.

Providing breast milk or breastfeeding at the time of discharge would be the goal, as this

would indicate that a mother succeeded with providing milk during the hospitalization

and would enable breastfeeding at discharge home. Without the funding, women were

left to find their own resources to access a hospital grade pump for milk expression.

Some may have accessed one for use at home while others may have used store bought

pumps or self expressed.

This study is a cohort study. Due to lack of randomization, there is the possibility

that there are unseen differences in the two groups that lead to differences or similarities

in breastfeeding practices that cannot be attributed to the intervention. Results from this

study can only give conclusions about the population of infants being studied at Shands

Hospital. In addition, the small sample size limited the ability to achieve adequate

power, which can affect being able to detect significant differences, if they exist. Based

on the utilized sample size, 38.6% power was achieved on the primary variable of breast

milk feeding initiation. This indicates that the probability of finding a difference in the

two groups was only 38.6%. The already high breast milk feeding initiation rate in the

pre-intervention group, 74.1%, gave less room for improvement in this variable.

Although comparable with other rates reported in the literature, the rate in this unit was

expected to be lower due to lack of lactation services in the unit and the large population









of Black mothers, mothers with low socioeconomic status and single mothers, which are

all predictive of decreased breastfeeding behaviors (Powers et al., 2003). The nature of

the population may have been a limitation, in that perhaps the effect of the intervention

would be different given a different demographic presence in the NICU. Given the effect

of the current intervention, a larger sample size would be needed to determine if there are

significant differences in the pre-intervention and post-intervention groups. Such sample

size was not feasible during the study period given the number of admissions and

discharges. Prolonging enrollment was not feasible given the risk of introducing

potential bias. To

Recommendations for Further Research

Women with infants in the NICU face unique challenges to the breastfeeding

process. These barriers result in decreased breastfeeding initiation rates and breast milk

feedings at discharge from the hospital. However, these fragile infants are at an increased

need for the benefits of breast milk. Therefore, further studies looking at trends in

breastfeeding and looking at interventions that can assist mothers in overcoming barriers

is necessary.

Future research could investigate the efficacy of educational interventions that

measure the intermediate effects of the intervention on health care professional behavior

as well as the effect on breastfeeding outcomes. With similar findings as this study, this

would make it easier to interpret if the educational interventions did not cause health care

professional behavior change and thus did not affect breastfeeding outcome or if the

educational interventions did affect health care professional behavior but that the

resultant effect of this was not strong enough to result in positive breastfeeding outcomes.









Future studies addressing an educational initiative for health care professionals

with more time to enable change would be appropriate to run in this population. If

change takes time to occur, a longer period between the intervention and the start of data

collection in the post intervention group would help identify change that occurred over a

longer period of time following the interventions. However, such a study would require

close control of the environment to prevent other confounds from being introduced as

time from the education initiative elapses.

Although an intervention with many parts that addresses milk expression and

breastfeeding over the course of hospitalization may be optimal in promoting

breastfeeding practices in the NICU, a look at intervention specific studies could help to

declare which interventions are successful and which are not. Such a study, although it

may require a much larger sample size and perhaps random assignment to groups and a

multi-center trial, could assist with understanding which, if any, and in what combination,

treatments) have an effect on breastfeeding practices in the NICU.

Further research could also investigate the specific results, amount of milk

expression and success with breastfeeding among women who had a nurse who was in an

action oriented stage of behavior change regarding support for lactation compared to

those receiving traditional NICU care. The effects of support and education for the

mother could further be investigated by including success and failure with breastfeeding

among mothers who demonstrated compliance with strategies in "A Mother's Gift"

compared to those who were not fully compliant.

With the paucity of breastfeeding (putting the infant to breast) in the NICU, it is of

great interest to determine the predictive effect of breastfeeding on breast milk feedings









at discharge, milk production, amount of breast milk feedings in the hospital, as well as

the effect on maternal and child health. Likewise, it would be of great interest to

determine if there are any negative effects of not enabling breastfeeding on the ability to

maintain the milk supply, success of breastfeeding at discharge and beyond and the

ability to transition from breast milk feeds to direct breastfeeding.

Conclusions

This study investigated change in breastfeeding practices following implementation

of an intervention plan with 3 parts; an educational booklet for new mothers, an

educational opportunity for health care professionals who serve mother-infant dyads in

the NICU, and changes to the individualized care plan that necessitate breastfeeding

practice documentation by nurses. There were general positive trends across all

variables, which make this type of intervention have some promise for affecting positive

changes in breastfeeding practices in the NICU, given the limitations of this study and the

difficulties that can be expected in the complex NICU environment. There were

statistically significant differences between groups in the rate of breastfeeding, with

infants in the post-intervention group being more than twice as likely to be breastfed in

the hospital.

While health care professionals may play a significant role in the breastfeeding

process, perhaps an educational plan and breastfeeding pathway as well as

complementary educational materials for mothers is not enough to promote full

participation and optimal breastfeeding practices in the NICU. This study demonstrated

limited, but encouraging, support for an intervention plan focused on education to

facilitate change in breastfeeding within the NICU, but perhaps a multifaceted

intervention plan, including hospital grade breast pump allocation may optimize positive






72


changes in breastfeeding practices. An important limitation of this study was lack of full

implementation of the intervention strategies. Future studies can be designed to better

measure compliance with educational interventions as well as the individual effect of

each intervention and the additive effect of multiple interventions implemented together.

Randomized multi-institutional studies will enable larger intervention trials as well as

generalization of findings.














APPENDIX A
OUTLINE OF THE EDUCATION MODULE

Breastfeeding in the Neonatal Intensive Care Unit
An Educational Module for Health Care Professionals


Introduction

Benefits of Breastfeeding for the Full Term Infant
Health Benefits
Developmental Benefits
Financial Benefits

Benefits for Preterm or High Risk Infants

Benefits for the Mother

Barriers to Breastfeeding
Full-term infants
High-risk infants
Contraindications

Breast Milk
Composition
Transition to Mature Milk

The Breast and Lactogenesis
Anatomy
Neuroendocrine control
Sucking pattern

General Interventions for Supporting Breastfeeding

Appropriate Timing of Interventions
Interventions in the NICU

Interventions Prior to Active Breastfeeding
Breast pumps
Maintaining a milk supply
Kangaroo Care
Non-nutritive suckling









Early initiation of direct breastfeeding

Monitoring Physiologic Responses
Transition to Direct Breastfeeding

Interventions for Supporting Breastfeeding
Privacy
Positioning
Timing
Special Techniques

Determining Adequacy of Intake

Conclusion

Post-test

References



For further information, contact the authors:
Roberta Gittens Pineda: jopineda@pol.net
Cammy Pane: cam92460@yahoo.com














APPENDIX B
OUTLINE OF ITEMS ADDED TO THE INDIVIDUALIZED CARE PLAN

Breastfeeding Guidelines:

Within 6 hours of delivery: Mom is given "A Mother's Gift"

Within 24 hours: Ensure proper pumping (with hospital grade pump) and storage

3 to 5 days of life: Assess Mom's milk production. Address problems

10 days of life: Mom's milk supply should be at least 350 ml per 24 hours

With each parent contact: Offer support and discuss any problems

Kangaroo care is encouraged as soon as possible

First oral feeing is at breast

Mom is encouraged to breastfeed at each visit with supplementation only when
medically indicated

At discharge: Encourage transition to full breastfeeding while monitoring weight
gain














APPENDIX C
OUTLINE OF THE EDUCATIONAL BOOKLET FOR MOTHERS

A Mother's Gift
Breastfeeding and Pumping for Your Baby in the NICU

Breast milk is the best food for your baby's start in life
Each mother's milk is prefect for her baby
Babies who drink breast milk are healthier and smarter
This makes breast milk even more important for premature babies
Breastfeeding is healthy for the mother too!

Providing breast milk for your baby is something very special you can do to help your
child
It is important to begin expressing your milk as soon after delivery as possible,
and to keep pumping
Pumping may seem complicated at first, but it is worth it for your baby
During pumping, relax and enjoy gentle thoughts about your baby
At first, you may not get any milk, or only a few drops
Here is how to store your milk
Every mother wants to know-"Am I making enough milk"
In order to maintain a good milk supply...

Most mothers really want to hold their babies
Kangaroo care is good for your baby
Kangaroo care is good for parents too

Privacy is more important to some than others...

Suckling at the breast is the best way to nourish your baby... and it takes practice
Positioning during breastfeeding will become easier with practice...
Is my baby really, really, really getting enough milk

Please take extra care of yourself too!
If you are taking medications
Making milk for twins or more...

Please ask if you have questions or need help

How to obtain a pump for home use...

Chart for tracking milk production






77



On-line resources

For more information, contact:
Roberta Gittens Pineda: jopineda@pol.net















LIST OF REFERENCES


Agostoni, C., & Haschke, F. (2003). Infant formulas. Recent developments and new
issues. Minerva Pediatr, 55(3), 181-194.

Allen, M. C., & Capute, A. J. (1990). Tone and reflex development before term.
Pediatrics, 85(3 Pt 2), 393-399.

Alm, J. S., Swartz, J., Bjorksten, B., Engstrand, L., Engstrom, J., Kuhn, I., et al. (2002).
An anthroposophic lifestyle and intestinal microflora in infancy. Pediatr Allergy
Immunol, 13(6), 402-411.

Arora, S., McJunkin, C., Wehrer, J., & Kuhn, P. (2000). Major factors influencing
breastfeeding rates: Mother's perception of father's attitude and milk supply.
Pediatrics, 106(5), E67.

Baker, R. (2003). Human milk substitutes. An American perspective. Minerva Pediatr,
55(3), 195-207.

Ball, T. M., & Wright, A. L. (1999). Health care costs of formula-feeding in the first year
of life. Pediatrics, 103(4 Pt 2), 870-876.

Berens, P. D. (2001). Prenatal, intrapartum, and postpartum support of the lactating
mother. Pediatr Cin North Am, 48(2), 365-375.

Beresford, H. J. (1984). The success of breast feeding. IPPFMedBull, 18(5), 3-4.

Bemaix, L. W. (2000). Nurses' attitudes, subjective norms, and behavioral intentions
toward support of breastfeeding mothers. JHum Lact, 16(3), 201-209.

Bick, D. E., MacArthur, C., & Lancashire, R. J. (1998). What influences the uptake and
early cessation of breast feeding? Midwifery, 14(4), 242-247.

Birch, E., Birch, D., Hoffman, D., Hale, L., Everett, M., & Uauy, R. (1993). Breast-
feeding and optimal visual development. JPediatr Ophthalmol Strabismus, 30(1),
33-38.

Black, K. A., & Hylander, M. A. (2000). Breastfeeding the high risk infant: implications
for midwifery management. JMidwifery Womens Health, 45(3), 238-245.

Blackburn, S. (1998). Environmental impact of the NICU on developmental outcomes. J
Pediatr Nurs, 13(5), 279-289.









Breastfeeding and the use of human milk. American Academy of Pediatrics. Work Group
on Breastfeeding. (1997). Pediatrics, 100(6), 1035-1039.

Bueno, M. B., de Souza, J. M., de Souza, S. B., da Paz, S. M., Gimeno, S. G., & de
Siqueira, A. A. (2003). [Risks associated with the weaning process in children born
in a university hospital: a prospective cohort in the first year of life, Sao Paulo,
1998-1999]. CadSaude Publica, 19(5), 1453-1460.

Byrne, B., & Hull, D. (1996). Breast milk for preterm infants. Prof Care Mother Child,
6(2), 39, 42-35.

Callen, J., & Pinelli, J. (2005). A review of the literature examining the benefits and
challenges, incidence and duration, and barriers to breastfeeding in preterm infants.
Adv Neonatal Care, 5(2), 72-88; quiz 89-92.

Chan, D. K. (2001). Enteral nutrition of the very low birth weight (VLBW) infant. Ann
AcadMed Singapore, 30(2), 174-182.

Chantry, C. J., Auinger, P., & Byrd, R. S. (2004). Lactation among adolescent mothers
and subsequent bone mineral density. Arch Pediatr Adolesc Med, 158(7), 650-656.

Chen, A., & Rogan, W. J. (2004). Breastfeeding and the risk of postneonatal death in the
United States. Pediatrics, 113(5), e435-439.

Chen, C. H., Wang, T. M., Chang, H. M., & Chi, C. S. (2000). The effect of breast- and
bottle-feeding on oxygen saturation and body temperature in preterm infants. J
Hum Lact, 16(1), 21-27.

Chua, S., Arulkumaran, S., Lim, I., Selamat, N., & Ratnam, S. S. (1994). Influence of
breastfeeding and nipple stimulation on postpartum uterine activity. Br J Obstet
Gynaecol, 101(9), 804-805.

Cohen, J. (1988). Statistical power analysisfor the behavioral sciences (2nd ed.).
Hillsdale, N.J.: L. Erlbaum Associates.

Coleman, S., Dracup, K., & Moser, D. K. (1991). Comparing methods of
cardiopulmonary resuscitation instruction on learning and retention. JNurs Staff
Dev, 7(2), 82-87.

Dai, D., & Walker, W. A. (1998). Role of bacterial colonization in neonatal necrotizing
enterocolitis and its prevention. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi,
39(6), 357-365.

Davis, M. K. (2001). Breastfeeding and chronic disease in childhood and adolescence.
Pediatr Clin North Am, 48(1), 125-141, ix.

Dewey, K. G., Heinig, M. J., & Nommsen, L. A. (1993). Maternal weight-loss patterns
during prolonged lactation. Am J Clin Nutr, 58(2), 162-166.









Diaz Gomez, M., Ramos Acosta, C. L., Rico Sevillano, J., Robayna Curbelo, M., &
Alvarez Alvarez, J. (1997). [Breast feeding and length of hospitalization]. Rev
Enferm, 20(231), 11-14.

do Nascimento, M. B., & Issler, H. (2004). [Breastfeeding in premature infants: in-
hospital clinical management]. JPediatr (Rio J), 80(5 Suppl), S163-172.

Docherty, S. L., Miles, M. S., & Holditch-Davis, D. (2002). Worry about child health in
mothers of hospitalized medically fragile infants. Adv Neonatal Care, 2(2), 84-92.

Ekstrom, A., Matthiesen, A. S., Widstrom, A. M., & Nissen, E. (2005). Breastfeeding
attitudes among counselling health professionals. ScandJPublic Health, 33(5),
353-359.

Ekstrom, A., Widstrom, A. M., & Nissen, E. (2005). Process-oriented training in
breastfeeding alters attitudes to breastfeeding in health professionals. ScandJ
Public Health, 33(6), 424-431.

Espy, K. A., & Senn, T. E. (2003). Incidence and correlates of breast milk feeding in
hospitalized preterm infants. Soc Sci Med, 57(8), 1421-1428.

Fergusson, D. M., & Woodward, L. J. (1999). Breast feeding and later psychosocial
adjustment. Paediatr Perinat Epidemiol, 13(2), 144-157.

Gartner, L. M., Morton, J., Lawrence, R. A., Naylor, A. J., O'Hare, D., Schanler, R. J., et
al. (2005). Breastfeeding and the use of human milk. Pediatrics, 115(2), 496-506.

Goldrick, B., Gruendemann, B., & Larson, E. (1993). Learning styles and
teaching/learning strategy preferences: implications for educating nurses in critical
care, the operating room, and infection control. Heart Lung, 22(2), 176-182.

Gomez-Sanchiz, M., Canete, R., Rodero, I., Baeza, J. E., & Avila, 0. (2003). Influence of
breast-feeding on mental and psychomotor development. Clin Pediatr (Phila),
42(1), 35-42.

Gray, L., Miller, L. W., Philipp, B. L., & Blass, E. M. (2002). Breastfeeding is analgesic
in healthy newborns. Pediatrics, 109(4), 590-593.

Gross, S. J., David, R. J., Bauman, L., & Tomarelli, R. M. (1980). Nutritional
composition of milk produced by mothers delivering preterm. JPediatr, 96(4),
641-644.

Hanson, L. A. (1998). Breastfeeding provides passive and likely long-lasting active
immunity. Ann Allergy Asthma Immunol, 81(6), 523-533; quiz 533-524, 537.

Hanson, L. A., Korotkova, M., Haversen, L., Mattsby-Baltzer, I., Hahn-Zoric, M.,
Silfverdal, S. A., et al. (2002). Breast-feeding, a complex support system for the
offspring. Pediatr Int, 44(4), 347-352.









Harrington, S. S., & Walker, B. L. (2004). The effects of computer-based training on
immediate and residual learning of nursing facility staff. J Contin Educ Nurs,
35(4), 154-163; quiz 186-157.

Harrold, J., & Schmidt, B. (2002). Evidence-based neonatology: making a difference
beyond discharge from the neonatal nursery. Curr Opin Pediatr, 14(2), 165-169.

Hart, S., Boylan, L. M., Carroll, S., Musick, Y. A., & Lampe, R. M. (2003). Brief report:
breast-fed one-week-olds demonstrate superior neurobehavioral organization. J
Pediatr Psychol, 28(8), 529-534.

Hill, P. D., Andersen, J. L., & Ledbetter, R. J. (1995). Delayed initiation of breast-feeding
the preterm infant. JPerinat Neonatal Nurs, 9(2), 10-20.

Hill, P. D., Hanson, K. S., & Mefford, A. L. (1994). Mothers of low birthweight infants:
breastfeeding patterns and problems. JHum Lact, 10(3), 169-176.

Holditch-Davis, D., & Miles, M. S. (2000). Mothers' stories about their experiences in the
neonatal intensive care unit. Neonatal Netw, 19(3), 13-21.

How breast-feeding postpones ovulation. (1985). Network, 7(1), 3.

Hwang, W. J., Chung, W. J., Kang, D. R., & Suh, M. H. (2006). [Factors affecting
breastfeeding rate and duration]. JPrev MedPub Health, 39(1), 74-80.

Hylander, M. A., Strobino, D. M., & Dhanireddy, R. (1998). Human milk feedings and
infection among very low birth weight infants. Pediatrics, 102(3), E38.

Hylander, M. A., Strobino, D. M., Pezzullo, J. C., & Dhanireddy, R. (2001). Association
of human milk feedings with a reduction in retinopathy of prematurity among very
low birthweight infants. JPerinatol, 21(6), 356-362.

Isaacson, L. J. (2006). Steps to successfully breastfeed the premature infant. Neonatal
Netw, 25(2), 77-86.

Jaeger, M. C., Lawson, M., & Filteau, S. (1997). The impact of prematurity and neonatal
illness on the decision to breast-feed. JAdv Nurs, 25(4), 729-737.

Jang, Y. S. (2005). [Effects of a workbook program on the perceived stress level,
maternal role confidence and breast feeding practice of mothers of premature
infants]. Taehan Kanho Hakhoe Chi, 35(2), 419-427.

Karl, D. J. (2004). Using principles of newborn behavioral state organization to facilitate
breastfeeding. MCNAm JMatern ChildNurs, 29(5), 292-298.

Kavanaugh, K., Mead, L., Meier, P., & Mangurten, H. H. (1995). Getting enough:
mothers' concerns about breastfeeding a preterm infant after discharge. JObstet
Gynecol Neonatal Nurs, 24(1), 23-32.









Kemp, A., & Kakakios, A. (2004). Asthma prevention: breast is best? JPaediatr Child
Health, 40(7), 337-339.

Kirsten, G. F., Bergman, N. J., & Hann, F. M. (2001). Kangaroo mother care in the
nursery. Pediatr Clin North Am, 48(2), 443-452.

Kronborg, H., & Vaeth, M. (2004). The influence of psychosocial factors on the duration
of breastfeeding. ScandJPublic Health, 32(3), 210-216.

Lanari, M., Papa, I., Venturi, V., Sermasi, S., Corvaglia, L., Faldella, G., et al. (2001).
[Neonatal sepsis]. Recent ProgMed, 92(11), 690-695.

Lang, S. (1996). Breastfeeding special care babies. ModMidwife, 6(11), 34-35.

Laubereau, B., Brockow, I., Zirngibl, A., Koletzko, S., Gruebl, A., von Berg, A., et al.
(2004). Effect of breast-feeding on the development of atopic dermatitis during the
first 3 years of life--results from the GINI-birth cohort study. JPediatr, 144(5),
602-607.

Li, R., Darling, N., Maurice, E., Barker, L., & Grummer-Strawn, L. M. (2005).
Breastfeeding rates in the United States by characteristics of the child, mother, or
family: the 2002 National Immunization Survey. Pediatrics, 115(1), e31-37.

Lucas, A., Morley, R., Cole, T. J., & Gore, S. M. (1994). A randomised multicentre study
of human milk versus formula and later development in preterm infants. Arch Dis
Child Fetal Neonatal Ed, 70(2), F 141-146.

Lucas, A., Morley, R., Cole, T. J., Lister, G., & Leeson-Payne, C. (1992). Breast milk
and subsequent intelligence quotient in children born preterm. Lancet, 339(8788),
261-264.

Lugo-Vicente, H. (2003). Necrotizing enterocolitis. BolAsoc MedP R, 95(2), 17-22.

Lupton, D., & Fenwick, J. (2001). 'They've forgotten that I'm the mum': constructing and
practising motherhood in special care nurseries. Soc SciMed, 53(8), 1011-1021.

Marild, S., Hansson, S., Jodal, U., Oden, A., & Svedberg, K. (2004). Protective effect of
breastfeeding against urinary tract infection. Acta Paediatr, 93(2), 164-168.

Matthews, K., Webber, K., McKim, E., Banoub-Baddour, S., & Laryea, M. (1998).
Maternal infant-feeding decisions: reasons and influences. Can JNurs Res, 30(2),
177-198.

McGrath, J. M., & Braescu, A. V. (2004). State of the science: feeding readiness in the
preterm infant. JPerinat Neonatal Nurs, 18(4), 353-368; quiz 369-370.

McVea, K. L., Turner, P. D., & Peppler, D. K. (2000). The role of breastfeeding in
sudden infant death syndrome. JHum Lact, 16(1), 13-20.









Meier, P. P. (1998). Strategies for assisting breatfeeding in preterm infants.

Meier, P. P. (2001). Breastfeeding in the special care nursery. Prematures and infants
with medical problems. Pediatr Clin North Am, 48(2), 425-442.

Meier, P. P., & Brown, L. P. (1996). State of the science. Breastfeeding for mothers and
low birth weight infants. Nurs Clin North Am, 31(2), 351-365.

Meier, P. P., Engstrom, J. L., Mingolelli, S. S., Miracle, D. J., & Kiesling, S. (2004). The
Rush Mothers' Milk Club: breastfeeding interventions for mothers with very-low-
birth-weight infants. J Obstet Gynecol Neonatal Nurs, 33(2), 164-174.

Mikiel-Kostyra, K. (2000). [Breast feeding as a component of reproductive health].
GinekolPol, 71(7), 641-647.

Miles, M. S., Funk, S. G., & Kasper, M. A. (1992). The stress response of mothers and
fathers of preterm infants. Res Nurs Health, 15(4), 261-269.

Mitra, A. K., Khoury, A. J., Hinton, A. W., & Carothers, C. (2004). Predictors of
breastfeeding intention among low-income women. Matern Child Health J, 8(2),
65-70.

Morley, R., Cole, T. J., Powell, R., & Lucas, A. (1988). Mother's choice to provide breast
milk and developmental outcome. Arch Dis Child, 63(11), 1382-1385.

Neiva, F. C., Cattoni, D. M., Ramos, J. L., & Issler, H. (2003). [Early weaning:
implications to oral motor development]. JPediatr (Rio J), 79(1), 7-12.

Nikolaj ski, P. Y. (1992). Investigating the effectiveness of self-learning packages in staff
development. JNurs StaffDev, 8(4), 179-183.

Nyqvist, K. H., Ewald, U., & Sjoden, P. O. (1996). Supporting a preterm infant's
behaviour during breastfeeding: a case report. JHum Lact, 12(3), 221-228.

Nyqvist, K. H., Sjoden, P. O., & Ewald, U. (1994). Mothers' advice about facilitating
breastfeeding in a neonatal intensive care unit. JHum Lact, 10(4), 237-243.

Nyqvist, K. H., Sjoden, P. O., & Ewald, U. (1999). The development of preterm infants'
breastfeeding behavior. Early Hum Dev, 55(3), 247-264.

Oddy, W. H., Sherriff, J. L., de Klerk, N. H., Kendall, G. E., Sly, P. D., Beilin, L. J., et al.
(2004). The relation of breastfeeding and body mass index to asthma and atopy in
children: a prospective cohort study to age 6 years. Am JPublic Health, 94(9),
1531-1537.

The optimal duration of exclusive breastfeeding: results of a WHO systematic review.
(2001). Indian Pediatr, 38(5), 565-567.









Owen, C. G., Whincup, P. H., Odoki, K., Gilg, J. A., & Cook, D. G. (2002). Infant
feeding and blood cholesterol: a study in adolescents and a systematic review.
Pediatrics, 110(3), 597-608.

Page, D. C. (2001). Breastfeeding is early functional jaw orthopedics (an introduction).
Funct Orthod, 18(3), 24-27.

Pantazi, M., Jaeger, M. C., & Lawson, M. (1998). Staff support for mothers to provide
breast milk in pediatric hospitals and neonatal units. JHum Lact, 14(4), 291-296.

Philipp, B. L., Malone, K. L., Cimo, S., & Merewood, A. (2003). Sustained breastfeeding
rates at a US baby-friendly hospital. Pediatrics, 112(3 Pt 1), e234-236.

Pinelli, J., Atkinson, S. A., & Saigal, S. (2001). Randomized trial of breastfeeding
support in very low-birth-weight infants. Arch Pediatr Adolesc Med, 155(5), 548-
553.

Piper, S., & Parks, P. L. (1996). Predicting the duration of lactation: evidence from a
national survey. Birth, 23(1), 7-12.

Powers, N. G., Bloom, B., Peabody, J., & Clark, R. (2003). Site of care influences
breastmilk feedings at NICU discharge. JPerinatol, 23(1), 10-13.

Premji, S. S., Paes, B., Jacobson, K., & Chessell, L. (2002). Evidence-based feeding
guidelines for very low-birth-weight infants. Adv Neonatal Care, 2(1), 5-18.

Prochaska, J. M., Prochaska, J. O., & Levesque, D. A. (2001). A transtheoretical
approach to changing organizations. Adm Policy Ment Health, 28(4), 247-261.

Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self-change of
smoking: toward an integrative model of change. J Consult Clin Psychol, 51(3),
390-395.

Rea, M. F. (2004). [Benefits ofbreastfeeding and women's health]. JPediatr (Rio J),
80(5 Suppl), S142-146.

Register, N., Eren, M., Lowdermilk, D., Hammond, R., & Tully, M. R. (2000).
Knowledge and attitudes of pediatric office nursing staff about breastfeeding. J
Hum Lact, 16(3), 210-215.

Riskin, A., & Bader, D. (2003). [Breast is best--human milk for premature infants].
Harefuah, 142(3), 217-222, 237, 236.

Rooney, B. L., & Schauberger, C. W. (2002). Excess pregnancy weight gain and long-
term obesity: one decade later. Obstet Gynecol, 100(2), 245-252.









Ryan, A. S., Wysong, J. L., Martinez, G. A., & Simon, S. D. (1990). Duration of breast-
feeding patterns established in the hospital. Influencing factors. Results from a
national survey. Clin Pediatr (Phila), 29(2), 99-107.

Schack-Nielsen, L., & Michaelsen, K. F. (2006). Breast feeding and future health. Curr
Opin Clin Nutr Metab Care, 9(3), 289-296.

Schanler, R. J., Hurst, N. M., & Lau, C. (1999). The use of human milk and breastfeeding
in premature infants. Clin Perinatol, 26(2), 379-398, vii.

Scott, J. A., Binns, C. W., Graham, K. I., & Oddy, W. H. (2006). Temporal changes in
the determinants of breastfeeding initiation. Birth, 33(1), 37-45.

Shadish, W. R., Cook, T. D., & Campbell, D. T. (2001). Experimental and quasi-
experimental designsfor generalized causal inference. Boston: Houghton Mifflin.

Siddell, E., Marinelli, K., Froman, R. D., & Burke, G. (2003). Evaluation of an
educational intervention on breastfeeding for NICU nurses. JHum Lact, 19(3),
293-302.

Singhal, A., Cole, T. J., Fewtrell, M., & Lucas, A. (2004). Breastmilk feeding and
lipoprotein profile in adolescents born preterm: follow-up of a prospective
randomised study. Lancet, 363(9421), 1571-1578.

Sisk, P. M., Lovelady, C. A., Dillard, R. G., & Gruber, K. J. (2006). Lactation counseling
for mothers of very low birth weight infants: effect on maternal anxiety and infant
intake of human milk. Pediatrics, 117(1), e67-75.

Slusher, T., Hampton, R., Bode-Thomas, F., Pam, S., Akor, F., & Meier, P. (2003).
Promoting the exclusive feeding of own mother's milk through the use of hindmilk
and increased maternal milk volume for hospitalized, low birth weight infants (<
1800 grams) in Nigeria: a feasibility study. JHum Lact, 19(2), 191-198.

Smith, M. M., Durkin, M., Hinton, V. J., Bellinger, D., & Kuhn, L. (2003). Influence of
breastfeeding on cognitive outcomes at age 6-8 years: follow-up of very low birth
weight infants. Am JEpidemiol, 158(11), 1075-1082.

Smithers, L. G., McPhee, A. J., Gibson, R. A., & Makrides, M. (2003). Characterisation
of feeding patterns in infants born < 33 weeks gestational age. Asia Pac J Clin
Nutr, 12 Suppl, S43.

Spicer, K. (2001). What every nurse needs to know about breast pumping: instructing and
supporting mothers of premature infants in the NICU. NeonatalNetw, 20(4), 35-41.

Stewart-Knox, B., Gardiner, K., & Wright, M. (2003). What is the problem with breast-
feeding? A qualitative analysis of infant feeding perceptions. JHum Nutr Diet,
16(4), 265-273.






86


Swanson, V., & Power, K. G. (2005). Initiation and continuation of breastfeeding: theory
of planned behaviour. JAdv Nurs, 50(3), 272-282.

Torgus, J., Gotsch, G., & La Leche League International. (1997). The womanly art of
breastfeeding (6th rev ed.). Schaumburg, Ill.: La Leche League International.

Vannuchi, M. T., Monteiro, C. A., Rea, M. F., Andrade, S. M., & Matsuo, T. (2004).
[The Baby-Friendly Hospital Initiative and breastfeeding in a neonatal unit]. Rev
Saude Publica, 38(3), 422-428.

Viggiano, D., Fasano, D., Monaco, G., & Strohmenger, L. (2004). Breast feeding, bottle
feeding, and non-nutritive sucking; effects on occlusion in deciduous dentition.
Arch Dis Child, 89(12), 1121-1123.

What is best birth control to use after having a baby? (1989). Contracept Technol Update,
10(10), 1S-2S.

Wheeler, J., Chapman, C., Johnson, M., & Langdon, R. (2000). Feeding outcomes and
influences within the neonatal unit. Int JNurs Pract, 6(4), 196-206.

Wheeler, J. L., Johnson, M., Collie, L., Sutherland, D., & Chapman, C. (1999).
Promoting breastfeeding in the neonatal intensive care unit. BreastfeedRev, 7(2),
15-18.

Wold, A. E., & Adlerberth, I. (2000). Breast feeding and the intestinal microflora of the
infant--implications for protection against infectious diseases. Adv Exp MedBiol,
478, 77-93.

Wolf, J. H. (2003). Low breastfeeding rates and public health in the United States. Am J
Public Health, 93(12), 2000-2010.

Yip, E., Lee, J., & Sheehy, Y. (1996). Breast-feeding in neonatal intensive care. J
Paediatr Child Health, 32(4), 296-298.

Ziemer, M. M., & George, C. (1990). Breastfeeding the low-birthweight infant. Neonatal
Netw, 9(4), 33-38.

Zimmerman, D. R., & Guttman, N. (2001). "Breast is best": knowledge among low-
income mothers is not enough. JHum Lact, 17(1), 14-19.















BIOGRAPHICAL SKETCH

Dr. Roberta Gittens Pineda received her doctor of philosophy degree at the

University of Florida. She received her Bachelor of Science in occupational therapy at

the Florida Agricultural and Mechanical University in 1992 and achieved her Master of

Health Science degree at the University of Florida in 1994. She has worked as an

occupational therapist, primarily in the inpatient pediatric setting, and has been a lecturer

at University of Florida as well as Washington University.

Dr. Pineda's primary clinical setting is the neonatal intensive care unit, where she

most recently has specialized in treatment of feeding and swallowing problems in these

complex, medically fragile infants. In addition, she suffered premature labor with her 3

pregnancies and learned, first hand, the difficulties associated with being a mother of an

infant hospitalized in the neonatal intensive care unit.

Dr. Pineda decided on her dissertation topic following an invitation by the chief of

neonatology at Shands Hospital to be part of a March of Dimes Advisory Committee.

The University of Florida at Shands Hospital had decided to participate in a program

sponsored by the March of Dimes aimed at making the unit more developmentally

supportive and family centered. Through this project, she joined efforts with a

pediatrician to design a plan aimed at making the neonatal intensive care unit (NICU)

more breastfeeding friendly. As she had frequently assisted mothers with breastfeeding

in the unit as part of occupational therapy intervention and had struggled with the issues

of nursing in the NICU when her own son was born at 29 weeks gestation, she found






88


herself very passionate about fostering changes and quickly decided to focus her research

around the changes that were set to be made. Dr. Pineda's dissertation topic is entitled

"Breastfeeding Practices in the Neonatal Intensive Care Unit before and after an

Intervention Plan". Dr. Pineda has thoroughly enjoyed the research process and looks

forward to a career in research and teaching.




Full Text

PAGE 1

BREASTFEEDING PRACTICES IN THE NEONATAL INTENSIVE CARE UNIT BEFORE AND AFTER AN INTERVENTION PLAN By ROBERTA GITTENS PINEDA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Roberta Gittens Pineda

PAGE 3

This dissertation is dedicated to all the mothers who have premature and medically fragile infants in the neonatal intensive care unit (NICU). It is hoped that this and other works with infants and mothers in the NICU will give you the hope, courage, and information needed to “mother” in the co mplex environment of the NICU, during your infant’s first precious days. This dissertati on is also dedicated to my husband, Jose, for his endless source of love and inspiration. It is also dedicated to my children, Alan and Marissa, whose early birth ma de me realize the importance of education and support for mothers, as well as to my daughter Abigail, whose premature birth in the middle of this research gave me the extra steam to see it through. This dissert ation is especially dedicated to my mom, Barbara Gittens Vale ntine, whose expertise with mothers and babies was critical during my own son’s hospitalization.

PAGE 4

iv ACKNOWLEDGMENTS This research would not have been possi ble without the supp ort and guidance of Dr. Lorie Richards, who has been my advisor, mentor, and friend. I want to thank Lorie for having faith in me and inspiring me to do my best. I would lik e to thank the nurses and health care professionals at Shands Ho spital who participated in this research endeavor. In particular, I give special th anks to Annmarie Brennan, who enabled this research project to occur in the neonatal intensive care unit (NICU) at Shands and supported the project every step of the way. I also would like to thank Cammy Pane, the co-author of the Educational Module; Ste phanie Meeks for your hours of work on “A Mother’s Gift”, and other members of the Lactation Committee at Shands who helped with my research: Elayne McNamara, Sandra Sullivan, Brenda Owens, Sheila Walker and Jeannette Sexton. I want to give special thanks to Susan Frazier from Medela for your support at the inservices. I also want to thank all those who provided donations as incentives for participation: Sonny’s Barb eque, Atlanta Bread Co mpany, Scholotsky’s Deli, and TGIF. I want to thank Dr. David Bu rchfield, the medical director of the NICU at Shands, for assisting with this project. I thank Sarah Boslaugh for guiding me through the statistics and for all y our patience from the many que stions that came up along the way. I would also like to thank my comm ittee for sticking with me through the years, and the move to St. Louis and the addition of the new baby. I a ppreciate your endless patience, high expectations and sincere enthus iasm for my interests and work. I want to thank Drs. Richards, Foss, Krueger, Seung, and Rosenbek!

PAGE 5

v I would finally like to thank my parents who always showed unc onditional love and always motivated me to strive to do better. I extend special thanks to my husband, Jose, for always being there when I needed you mo st and giving me patience and love every step of the way. You enabled me to go back to school and were there when it came to crunch time. You have made this all possibl e and I am eternally grateful for your love and support.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT....................................................................................................................... ..x CHAPTER 1 INTRODUCTION........................................................................................................1 The Importance of Breast Milk and Breastfeeding.......................................................2 Health Benefits of Breastfeed ing for the Full Term Infant...................................3 Health Benefits of Breastfeed ing for Premature Infants.......................................4 Long Term Benefits of Breastfeeding...................................................................5 Developmental Benefits of Breastfeeding.............................................................5 Benefits of Breastfeeding for the Mother..............................................................6 Current Breastfeeding Recommendations....................................................................7 Why More Women Are Not Giving Their In fants the Benefits of Breast Milk...........8 General Breastfeeding Barriers.............................................................................9 Barriers to Breastfeeding Premature Infants.........................................................9 Health Care Professionals Can Hinde r the Breastfeeding Process in the Neonatal Intensive Care Unit...........................................................................14 Treatments to Foster Improved Breastfeeding Rates..................................................16 Need for an Educational P ackage for Health Care Professionals and Mothers of Infants in the Neonatal Intensive Care Unit............................................................19 Synactive Theory and Breastfeeding Interventions in the Neonatal Intensive Care Unit..........................................................................................................................2 0 Theory Governing the Behavior of Health Care Professionals..................................25 Transtheoretical Model of Behavior Change and Methods of Education..................31 Summary and Research Questions.............................................................................33 2 METHODOLOGY.....................................................................................................35 Participants.................................................................................................................35 Research Interventions................................................................................................36 Intervention 1: Breast Pump Loaner Closet.......................................................36

PAGE 7

vii Intervention 2: Health Prof essional Education Initiative...................................37 Intervention 3: Br eastfeeding Guideline............................................................38 Intervention 4: Educational Pam phlet for New Mothers of Neonatal Intensive Care Unit Infants..............................................................................38 Intervention Plan Modification...................................................................................39 Design......................................................................................................................... 39 Procedures...........................................................................................................40 Program Evaluation.............................................................................................41 Data Collection...........................................................................................................42 Data Analysis..............................................................................................................42 Hypotheses..........................................................................................................43 Adjusting the Alpha Level...................................................................................44 3 RESULTS...................................................................................................................46 Intervention Implementation......................................................................................46 The Sample.................................................................................................................48 Inter-Rater Agreement................................................................................................50 Demographics.............................................................................................................50 Investigation for Selection Differences......................................................................52 Results Per Research Question...................................................................................52 4 DISCUSSION.............................................................................................................59 The Effect of the Interventions on Br eastfeeding Practices in the Neonatal Intensive Care Unit.................................................................................................59 Limitations..................................................................................................................64 Recommendations for Further Research....................................................................69 Conclusions.................................................................................................................71 APPENDIX A OUTLINE OF THE EDUCATION MODULE..........................................................73 B OUTLINE OF ITEMS ADDED TO TH E INDIVIDUALIZED CARE PLAN.........75 C OUTLINE OF THE EDUCATI ONAL BOOKLET FOR MOTHERS......................76 LIST OF REFERENCES...................................................................................................78 BIOGRAPHICAL SKETCH.............................................................................................87

PAGE 8

viii LIST OF TABLES Table page 1-1 Stages of change in which particul ar processes of change are emphasized.............28 2-1 Hypothesis testing accordi ng to dependent variable................................................43 3-1 Demographics of the pre-interventio n and post-intervention groups with test statistics for selection differences.............................................................................51 3-2 Breast milk feeding initiation rates...........................................................................53 3-3 Comparison of rates of breastf eeding after 30 we eks gestation................................55 3-4 Rates of ever breastfed in the neonatal inte nsive care unit.......................................56 3-5 Rates of breast milk feedings at discharge................................................................57

PAGE 9

ix LIST OF FIGURES Figure page 3-1 Number of times breastfed per day after 30 weeks gestation..................................54 3-2 Proportion of the stay that breast milk was provided...............................................58

PAGE 10

x Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy BREASTFEEDING PRACTICES IN THE NEONATAL INTENSIVE CARE UNIT BEFORE AND AFTER AN INTERVENTION PLAN By Roberta Gittens Pineda August 2006 Chair: Lorie Richards Major Department: Rehabilitation Science The benefits of breastfeeding for both mother and infant ar e cited extensively in the literature. Premature infants hospitalized in the neonatal intensive care unit (NICU) have a great need for the benefits that breast milk offers, due to their fragile health states. However, mothers of very low birth weight infants hospitalized in the NICU have magnified barriers to th e breastfeeding process due to the complexity of medical conditions that warrant admission to the NICU and the separation of the infant from the mother to enable medical care. Studies ha ve cited lack of edu cation about lactation among health care professionals and discrepanc ies in education dissemination to mothers as a major barrier to the breastfeeding proce ss. A three-part intervention within the NICU was implemented that consisted of an educational initiative for health care professionals who instruct and support mother s, modifications to th e individualized care plan that included a new breastfeeding path way, and an educational booklet for mothers with infants hospitalized in the NICU. Cha nge in breast milk feeding initiation rates,

PAGE 11

xi breastfeeding rates, breast milk at discharge rates, and propor tion of the hosp ital stay that breast milk was provided was investigated be tween pre-intervention and post intervention groups. Results indicated general positive tre nds in all variables, but only one variable achieved statistical significance. The percen tage of infants who were ever breastfed while in the hospital increased from 25.9% before the interventi on to 44.4% after the intervention, and this reached statistical significance with a p value of .025. Full implementation of strategies l earned in the interventions wa s questionable. This study provides partial support of the three-part intervention in facilitating breast feeding in the NICU. Possible reasons for lack of change across all variables, as well as other possible interventions that could aff ect change, are explored.

PAGE 12

1 CHAPTER 1 INTRODUCTION Breastfeeding is an important part of the occupation of mothering. However, mothers of infants admitted into the neonatal intensive care unit (NICU) are not able to function in the traditional role of mother. They are usually separated from their infants, and the role of caregiver shif ts to health care professiona ls. In addition, many infants may be attached to life-saving or monitori ng equipment, which can be intimidating for new parents. Even more intimidating is that many of these infants are fragile or lack neurological maturity, which a ffects how the mother will in teract with and care for her infant. This environment presents significan t barriers to the pr ovision of breast milk, including the fragility of the infant, the separation of the in fant from the mother, and the behavior of the health care professionals who are focused on the medical interventions necessary for these infants. This is unfortunate because the established benefits of breast milk may be even greater in these medically fragile and matu rationally immature infants. The rate of breastfeeding in the United St ates, despite repeated advertisement of its benefits, is only 71.4% (Li, Darling, Maurice, Barker, & Grummer-Strawn, 2005). Unfortunately, due to the many barriers to br eastfeeding in the NICU environment, this percentage is significantly lower for infants discharged from the NICU, with breast milk feedings in premature infants reaching only approximately 50% (Espy & Senn, 2003). However, despite the medical complexities of the NICU and the shift of care to health care professionals, with adequate circumventi on of barriers, mothers can be supported in the occupation of moth ering through support of breastfeeding.

PAGE 13

2 Studies have identified that health care professional support is predictive of success with breastfeeding (Swanson & Power, 2005). However, health care professionals must be given the tools to foster breastfeeding in the complex NICU environment. Therefore, the aim of this study was to test the efficacy of an intervention to support breastfeeding practices in the NICU. The intervention centered on health care professional behavior change through an educational initiative for health care professionals, modifications to the individualized care plan (ICP) with a breastfeeding protocol, and educational materials for mothers with infants in the NI CU. It was hypothesized that the intervention plan would foster change in health care pr ofessionals, which would then enable positive changes in breastfeeding pr actices in the NICU. The Importance of Breast Milk and Breastfeeding Breast milk can be provided to the infant either directly through infant suckling at the breast (breastfeeding) or by having the mother express the breast milk with a pump and providing the milk via enteral feedings or bottle (breast milk feedings). The health benefits of breastfeeding for the infant are cited in the literatu re extensively (Wolf, 2003). Breast milk has a protective effect against many childhood health problems. Breast milk differs from formula in that it has unique in gredients that are difficult, if not impossible, to duplicate. Important components of breas t milk are IgA antibodies, which aid in preventing infection by creating a non-inflammatory response in body cells. This enables a more active immune system, which demonstr ates better defense against infection. Other factors in breast milk, such as lactof errin and oligosaccharides have also been isolated and are believed to prevent mucous attachment, the origin of most infections (Hanson, 1998; Hanson et al., 2002).

PAGE 14

3 Infant formulas continue to strive to be similar to breast milk and have become nutritionally advanced in the la st decade, however, research c ontinues to illustrate that breast milk is far superior to formula (A gostoni & Haschke, 2003; Baker, 2003; Wold & Adlerberth, 2000). Thus far, formula companie s have been unable to replicate the exact ingredients of breast milk. Perhaps predom inantly due to the IgA antibodies found in human milk, breastfed infants have superi or protection from many ailments that compromise health and pr event optimal functioning. Health Benefits of Breastfeeding for the Full Term Infant When comparing babies who are fed breas t milk to those who are formula fed, there is a significant reduction in resp iratory infections, di arrhea, necrotizing enterocolitis, meningitis, sepsis, urinary tract infections, atopic dermatitis, celiac disease, and inflammatory bowel disease in the breas tfed babies (Dai & Walker, 1998; Hanson, 1998; Hylander, Strobino, & Dhanireddy, 1998; Laubereau et al., 2004; Marild, Hansson, Jodal, Oden, & Svedberg, 2004; Wold & Adlerberth, 2000). Although preliminary studies have not been conclusive, it is also suggested that allergie s and asthma are also diminished among breastfed babies (Kem p & Kakakios, 2004; Oddy et al., 2004). Breastfed babies have a diminished risk of sudden infant death syndrome (Alm et al., 2002; McVea, Turner, & Peppler, 2000), as well as a significantly lower risk of mortality after the neonatal peri od (Chen & Rogan, 2004). Because it is associated with less infant illness, breast feeding may cut medical expenses for the infant. Ball and Wright (1999) addressed excess medical costs for 3 common childhood illnesses: gast rointestinal infection, resp iratory tract infection and otitis media among breast fed versus formula fed infants in the first year of life. There was evidence that children who were never breast fed incurred significantly more office

PAGE 15

4 visits, hospitalizations, prescr iptions and subsequently had higher health care costs (Ball & Wright, 1999). Thus, the health advantages associated with breas tfeeding create less financial burden as health care costs diminish (Ball & Wright, 1999) and, more importantly, they improve the quality of lif e and health status among mother-infant dyads. There have been a multitude of studies th at have also investigated health and developmental benefits of breast milk for premature and high risk neonates (Callen & Pinelli, 2005). The fragile health states of these infants make them more susceptible to infection, gastrointestinal probl ems, and life threaten ing illnesses than full term infants (Lanari et al., 2001; Lugo-Vicente, 2003). Ther efore, breast milk is perhaps more important in this fragile population, because it diminishes the risk of multiple medical problems, which can complicate the medical co urse and put them at a higher risk of developmental sequelae. Health Benefits of Breastfeeding for Premature Infants Breast milk fed infants from the NICU di ffer significantly from formula fed infants in incidence of infection and diagnosis of sepsis/meningitis (Hylander et al., 1998), necrotizing enterocolitis, and retinopathy of prematurity (Hyl ander et al., 1998; Hylander, Strobino, Pezzullo, & Dhanireddy, 2001; Scha nler, Hurst, & Lau, 1999). Breast fed premature babies have been noted to expe rience less stress than bottle fed infants as evidenced by fewer episodes of oxygen desatu ration and temperature instability (C. H. Chen, Wang, Chang, & Chi, 2000). Breastfeeding has been cited as an intervention that has lasting, long term benef it beyond discharge from the hos pital (Harrold & Schmidt, 2002), and studies have detected significant reductions in le ngth of stay among breastfed premature infants (Gomez, Acosta, Sevillano, Curbelo, & Alvarez, 1997).

PAGE 16

5 Long Term Benefits of Breastfeeding More recent studies are sugge sting that the effects of breast milk extend beyond the period of infancy and early childhood, and pr omote long term immunity and protection from chronic diseases. Lower risk of de veloping childhood cancers, obesity, type I diabetes, and cardiovascular disease have been cited as long term benefits of breastfeeding (Davis, 2001; Hanson, 1998; Schack-Nielsen & Michaelsen, 2006; Singhal, Cole, Fewtrell, & Lucas, 2004). Studies of long term effects of breast milk on premature infants have also found benefits with lowe r blood pressure readings in adolescence (Owen, Whincup, Odoki, Gilg, & Cook, 2002). Despite concerns that breastfeeding results in suboptimal growth in infancy, studi es have demonstrated an increased growth velocity in late childhood in breastfed groups (Schack-N ielsen & Michaelsen, 2006). Additionally, improved parental attachment in the teenage years has been linked to breastfeeding (Fergusson & Woodward, 1999). Developmental Benefits of Breastfeeding Research also points to the importance of breastfeeding on infant development. Breastfeeding results in improved oral motor development and orthodontics (Page, 2001), with early weaning increasi ng the risk of malocclusion, mouth breathing, dysfunctional oral motor development and subsequent s uboptimal speech development (Neiva, Cattoni, Ramos, & Issler, 2003; Viggiano, Fasano, M onaco, & Strohmenger, 2004). Breastfed infants have improved visual motor skills (Bir ch et al., 1993), have better responses to pain (Gray, Miller, Philipp, & Blass, 2002) with improved neurobeha vioral organization (Hart, Boylan, Carroll, Musick, & Lampe, 200 3) and have demonstrated improved scores on mental functioning (Gomez-Sanchiz, Canete, Rodero, Baeza, & Avila, 2003).

PAGE 17

6 Premature infants have a greater risk of poor neurological outcome, which suggests that breast milk may be critical to enable optimal developmental functioning. Research has demonstrated improved cognitive and motor functioning scores among premature infants who had breast milk feedings (Lanar i et al., 2001; Schanler et al., 1999). Studies demonstrate improved cognitive scores and inte lligence quotients that continued to be evident through middle childhood among breastf ed infants (Lucas, Morley, Cole, & Gore, 1994; Lucas, Morley, Cole, Lister, & L eeson-Payne, 1992; Morley, Cole, Powell, & Lucas, 1988; Smith, Durkin, Hinton, Bellinger, & Kuhn, 2003). Benefits of Breastfe eding for the Mother In addition to the benefits given to th e baby through breastfeeding, there are also benefits for the mother. Women who succeed with breastfeeding comment on the special bonding experience (Torgus, Gotsch, & La Lech e League International., 1997). Women who breastfeed have less postpartum bleed ing (Chua, Arulkumaran, Lim, Selamat, & Ratnam, 1994) and have a faster rate of pregnancy related we ight loss (Dewey, Heinig, & Nommsen, 1993). Women who breastfeed al so postpone ovulation (Rea, 2004), and breastfeeding has been demonstrated to serv e as a natural and effective birth control method in the postpartum period ("How breast-feeding postpones ovulation," 1985; "What is best birth control to use after ha ving a baby?," 1989). In addition, women who succeed with breastfeeding lower their risk of osteoporosis (Chantry, Auinger, & Byrd, 2004), obesity (Rooney & Schauberger, 2002), ov arian cancer, breas t cancer (MikielKostyra, 2000), diabetes and rheumatoid arthritis (Rea, 2004). Mothers of preterm infants have additiona lly reported an improved sense of well being, as they feel that they are actively contributing to the health of their babies (Schanler et al., 1999). Having an infant in the NICU is a difficult challenge. Mothers

PAGE 18

7 may feel shut off from their in fant as the nurses take on the role of primary caregiver. Being able to provide the best source of nutrition can be one task that embraces the mother in her role and fosters parental involvement, as it is something only she can do for her baby. Current Breastfeeding Recommendations The health benefits of breastfeeding for mother, baby and health care systems are evident and extensive. Therefore, the American Academy of Pediatrics (AAP), as well as the American Dietetic Association, have res ponded to the benefits of breastfeeding for mother and baby by recommending exclusive br eastfeeding for the first 6 months with breastfeeding and supplemental solids until the in fant is 1 year old ("Breastfeeding and the use of human milk. American A cademy of Pediatrics. Work Group on Breastfeeding," 1997). The World Health Orga nization recommends breastfeeding for at least 2 years ("The optimal duration of exclusive breastfeeding: results of a WHO systematic review," 2001). However, breastf eeding statistics conti nue to demonstrate a gap between these recommendations and how the general population of mothers in the United States chooses to feed th eir infants (Li et al., 2005). Some women never breastfeed, some br eastfeed exclusively, some supplement breastfeeding with bottle feed s of human milk, some supplement breastfeeding with bottle feeds of formula, some bottle feed fo rmula only, some bottle feed breast milk only, and some women start out breastfeeding and completely wean once formula is introduced. Statistics from the year 2003 indicated that 71.4. % of women in the general population initiated breastfeeding while in th e hospital, and 35.1% of mothers were still breastfeeding when their babies turned 6 m onths of age. At one year of age, 16.1% continued to provide some breast milk for thei r infants (Li et al., 2005) Variable rates of

PAGE 19

8 breastfeeding have been reporte d for infants in the NICU. Th e rates for infants receiving some breast milk at some point range from 50% to 83% (Byrne & Hull, 1996; Espy & Senn, 2003; Meier, Engstrom, Mingolelli, Mira cle, & Kiesling, 2004; Smithers, McPhee, Gibson, & Makrides, 2003; Yip, Lee, & Sheehy, 1996). However, studies have found that the rates of breast milk feeds at discharge ar e 64%, with the rate of breastfeeding being 38% (Yip et al., 1996). One study found that at 4 months of age, only 24% of infants born at less than 33 weeks gestation continue to re ceive some breast milk feedings (Smithers et al., 2003). Subsequently, breastfeeding rates at hospital discharge for infants born prematurely are significantly lo wer than those of full term, healthy infants (Yip et al., 1996). To understand the suboptimal breastfeedi ng rates for premature and high risk neonates, it is beneficial to investigate the barriers to breastfeeding. By understanding the barriers to breastfeeding, appropriate interventions can be developed and implemented to facilitate im proved breastfeeding practices. Why More Women Are Not Giving Their Infa nts the Benefits of Breast Milk Maternal demographics are st rong predictors of breastfeed ing. Women with higher socioeconomic status, more education, prev ious children but smaller family size, Caucasian race, and women who are marri ed are more likely to succeed with breastfeeding (Bueno et al ., 2003; Kronborg & Vaeth, 2004; Mitra, Khoury, Hinton, & Carothers, 2004). However, perinatal medical condition is also an important predictor of successful breastfeeding (Espy & Senn, 2003; Powers, Bloom, Peabody, & Clark, 2003). Scott (2006) discovered that the infant bei ng admitted to the intensive care unit was the strongest predictor of not bei ng exclusively breastfed at di scharge (Scott, Binns, Graham, & Oddy, 2006). Other studies have concluded that having a cesa rean section, as well as

PAGE 20

9 having a low birth weight infant, makes a woman less likely to breastfeed (Hwang, Chung, Kang, & Suh, 2006). Demographic factor s as well as medical condition and type of delivery have been shown to be strong in fluences on the decision to breastfeed and the success of breastfeeding. General Breastfeeding Barriers The barriers to breastfeeding full term, healthy infants include lack of family and spouse support and perceptions of lack of support; (Arora McJunkin, Wehrer, & Kuhn, 2000; Matthews, Webber, McKim, Banoub-Baddour & Laryea, 1998; Scott et al., 2006), social withdrawal and isol ation (Stewart-Knox, Gardiner & Wright, 2003), perceived inconvenience (Zimmerman & Guttman, 2001) perceived inadequacy to provide adequate nutrition (Arora et al., 2000; Matt hews et al., 1998), early supplementation or first feeding of formula (Wheeler, Chapman, Johnson, & Langdon, 2000), lack of appropriate education (Arora et al., 2000), functional probl ems with the process of breastfeeding; (Bick, MacA rthur, & Lancashire, 1998), intent to return to work (Arora et al., 2000; Matthews et al., 1998; Piper & Parks, 1996; Ryan, Wysong, Martinez, & Simon, 1990), and maternal illness (Black & Hylander, 2000; Riskin & Bader, 2003). Barriers to Breastfeeding Premature Infants Breastfeeding challenges are stronger and even more numerous for the high risk neonate, despite these babies having an even greater need for human milk. Mothers of infants who are born prematurely have unique challenges to succe ssful breastfeeding. One barrier to breastfeeding the premature in fant is that when an infant is born prematurely and warrants admission into the NICU, the mother is separated from her baby (Black & Hylander, 2000). The time after bi rth is very different for these mothers compared to those with full term infants. Ther e is usually not a period of being able to put

PAGE 21

10 the baby to breast immediately after birth, and breastfeeding may not be possible for several weeks or months, depending on the in fantÂ’s level of prematurity and medical instability. When visitation is possible, mothers may vis it their baby in the intensive care unit. Here, they may have difficulty with the transition to motherhood as the doctors and nurses make decisions related to the care of the baby, including whether or not the mother may hold her new baby (Holditch-Davis & Miles, 2000; Lupton & Fenwick, 2001). The machines and equipment present and being utilized by the baby in the NICU can be overwhelming for many parents (Wheeler et al., 2000), and this environment is very different from the quiet, home -like environment one would typically envision during the first days of the babyÂ’s life. An additional barrie r is that the ability to achieve let down, in which breast milk begins to flow during in fant feeding and pumping, is hindered by the inability to relax in this stressful envi ronment (Beresford, 1984; Nyqvist, Ewald, & Sjoden, 1996; Wheeler, Johnson, Collie, Sutherland, & Chapman, 1999). Many low birth weight infants are unable to breast feed for several weeks or months following birth (Hill, Andersen, & Ledbetter, 1995). Their gastrointestinal systems are immature and feedings can be da ngerous or life threaten ing. During the first days, a baby may be fed intravenously or th rough an orogastric or nasogastric tube, in which feeds may be slowly introduced and adva nced. When the gastrointestinal system is ready for bolus feeds direc tly into the stomach, the ba byÂ’s immature central nervous system may not enable consistent presenta tion of sucking and swallowing responses to enable safe oral feeding (Nyqvist, Sjoden, & Ewald, 1999; Ziemer & George, 1990). Although breastfeeding may not be possible initial ly, breast milk can be expressed by the

PAGE 22

11 mother with a breast pump, and the infant can be advanced on gastric feeds with breast milk. Just as the first feeding by breast is a good predictor of sustained breastfeeding in full term infants, timely pumping for those mothers who are unable to put the baby to breast is an important predicto r of sustained breastfeeding in the premature baby (Jaeger, Lawson, & Filteau, 1997). Women of premature babies may express their breast milk and supply it to hospital staff so that the baby ma y be tube fed with human milk instead of infant formula (Meier & Brown, 1996). Add itionally, this process establishes and maintains a milk supply so that the mother will not have diminished or absent milk supply, when the baby is stable enough to engage in the breastfeeding pr ocess. Barriers to breastfeeding related to this early process include incr eased amounts of stress (Docherty, Miles, & Holditch-Davis, 2002; Miles, Funk, & Kasper, 199 2) and time constraints placed on these new mothers, difficulty in acq uiring hospital grade br east pumps for milk expression, lack of special bonding and emoti onal feedback received from using a pump, delayed initiation of milk expression, se paration from the infant, reliance on medical technology to feed the baby, and psychological adju stment to the idea of not being able to breastfeed for weeks or even months (Byrne & Hull, 1996). With the mother experiencing stress associated with coping w ith her sick baby (Miles et al., 1992) and a shift of care from the mother to the baby afte r the birth, there may be delayed initiation of pumping and lack of accessibility of hospita l grade pumps to promote milk supply in an efficient manner. New studies are highlighting the importance of investig ating barriers at different time periods during an infantÂ’s hospitalizati on (Callen & Pinelli, 2005). If a mother

PAGE 23

12 successfully overcomes the chal lenge of maintaining her milk supply, there are additional challenges as an infant approaches discharg e from the hospital. Poor central nervous system maturity may initially prevent complete success with breastfeeding, and dysphagia is common in this population (H ill, Hanson, & Mefford, 1994). The literature cites problems with the mechanics of breas tfeeding a premature baby as a barrier to breastfeeding (Kavanaugh, M ead, Meier, & Mangurten, 1995). Once discharge is approaching, there frequently is little time to enable a mom and baby to achieve successful breastfeeding (Meier & Brown, 1996). Bottle feeding is often preferred as it allows nurses to orally feed the baby when th e mother is not present and the exact amount ingested can be accurately measured (McG rath & Braescu, 2004). Additionally, infants can be fed more passively with bottle f eeding compared to the active process of breastfeeding. Although this can have negati ve side effects of desaturations and bradycardic events as well as increased risk of gastroesoph ageal reflux, it is frequently preferred because of the efficiency of or al feeding. However, breastfeeding can be achieved in this population despite the preferences of health care staff and the challenges that must be overcome. Infants in the NICU are fed according to a schedule, typically every 3 to 4 hours, and may be fed via bottle, tube or breast to optimize the nutritional status. The inability of a preterm baby to breastfeed on demand in an environment with scheduled feedings via different modes is a significant barrier to breastfeeding (Black & Hylander, 2000). It undermines the typical procedures associated with breastfeeding a full term infant which involves feeding a baby when he/she s hows hunger signs and not supplementing until breastfeeding is well established. This allows for the infant to ingest a smaller feeding

PAGE 24

13 and thus become hungrier and to have a more rigorous, larger feeding for the next one. However, scheduled feedings of specific am ounts can affect the transition to active breastfeeding in the NICU. If the infant typi cally receives a prescr ibed amount of breast milk by bottle or nasogastric tube and the mo ther attempts to breastfeed, not knowing the exact amount of breast milk ingested by br eastfeeding may result in the health care professional doubting if there was adequate intake. Thus supplementation frequently occurs, which inhibits the next breastfeed ing session, decreases the demand for breast milk produced by the mother and diminishes milk supply, and thus becomes a cyclic problem. Diminished milk supply is cited extensiv ely in the literatu re as one of the significant barriers to breastfeeding in th e NICU (Callen & Pine lli, 2005). Among the earliest of premature infants, the average dur ation of providing breast milk is 4-5 weeks (Byrne & Hull, 1996). Lack of ability to enga ge in active breastfeeding due to the health status of the infant, lack of presentation of sucking and swallowing capabilities of the infant and the need for the mother to demons trate consistent milk expression via a breast pump to establish and maintain a milk supply all contribute to diminished milk supply in mothers of infants hospitalized in the NICU. Infants who are born prematurely have diffe rent nutritional need s than full term infants. Thus, there are premature infant form ulas that are utilized in the neonatal period and many are used until one year of life. When gastric feeds are being established, physicians are concerned with establishing a good weight gain trend. If an infant is not gaining weight as desired, the physician may increase caloric density or add lipids to infant formula or expressed breast milk to fo ster weight gain. Fre quently nutrients and

PAGE 25

14 calories are added to human milk by the way of human milk fortifiers (Chan, 2001), which promote establishment of a good weight ga in curve. However, this is a barrier to breastfeeding as the mother perceives that the composition of her breast milk is not adequate to promote the health of her child. She may perceive that formula or fortified breast milk by bottle is essential to enable the appropriate milk composition (Kavanaugh et al., 1995). However, studies suggest that motherÂ’s milk of premature babies differs from that of full term infants with the mo st notable differences evident between 4-6 weeks after delivery (Gross, David, Bauman, & Tomarelli, 1980). Additionally, the use of hind milk, the milk at the e nd of a breastfeeding se ssion that is very high in fat content, has been shown to facilitate weight gain in premature babies (Slusher et al., 2003). Research is identifying that there are fact ors in the hospital setting that influence breastfeeding decisions. The site of care is a strong predictor of c hoice and success with breastfeeding (Powers et al., 2003) In addition, literature is highlighting the important role of health care professionals on the deci sion to initiate and continue breastfeeding (Nyqvist, Sjoden, & Ewald, 1994; Swanson & Power, 2005). Health Care Professionals Can Hinder the Breastfeeding Process in the Neonatal Intensive Care Unit There are many inconsistencies in what parents are educated about and many disparities in what parents are instructed to do by health care professionals, and this can be confusing and frustrating for new mothers (Byrne & Hull, 1996; Nyqvist et al., 1994). One study identified that 48% of mothers reported receiv ing conflicting advice about breastfeeding in the NICU (Jaeger et al., 1997). Mothers rely on health care professiona ls in the NICU to provide accurate, complete, and consistent information about br eastfeeding their high risk infant. Many of

PAGE 26

15 the mothers of preterm infants have not had an opportunity to fully prepare for motherhood before the birth of their ba by. Some may have planned on taking a breastfeeding or parentin g class, but the early arrival dampened these plans. In addition, the NICU is a medically complex environm ent, and parents need guidance on how to function in their role as mother with the environmental constraints. The literature suggests that there is a lack of health care professionals who are tr ained in lactation and breastfeeding with premature babies, and th at many health care professionals have incorrect knowledge and negative beliefs about lactation (Berens, 2001; Pantazi, Jaeger, & Lawson, 1998; Register, Eren, Lowdermil k, Hammond, & Tully, 2000; Spicer, 2001). Yet, it is the health care professionals in the NICU, despite lack of education, who are teaching and instructing these new mothers on breast milk feedings and breastfeeding. Health care professionals can influence breastfeeding behaviors, and their own values and beliefs concerning breastfeeding can have supporting or damaging results on the breastfeeding process (Ekstrom, Matth iesen, Widstrom, & Nissen, 2005). Studies have identified that education and traini ng can affect attitude s and knowledge about breastfeeding (Bernaix, 2000; Siddell, Marinelli, Froman, & Burke, 2003; Swanson & Power, 2005). However, to date, there are no st udies that have invest igated the indirect impact of health care professionalsÂ’ behavi or change with acqui sition of knowledge and attitude change on breastfeed ing outcomes in mothers and infants in the NICU. The American Academy of Pediatrics issu ed a statement in February 2005 that stated that breastfeeding or human milk feedings are recommended for all healthy, premature and high-risk infants for wh om breastfeeding is not specifically contraindicated. It further st ates that health care professi onals should provide complete,

PAGE 27

16 current and accurate information to parents on the benefits and techniques involved with breastfeeding (Gartner et al ., 2005). However, to date there has been no specific, standard set of tools devel oped and utilized to achieve education of health care professionals to enable consistent info rmation dissemination to parents. Treatments to Foster Improved Breastfeeding Rates The United States Surgeon General, Davi d Satcher, identified breastfeeding as a national health priority and released the “Health and Hu man Services, Blueprint for Action on Breastfeeding” in October 2000. In response to this, The World Health Organization in conjunction w ith UNICEF is promoting breastfeeding through the baby friendly hospital initiative. To be designated as “baby friendly,” the hospital must follow the ten steps to successful breastfeeding: Every facility providing maternity services and care for newborn infants should: Have a written breastfeeding policy that is routinely communicat ed to all health care staff. Train all health care staff in skills necessary to implement this policy. Inform all pregnant women about the bene fits and management of breastfeeding. Help mothers initiate breastfeedi ng within half an hour of birth. Show mothers how to breastfee d, and how to maintain lact ation even if they should be separated from their infants. Give newborn infants no food or drink other than breast milk, unless medically indicated. Practice rooming-in (i.e., allowing mother s and infants to remain together) 24 hours a day. Encourage breastfeeding on demand. Give no artificial teats or pacifiers (also called dummies or soothers) to breastfeeding infants.

PAGE 28

17 Foster the establishment of breastfeeding support groups and refer mothers to them on discharge from the hospital or clinic. Source: Protecting, Promoting and Supporting Breastfeeding: The Special Role of Maternity Services, a joint WHO/UNICEF statement published by the World Health Organization These ten steps specifically involve how h ealth care professionals and the hospital system will deal with mother-infant dyads on regular maternity floors within a hospital. Once these ten steps are put into practice, the hospital may apply for designation as a baby friendly hospital. Hospitals that have been through the pro cess of baby friendly designation have demonstrated improvement with breastfeeding ra tes (Philipp, Malone, Cimo, & Merewood, 2003). For example, Bo ston Medical Center was designated as baby friendly in 1999, with increased breastf eeding rates of 58% in 1995 to 86.5% in 1999. Breastfeeding rates were maintained at this high rate from 1999 to 2001. Although the baby friendly designation is sp ecifically for the maternity floors of a hospital system and not designed for the unique needs of the high risk population, there have been positive effects on breastfeeding practices in the NICU following designation (Vannuchi, Monteiro, Rea, A ndrade, & Matsuo, 2004). However, a program specifically designed for the high risk population with its si gnificant barriers to breastfeeding, could have the potential for greater enhancement in breastfeeding rates in the NICU Premature infants are a unique populati on and warrant individualized breastfeeding strategies and interventions (Kavanaugh et al., 1995; Meier, 2001) Many hospitals have implemented pract ices that will educate and promote breastfeeding practices for infants within the NICU. However, only a few have evaluated their programs for effectiveness. The Rush MotherÂ’s Milk Club has proven to be effective in increasing breastf eeding rates (Meier et al., 2004) by enabling free access to

PAGE 29

18 hospital grade breast pumps, by offering lact ation support 24 hours a day, by use of cue based feeding when an infant consumes at l east 50% of feeds orally, and by providing of breastfeeding peer support. Ot her studies have found positive increases in breastfeeding initiation with the introduction of counseling as well as contact with lactation consultants among mothers with low birth weight infant s (Pinelli, Atkinson, & Saigal, 2001; Sisk, Lovelady, Dillard, & Gruber, 2006). A workbook pr ogram introduced at 2 different time periods during the hospital stay was also s hown to have positive effects on breastfeeding with premature infants in one hospital setting (Jang, 2005). Many papers have documented specific prot ocols to instruct mothers and promote breastfeeding (Isaacson, 2006; Premji, Paes, Jacobson, & Chessell, 2002; Spicer, 2001), but no research has been conducted to determine the effectiveness of such recommendations. Although studies have shown increased knowledge acquisition by health care professionals in the NICU followi ng an education plan (Siddell et al., 2003), there are no studies that have investigated the effect of education of health care professionals coupled with protocols and st andard written information for parents on changes in breastfeeding pr actices in the NICU. Of all interventions for breastfeeding with the high risk neonate, the Rush MotherÂ’s Milk Club is probably the most well known. The health care providers in the NICU at Rush University have an increased level of knowledge regarding breastfeeding in the NICU. With this knowledge, they are able to implement advanced strategies, such as putting breast milk in a centr ifuge to modify the fat cont ent to promote weight gain (Meier, 1998). Many studies have been conduc ted to evaluate the effectiveness of the Rush MotherÂ’s Milk Club with positive resu lts (Meier et al., 2004). However, the high

PAGE 30

19 level of education about breastfeeding among health car e professionals at Rush University likely underlies the capability to im plement the more advanced interventions. For many hospitals, basic education on lactati on with high risk infants is lacking. In addition, many interventions that have proven to be effective, includ ing the use of free access to pumps and accessibility to lactation counseling, have associated costs, which many hospitals do not have budgets to support. Need for an Educational Package for He alth Care Professionals and Mothers of Infants in the Neonatal Intensive Care Unit Breastfeeding and human milk feedings ar e possible and benefi cial in the NICU, however, there is sign ificant support and education that must occur to enable success among mothers in the NICU (do Nascimento & Issler, 2004). Education and treatments should be based on research with premature a nd high risk infants, as they have unique needs in the breastfeeding pr ocess (Meier, 2001). The use of developmental care practices can drive the unders tanding of appropriate breastf eeding interventions based on infant readiness cues (Karl, 2004). Although there have been many articles and books written on the subject of breastfeeding with the high ri sk neonate, there was no up to date, comprehensive, evidence based education packet with comp lementary information for both parents and health care professionals available on the market. By understanding each of the challenges to breastfeeding the medically frag ile infant and the specific developmental and nutritional needs of the high risk infant, an education initiative can be developed and then evaluated for efficacy.

PAGE 31

20 Synactive Theory and Breastfeeding Interv entions in the Neonatal Intensive Care Unit While investigating, developing and im plementing appropriate interventions targeted to improve breastfeedi ng rates, it is important to cons ider the vulnerability of the special population in the NICU and the imp act of environmental stressors on this population. One theory that can be used to guide appropriate interv entions in the NICU is the synactive theory, which identifies the process of neurobehavioral maturation of the infant. Breastfeeding interventions should be individualized, based on infant readiness cues and tailored to the responses of th e infant (Blackburn, 1998). Review of the synactive theory and its app lication to breastfeeding should be part of any education initiative for health care prof essionals who serve the vulnera ble infants in the NICU. The synactive theory was developed by Heidelese Als in the early 1980Â’s (Als, 1982). The process of developmental care, rela ted to the synactive theory, is intended to facilitate a well organized, stable infant who may optim ally grow and develop. Developmental care has been instituted in many neonatal intensive care units around the country as a developmental care initiative. It provides a framework for interacting with these fragile infants without jeopardizing health. The synactive theory of development describes the process of neurobehavioral matu ration related to an infantÂ’s internal and external environment. As the infant attempts to interact with the external environment, a dynamic process occurs internally among 5 di fferent subsystems. The dynamic process among the 5 subsystems can explain the be haviors and responses exhibited by the premature infant and can guide appropriate interventions. The 5 distinct, yet interdependent subsystems ar e physiological or autonomic, motor, state organization, attention and inte raction, and the state regulation subsystems.

PAGE 32

21 These subsystems are believed to impact th e functional organization of the infantÂ’s system in an ordered fashion. The subsystems are not hierarchical, but they are believed to be ordered and interdepe ndent (Als, 1982; Als, 1994). The physiological subsystem is considered the core of the system. It is the foundation for which all of the other systems ga in stability. This physiological subsystem allows the infant to have control over autonomic functions such as voiding, breathing, maintaining steady vital signs and processing nutrition. The motor system provides control over movement, muscle tone, and postu re. The state subsystem gives the infant control over his/her level of consciousness. It enables the infant to move through identifiable states and move smoothl y from one state to another. The attentional/interactive subsystem enables control over functional responses to stimulation in the environment and governs the ability to interact. The stat e regulation subsystem gives the infant the ability to balance environmental stresso rs and recover by modulating all the other systems (Als, et al, 1982). While the autonomic subsystem serves as the foundation of the system, the state regulation s ubsystem serves as the gate keeper and is achieved with increasing maturity. A cone shaped diagram is used to represent the complex development of the infant as it relates to the five subsystems (Als, 1982). The cone has its tip at the bottom with the funnel going upward. The five subsystems can be viewed at the top of the open cone. At the smallest center is the physiological subs ystem with the remaining (motor, state, attentional/interaction) forming layers out side the center, much like an onion. The youngest fetus is represented at the bottom of the cone and has with it only components (not yet a fully developed system) from the autonomic subsystem. This indicates that

PAGE 33

22 infants who are born early are unable to inte grate the higher order systems. Stressors within the system interfere w ith the physiologic capabilities of the infant. Subsequently, early premature infants are incapable of a ny interaction and need all their energy to maintain homeostasis of the system to sustai n life. There is also instability in the physiologic system, which is why prematur e infants frequently have medical or physiological problems when born early and have to contend with the stressful environment. With the earliest fetus at the bottom of the cone, increasing gestational age is associated with increasing maturity spreading out to the other layers of the system. With increasing gestational age and thus maturity, the infant may extend its control out to the next level, the motor subsystem. The infa nt may demonstrate improved muscle tone and postural control. This concept parallels th e literature, which demonstrates improvement in muscle tone and reflex development with increasing gestational age (Allen & Capute, 1990). Further maturity may extend the infant Â’s control out to the state subsystem enabling the infant to demonstrate some aw ake periods and to smoothly transition from one state to another. As maturity continue s, the infant may be able to achieve some attention and interaction with caregivers and the environment. Lastly, as the infant approaches term and achieves more maturity, he/she will be able to tolerate stressors, cope with them, reorganize and continue interaction without being knocked down to functioning at the lower subsystems. The term synaction refers to the relati onship between all the subsystems and how instability in one system has the potential to affect all the other subsystems and thus the integrity of the childÂ’s health and well-being (Als, 1982). On the right side of the cone

PAGE 34

23 are gestational ages that reflect the increasi ng maturity of the system (dependent on the subsystems) with increasing gestational age. On the left side of the c one are influences of the environment on the maturation of the syst em, with a break in the intrauterine and extrauterine environment before term to i ndicate the premature birt h, thus representing the role of environmental stresses before full maturity occurs. With earlier birth and more stressors from the environment, there will be a resultant decrease in neurobehavioral maturity. When an infant is stressed from the environment, he/she may initially demonstrate stress reactions based on the predominant leve l of neurobehavioral maturation. If he/she is primarily functioning in th e physiological state, he/she may demonstrate bradycardic events, oxygen desaturations, hicc ups, stooling, or spitting up. If he/she is primarily in the motor state, he/she may demonstrate gr imacing, arching, saluting, finger splaying, or sitting on air. If he/she is in the state s ubsystem he may shut down or move to a light sleep state. If he/she is in the attenti onal/interactive subsystem, he/she may avoid interaction by turning away. The infant has th e capability to re-ach ieve organization with time outs or specific strategies designed to he lp him cope. Interventions designed to help infants cope with stressors include pr oviding boundaries, swaddling, positioning in flexion, bringing hands to mouth, minimizi ng environmental stimulation, non nutritive sucking, and enabling grasping. Infants addi tionally will demonstrate approach signals such as smiling, mouthing, ooh face, cooing, quiet and alert state, and soft and relaxed facial expressions when they are ready fo r interaction (Hussey-Gardner, 1996). Once reorganized and demonstrating approach signals the stressor may be reintroduced slowly.

PAGE 35

24 The synactive theory defines the subsys tem along with stress and coping signs consistent with each of the subsystems to enable caregivers to identify and respond to behaviors appropriately. Wh en stress signs are recogni zed, the caregiver can then withdraw the stressor th at contributed or help facilitate the infant to cope. Once a time out is given and the infant reorganizes, the treatment or interacti on can continue. This “give and take” enables the infant to func tion optimally within the environment and allows him/her to continue to benefit from interaction and stimulation, including breastfeeding, as he/she tolerate s. The synactive theory proposes an approach for each individual child that is adapted to fit the needs of that infant. It promotes infant development to occur as normally as possi ble, despite medical complications and immaturity brought on by an early birth. Infants born at earlier gestational ages and with decreased neurobehavioral maturation are not capable of handling envir onmental stressors typically experienced by newborns. When interventions for these neurobehaviorally immature infants are done without respect for readiness cues the infant is at risk for regressing to one of the more primitive states, putting them at risk of developmental and medical sequelae. Breastfeeding is an environmental stressor. Without observing infant readiness cues and introducing breastfeeding at the appropriate ti me individualized for each infant, optimal responses to the environment as well as op timal neurological ma turation are delayed. Therefore, breastfeeding cannot be introduced at a prescribed time or introduced in the same way that it would be for a full te rm infant, but it must be based on the neurobehavioral maturation of the infant and advanced according to stress and readiness cues. These concepts need to be in any educational program fo r both health care

PAGE 36

25 professionals and for mothers of infants in the NICU to help them implement the best breast milk feeding program for these infants. While the synactive theory defines the appropriate time and way to introduce interventions, it also assists with underst anding that many extremely low birth weight infants and low birth weight infants are t oo neurologically immature and fragile to engage in any breastfeeding. Subsequently, mothers of infants in the NICU need equipment that will enable them to achieve and maintain a milk supply, in the absence of infant suckling at the breast, until the infant is appropriate for nut ritional brea stfeeding. Hospital grade breast pumps that will enable long term milk expression are necessary for mothers with infants in the NICU to mainta in adequate milk supplies while they are waiting for their babies to become medica lly and developmentally stable enough to engage in feeding at the breast. Understanding the synactive theory and implementing develo pmentally supportive care can instruct and guide interventions and NICU practices as they relate to breastfeeding the premature infant. One cas e study in the literat ure highlighted the significant benefits of a developmentally s upportive plan on the breastfeeding process in a premature infant (Nyqvist et al., 1996) The synactive theo ry should guide the development of any educational module and in service for health care professionals who serve infants in the NICU. Theory Governing the Behavior of Health Care Professionals Studies have identified that there is a lack of education about la ctation and lack of consistent support and instruction about breas tfeeding among health care professionals in the NICU (Ekstrom, Widstrom, & Nissen, 2005; Pantazi et al., 1998) Having educated health care professionals will not necessarily impact breastfeeding practices. It is how

PAGE 37

26 those health care professiona ls respond and utilize that education to execute new interventions that will foster change a nd subsequent improvement in human milk feedings. The behavior of health care prof essionals in the NICU needs to change to support the breastfeed ing process. Studies have shown that behavior change is much more successful when interventions are matc hed to the stage of readiness to change (Prochaska, Prochaska, & Levesque, 2001). The transtheoretical model (TTM) provides a description of how the indi vidualÂ’s state of readiness to change translates into behavioral change. The premise of the TTM is that there are several stages associated with behavior change. Individuals go through these stages on their way to making a change. They may start anywhere along the con tinuum of the 5 stages and they may move forward or backward or skip stages, but there is some progression through the stages on their way to behavior change. The TTM has been used to describe many health behavior changes, such as use of sunscreen, use of condoms self examination breast checks, smoking cessation and initiating an exercise plan. Appr opriate interventions can be tailored to the stage of readiness to change. The five stages of the TTM are prec ontemplation, contemplation, preparation, action and maintenance. Each stage identifi es if the individual has an intention of changing behavior and identifies how significan t the intention to ch ange behavior is (Prochaska & DiClemente, 1983). Someone in the precontemplation stage does not intend to take action within 6 months. A pe rson in the contemplation stage intends to take action within the next 6 months. Someone in the preparation phase intends to take action in the next 30 days. The action stag e refers to persons who have made obvious

PAGE 38

27 changes less than 6 months ago. The maintena nce stage refers to individuals who have made significant changes more than 6 months ago. Interventions that are impl emented to enable behavior change should be conducted in such a way that the intervention matches the stage an individual is in, or should be tailored to how ready the person is for change Ten fundamental proc esses that can affect change have been identified along with interven tions that can be matched to the stage of readiness to change (Table 1-1) (Prochaska et al., 2001). Certain interventions will only be effective if they are appropriate for the st age of readiness for behavioral change of the individual. For example if someone is in the precontemplation stage, interventions should be focused on educational ini tiatives and strategies to pr omote reflection about how the change may impact the personÂ’s situation and how it will benefit others, while interventions for the person in the contemplative stage s hould be about supporting and motivating the person to actually initiate th e intended behavioral change. These latter interventions provided to the precomtemplate r would not be effective because the person has not yet formed the conviction that change is needed or desire d and could actually create significant resistance and prevent be havioral change from succeeding (Prochaska et al., 2001). By enabling stage matched inte rventions, education and strategies can be implemented to foster movement across stag es to enable successful change. The TTM has been used to address beha vior change within organizations by targeting employees. By providing stage matc hed interventions, all employees can be given opportunities to participate in the change process. Although stage matched interventions have been show n to facilitate movement toward action, not all employees may achieve action. Change at the organizati onal level should include interventions that

PAGE 39

28 are stage matched to each stag e of change to give all employees the opportunity for participation in the ch ange initiative (Prochas ka et al., 2001). Table 1-1. Stages of change in which part icular processes of change are emphasized Stages Precontemplation Contemplation Preparation Action Maintenance Process Conscious Raising, Dramatic Relief, Environmental Reevaluation Self Reevaluation Self Liberation ContingencyManagement, Helping Relationship, Counter Conditioning, Stimulus Control To foster behavior change regarding support and information dissemination about breastfeeding within the NICU, the TTM can be utilized to structure an intervention program. By matching breastfeed ing interventions to stages of readiness to change, all health care professionals can ha ve the opportunity to participat e in the change process. By introducing interventions that can target health care professionals in each stage of readiness to change, a meaningful intervention plan can be implemented to foster change at the organizational level. A breastfeeding intervention for the NICU with 4 parts could theoretically target individuals in all of the stages of readiness to change. For those health care professionals in the pre-contemplation stage, there is no intention of making a behavior change. Health care profes sionals in the contemplation stage intend to make a change within the next 6 months. Both of these stages describe

PAGE 40

29 individuals who are not yet rea dy for action, and interventions fo r those in either of these stages would be the same. Interventions for those in these stages should be two-fold. One intervention, education, would be aimed di rectly at the health care professional. With education, conscious raising can be foster ed. With it, it is hope d that health care professionals will have the resources needed to become aware of the need for breastfeeding support in premature infants and will start to see solutions to the problem. However, because those in the pre-contempl ation stage have no intention of making a behavior change, there would be no motivation to participate in an educational initiative. Likewise, those in the contem plation stage also may need incentives to push them to participate. Therefore, incentives on annual review, food, prizes and continuing education units could serve as motivati on for participation among those who lack motivation to attend without some perceived pe rsonal benefit. With the participation in the educational initiat ive, they would be exposed to content of the education that highlights the great benefits of breast milk and the need for support and education among mothers. With this increased awareness of the problem and possible solutions, behavior change could be fostered. The other intervention shoul d be aimed at achieving some level of support and education for the mothers. Materials that provide consistent a nd thorough information could be issued to all new parents in the NICU to ensure that all mothers receive information about initiating and sustaining br east milk feedings in the NICU. Although this intervention would serve to enable educ ation of mothers, it also may serve as a conscious raising effort for the health care professional, who may be asked for guidance and support by the mother on information contai ned in the educational materials. The

PAGE 41

30 health care professional may then better understand the problem and the need for behavior change to facilitate succ ess with the breastfeeding process. Individuals in the preparati on phase intend to make a change in the next 30 days. These individuals are ready for action orie nted interventions. Therefore, clinical pathways or protocols could fo ster change in how they deal with breastfeeding mothers. Protocols or pathways, which become a part of required paperwor k, could theoretically facilitate professionals to make a commit ment to change by giving protocols that necessitate action. Those in the action and maintenance stag es have already made changes. The motherÂ’s positive experiences could serve as motivation from the environment, and there could be other motivators for continued compliance from within the organizational structure, such as acknowledgement on the a nnual review and identification as one who has expertise in breastfeeding with high risk infants. Those in the action and preparation phase also may take an active part in motiv ating others and facilitating more positive change related to breastfeeding in terventions in the NICU. Thus based on the TTM, an educational initiative that includes educational materials to parents, opportunities for e ducation with incenti ves for health care professionals, and protocols or pathways of car e could be effective in facilitating change in breastfeeding practices in the NICU, and each is theoretically matched to all of the stages of readiness to change. Although the primary focus of the interven tion for this research is health care professional behavior change, in terventions structured to enab le change in breastfeeding practices also must target be havior change in the mother s. Theoretically, developed

PAGE 42

31 interventions for the health ca re professionals can also be matched to mothers in each stage of readiness to change. An educa tional booklet for new mothers could enable conscious raising for mothers in the prec ontemplation and contemplation stages. A breast pump loaner closet as well as milk expression guidelines and a breast milk log (that could be included in the educational booklet) could be appropria te interventions for those in the preparation and act ion stages. Appropriate interv entions for those mothers in the action and maintenance stages would include concepts such as the first feeding being at the breast and enabling breastfeeding while in the hospital to support continued breast milk feedings in the presence of the d ecision to initiate breast milk feedings. Interventions targeted at health care pr ofessional change can be structured to move mothers to decide to br eastfeed and help them maintain that behavior once they start. Theoretically, interv entions including a breast pump loaner closet, an educational booklet for new mothers, a breastfeeding pathwa y, and an educational initiative for health care professionals can support behavior change in two different gr oups, the health care professionals as well as the mothers. The health care professiona l group functions as a primary support for the mothers to initiate and sustain breastfeeding. Transtheoretical Model of Behavior Change and Methods of Education Equally important with providing stage matched interventions for health care professionals is consideration of what mode of learning to uti lize for the conscious raising strategy. Self learning modules can be considered easy to implement and enables staff to participate in the learning initiative s at their own pace, enables them to take modules home if work responsibilities preven t participation during working hours, and is rather inexpensive when compared to othe r modes. The literature reflects good success with self-directed learning modules, with good performance on post tests and learning

PAGE 43

32 retention (Coleman, Dracup, & Moser, 1991). When compared with lecture-discussion formats, self learning methods were compar able in achieving the educational objectives being targeted within the nursi ng field (Nikolajski, 1992). Computer based training is another method of education gaining increased acceptance and use in the last several years w ith the increasing capab ilities of technology. Harrington and Walker (2004) discovered that, although both groups significantly improved their post test scores a group of individuals who engaged in computer based training did significantly better th an individuals in an instructor led course on fire safety (Harrington & Walker, 2004). Research and e xperience are beginning to define computer based learning as a viable option for e ducational purposes. However, the access to technology and to the people who format and de sign the computer systems is a significant barrier to widespread use of such learning practices today. Not all individuals have the same learning style. While some may prefer self paced methods, others may be more motivated a nd embraced in a face to face lecture and discussion with peers. Goldrick, Gruende mann, and Larson (1993) found that 64% of nurses in a pediatric intensive care unit had an abstract lear ning style and preferred self learning modules. However, there remained 36% who preferred more traditional methods (Goldrick, Gruendemann, & Larson, 1993). Self directed learning, through an educatio nal module, is an effective form of educating health care professionals. Howeve r, not all individuals possess the learning style necessary for successful completion of self learning modules and prefer lecturediscussion formats. By providing both forms of educational opportunities, more health

PAGE 44

33 care professionals could be encouraged to pa rticipate in an educational initiative. Computer based training could also be effec tive if access to technology can be achieved. Summary and Research Questions The provision of breast milk has important be nefits to infants, especially those in the NICU who are less healthy and less mature at birth than full term infants. Yet significant, but not insurmount able, barriers to breast milk feedings and breastfeeding exist in the NICU. Health care professi onals are a powerful influence (Swanson & Power, 2005). It is proposed that change fr om the health care professionals can enable mothers and infants to overcome many of the barriers to breastfeeding, and subsequently, positive changes in breastfeeding practices can occur in the NICU. Therefore, in this study, the effect of a 4-pronged educati on and support interv ention to promote breastfeeding in the NICU, based on the synac tive theory and transtheoretical model of behavior change, will be explored. The four parts of the proposed intervention are a breast pump loaner closet, a breastfeeding pa thway on the individualized care plan, an educational booklet for mothers, and an educational initiativ e for health care professionals who work with infants and moth ers in the NICU. The research questions are Is there a significant difference in breast milk feeding initiation in very low birth weight (VLBW) infants admitted to the NICU before and after implementation of the intervention plan? Is there a significant difference in the rate of breastfeeding in the hospital among women with VLBW infants hospitalized in the NICU before and after the implementation of the intervention plan? Is there a significant difference in breas t milk feedings at discharge in VLBW infants admitted to the NICU before and after the implementation of the intervention plan?

PAGE 45

34 Is there a significant difference in the pr oportion of the hospital stay that breast milk is provided in VLBW infants admitted to the NICU before and after the implementation of the intervention plan?

PAGE 46

35 CHAPTER 2 METHODOLOGY The purpose of this study was to test the efficacy of a 4-part intervention on improving breastfeeding practices in the ne onatal intensive care unit (NICU). The overall goal was to attempt to develop an effective intervention to assist mother-infant dyads in the complex NICU environment. Participants Very low birth weight (VLBW) infants (< 1500 grams) were included in the study if they were 1) admitted to the Level II or II I nursery at Shands Hospital during the study periods and 2) had a length of stay greater than or equal to 7 days, 3) were admitted to the NICU within the first 3 days of life, 4) were hospitalized less than 4 months, 5) achieved full gastric feeds during their stay, and 6) had a hospital stay that di d not cross over from the pre-intervention group time period into th e education initiative time period. Very low birth weight infants were exclude d from the study if they 1) ha d a length of stay less than 7 days, 2) were transferred to Shands Hosp ital after the third day of life, 3) were hospitalized greater than 4 m onths, 4) did not achieve full gastric feeds during the hospital stay, 5) had a hospital stay that crossed over from the pre-intervention group time period to the education plan time fram e, or 6) had conditions that would make breastfeeding contraindicated as established by the physician. Power indicates the probability of rej ecting the null hypothesis, if a condition exists. With a power of 80%, which is freque ntly used in the literature, there is a 20% chance of failing to reject the null hy pothesis when it should be rejected.

PAGE 47

36 Prior to conducting the study, a power an alysis was conducted to determine the appropriate sample size. The mean and st andard deviations of breast milk feeding initiation were unavailable from other studies to compute an effect size and subsequently a sample size. Therefore CohenÂ’s Criteria wa s utilized to make sample size estimations (Cohen, 1988). According to Cohen, a .2 standard deviation change is a small effect, a .5 is a medium effect, and a .8 is a large eff ect. For the purposes of this study, a medium size effect was selected. By using CohenÂ’ s criteria and determining the sample size necessary with a power of 80%, alpha of .05, and looking for a medium size effect of .5, CohenÂ’s Criteria indicated a needed sample size of 82 per group. Th erefore the research plan consisted of intent to conduct quota sa mpling with participants enrolled from the beginning study dates for both the pre-interv ention and post-interv ention groups until 82 were achieved in each group. The planned pre-intervention group consisted of all very VLBW infants admitted to Shands Hospital NICU from April 15, 2004 forward until 82 participants were enrolled in the study. The intervention started on March 1, 2005 with conclusion of the educational initiative on April 15, 2005. The planned pos t-intervention group cons isted of all very low birth weight infants admitted to Shands Hospital NICU after implementation of the intervention plan, from April 15, 2005 until 82 were admitted into the study. Data from participants were collected from the same time of year to account for seasonal confounds. Research Interventions Intervention 1: Breast Pump Loaner Closet Intervention 1 consisted of the developmen t of a breast pump loaner closet for use by mothers with infants hospitalized in the NICU Hospital grade breast pumps could be checked out by mothers who had infants in the NI CU to enable them to express their milk

PAGE 48

37 the recommended 8 to 12 times per 24 hour period. This would enable a supply of expressed breast milk for initiation of breast m ilk feedings in the infant and would enable the mothers to establish and maintain a m ilk supply until the infant was able to go directly to breast. Intervention 2: Health Prof essional Education Initiative Although there are many different recomm endations and published articles about breast milk feedings and breas tfeeding in the NICU, there was no up to date, available educational plan that could be utilized for staff education. Therefore, an education initiative encompassing key areas of educati on on breastfeeding special care babies was developed to educate as many of existing staff in the NICU as possible. The initiative consisted of education to staff on breastfeeding to enable health care providers to have the education and tools to support mothers in the breastfeeding process. The education was offered through completion of a self st udy educational module on breastfeeding in the NICU or through attendan ce at an inservice on breas tfeeding in the NICU. Education topics contained in the self st udy module and discusse d in the inservice included the benefits of br eastfeeding, the barriers to breastfeeding, the physiology of lactation, use of breast pumps, pre feeding interventions based on the synactive theory and breastfeeding interventions that acknowledge the readine ss of the infant. All the information contained in the module was base d on an extensive literature review to represent evidence based practice and was desi gned to foster success with breastfeeding in the high risk neonate population while acknowledging their un ique needs. The educational module was reviewed by two individuals considered to be experts in the area of breastfeeding for establishing validity of information provided. Minor adjustments were made to the education plan based on the expert feedback. Refer to the outline of the

PAGE 49

38 educational module, appendix A, or contact the author for further details. The successful completion of the health prof essional education was defined as completion of the module or attendance at one of the inservices and a passing score of at le ast 80% on a post test that was identical for either form of the education. Intervention 3: Breastfeeding Guideline Each medical chart contains an individua lized care plan (ICP) for documentation by nurses. This ICP was modified to al so have a pathway of care for providing breastfeeding support to new mothers (a ppendix B). This ICP necessitated documentation of education and support by nurse s at critical times in the breastfeeding process. The guideline called attention to and necessitated documentation on specific key points that were identified in the literature to be predic tive of success: achieving and maintaining a milk supply, timely pumping, skin to skin contact, and first feeding being at the breast. It also included areas to check off, date, and sign at the following critical times in the breastfeeding process: within 6 hou rs of delivery, issue a nd instruct in proper pumping and breast milk storage techniques; within 24 hours, ensure proper pumping and storage technique; on day 3 to 5, ensure that the milk has come in and trouble shoot any problems; weekly, foster continued pumping a nd skin to skin care; first oral feeding, ensure that it is a breastfeeding session; 10 days, monitor milk supply and make referrals as appropriate. Intervention 4: Educational Pamphlet fo r New Mothers of Neonatal Intensive Care Unit Infants An educational pamphlet, “A Mother’s Gift”, for mothers who had an infant admitted to the NICU was developed. The outline of the educ ational booklet (see appendix C) addressed the following key poi nts: benefits of breastfeeding, how to

PAGE 50

39 express and store human milk, pre-breastfeed ing strategies, and cu e based breastfeeding interventions. The back of this pamphlet also included a place for mothers to document breast milk production to facilitate communi cation with nurses about their milk supply. This educational pamphlet was developed to en sure that all mothers received a standard set of educational points dur ing their infantÂ’s hospitaliza tion, and that the information contained in it was consistent with the e ducation that the health care professionals received. Intervention Plan Modification The original intervention plan consisted of 4 parts: a breast pump loaner closet, an education module and inservic ing, changes to the individu alized care plan and an educational booklet. Prior to the initiation date of March 1, 2005, it was learned that external funding for the breast pump loaner clos et could not be obtai ned. Therefore, this prong of the intervention had to be deleted from the intervention program. The study was then conducted with the followi ng being the intervention/educ ation plan: the education initiative, the motherÂ’s edu cational booklet, and the breastf eeding pathway addition to the individualized care plan. Design This study was a quasi experimental, ma tched through cohort controls, design (Shadish, Cook, & Campbell, 2001), investigat ing indirect changes in breastfeeding practices following a 3-part breastfeeding in tervention in the NICU. Through this design the pre-intervention group consisted of a group of VLBW infants hospitalized in the NICU before the implementation of the interv ention plan. This group was then compared to the post-intervention group, which consisted of a group of VLBW infants who were hospitalized in the NICU after the im plementation of the intervention plan.

PAGE 51

40 The independent variable was the implemen tation of the intervention as described above. Dependent variables included 1) breast milk feeding initiati on rate (was breast milk ever consumed/breast milk feeds initiated? (yes/no)), 2) br eastfeeding rate (number of times the infant was put to the mother’s breast after 30 weeks ge station divided by the number of days hospitalized after 30 weeks gest ation), 3) breast milk feeding at discharge rate (did the infant continue to have breast m ilk feedings at discharge? (yes/no)), and 4) the proportion of the hospital stay that breast milk was provi ded (total number of days into the hospitalization that breast milk wa s provided divided by the length of stay). Procedures The educational intervention was imple mented March 1, 2005 to April 15, 2005 with opportunities for health care professionals to complete the self study educational module or participate in an in service. “A Mother’s Gift ”, the educational booklet for mothers was issued to all new mothers with infants admitted to the NICU on or after March 1, 2005. Last, the modified individualized care plan was used in the medical chart on all new admissions after March 1, 2005. To promote completion of the educational in itiative, incentives were given to those who participated in the br eastfeeding education initiativ e by way of food, prizes, continuing education credits and docume ntation on the annual review of their performance. Following the six week educational initiative, completion of the self study educational module on breastfeeding in the NICU became part of the orientation process to enable the same education for those sta ff who were not employed at Shands Hospital during the six week educa tional initiative. The educational opportunities during the initiative date s included a self study module and/or inservices. A breastfeeding module was available for health professionals

PAGE 52

41 to check out and complete at home or wor k. Food and prize incentives as well as 2 continuing education credits were awarded for those who completed the educational module. For those who preferred lectur e-discussion formats for learning, 1 hour inservices were offered at least one time per week throughout the education initiative period. Those who attended the inservices were educated on the same information contained in the education m odule, however in a condensed amount of time. Therefore one continuing education credit, in addition to food and prizes, were awarded to those who attended an inservice duri ng the initiative dates. The booklet entitled “A Mother’s Gift” was issued to mothers with infants admitted into the NICU after March 1, 2005. There was a central location at the reception desk where nurses who had new admissions could acce ss and issue them to mothers. Nurses were instructed to issue these booklets during staff meetings through the monthly bulletin and in the breastfeed ing inservice that occurred over the six week period. The modified individualized care plan w ith the breastfeeding pathway replaced the old ICPs and were placed in the chart as ro utine paperwork as of March 1, 2005. Nursing staff were instructed to use it by way of a m onthly written bulletin. It was also discussed in staff meetings and further reminders were given to document on it during the breastfeeding inservices that occurred over the six week period. Program Evaluation The desired impact of this program wa s increased breastfeeding in the NICU. However, the intervention strategies used in the current study can only be effective if they are implemented. The full implemen tation of the 3-pronged intervention was evaluated in four ways. All the educational tools (the educational module, the inservice, the educational booklet for he mothers and the m odified ICP) stressed that the first oral

PAGE 53

42 feeding should be at the breast. Therefore the primary outcome measure to determine implementation of the intervention was whether the first oral feeding was at the breast. Second, attempts were made to track the percen tage of mothers of infants newly admitted to the NICU to whom educational booklets we re issued to determine if, in fact, most mothers were being issued th is educational booklet. Last, weekly communications with the nursing administrator indicated the degree of compliance with educational key points based on her monthly experiences as a bedside nurse, in which she worked directly with mothers and their babies in the NICU. Data Collection Participants were recruited by way of a data base containing all admissions and discharges from the NICU during the two different time periods. For each infant admitted to the hospital during the applicable time periods, an extensive retrospective chart review was conducted. Each identified chart was fi rst investigated to ensure that the infant did not have any exclus ion criteria. Given that incl usion criteria were met, the dependent variables as well as demographics were collected and recorded on a laptop computer. Inter-rater reliability was determined in 3 different participants to ensure that accurate variables were collected from the charts. This occurred by having another researcher collect data on the same particip ants following data coll ection by the principal investigator and comparing if the variable s collected by the two different researchers were in agreement. Data Analysis Retrospectively, charts were reviewed and data was analyzed for significant differences in the proportion of mothers who in itiated breast milk feedings, the number of

PAGE 54

43 times per day the mother breastfed after 30 weeks gestation, the proportion of mothers who provided breast milk at discharge, and th e proportion of the hospital stay that breast milk feedings occurred. Table 2-1 su mmarizes the dependent variables and null hypotheses. Table 2-1. Hypothesis testing acco rding to dependent variable Group Breast milk feedings initiated (yes, no) The Number of Time s the Mother Breastfed Per Day After 30 EGA Breast milk provided at discharge (yes, no) Proportion of the hospital stay that breast milk was provided -------------PreIntervention Group --------------A1 ---------------B1 --------------C1 --------------------------------D1 -------------Post – Intervention Group --------------A2 ---------------B2 --------------C2 --------------------------------D2 Hypotheses The following hypotheses and data an alysis plan guided this study. The rate of breast milk feeding in itiation will be higher in the post intervention group (A2>A1). Data Analysis Plan : A Pearson’s Chi Square was us ed to test two proportions for significant differences between the two groups. The number of times per day that an infant is breastfed after 30 weeks gestation will be higher in the post-intervention group (B2>B1). Data Analysis Plan: A one-way analysis of va riance (ANOVA) was not possible secondary to a violation of the assumption of normality. Therefore, the nonparametric Mann Whitney was used to test for differences between the 2 groups.

PAGE 55

44 The rate of breast milk feedings at discharge will be higher in the post intervention group (C2>C1). Data Analysis Plan : A PearsonÂ’s Chi Square was us ed to test two proportions for significant differences between the two groups. The proportion of the hospital stay that breast milk was provided will be higher in the post intervention group (D2>D1). Data Analysis Plan : A one-way analysis of va riance (ANOVA) was not possible secondary to a violation of the assumption of normality. Therefore, the nonparametric Mann Whitney was used to test for differences between the 2 groups. In this study, the pre-intervention group and post-intervention group were compared for significant differences in four different variables. For the purposes of this study, an alpha level of .05 was chosen, which is standard throughout the literature. Adjusting the Alpha Level There are no statistical procedures that can simultaneously test multiple outcomes, some of which are continuous and some of which are dichotomous. Therefore, the significance levels of the i ndividual tests were adjust ed by the ranked Bonferroni adjustment. There has been criticism of th e standard Bonferroni adjustment being too conservative and that, in theory, if many test s were run, the level of significance would be so low that no differences c ould be detected. The ranked Bonferroni adjustment was preferred over a standard B onferroni adjustment to enable maximum power in initial comparisons, by adjusting the alpha level w ith each additional comparison to prevent inflation of the type I error rate. This would help to prevent the researcher from rejecting the null hypothesis inappropr iately while minimizing inappropriate stringent p value

PAGE 56

45 constraints (Benjamini & Hochberg, 1995). Fo r this study, the questions were ranked in order of importance. The first question, whet her or not there was a difference in breast milk feeding initiation, was tested at an al pha of .05. The second question, whether or not there was a difference in number of tim es breastfed after 30 weeks gestation, was tested at an alpha of .025 (.05/2). The th ird question was tested at an alpha of .017 (.05/3). The fourth question was tested at an alpha of .013. Each statistical analysis was conducted as a one sided test as it was assumed that trends would be toward increased rates of br eastfeeding with the interventions that were implemented.

PAGE 57

46 CHAPTER 3 RESULTS Intervention Implementation One hour inservices were conducted 1 to 3 times per week for a total of 10 inservices during the intervention pe riod of March 1, 2005 through April 15, 2005. General attendance at each inse rvice was low with approximately 2 to 5 participants at each one. Self-study modules were also availa ble for check out during this time. Overall response to complete the self-study modules was also low in the month of March. Therefore, in April, the researcher started directly asking health care professionals to complete the modules and offered food prizes for those who did. It appeared that directly requesting participati on was beneficial in promoti ng participation by the health care professionals. There were 11 health care professionals who completed the self study education modules from March 1 through Ma rch 31, 2006, and there were 45 health care professionals who completed the educatio nal module from April 1 through April 15, 2006. The total number of health care professionals who partic ipated in the educational initiative was 88, which was 63% of health care professionals working in the neonatal intensive care unit (NICU). The total number of nurses who participat ed in at least one of the methods of education was 75, which was 77% of all nurses who care for infants in the NICU. There were 3 rehabilitation ther apists (100%), 1 nurse practitioner (9%), 2 neonatologists (20%), 2 social workers (100%), 1 respiratory therapist (10%), and 5 other

PAGE 58

47 health professionals (83%). All those who participated in the education achieved a passing score of 80% on a post test. Nursing managers reported variable levels of compliance with the new strategies presented in the educational in itiative, contained in the educational booklet for mothers, and on the modified individualized care plan (ICP). Starting on March 1, 2005 the nurses initiated use of a new, revised individualiz ed care plan (ICP) for documentation. The revised ICP was supposed to replace the old one. However, in mid April, it was realized that some old stores of the previous ICP, that did not include th e breastfeeding pathway, had been pulled from the shelf and were being utilized. Acco rding to the nursing manager, this problem was resolved with full use of the new ICP by May 1, 2005. Although all nurses were expected to fo llow the established guidelines on the breastfeeding pathway, during da ta collection it was observed that the new ICP was not utilized fully. One example of the lack of full implementation of the new pathway concerned whether the first oral feeding was at the breast. On the breastfeeding pathway, all mothers should have been encouraged to have the first feeding at the breast with documentation accordingly or documentation stating why care deviated from the pathway. However, the first feeding being at the breast occurred in only 25% of mothers in the post intervention group, and with full implementation it should have approached 100%. Although it is possible that mothers were encouraged, but declined to participate in the first feeding at breast, it is more like ly that there was lack of full compliance with the educational key points and the modi fied individualized care plan. Starting on March 1, 2005 “A Mother’s Gift ”, the educational booklet for mothers, was available to be issued to new mothers wi th infants in the NICU. Initial “Mother’s

PAGE 59

48 Gift” educational booklets were tracked to be able to determine if the number of booklets that were issued matched the number of admissions. Not all mothers were given the pamphlet over the first few weeks of the inte rvention. There were reports of running out of the booklets and not being able to find them. Multiple copies of these were distributed during and after the educational initiative, bu t they became impossible to track as they were frequently misplaced, redirected to the maternity floor rather than remaining in a central location in the NICU, and others outside of the research initiative made copies of the booklet for distribution. Nursing managers reported variable levels of compliance with the new strategies presented in the educational initiative, contained in the educational booklet for mothers and on the modified ICP. One nursing administ rator, who would func tion in the role of bedside nurse approximately once a month a nd would work directly with mothers and their babies during this time, reported certa in personnel to be imp lementing strategies while others, even those who par ticipated in the educational in itiative, to be consistently ignoring the pathway of care contained in th e medical chart. The nursing administrator’s occasional role of bedside nurse revealed th at there were mothers who never received the educational materials and that ICPs in the medical chart had inadequate documentation. The Sample The pre-intervention sample data was obtained before the education plan implementation using quota sampling from th e beginning study date of April 15, 2004. The post-intervention group was obtained afte r the intervention pe riod implementation from April 15, 2005 onward. Eighty one part icipants were obta ined for the preintervention group from April 15, 2004 through discharges on December 7, 2004. Data collection in the preintervention group was stopped at 81, because the subsequent 2

PAGE 60

49 admissions crossed into the treat ment period. Data from only fift y four participants in the post-intervention group was collected fr om April 15, 2005 through discharges on November 29, 2005. There were no discharges from the NICU of participants who met inclusion criteria from November 29, 2005 to December 7, 2005. This sample included all admissions of VLBW infants admitted during the preintervention study dates except for 17 infant s who did not meet inclusion criteria. Among the 17 infants who were excluded, 13 of them were extremely low birth weight and expired shortly after birt h, thus never achieved full gast ric feeds. Two of them did not achieve full gastric feeds before being tran sferred to another hospi tal, and 2 of them had a length of stay that ex tended into the treatment peri od. The pre-intervention group consisted of 83% of all admissions of VLBW infants admitted to the NICU at Shands during the study dates. The sample include d all admissions of VLBW infants admitted during the post-intervention study dates exce pt for 11 infants. Among those 11 infants who were excluded were 9 infants who neve r achieved full gastric feeds and expired shortly after birth and 2 who had genetic disorders that made eventual oral feeding contraindicated. The post-int ervention group also consisted of 83% of all admissions of VLBW infants admitted to the NICU at Shands during the study dates. The data collection period was not exte nded in order to capture the remaining 17 participants for two reasons: the first is that a long period of tim e had passed since the intervention plan, and new interv entions were scheduled to be implemented in the NICU. These would have introduced significant additional conf ounds into the study. Secondly, a new power analysis based on actual effect sizes of this partial sample indicated a need for data from an additional 124 participants in the post-intervention group and 95 in the pre-

PAGE 61

50 intervention group to achieve 80% power b ecause of the already high breast milk initiation rate (74.1%). Continuing data co llection to enroll 82 in each group based on the original research plan would have incr eased power from 38.6% to 45.6%, an increase that was considered to not be feasible gi ven the potential confounde rs listed above, or likely to change the statistical outcomes. Inter-Rater Agreement To ensure accurate documentation of the re search variables, inter-rater agreement was tested on the chart review procedures. Another researcher c onducted data collection on 3 charts that the principal investigator had already coll ected data from. There was 100% agreement in 2 out of 3 of the charts. However, one chart revealed agreement of 92%, for a total inter-rater ag reement of 97% for this study. The principal investigator reviewed the chart that did not have complete agreement to find 100% agreement with her initial findings. Demographics Table 3-1 includes sample demographics and p values for statistical tests to rule out selection differences. All demographics we re collected as continuous or dichotomous variables, with the exception of race. Race in the medical chart was classified as White, Black, Asian, Hispanic or Other. Theref ore, race is documented with the same classifications. The pre-intervention gr oup was 4% Hispanic, 42% Black, and 54% White. The post-intervention group was 2% Asian, 3% Hispanic, 49% Black, 42% White and 4% with undocumented race in the medical char t. Due to the majority of participants being Black or White, with minimal representa tion of other racial backgrounds, and due to Black being a known predictiv e factor in the literature, race was dichotomized into Black and not Black for statistical purposes to rule out selection differences.

PAGE 62

51 Table 3-1. Demographics of the pre-interv ention and post-interven tion groups with test statistics for selection differences Low SES Race (Black) Maternal Age Marital Status (not married) Transferred Instead of DC Home Length of Stay Birth Weight EGA No. of Sibs PreIntervention Group .775 .42 25.46 .56 .432 50 1074 28.57 1.01 PostIntervention Group .70 .49 25.62 .57 .327 54 1114 28.7 .86 p Value to Investigate Selection Differences .339 .256 .899 .860 .225 .534 .368 .762 .297 There was a large percentage of participan ts of low socioeconomic status (77.5% in pre-intervention group a nd 70% in the post-intervention gr oup), Black race (42% in the pre-intervention group a nd 48% in the post-intervention group), and unmarried mothers (56% in the pre-interventi on group and 57% in the post-in tervention group). Average maternal age in the pre-in tervention group was 25.46 and in the post-intervention group was 25.62 years. The average birth weight in the pre-intervention group was 1074 grams, and the average birth weight in the post-i ntervention group was 1114 grams. The average gestational age (abbreviated EGA) at bi rth was 28.57 weeks gestation in the preintervention group and 28.7 week s in the post-intervention group. The average number of siblings (abbreviated No. of Sibs) in th e pre-intervention group was 1.01 and in the postintervention group was .86. Eighty four percen t of the pre-interv ention group consisted of single births, and 83.3% of the post-intervention group consisted of single births. In the pre-intervention group there were 43.2% of participants who were transferred to another hospital instead of discharge home, and in the post-intervention group there were 32.7% who were transferred to another hospita l. Average length of stay in the preintervention group was 50 days and in th e post-intervention group was 54 days.

PAGE 63

52 Investigation for Selection Differences Due to the matching through cohort controls research design, it was important to first determine if there were selection diffe rences in the two groups being compared. Socioeconomic status was categorized in to Women, Infants and Children (WIC) or Medicaid eligibility or not WIC/Medicaid eligible. Difference in this variable between the two groups was investigated by use of a z test for 2 proportions. Hypothesis testing of two proportions with a z test was used to test for group differences in maternal race, which was dichotomized as Black or not Black. Group dissimilarity based on maternal age was investigated through an independent sa mples t test, while differences in marital status (married, not married) and sex of the infa nt were investigated by use of a z test for two proportions. Gestational age at birth, bi rth weight and number of siblings was investigated by use of an independent sample s t test. Discharge status was investigated with a z test of 2 proportions a nd length of stay with an i ndependent samples t test. By testing each of the demographic variables at an alpha of .05, none of the demographic variables were significan tly different between the two groups (see table 3-1). Subsequently, having no selection differences supports the ability to use the matching through cohort controls design. Results Per Research Question The primary aim of this study was to im plement a breastfeeding intervention that would improve breastfeeding practices in th e NICU. The results of this study are provided per research question. Is there a significant difference in breast milk feeding initiation in very low birth weight (VLBW) infants admitted to the NICU before and after implementation of the intervention plan?

PAGE 64

53 Table 3-2 summarizes breast milk feeding initiation results. The breast milk feeding initiation rate in the pre-interventi on group was 74.1%. The breast milk feeding initiation rate in the post-i ntervention group was 85.2%. This represents an increase of 11.1%. However, through a PearsonÂ’s Chi Squa re Test of 2 proportions, the p value is .124, indicating no significant difference between groups when tested at an alpha of .05. The odds ratio of breast milk feeding initiati on is 2.013 with a confid ence interval of .818 to 4.952. Table 3-2. Breast milk feeding initiation rates Was Brest Milk Ever Provided? Total Test Statistic No Yes Count 21 60 81 Pre-Intervention % within subject 25.9% 74.1% 100.0% Count 8 46 54 Group PostIntervention % within subject 14.8% 85.2% 100.0% Count 29 106 135 Total % within subject 21.5% 78.5% 100.0% PearsonÂ’s Chi-Square Significance .124 Odds Ratio 2.013 Odds Ratio Confidence Interval .818 to 4.952 Is there a significant difference in the rate of breastfeeding in the hospital among women with VLBW infants hospitalized in the NICU before and after the implementation of the intervention plan? For the continuous variable of number of times breastfed per day after 30 weeks estimated gestational age (EGA), see figure 3-1. The graph is clearly skewed toward 0. Due to the violation of normality, a Mann Whitn ey nonparametric test was used to test significance of this variable. Interpretation of this graph a nd variable is difficult as the

PAGE 65

54 rate of breastfeeding in the NICU is si gnificantly low at .059 in the pre-intervention group, which is once every 17 days, and .139 in the post-intervention group, which is once every 7 days (see Table 3-3). This va riable proved to be significantly different between the two groups with a p value of .011. Figure 3-1. Number of times breastf ed per day after 30 weeks gestation

PAGE 66

55 Table 3-3. Comparison of rates of breastfeeding after 30 weeks gestation subject Number of Times Breastfed Per Day After 30 EGA Test Statistic Pre-Intervention Group Mean .0593937 N 81 Std. Deviation .18818812 Post-Intervention Group Mean .1389242 N 54 Std. Deviation .24433376 Total Mean .0912059 N 135 Std. Deviation .21513898 Mann Whitney Significance .011 To enable easier interpretation of this va riable, it was dichotomized into whether a mother ever participated in breastfeeding while in the hospita l. In the pre-intervention group, there were 25.9% of mothers who ever breastfed their infa nt in the hospital. In the post-intervention group, there we re 44.4% of mothers who ever breastfed their infants in the hospital (see Table 3-4). This represen ted an increase of 18.5%, which achieved a p value of .025 through a chi-square test of 2 pr oportions. Therefore, there were significant differences in proportion of wo men who ever breastfed in th e two groups, using an alpha of .025. The odds ratio of ever breastfed in the hospital was 2.286 with a confidence interval of 1.1 to 4.750.

PAGE 67

56 Table 3-4. Rates of ever breastfed in the neonatal intensive care unit Was the Infant Ever Breastfed While in the Hospital? Total Test Statistic Frequency No Yes Count 60 21 81 PreIntervention % within subject 74.1% 25.9% 100.0% Count 30 24 54 Group PostIntervention % within subject 55.6% 44.4% 100.0% Count 90 45 135 Total % within subject 66.7% 33.3% 100.0% PearsonÂ’s Chi Square Significance .025 Odds Ratio 2.286 Odds Ratio Confidence Interval 1.1 to 4.750 Is there a significant difference in breas t milk feedings at discharge in VLBW infants admitted to the NICU before and after the implementation of the intervention plan? There were 35.8% of infants who were pr ovided with breast milk at the time of discharge in the pre-intervention group. There were 40.7% of infants in the postintervention group who were provided with breas t milk at discharge. This 4.9% increase resulted in a p value of .562 through a chi-sq uare test of 2 pro portions, indicating no statistically significant difference among groups. The odds ratio was 1.233 with a confidence interval of .607 to 2.502.

PAGE 68

57 Table 3-5. Rates of breast milk feedings at discharge Was Breast Milk Provided at Discharge Total Test Statistic No Yes Count 52 29 81 PreIntervention % within subject 64.2% 35.8% 100.0% Count 32 22 54 Group PostIntervention % within subject 59.3% 40.7% 100.0% Count 84 51 135 Total % within subject 62.2% 37.8% 100.0% Pearson Chi-Square Significance .344 Odds Ratio 1.233 Odds Ratio Confidence Interval .607 to 2.502 Is there a significant difference in the pr oportion of the hospital stay that breast milk is provided in VLBW infants admitted to the NICU before and after the implementation of the intervention plan? Looking at figure 3-2, both groups have peaks at 0 and 1. However, there is a larger peak at 0 in the pr e-intervention group. The variab le, proportion of the hospital stay that breast milk was provided, did not achieve the assumption of normality as the graphs are u-shaped. Therefore, an ANOVA co uld not be run on this variable without violating assumptions. The nonparametric Ma nn-Whitney test was used to test for significant differences. This test indicated that the p value was .108, therefore there were not significant differences between the two gr oups in proportion of th e hospital stay that breast milk was provided.

PAGE 69

58 Figure 3-2. Proportion of the stay that breast milk was provided To look at effect size in a variable th at does not have a normal distribution, the proportion of the hospital stay was dichotomi zed into breast milk provided for most of the hospitalization or not. This variable represented whether breast milk feeds were supplied to the infant more than 50% of the length of stay not. Th ere were 51% of the pre-intervention group who provide d breast milk for most of the hospital stay, and there were 57% of the post-intervention group who provided breast milk for most of the hospital stay. This gives an odds ratio 1.219 with a confidence interval of .608 to 2.444 for breast milk feeds being provided for most of the hospital stay.

PAGE 70

59 CHAPTER 4 DISCUSSION The Effect of the Interventi ons on Breastfeeding Practice s in the Neonatal Intensive Care Unit This research investigated if a three-pa rt intervention plan designed to promote increased breastfeeding in premature infant s would have an effect on breastfeeding practices in the neonatal intensive care un it (NICU). The three-part intervention consisted of opportunities for education of health care professionals over a 6 week educational initiative, an educational booklet for new mothers in the NICU, and modifications to the individualized care plan (ICP) with a pathway of care for breastfeeding. This interven tion did have an effect on br eastfeeding practices in the NICU, but it did not result in changes across all breastfeeding variab les as hypothesized. Evidence of improved breastfeeding practi ces was that rates of breastfeeding (mothers putting their infants directly at the breast) in the NICU improved following intervention. The number of times infants were breastfed per day after 30 weeks gestation was significantly greater in the post-interv ention group than in the pre-intervention group. Mothers in the pre-intervention group breastf ed their infants after 30 weeks gestation .059 times per day, and in the post interven tion group they breastfed .139 times per day. This works out to an average of once ever y 17 days in the preintervention group and once every 7 days in the post intervention group. Although a stat istically significant increase was observed following the interventi on, the resultant rate of breastfeeding in the NICU remained low. After dichotomizing this variable, it was noted that there were

PAGE 71

60 25.9% of mothers who ever breastfed their infa nt in the hospital in the pre-in tervention group. In contrast, there were 44.4% of mothers who ever breas t fed their infants in the post intervention group. This represented a si gnificant increase of 18.5%. This provides some support that the program had a positive effect with increase d participation in breastfeeding among mothers in the NICU. However, the ov erall rate of breastfeeding participation remains low in the NICU. The variable of ever breastfed while in the hospital is an important one, as it requires active participation by the mother and infant. Diminished milk supply is cited as one of the most significant barriers to breastf eeding in the NICU population. Maternal stress has been linked to inhibition of oxyt ocin, which is responsible for the let down response during pumping and breastfeeding (Lang, 1996). However, physiologically, oxytocin is facilitated with in creased mother-infant contact an d environments that foster breastfeeding ("How brea st-feeding postpones ovulation," 1 985). Mothers with infants in the NICU typically experience high leve ls of stress and anxiety and many are overwhelmed by the NICU environment (Nyqvi st et al., 1994). Many also comment on the loss of control of their infant to othe rs during stays in the NICU (Lupton & Fenwick, 2001). Close contact, as in skin to skin, as well as breastfeeding, are important in maintaining bonding in a difficult environment, promoting the milk supply and providing some control over care for mothers (Kirsten, Bergman, & Hann, 2001). This direct interaction of the mother and infant can be assumed to be critical to the other breastfeeding outcomes. Although positive trends in favor of the post-intervention group were observed across all variables, the inte rvention did not have a str ong enough impact to result in

PAGE 72

61 significant changes in breast milk feeding initiation rates, breast milk feeding at discharge rates and proportion of the hospital stay that breast milk was provided. This could be due to lack of a strong enough impact of the intervention with a need for a more extensive list of interventions or adjustment of the interv entions proposed in this study, need for consistent compliance with the rese arch interventions to promote change, or a different time frame for the study to detect changes. The breast milk feeding initiation rate (was breast milk ever provided) in the preintervention and post-intervention groups app ears to be fairly comparable to other research findings, which have documented rate s of breast milk feeding initiation in the NICU at 64% (Byrne & Hull, 1996), 72.9% (Mei er et al., 2004), a nd 83% (Yip et al., 1996). Although breast milk feeding initiation failed to reach statistical significance, there was an increase in breast milk feed ing initiation of 11.1%, which was a positive change in the right direction as more infants received some breast milk following the intervention. Scientists are beginning to refer to breast milk as medicine and have initiated discussing breast milk in terms of a dose (Meier et al., 2004). Thus following the intervention, 11.1% more infants in the post-intervention group re ceived breast milk at the most critical stage of their recover y. However, this va riable did not achieve statistical significance, which c ould indicate that the impact of the interventions was not strong enough or that there wa s not enough compliance with th e interventions to promote change. It is important to note that breast m ilk feeding initiation re flected whether breast milk was ever provided and not if breastfeeding ever occurred or if br east milk continued to be supplied after the 2nd day of life or at discharge.

PAGE 73

62 Although the rate of breast milk feedings at discharge (was breast milk provided to the infant at discharge) increased slightly by 4.9%, this difference failed to reach statistical significance. Succeeding with breast milk feedings until discharge in the NICU population is a significant challenge for mother s with infants in the NICU. Discharge for many of these infants did not occur until they were 1 to 4 months old. This supports other research findings that state that so me of the strongest predictors of not breastfeeding by discharge is being low birt h weight, having decrea sed gestational age and being admitted into the NICU (Hwang et al., 2006; Li et al., 2005; Powers et al., 2003; Scott et al., 2006). The findings of th is study in both the pre-intervention (35.8%) and post-intervention (40.7%) groups are compar able to other research findings, which document breast milk feedings at discharge to be 38% (Yip et al., 1996). The low rate of breast milk feeding in the NICU at discharg e gives some insight into the significant challenges that women encounter with succeeding with breast milk feedings through a NICU hospitalization. However, to enable moth ers with this process, it is critical that strategies to promote long term success are implemented. Although lack of full implementation and compliance with intervention strategies may have factored into this research study, these results suggest that th is 3-pronged intervention was not sufficient to enable mothers to overcome the barriers to maintaining breastfeeding until the time of hospital discharge in this complex and challenging environment. There was no difference between the groups in the proportion of the hospital stay that breast milk was provided. These re sults again highlight the importance of determining strategies for long term success to enable mothers to succeed with the

PAGE 74

63 breastfeeding process fo r a larger proportion of the hospita lization. Such strategies could enable breast milk feedings until discharge and beyond. This study complements conclusions from other studies that attention to and education about lactation affects health car e professional knowledge and support of the breastfeeding process (Siddell et al., 2003). Many studies c ite lack of health care professional education as a significant barrie r to the breastfeeding process in the NICU (Register et al., 2000). Although improvement in breastfeeding rate s in the NICU was observed in the post education group, there is no way to de termine which part of the intervention plan may have had an effect on th is variable. Other studies have concluded that health care professionals play a signi ficant role in breastfeeding practices in the NICU (Swanson & Power, 2005) as does in creased knowledge about breastfeeding among mothers in the NICU (Bernaix, 2000). However, because all the interventions were implemented together as a 3-pronged a pproach, it is unclear if the educational materials for mothers, the educational in itiative, or the modifications to the ICP/breastfeeding pathway resulted in th e effect on breastfee ding in the NICU. According to the transtheoretical model of behavior change, change is a process and sometimes requires multiple approaches th at are stage matched in addition to the passage of time. The real impact of the e ducation perhaps could not be fully observed in assessing the indirect impact on breastfeeding practices am ong mothers and infants. By assessing the indirect impact of the education interventi on on breastfeedi ng strategies without measuring the knowledge and behavior change of th e health care professionals, it cannot be determined if there was a change in the health care profe ssionals and what the magnitude of such a change was. This calls into question whether the educational

PAGE 75

64 initiative did what it was intended, which was to change health care professional behavior. With knowledge of a change in health care professional behavior, the true impact of education and inte rvention strategies on breastf eeding outcomes in the NICU could be assessed. In addition, there were 2 behavior changes that could have been assessed, the health care professional as we ll as the mother. Further measuring the mothersÂ’ behavior change, in the presence of education from the health care professional, would have provided useful information on whether the intervention was strong enough to elicit positive changes in the mothers. Although positive changes are evident, the re ported compliance with strategies and follow through of education key points was called into question by the nursing administration. Other studies have determined that education about breastfeeding has had an effect on breastfeeding knowledge and s upportive behaviors (Ekstrom, Widstrom et al., 2005). This research study di d not measure the behaviors of health care professionals following education. Therefore, it could be that the 3 part intervention plan was not strong enough to elicit changes in health ca re professional behavior that not enough time passed post intervention to enable successful change or that the health care professional change elicited following the intervention was not enough to result in positive outcomes across all proposed breastfeed ing variables. Subsequently, this study provides only partial support for positive changes in breas tfeeding practices in the NICU following the 3 part intervention. Limitations This study is not without limitations. Limitations included inadequate implementation of all the intervention strategi es by the health care professionals, lack of methods to determine behavior change and im plementation by health care professionals,

PAGE 76

65 lack of ability to give ample time for ch anges to be implemented without introducing other confounds, lack of participation by key decision makers in the NICU, the inability to control for other changes in the NICU environment, lack of a more comprehensive breastfeeding intervention plan, and lack of a randomized sample. An important limitation of this study was the questionable full implementation of the strategies learned in the educational in itiative as well as inconsistent use of the modified ICP and inconsistent distribution of “A Mother’s Gift”. Although there was a high health care professional par ticipation rate in the education initiative of 63%, health care professional behavior change and attitude s were not measured. Therefore, there is no way to know the direct effect of the edu cation and placement of the modified ICP in the medical chart on health care behavior and attitudes. It can be assumed that although there was good participation in the educational initiative by bedside nurses and the breastfeeding pathway on the ICP was added to the medical record, there was a lack of movement to action among many health care professionals based on the observations by nursing administration as well as by the first f eeding at the breast variable remaining low in the post-intervention group. Despite education about promoting br eastfeeding in the NICU and how to introduce such practices, there remained a la rge percentage of women (56%) who never breastfed while in the hospital. This demons trates that although the education may have occurred, change was not fully embraced and implemented. Health care professionals care for infants, including caring for nutriti onal feeds, in the NI CU 24 hours a day, and parents are visitors to the NICU when they come to see their infant. Breastfeeding is typically something that hea lth care professionals would en able, and thus its low rate

PAGE 77

66 could be attributed to low hea lth care professional support of the breastfeeding process. This study investigated breas tfeeding outcomes in the natu ral NICU environment in which health care professionals may have ha d other priorities, may have had negative beliefs about breastfeeding in th e NICU, may have had control issues that interfered with promoting a feeding method that they eventua lly cannot participate in, or may have been suffering from burn out. However, without the intermediate meas ure of health care behavior, it is difficult to determine if the in tervention could be effective if health care professionals demonstrated an appropriate beha vior change and were called to action. According to the trantheoretial model of behavior change (TTM), change is a process with people moving through different stages of readiness to change (Prochaska et al., 2001). The intervention plan was desi gned to provide interventions that are appropriate for individuals in each stage of readiness to change. However, the educational initiative and interventions were introduced with data collection occurring in the post intervention group 6 s hort weeks later. This may not have been a long enough period of time to enable individuals to re spond to the intervention and move them to action. According to the TTM, those in the pre-action stages are the precontemplators (no intention of making a change), the contem plators (intention of making a change in the next 6 months) and those in the preparation stage (intend on making a change in the next 30 days). Theoretically a six week period may not have been enough to enable movement through the stages of change to promote action toward increasing breastfeeding support and implementing strategies learned in the educ ational initiative. In contrast, other confounds were being intr oduced in the NICU in the months following the conclusion of this study, as more inte rventions were set to be implemented.

PAGE 78

67 Therefore, this study could not account for th e possible inadequate time frame due to the potential introduction of confounds th at could bias the results. Another limitation was the weak partic ipation in the education modules by physicians and nurse practitioners, who are ke y decision makers in the NICU. Prior to the educational initiative dates, they participated in a short, general inservice about breastfeeding to f acilitate discussion and direction. Due to this recent meeting, it was difficult to get them to participate in the research educational in itiative. Therefore, participation was low with only 20% of neona tologists and 9% of nurse practitioners attending an inservice or co mpleting the self study edu cational module. Although the researcher did meet with the physician medical director of the unit to discuss key points of the educational initiative, the initiative ma y have been much more successful if the neonatologists and nurse practiti oners had higher levels of pa rticipation. Nursing staff and parents look to physicians and neonatal nu rse practitioners for direction, and their lack of participation was a signi ficant limitation to this study. The NICU is a constantly changing environm ent that can not be fully controlled in a study like this, where all infants during a sp ecific time frame are being enrolled. This study does not account for other changes that may have occurred, such as staff changeover, other education that health care professionals may be receiving and implementing, and changes made to the physical NICU environment. Cosmetic changes were made to the NICU environment duri ng the course of this study, and it was impossible to determine if this may have had an effect on the results. The lack of optimal resources that may facilitate breastfeedin g may also pose a limitation to this study. During the early stages of planning this res earch project, funds

PAGE 79

68 were applied for to provide a breast pump loaner closet that would provide a needed resource for long term maintenance of the m ilk supply (Meier et al., 2004). These funds were not achieved, making it impossible to pr ovide this resource to enable long term success with maintaining the milk supply, and subsequently with addressing the idea of providing breast milk all the way until the in fant was discharged from the NICU. Providing breast milk or breastfeed ing at the time of discharge would be the goal, as this would indicate that a mother succeeded w ith providing milk during the hospitalization and would enable breastfeeding at discharg e home. Without the funding, women were left to find their own resources to access a hospital grade pump for milk expression. Some may have accessed one for use at home while others may have used store bought pumps or self expressed. This study is a cohort study. Due to lack of randomization, there is the possibility that there are unseen differences in the two groups that lead to differences or similarities in breastfeeding practices that cannot be attributed to the in tervention. Results from this study can only give conclusions about the popu lation of infants being studied at Shands Hospital. In addition, the small sample si ze limited the ability to achieve adequate power, which can affect being able to detect si gnificant differences, if they exist. Based on the utilized sample size, 38.6% power was ac hieved on the primary variable of breast milk feeding initiation. This indicates that the probability of finding a difference in the two groups was only 38.6%. The already high br east milk feeding ini tiation rate in the pre-intervention group, 74.1%, ga ve less room for improvement in this variable. Although comparable with other ra tes reported in the literature, the rate in this unit was expected to be lower due to lack of lactation se rvices in the unit a nd the large population

PAGE 80

69 of Black mothers, mothers with low socioec onomic status and single mothers, which are all predictive of decreased br eastfeeding behaviors (Powers et al., 2003). The nature of the population may have been a limitation, in th at perhaps the effect of the intervention would be different given a diffe rent demographic presence in the NICU. Given the effect of the current intervention, a larger sample si ze would be needed to determine if there are significant differences in the pre-interventi on and post-intervention groups. Such sample size was not feasible during the study pe riod given the number of admissions and discharges. Prolonging enroll ment was not feasible give n the risk of introducing potential bias. To Recommendations for Further Research Women with infants in the NICU face unique challenges to the breastfeeding process. These barriers result in decreased breastfeeding initiation rates and breast milk feedings at discharge from the hospital. Howe ver, these fragile infants are at an increased need for the benefits of breast milk. Ther efore, further studies looking at trends in breastfeeding and looking at in terventions that can assist mothers in overcoming barriers is necessary. Future research could investigate the e fficacy of educational interventions that measure the intermediate effects of the inte rvention on health care professional behavior as well as the effect on breastf eeding outcomes. With similar findings as this study, this would make it easier to interpret if the educat ional interventions did not cause health care professional behavior change and thus did not affect breastfeeding outcome or if the educational interventions did affect health care professi onal behavior but that the resultant effect of this was not strong enough to result in positi ve breastfeeding outcomes.

PAGE 81

70 Future studies addressing an educationa l initiative for health care professionals with more time to enable change would be appropriate to run in this population. If change takes time to occur, a longer period be tween the intervention a nd the start of data collection in the post interven tion group would help identify ch ange that occurred over a longer period of time following the interventions. However, such a study would require close control of the environment to preven t other confounds from being introduced as time from the education initiative elapses. Although an intervention w ith many parts that addre sses milk expression and breastfeeding over the course of hospitalization ma y be optimal in promoting breastfeeding practices in the NICU, a look at intervention sp ecific studies could help to declare which interventions are successful and which are not. Such a study, although it may require a much larger sample size and perhaps random assignment to groups and a multi-center trial, could assist with understa nding which, if any, and in what combination, treatment(s) have an effect on brea stfeeding practices in the NICU. Further research could also investigat e the specific results, amount of milk expression and success with breastfeeding am ong women who had a nurse who was in an action oriented stage of beha vior change regarding support for lactation compared to those receiving traditional NICU care. The effects of support and education for the mother could further be investigated by in cluding success and failure with breastfeeding among mothers who demonstrated compliance wi th strategies in “A Mother’s Gift” compared to those who were not fully compliant. With the paucity of breastfeed ing (putting the infant to breast) in the NICU, it is of great interest to determine the predictive e ffect of breastfeeding on breast milk feedings

PAGE 82

71 at discharge, milk production, amount of breast m ilk feedings in the hospital, as well as the effect on maternal and child health. Likewise, it would be of great interest to determine if there are any nega tive effects of not enabling brea stfeeding on the ability to maintain the milk supply, success of breastfeeding at discharge and beyond and the ability to transition from breast milk feeds to direct breastfeeding. Conclusions This study investigated change in breastf eeding practices following implementation of an intervention plan with 3 parts; an educational booklet for new mothers, an educational opportunity for health care profe ssionals who serve mother-infant dyads in the NICU, and changes to the individualized care plan that necessitate breastfeeding practice documentation by nurses. There we re general positive trends across all variables, which make this type of interven tion have some promise for affecting positive changes in breastfeeding practices in the NICU given the limitations of this study and the difficulties that can be expected in the complex NICU environment. There were statistically significant differe nces between groups in the rate of breastfeeding, with infants in the post-intervention group being more than twice as likely to be breastfed in the hospital. While health care professionals may play a significant role in the breastfeeding process, perhaps an educational plan and breastfeeding pathway as well as complementary educational materials for mothers is not enough to promote full participation and optimal breas tfeeding practices in the NICU This study demonstrated limited, but encouraging, support for an in tervention plan focused on education to facilitate change in breastfeeding wi thin the NICU, but perhaps a multifaceted intervention plan, including hospital grade br east pump allocation may optimize positive

PAGE 83

72 changes in breastfeeding practices. An important limitation of this study was lack of full implementation of the intervention strategies. Future studies can be designed to better measure compliance with educational interven tions as well as the individual effect of each intervention and the additive effect of multiple interventions implemented together. Randomized multi-institutional studies will enab le larger intervention trials as well as generalization of findings.

PAGE 84

73 APPENDIX A OUTLINE OF THE EDUCATION MODULE Breastfeeding in the Neonatal Intensive Care Unit An Educational Module for H ealth Care Professionals Introduction Benefits of Breastfeeding for the Full Term Infant Health Benefits Developmental Benefits Financial Benefits Benefits for Preterm or High Risk Infants Benefits for the Mother Barriers to Breastfeeding Full-term infants High-risk infants Contraindications Breast Milk Composition Transition to Mature Milk The Breast and Lactogenesis Anatomy Neuroendocrine control Sucking pattern General Interventions for Supporting Breastfeeding Appropriate Timing of Interventions Interventions in the NICU Interventions Prior to Active Breastfeeding Breast pumps Maintaining a milk supply Kangaroo Care Non-nutritive suckling

PAGE 85

74 Early initiation of direct breastfeeding Monitoring Physiologic Responses Transition to Direct Breastfeeding Interventions for Supporting Breastfeeding Privacy Positioning Timing Special Techniques Determining Adequacy of Intake Conclusion Post-test References For further information, contact the authors: Roberta Gittens Pineda : jopineda@pol.net Cammy Pane: cam92460@yahoo.com

PAGE 86

75 APPENDIX B OUTLINE OF ITEMS ADDED TO TH E INDIVIDUALIZED CARE PLAN Breastfeeding Guidelines: Within 6 hours of delivery: Mom is given “A Mother’s Gift” Within 24 hours: Ensure proper pumping (w ith hospital grade pump) and storage 3 to 5 days of life: Assess Mom’ s milk production. Address problems 10 days of life: Mom’s milk supply should be at least 350 ml per 24 hours With each parent contact: Offer support and discuss any problems Kangaroo care is encouraged as soon as possible First oral feeing is at breast Mom is encouraged to breastfeed at each visit with supplementation only when medically indicated At discharge: Encourage tr ansition to full breastfeedi ng while monitoring weight gain

PAGE 87

76 APPENDIX C OUTLINE OF THE EDUCATI ONAL BOOKLET FOR MOTHERS A Mother’s Gift Breastfeeding and Pumping for Your Baby in the NICU Breast milk is the best food for your baby’s start in life Each mother’s milk is prefect for her baby Babies who drink breast m ilk are healthier and smarter This makes breast milk even more important for premature babies Breastfeeding is healthy for the mother too! Providing breast milk for your baby is something very special you can do to help your child It is important to begin expressing your milk as soon after delivery as possible, and to keep pumping Pumping may seem complicated at fi rst, but it is worth it for your baby During pumping, relax and enjoy gentle thoughts about your baby At first, you may not get any milk, or only a few drops Here is how to store your milk Every mother wants to know -“Am I making enough milk” In order to maintain a good milk supply… Most mothers really want to hold their babies Kangaroo care is good for your baby Kangaroo care is good for parents too Privacy is more important to some than others… Suckling at the breast is the best way to nourish your baby… and it takes practice Positioning during breastfeeding will become easier with practice… Is my baby really, really, really getting enough milk Please take extra care of yourself too! If you are taking medications Making milk for twins or more… Please ask if you have questions or need help How to obtain a pump for home use… Chart for tracking milk production

PAGE 88

77 On-line resources For more information, contact: Roberta Gittens Pineda : jopineda@pol.net

PAGE 89

78 LIST OF REFERENCES Agostoni, C., & Haschke, F. (2003). Infant formulas. Recent developments and new issues. Minerva Pediatr, 55 (3), 181-194. Allen, M. C., & Capute, A. J. (1990). Tone and reflex development before term. Pediatrics, 85 (3 Pt 2), 393-399. Alm, J. S., Swartz, J., Bjorksten, B., Engstra nd, L., Engstrom, J., Kuhn, I., et al. (2002). An anthroposophic lifestyle and in testinal microflora in infancy. Pediatr Allergy Immunol, 13 (6), 402-411. Arora, S., McJunkin, C., Wehrer, J., & K uhn, P. (2000). Major factors influencing breastfeeding rates: Mother's perception of father's attitude and milk supply. Pediatrics, 106 (5), E67. Baker, R. (2003). Human milk subst itutes. An American perspective. Minerva Pediatr, 55 (3), 195-207. Ball, T. M., & Wright, A. L. (1999). Health care costs of formula-feed ing in the first year of life. Pediatrics, 103 (4 Pt 2), 870-876. Berens, P. D. (2001). Prenatal, intrapartu m, and postpartum support of the lactating mother. Pediatr Clin North Am, 48 (2), 365-375. Beresford, H. J. (1984). The success of breast feeding. IPPF Med Bull, 18 (5), 3-4. Bernaix, L. W. (2000). Nurses' attitudes, subj ective norms, and behavioral intentions toward support of breastfeeding mothers. J Hum Lact, 16 (3), 201-209. Bick, D. E., MacArthur, C., & Lancashire, R. J. (1998). What influences the uptake and early cessation of breast feeding? Midwifery, 14 (4), 242-247. Birch, E., Birch, D., Hoffman, D., Hale, L ., Everett, M., & Uauy, R. (1993). Breastfeeding and optimal vi sual development. J Pediatr Ophthalmol Strabismus, 30 (1), 33-38. Black, K. A., & Hylander, M. A. (2000). Breas tfeeding the high risk infant: implications for midwifery management. J Midwifery Womens Health, 45 (3), 238-245. Blackburn, S. (1998). Environmental impact of the NICU on developmental outcomes. J Pediatr Nurs, 13 (5), 279-289.

PAGE 90

79 Breastfeeding and the use of human milk. American Academy of Pediatrics. Work Group on Breastfeeding. (1997). Pediatrics, 100 (6), 1035-1039. Bueno, M. B., de Souza, J. M., de Souza, S. B., da Paz, S. M., Gimeno, S. G., & de Siqueira, A. A. (2003). [Risks associated with the weaning process in children born in a university hospital: a prospective cohor t in the first year of life, Sao Paulo, 1998-1999]. Cad Saude Publica, 19 (5), 1453-1460. Byrne, B., & Hull, D. (1996). Breast milk for preterm infants. Prof Care Mother Child, 6 (2), 39, 42-35. Callen, J., & Pinelli, J. (2005). A review of the literature examining the benefits and challenges, incidence and duration, and barrier s to breastfeeding in preterm infants. Adv Neonatal Care, 5 (2), 72-88; quiz 89-92. Chan, D. K. (2001). Enteral nu trition of the very low bi rth weight (VLBW) infant. Ann Acad Med Singapore, 30 (2), 174-182. Chantry, C. J., Auinger, P., & Byrd, R. S. (2004). Lactation among adolescent mothers and subsequent bone mineral density. Arch Pediatr Adolesc Med, 158 (7), 650-656. Chen, A., & Rogan, W. J. (2004). Breastfeeding and the risk of postneonatal death in the United States. Pediatrics, 113 (5), e435-439. Chen, C. H., Wang, T. M., Chang, H. M., & Chi, C. S. (2000). The effect of breastand bottle-feeding on oxygen satu ration and body temperature in preterm infants. J Hum Lact, 16 (1), 21-27. Chua, S., Arulkumaran, S., Lim, I., Selamat, N., & Ratnam, S. S. (1994). Influence of breastfeeding and nipple stimulati on on postpartum uterine activity. Br J Obstet Gynaecol, 101 (9), 804-805. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, N.J.: L. Erlbaum Associates. Coleman, S., Dracup, K., & Moser, D. K. (1991). Comparing methods of cardiopulmonary resuscitation instru ction on learning and retention. J Nurs Staff Dev, 7 (2), 82-87. Dai, D., & Walker, W. A. (1998). Role of b acterial colonization in neonatal necrotizing enterocolitis and its prevention. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi, 39 (6), 357-365. Davis, M. K. (2001). Breastfeeding and ch ronic disease in chil dhood and adolescence. Pediatr Clin North Am, 48 (1), 125-141, ix. Dewey, K. G., Heinig, M. J., & Nommsen, L. A. (1993). Maternal weight-loss patterns during prolonged lactation. Am J Clin Nutr, 58 (2), 162-166.

PAGE 91

80 Diaz Gomez, M., Ramos Acosta, C. L., Rico Sevillano, J., Robayna Curbelo, M., & Alvarez Alvarez, J. (1997). [Breast feeding and length of hospitalization]. Rev Enferm, 20 (231), 11-14. do Nascimento, M. B., & Issler, H. (2004). [Breastfeeding in premature infants: inhospital clinical management]. J Pediatr (Rio J), 80 (5 Suppl), S163-172. Docherty, S. L., Miles, M. S., & Holditch-D avis, D. (2002). Worry about child health in mothers of hospitalized medically fragile infants. Adv Neonatal Care, 2 (2), 84-92. Ekstrom, A., Matthiesen, A. S., Widstrom A. M., & Nissen, E. (2005). Breastfeeding attitudes among counselling health professionals. Scand J Public Health, 33 (5), 353-359. Ekstrom, A., Widstrom, A. M., & Nissen, E. (2005). Process-or iented training in breastfeeding alters attitudes to br eastfeeding in health professionals. Scand J Public Health, 33 (6), 424-431. Espy, K. A., & Senn, T. E. (2003). Incidence a nd correlates of breast milk feeding in hospitalized preterm infants. Soc Sci Med, 57 (8), 1421-1428. Fergusson, D. M., & Woodward, L. J. (1999) Breast feeding and later psychosocial adjustment. Paediatr Perinat Epidemiol, 13 (2), 144-157. Gartner, L. M., Morton, J., Lawrence, R. A., Na ylor, A. J., O'Hare, D ., Schanler, R. J., et al. (2005). Breastfeeding and the use of human milk. Pediatrics, 115 (2), 496-506. Goldrick, B., Gruendemann, B., & Larson, E. (1993). Learning styles and teaching/learning strategy preferences: imp lications for educating nurses in critical care, the operating room, and infection control. Heart Lung, 22 (2), 176-182. Gomez-Sanchiz, M., Canete, R., Rodero, I., Ba eza, J. E., & Avila, O. (2003). Influence of breast-feeding on mental and psychomotor development. Clin Pediatr (Phila), 42 (1), 35-42. Gray, L., Miller, L. W., Philipp, B. L., & Bla ss, E. M. (2002). Breas tfeeding is analgesic in healthy newborns. Pediatrics, 109 (4), 590-593. Gross, S. J., David, R. J., Bauman, L., & Tomarelli, R. M. (1980). Nutritional composition of milk produced by mothers delivering preterm. J Pediatr, 96 (4), 641-644. Hanson, L. A. (1998). Breastfeeding provide s passive and likely long-lasting active immunity. Ann Allergy Asthma Immunol, 81 (6), 523-533; quiz 533-524, 537. Hanson, L. A., Korotkova, M., Haversen, L ., Mattsby-Baltzer, I., Hahn-Zoric, M., Silfverdal, S. A., et al. (2002). Breastfeeding, a complex support system for the offspring. Pediatr Int, 44 (4), 347-352.

PAGE 92

81 Harrington, S. S., & Walker, B. L. (2004). Th e effects of computer-based training on immediate and residual learning of nursing facility staff. J Contin Educ Nurs, 35 (4), 154-163; quiz 186-157. Harrold, J., & Schmidt, B. (2002). Eviden ce-based neonatology: making a difference beyond discharge from the neonatal nursery. Curr Opin Pediatr, 14 (2), 165-169. Hart, S., Boylan, L. M., Carroll, S., Musick, Y. A., & Lampe, R. M. (2003). Brief report: breast-fed one-week-olds demonstrate s uperior neurobehavi oral organization. J Pediatr Psychol, 28 (8), 529-534. Hill, P. D., Andersen, J. L., & Ledbetter, R. J. (1995). Delayed initiation of breast-feeding the preterm infant. J Perinat Neonatal Nurs, 9 (2), 10-20. Hill, P. D., Hanson, K. S., & Mefford, A. L. (1994). Mothers of low birthweight infants: breastfeeding pattern s and problems. J Hum Lact, 10 (3), 169-176. Holditch-Davis, D., & Miles, M. S. (2000). Moth ers' stories about thei r experiences in the neonatal intensive care unit. Neonatal Netw, 19 (3), 13-21. How breast-feeding postpones ovulation. (1985). Network, 7 (1), 3. Hwang, W. J., Chung, W. J., Kang, D. R., & Suh, M. H. (2006). [Factors affecting breastfeeding rate and duration]. J Prev Med Pub Health, 39 (1), 74-80. Hylander, M. A., Strobino, D. M., & Dhanir eddy, R. (1998). Human milk feedings and infection among very low birth weight infants. Pediatrics, 102 (3), E38. Hylander, M. A., Strobino, D. M., Pezzullo, J. C., & Dhanireddy, R. (2001). Association of human milk feedings with a reduction in retinopat hy of prematurity among very low birthweight infants. J Perinatol, 21 (6), 356-362. Isaacson, L. J. (2006). Steps to successfu lly breastfeed the premature infant. Neonatal Netw, 25 (2), 77-86. Jaeger, M. C., Lawson, M., & Filteau, S. (1997) The impact of prematurity and neonatal illness on the decision to breast-feed. J Adv Nurs, 25 (4), 729-737. Jang, Y. S. (2005). [Effects of a workbook program on the perceived stress level, maternal role confidence and breast f eeding practice of mothers of premature infants]. Taehan Kanho Hakhoe Chi, 35 (2), 419-427. Karl, D. J. (2004). Using principles of newbor n behavioral state orga nization to facilitate breastfeeding. MCN Am J Matern Child Nurs, 29 (5), 292-298. Kavanaugh, K., Mead, L., Meier, P., & Mangurten, H. H. (1995). Getting enough: mothers' concerns about breastfeeding a preterm infant after discharge. J Obstet Gynecol Neonatal Nurs, 24 (1), 23-32.

PAGE 93

82 Kemp, A., & Kakakios, A. (2004). Asth ma prevention: breast is best? J Paediatr Child Health, 40 (7), 337-339. Kirsten, G. F., Bergman, N. J., & Hann, F. M. (2001). Kangaroo mother care in the nursery. Pediatr Clin North Am, 48 (2), 443-452. Kronborg, H., & Vaeth, M. (2004). The influenc e of psychosocial factors on the duration of breastfeeding. Scand J Public Health, 32 (3), 210-216. Lanari, M., Papa, I., Venturi, V., Sermasi, S ., Corvaglia, L., Faldella, G., et al. (2001). [Neonatal sepsis]. Recenti Prog Med, 92 (11), 690-695. Lang, S. (1996). Breastfeeding special care babies. Mod Midwife, 6 (11), 34-35. Laubereau, B., Brockow, I., Zi rngibl, A., Koletzko, S., Gruebl, A., von Berg, A., et al. (2004). Effect of breast-feeding on the de velopment of atopic dermatitis during the first 3 years of life--results fr om the GINI-birth cohort study. J Pediatr, 144 (5), 602-607. Li, R., Darling, N., Maurice, E., Barker, L., & Grummer-Strawn, L. M. (2005). Breastfeeding rates in the United States by characteristics of the child, mother, or family: the 2002 National Immunization Survey. Pediatrics, 115 (1), e31-37. Lucas, A., Morley, R., Cole, T. J., & Gore S. M. (1994). A randomised multicentre study of human milk versus formula and la ter development in preterm infants. Arch Dis Child Fetal Neonatal Ed, 70 (2), F141-146. Lucas, A., Morley, R., Cole, T. J., Lister, G., & Leeson-Payne, C. (1992). Breast milk and subsequent intelligence quotie nt in children born preterm. Lancet, 339 (8788), 261-264. Lugo-Vicente, H. (2003). Necrotizing enterocolitis. Bol Asoc Med P R, 95 (2), 17-22. Lupton, D., & Fenwick, J. (2001). 'They've forg otten that I'm the mum': constructing and practising motherhood in sp ecial care nurseries. Soc Sci Med, 53 (8), 1011-1021. Marild, S., Hansson, S., Jodal, U., Oden, A., & Svedberg, K. (2004). Protective effect of breastfeeding against ur inary tract infection. Acta Paediatr, 93 (2), 164-168. Matthews, K., Webber, K., McKim, E., Banoub-Baddour, S., & Laryea, M. (1998). Maternal infant-feeding decisi ons: reasons and influences. Can J Nurs Res, 30 (2), 177-198. McGrath, J. M., & Braescu, A. V. (2004). State of the science: feeding readiness in the preterm infant. J Perinat Neonatal Nurs, 18 (4), 353-368; quiz 369-370. McVea, K. L., Turner, P. D., & Peppler, D. K. (2000). The role of breastfeeding in sudden infant death syndrome. J Hum Lact, 16 (1), 13-20.

PAGE 94

83 Meier, P. P. (1998). Strategies for assi sting breatfeeding in preterm infants. Meier, P. P. (2001). Breastfeeding in the sp ecial care nursery. Prematures and infants with medical problems. Pediatr Clin North Am, 48 (2), 425-442. Meier, P. P., & Brown, L. P. (1996). State of the science. Breastfeeding for mothers and low birth weight infants. Nurs Clin North Am, 31 (2), 351-365. Meier, P. P., Engstrom, J. L., Mingolelli, S. S., Miracle, D. J., & Kiesling, S. (2004). The Rush Mothers' Milk Club: breastfeeding in terventions for mothers with very-lowbirth-weight infants. J Obstet Gynecol Neonatal Nurs, 33 (2), 164-174. Mikiel-Kostyra, K. (2000). [Breast feeding as a component of reproductive health]. Ginekol Pol, 71 (7), 641-647. Miles, M. S., Funk, S. G., & Kasper, M. A. (1992). The stress response of mothers and fathers of preterm infants. Res Nurs Health, 15 (4), 261-269. Mitra, A. K., Khoury, A. J., Hinton, A. W ., & Carothers, C. (2004). Predictors of breastfeeding intention among low-income women. Matern Child Health J, 8 (2), 65-70. Morley, R., Cole, T. J., Powell, R., & Lucas, A. (1988). Mother's choice to provide breast milk and developmental outcome. Arch Dis Child, 63 (11), 1382-1385. Neiva, F. C., Cattoni, D. M., Ramos, J. L., & Issler, H. (2003). [Early weaning: implications to oral motor development]. J Pediatr (Rio J), 79 (1), 7-12. Nikolajski, P. Y. (1992). Invest igating the effectiveness of se lf-learning packages in staff development. J Nurs Staff Dev, 8 (4), 179-183. Nyqvist, K. H., Ewald, U., & Sjoden, P. O. (1996). Supporting a preterm infant's behaviour during breastf eeding: a case report. J Hum Lact, 12 (3), 221-228. Nyqvist, K. H., Sjoden, P. O., & Ewald, U. (1994). Mothers' advice about facilitating breastfeeding in a neonatal intensive care unit. J Hum Lact, 10 (4), 237-243. Nyqvist, K. H., Sjoden, P. O., & Ewald, U. ( 1999). The development of preterm infants' breastfeeding behavior. Early Hum Dev, 55 (3), 247-264. Oddy, W. H., Sherriff, J. L., de Kl erk, N. H., Kendall, G. E., Sly, P. D., Beilin, L. J., et al. (2004). The relation of breastfeeding and body mass index to asthma and atopy in children: a prospective c ohort study to age 6 years. Am J Public Health, 94 (9), 1531-1537. The optimal duration of exclusive breastfeed ing: results of a WHO systematic review. (2001). Indian Pediatr, 38 (5), 565-567.

PAGE 95

84 Owen, C. G., Whincup, P. H., Odoki, K., G ilg, J. A., & Cook, D. G. (2002). Infant feeding and blood cholestero l: a study in adolescents and a systematic review. Pediatrics, 110 (3), 597-608. Page, D. C. (2001). Breastfeeding is early functional jaw orthopedics (an introduction). Funct Orthod, 18 (3), 24-27. Pantazi, M., Jaeger, M. C., & Lawson, M. ( 1998). Staff support for mothers to provide breast milk in pediatric ho spitals and neonatal units. J Hum Lact, 14 (4), 291-296. Philipp, B. L., Malone, K. L., Cimo, S., & Merewood, A. (2003). Sustained breastfeeding rates at a US baby-friendly hospital. Pediatrics, 112 (3 Pt 1), e234-236. Pinelli, J., Atkinson, S. A., & Saigal, S. (2001). Randomized trial of breastfeeding support in very low-birth-weight infants. Arch Pediatr Adolesc Med, 155 (5), 548553. Piper, S., & Parks, P. L. (1996). Predicting the duration of lactation: evidence from a national survey. Birth, 23 (1), 7-12. Powers, N. G., Bloom, B., Peabody, J., & Clar k, R. (2003). Site of care influences breastmilk feedings at NICU discharge. J Perinatol, 23 (1), 10-13. Premji, S. S., Paes, B., Jacobson, K., & Ch essell, L. (2002). Evidence-based feeding guidelines for very low-birth-weight infants. Adv Neonatal Care, 2 (1), 5-18. Prochaska, J. M., Prochaska, J. O., & Le vesque, D. A. (2001). A transtheoretical approach to changing organizations. Adm Policy Ment Health, 28 (4), 247-261. Prochaska, J. O., & DiClemente, C. C. (1983) Stages and processes of self-change of smoking: toward an integrative model of change. J Consult Clin Psychol, 51 (3), 390-395. Rea, M. F. (2004). [Benefits of br eastfeeding and women's health]. J Pediatr (Rio J), 80 (5 Suppl), S142-146. Register, N., Eren, M., Lowdermilk, D., Hammond, R., & Tully, M. R. (2000). Knowledge and attitudes of pediatric o ffice nursing staff about breastfeeding. J Hum Lact, 16 (3), 210-215. Riskin, A., & Bader, D. (2003). [Breast is best--human milk for premature infants]. Harefuah, 142 (3), 217-222, 237, 236. Rooney, B. L., & Schauberger, C. W. (2002) Excess pregnancy weight gain and longterm obesity: one decade later. Obstet Gynecol, 100 (2), 245-252.

PAGE 96

85 Ryan, A. S., Wysong, J. L., Martinez, G. A., & Simon, S. D. (1990). Duration of breastfeeding patterns established in the hospita l. Influencing factors. Results from a national survey. Clin Pediatr (Phila), 29 (2), 99-107. Schack-Nielsen, L., & Michaelsen, K. F. (2006). Breast feeding and future health. Curr Opin Clin Nutr Metab Care, 9 (3), 289-296. Schanler, R. J., Hurst, N. M., & Lau, C. ( 1999). The use of human milk and breastfeeding in premature infants. Clin Perinatol, 26 (2), 379-398, vii. Scott, J. A., Binns, C. W., Graham, K. I., & Oddy, W. H. (2006). Temporal changes in the determinants of breastfeeding initiation. Birth, 33 (1), 37-45. Shadish, W. R., Cook, T. D., & Campbell, D. T. (2001). Experimental and quasiexperimental designs for gene ralized causal inference Boston: Houghton Mifflin. Siddell, E., Marinelli, K., Froman, R. D ., & Burke, G. (2003). Evaluation of an educational intervention on br eastfeeding for NICU nurses. J Hum Lact, 19 (3), 293-302. Singhal, A., Cole, T. J., Fewtrell, M., & Lucas, A. (2004). Breastmilk feeding and lipoprotein profile in adol escents born preterm: fo llow-up of a prospective randomised study. Lancet, 363 (9421), 1571-1578. Sisk, P. M., Lovelady, C. A., Dillard, R. G., & Gruber, K. J. (2006). Lactation counseling for mothers of very low birth weight infant s: effect on maternal anxiety and infant intake of human milk. Pediatrics, 117 (1), e67-75. Slusher, T., Hampton, R., Bode-Thomas, F., Pam, S., Akor, F., & Meier, P. (2003). Promoting the exclusive feeding of own mo ther's milk through the use of hindmilk and increased maternal milk volume for hos pitalized, low birth weight infants (< 1800 grams) in Nigeria: a feasibility study. J Hum Lact, 19 (2), 191-198. Smith, M. M., Durkin, M., Hinton, V. J., Be llinger, D., & Kuhn, L. (2003). Influence of breastfeeding on cognitive outcomes at age 68 years: follow-up of very low birth weight infants. Am J Epidemiol, 158 (11), 1075-1082. Smithers, L. G., McPhee, A. J., Gibson, R. A., & Makrides, M. (2003). Characterisation of feeding patterns in infants born < 33 weeks gestational age. Asia Pac J Clin Nutr, 12 Suppl S43. Spicer, K. (2001). What every nurse needs to know about breast pumping: instructing and supporting mothers of premature infants in the NICU. Neonatal Netw, 20 (4), 35-41. Stewart-Knox, B., Gardiner, K., & Wright, M. (2003). What is the problem with breastfeeding? A qualitative analysis of infant feeding perceptions. J Hum Nutr Diet, 16 (4), 265-273.

PAGE 97

86 Swanson, V., & Power, K. G. (2005). Initiati on and continuation of breastfeeding: theory of planned behaviour. J Adv Nurs, 50 (3), 272-282. Torgus, J., Gotsch, G., & La Lech e League International. (1997). The womanly art of breastfeeding (6th rev ed.). Schaumburg, Ill.: La Leche League International. Vannuchi, M. T., Monteiro, C. A., Rea, M. F., Andrade, S. M., & Matsuo, T. (2004). [The Baby-Friendly Hospital Initiative a nd breastfeeding in a neonatal unit]. Rev Saude Publica, 38 (3), 422-428. Viggiano, D., Fasano, D., Monaco, G., & Str ohmenger, L. (2004). Breast feeding, bottle feeding, and non-nutritive sucking; eff ects on occlusion in deciduous dentition. Arch Dis Child, 89 (12), 1121-1123. What is best birth control to use after having a baby? (1989). Contracept Technol Update, 10 (10), 1S-2S. Wheeler, J., Chapman, C., Johnson, M., & Langdon, R. (2000). Feeding outcomes and influences within the neonatal unit. Int J Nurs Pract, 6 (4), 196-206. Wheeler, J. L., Johnson, M., Collie, L., Sutherland, D., & Chapman, C. (1999). Promoting breastfeeding in the neonatal intensive care unit. Breastfeed Rev, 7 (2), 15-18. Wold, A. E., & Adlerberth, I. (2000). Breast f eeding and the intestinal microflora of the infant--implications for protecti on against infect ious diseases. Adv Exp Med Biol, 478 77-93. Wolf, J. H. (2003). Low breastf eeding rates and pub lic health in the United States. Am J Public Health, 93 (12), 2000-2010. Yip, E., Lee, J., & Sheehy, Y. (1996). Breas t-feeding in neonata l intensive care. J Paediatr Child Health, 32 (4), 296-298. Ziemer, M. M., & George, C. (1990). Br eastfeeding the low-birthweight infant. Neonatal Netw, 9 (4), 33-38. Zimmerman, D. R., & Guttman, N. (2001). "Breast is best": knowledge among lowincome mothers is not enough. J Hum Lact, 17 (1), 14-19.

PAGE 98

87 BIOGRAPHICAL SKETCH Dr. Roberta Gittens Pineda received her doctor of philosophy degree at the University of Florida. She received her B achelor of Science in occupational therapy at the Florida Agricultural and Mechanical Univ ersity in 1992 and ach ieved her Master of Health Science degree at the University of Florida in 1994. She has worked as an occupational therapist, primarily in the inpati ent pediatric setting, a nd has been a lecturer at University of Florida as well as Washington University. Dr. PinedaÂ’s primary clinical setting is the neonatal intensive care unit, where she most recently has specialized in treatment of feeding and swallowing problems in these complex, medically fragile infants. In addi tion, she suffered premature labor with her 3 pregnancies and learned, first hand, the difficulties associated with being a mother of an infant hospitalized in the neonatal intensive care unit. Dr. Pineda decided on her di ssertation topic following an invitation by the chief of neonatology at Shands Hospital to be part of a March of Dimes Advisory Committee. The University of Florida at Shands Hospita l had decided to participate in a program sponsored by the March of Dimes aimed at making the unit more developmentally supportive and family centered. Through this project, she joined efforts with a pediatrician to design a plan aimed at making the neonatal intensive care unit (NICU) more breastfeeding friendly. As she had fre quently assisted mothers with breastfeeding in the unit as part of occupa tional therapy intervention and had struggled with the issues of nursing in the NICU when her own s on was born at 29 weeks gestation, she found

PAGE 99

88 herself very passionate about fo stering changes and quickly d ecided to focus her research around the changes that were set to be made. Dr. Pineda’s disserta tion topic is entitled “Breastfeeding Practices in the Neonatal Intensive Care Unit before and after an Intervention Plan”. Dr. Pi neda has thoroughly enjoyed the research process and looks forward to a career in re search and teaching.