Group Title: cumulative effects of bupivacaine epidural anesthesia and obstetric variables on neonatal behavior /
Title: The Cumulative effects of bupivacaine epidural anesthesia and obstetric variables on neonatal behavior /
Full Citation
Permanent Link:
 Material Information
Title: The Cumulative effects of bupivacaine epidural anesthesia and obstetric variables on neonatal behavior /
Physical Description: xi, 171 leaves : ; 28 cm.
Language: English
Creator: Sepkoski, Carol Marie, 1951-
Publication Date: 1984
Copyright Date: 1984
Subject: Obstetrical pharmacology   ( lcsh )
Anesthesia in obstetrics   ( lcsh )
Newborn infants   ( lcsh )
Psychology thesis Ph. D
Dissertations, Academic -- Psychology -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis (Ph. D.)--University of Florida, 1984.
Bibliography: Bibliography: leaves 134-151.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Carol Marie Sepkowski.
 Record Information
Bibliographic ID: UF00099595
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000468626
oclc - 11627919
notis - ACN3312


This item has the following downloads:

PDF ( 6 MBs ) ( PDF )

Full Text








This research was supported by grants from the Spencer

Foundation and the March of Dimes. It was conducted while I

was a research associate in the Child Development Unit, at

Children's Hospital, Boston, Massachusetts.

I am deeply indebted to Dr. T. Berry Brazelton for the

opportunity to do research at the Child Development Unit and

for the depth and quality of the training I received while

working there. Berry's model of child development and his

Neonatal Behavioral Assessment Scale have been critical to

my work. His willingness to ponder over research questions

and share his insights into baby behavior made my years at

the Unit especially meaningful and enjoyable.

The contributions of my good friend, Dr. Barry Lester,

to my graduate education have been innumerable. Barry has

been my inspirator; he both started me down the path of

neonatal studies and taught me much of what I know about

conducting research. His guidance throughout this project

was invaluable. Whenever the going got rough, Barry was

always available with lots of new ideas, optimism and humor.

His continual enthusiasm and encouragement kept me plugging

along. I can't thank Barry enough for the opportunities he

has made available to me and for his part in making my

graduate career a rewarding experience.

I would also like to express my sincere appreciation to

my chairperson, Dr. Keith Berg, who provided me with some

continuity between my first and last stints at the

University of Florida. Keith's excellent advice and

enthusiasm helped get me through all of my requirements. He

patiently supported my unusual approach to graduate work and

willingly arranged committee meetings from San Francisco to

Providence and from Boston to Florida.

I feel fortunate to have had the input of Drs. Pat

Ashton, Yvonne Brackbill, Jacque Goldman and Pat Miller

while serving on my doctoral committee. Each member

contributed to my developmental outcome with many helpful

suggestions and with much interest in my work.

Dr. Gerry Ostheimer provided me with many of the

resources to do the project, including the opportunity to

present my work to anesthesiologists (in the Grand Tetons!).

I would like to express my sincere gratitude to him for all

he taught me about obstetric anesthesia and perinatal

pharmacology and for his continued support of my project.

Many thanks go to Dr. Joel Hoffman for diligently

performing the data analyses and to both Joel and Gaye for

their friendship and for typing my dissertation, even in the

face of sleepless nights with their new baby.

Jude Morrison persevered for two years to recruit

babies for me. She patiently stood by the deliveries of all

babies in the project. A woman of many trades, Jude drew

blood samples, administered Brazeltons and had a knack for

recruiting subjects who lived near the ocean. She deserves

many thanks for doing a marvelous job.

The labor of collecting data was also made possible

with the assistance of Nancy Poland and Harri Thibeault.

Collette LaVoie analyzed blood samples. I am grateful to all

of them for their part in making this project a reality.

I would also like to express my appreciation to the

staffs of labor and delivery and of the newborn nurseries at

the Brigham and Women's Hospital, Boston, Massachusetts, for

their full cooperation with the project. Of course, my

gratitude is extended to all of the families who so

willingly participated in this study and allowed me to share

part of a very special time with them, i.e., the first month

of their babies' lives.

Now I'd like to thank many friends for providing a

constant source of support and interest throughout the

duration of my endeavors.

Dr. Freda Rebelsky first led me astray at a young age

down the road of developmental psychology. Much appreciation

goes to her for the tremendous influence she has had on my


Thanks go to my two good friends and colleagues, Drs.

Cindy Garcia Coll and Fonda Eyler, who I started this whole

process with and couldn't have finished without (too bad I

didn't finish with!). To another friend and psychologist,

Dr. Milt Kotelchuck, my thanks are extended for his many

thoughtful suggestions and for always telling me I could do it.


Several special friends have been a part of my women's

group at one time or another over the past 12 years. They

have always been excited about my work and career. Our many

conversations regarding both those untraditional and

traditional aspects of our lives as women have helped me to

grow as an individual and given me a better understanding of

much of what I've been doing. Thanks go to Annie, Brandy,

Cindy, JoAnne, Kathy, Lucy, Patty and Priscilla for their

invaluable friendships. Much appreciation is extended to

them and to my other long-time friends for being warm,

supportive and always interested--David, Henry, Jan, Karl

(both of you), Molly, and Peter F.

I can't forget to thank the numerous babies born to

these friends during the final stages of my graduate school

career for cheering me up when I needed it and reminding me

that there would be plenty more work on down the road--Abbe,

Adrian, Anna T-C., Anna C-K., Devon, Jan Jr., Luke,

Nataniel, Nicolas and Silas.

Thanks go to my other friends at the Child Development

Unit who I have not yet mentioned, Debbie, Kate, Kevin and

Zach, for their help and caring.

Much appreciation also goes to my friends in the second

round of graduate school who helped make my stay in

Gainesville more enjoyable-- Bob, Darlene, Dee Jay, Fonda,

John, Kim, Maria, Marite, Nancy, Pippa, Terry, and Vern.

Finally, extra special thanks are due to Peter and his

children, Johanna and Nicole, who took good care of me

throughout this endeavor giving me their unending love and

support. Thanks go to them for understanding me and standing

by me through long periods of unavailability in New

Hampshire and Florida. I couldn't have made it without them.

Lastly, to the Sepkoski family, thanks are extended to

my siblings, Diane, Mary and, especially, Jack, for their

contributions to making this a possible and meaningful

experience for me. Thanks also go to Maureen for always

being interested and to David for teaching me a lot about

growing up.

I wish to dedicate this work to my parents, Sally and

Joe Sepkoski. It was their continued faith in me which

allowed me to believe in myself and make it through graduate

school. I can't thank them enough for everything.


ACKNOWLEDGMENTS. ........... .................... ii

ABSTRACT................... ................... ix

INTRODUCTION. .......... ......................... 1

Types of Obstetric Medication............... 6
Perinatal Pharmacology ......................9
Empirical Studies of the Behavioral
Effects of Obstetric Medication........... 14
Lack of Significant Behavioral Effects......45
Methodological Considerations................54
Purpose of the Study ....................... 65

METHODS ..................................... 72

Study Site .......... .......... ............ 72
Subject Recruitment ........................ 72
Sample. ........... ......................... 75
Instruments. .......... ...................... 76
Procedure. ............ ..................... 82

RESULTS.. ....... .................... ......... 85

Characteristics of the Sample................85
Relationship Among Variables:
Bivariate Correlations ...................94
Cumulative Effects of Bupivacaine and
Obstetric Variables: Multiple
Regression ................... ............96
Comparisons of Epidural and
Nonmedicated Groups ..................... 102

DISCUSSION ............ ........... .......... 109

Cumulative Effects of Bupivacaine
and Other Obstetric Variables............109
Comparison of Bupivacaine Epidural
and Nonmedicated Groups..................118
Comparisons with other Bupivacaine
Epidural Studies. ....................... 120
Conclusions. .......... ...................... 123
Implications ............................... 129

REFERENCES .......... .......................... 134


1 SUMMARY OF OBSTETRICIANS................. 152
2 PARENTAL CONSENT FORM ....................153
ASSESSMENT SCALE ITEMS................... 156
DURING BRAZELTON EXAM. ................... 157
SCORING CRITERIA ........................158
6 MATERNAL INTERVIEW.......................159
LABOR REMOVED ............................165
LABOR REMOVED............................ 166

BIOGRAPHICAL SKETCH................. ........... 167


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



Carol Marie Sepkoski

April 1984

Chairperson: W. Keith Berg

Major Department: Psychology

This study was designed to investigate the cumulative

behavioral effects of maternal epidural anesthesia with

bupivacaine and the constellation of nonoptimal obstetric

conditions found in a relatively healthy sample of newborns

with varying birth histories. It was hypothesized that

obstetric variables which may influence the rate of

placental transfer of local anesthetics potentiate the

effects of bupivacaine on behavior over the first month of


The sample was 60 healthy fullterm infants selected

randomly from a population of mothers delivering without

pain-relieving medication or with bupivacaine epidural

anesthesia only. The Brazelton Neonatal Behavioral

Assessment Scale was administered by trained examiners at 3

hours and at 3, 7, and 28 days after birth.


The variables entered as predictors into stepwise

multiple regressions performed on six Brazelton clusters

were dose of bupivacaine, hours of administration of

oxytocin, umbilical arterial pH, ponderal index, maternal

weight:height ratio, lengths of the first and second stages

of labor and Obstetric Comolication Scale score. Eleven of

24 multiple correlations were significant. More behavioral

clusters were predicted on the first day than on any other

testing day. The motor and autonomic regulation clusters

were predicted more frequently than the other clusters over

the entire month.

All independent variables contributed to some of the

regressions. Dose of bupivacaine was the most important

predictor followed by hours of oxytocin. Comparisons of

multiple and bivariate correlations indicated that the

effects of bupivacaine on behavior were potentiated by the

other nonoptimal obstetric conditions.

The sample was divided into epidural and nonmedicated

groups matched on obstetric variables. Analyses of variance

indicated that the two groups differed in Brazelton

performance on each testing day and on six of seven

clusters. Although all babies performed within a normal

range of behavior, the nonmedicated group consistently

performed better than the epidural group.

Post hoc analyses illustrated possible indirect effects

of bupivacaine. The epidural group had more force

deliveries, longer labors and more oxytocin than the

nonmedicated group. Also, these mothers spent significantly

less time with their infants while in the hospital.

Limitations of the study design and recommendations for

future research were discussed.


Over the past twenty-five years, a new picture of the

capabilities of the newborn has emerged. With this, the

neonatal period has become a focus of developmental


During the first half of this century, the neonate was

conceptualized as a passive and helpless organism who could

not see or hear (Kessen, 1963). It was thought that newly

born babies could not learn and that their behavior was

totally reflexive. Scientists and physicians thought that

infants were born essentially alike, having been immune to

environmental forces inutero. After birth, the infant was

considered to be a behavioral tabula rasa, ready to be

completely shaped by the outside world.

The focus of the earliest infant research was to

document normative behavioral development (Bayley, 1933;

Buhler & Hetzer, 1935; Gesell, 1928). Illuminating the

abilities of the baby led to an explosion of research in the

area and thus, the newly emerging picture of the newborn.

Scientists discovered that neonates could not only see and

hear but that they could actively select input from the

environment by the mechanisms of habituation and orientation

(Tronick, Als, & Brazelton, 1979). They seemed to prefer

and respond differently to animate stimuli such as the human

face and voice than to inanimate stimuli (Brazelton,


Koslowski, & Main, 1974; Carpenter & Stechler, 1967; Condon

& Sander, 1974; Eimas, Siqueland, Juzczyk, & Vigorito,

1971). It was also discovered that newborns could be

conditioned not only to modify their sucking in response to

stimuli (Kron, 1966; Lipsitt & Kaye, 1964) but that they

could condition adults to modify responses to their sucking.

Kaye and Brazelton (1971) investigated the effects of

mothers' jiggling behavior on infants' sucking responses

during feeding. They found that the jiggling did not

stimulate infants to resume sucking but that the infants

stopped sucking to provoke the stimulation from their

mothers. Blauvelt and McKenna (1961) found that rooting is

not merely a reflexive action of the infant's but that it is

used to change or to control the mother's activity in order

to elicit help attaining the nipple. Thus, the picture of

the normal newborn which has emerged from this research is

one of a competent organism who is skilled, selective,

socially influential and capable of actively interacting

with and making demands on its environment (Lester, 1979).

Standardized assessment tools which were developed with

the growth in understanding of the complexity of the

behavioral repertoire of the neonate have led to the

recognition of the existence of individual differences in

behavior at birth. The present task of developmental

research is to investigate the origins of these differences

and their effect on developmental outcome (Kopp & Parmalee,



The study of the relationship between perinatal events

and behavioral outcome has become a major area of importance

in developmental research. The individual goes through more

physiological, behavioral and environmental changes during

the perinatal period than at any other time in life. Since

the neonate is still morphologically and functionally

immature, it is especially vulnerable to insult from the

host of variables surrounding the birth process. Advances

in perinatal technology have reduced the mortality rate for

mothers and babies so that now it has become increasingly

important to investigate the perinatal conditions which may

be related to morbidity and poor developmental prognosis

(Lipsitt & Field, 1982).

One variable introduced into the delivery room with

medical technology was pain relieving medication. Although

various forms of analgesia had been used for centuries,

anesthesia was first used during labor and delivery by

Simpson, a Scottish obstetrician, in 1847 (Brackbill, 1979).

Childbirth without pain was such an attractive option that

the use of obstetric medication spread rapidly. Today, its

use is so widespread that it is considered part of routine

medical practice.

When Simpson first introduced the use of anesthesia to

obstetrics, scientific opinion still promoted the idea that

the uterus and placental barrier protected the fetus from

foreign substances. However, as early as 1889, Hirsh had

discovered that many substances, including opiates and


ether, flow from the mother's bloodstream through the

placenta and into the baby (MacFarlane, 1977). It was not

until the thalidomide tragedy of the early 1960's that the

potential consequences of placental transfer of drugs were

fully recognized. The use of this tranquilizer during the

second month of pregnancy was implicated as the cause of

severe deformities of infants' limbs (Moya & Thorndike,

1963). Since this finding, the teratogenic effects of many

pharmacologic agents have been studied. The exposure of

rodents to inhalant anesthesia early in gestation has been

related to degeneration of the liver (Chang, Dudley, Lee, &

Katz, 1975), pathological development of the central nervous

system (Chang, Dudley, Lee, & Katz, 1974; Quimby,

Aschkenase, Bowman, Katz, & Chang, 1974), morphologic

changes (Lund, Owen, & Linde, 1981; Smith, Gaub, & Moya,

1965) and fetal death (Corbett, Cornell, Endres, & Millard,

1973). Studies of teratogenic action in humans are limited

by ethical considerations to naturally occurring situations.

Several investigators have found a higher incidence of birth

defects among children of operating room personnel (Cohen,

Brown, Bruce, Cascorbi, Corbett, Jones, & Whitcher, 1974;

Corbett, Cornell, Endres, & Lieding, 1974; Knill-Jones,

Newman, & Spence, 1975). Chronic exposure to low levels of

inhalation anesthesia has also been related to infertility

and miscarriage (Cohen, Bellville, & Brown, 1971; Knill-

Jones, Moir, Rodriguez, & Spence, 1972).


Coyle, Wayner and Singer (1975) have suggested that

behavioral changes in the offspring of exposed mothers might

be a more sensitive indicator of the teratogenic action of

drugs. Behavioral teratogenic deficits may be induced by

smaller doses of a drug than those necessary to produce

morphological abnormalities. These effects may be further

mediated through the developing relationship between the

mother and baby, both of whom are recipients of the

medication (Brazelton, 1971; Coyle et al., 1975).

Unlike most other perinatal variables or potential

teratogens, exposure to drugs during delivery is under the

direct control of the obstetrician and/or anesthesiologist.

The safety of differing agents can be monitored and if

harmful effects are found, their use can be discontinued or

their dosage decreased (Scanlon & Hollenbeck, 1982). With

this in mind, the assessment of the neonatal effects of

drugs used routinely during labor and delivery deserves

special attention. The present study focuses on the

potential behavioral effects of obstetric medication on the

newborn. Before reviewing the literature which has

investigated these effects, I will first briefly discuss the

types and routes of parturitional medicine and the

principles of perinatal pharmacology.


Types of Obstetric Medication

Medications given to relieve pain or anxiety during

labor are divided into two general categories:

premedication and anesthesia. Either may be given alone or

in conjunction with a drugs) from the other general


Premedication is usually administered early in labor

before anesthesia. It includes the narcotic-analgesics,

sedative-hypnotics and tranquilizers. These medications may

be administered orally, intravenously or intramuscularly.

Narcotics include both the opium alkaloids such as

morphine and codeine and their synthetic analogues such as

meperidine and alphaprodine. They are used as analgesics to

reduce the sensation of pain and work by depressing central

nervous system (CNS). When taken in large doses, they also

produce euphoria and drowsiness. Narcotics are commonly

administered intramuscularly early in labor and may be used

as the only form of analgesic or, more frequently, in

conjunction with anesthesia.

Sedatives include the barbiturates such as

pentobarbitol and secobarbitol and nonbarbiturates such as

scopolamine and diazepam (Valium). They are given to calm

the parturient and may induce drowsiness or sleep. Diazepam

is typically given to relax muscles. High doses of some

sedatives will produce general anesthesia. Like narcotics,

sedatives work by depressing the CNS. They are used much

less frequently today in obstetric practices than they were

in the past.

Tranquilizers most commonly used during parturition

include the phenothiazine derivatives: chloropromazine,

promazine and promethazine, and the benzodiazepines. They

are given to relieve anxiety, prevent nausea and vomiting

during labor, and provide sedation. When given in small

doses, tranquilizers do not work as general CNS depressants.

They produce a alpha-adrenergic blockade leading to an

epinephrine reversal and norepinephrine unresponsiveness

(Cohen & Olsen, 1970). Tranquilizers are frequently used

today both in early and later stages of labor.

Anesthesia is the second major category of medication

commonly used during childbirth. It produces not only

relief from pain, but a loss of sensation. General

anesthesia depresses the CNS and produces unconsciousness.

The agents used to produce this effect include inhalants

such as nitrous oxide, enflurane, isoflurane and

methoxyflurane and intravenous agents such as thiopental and

ketamine. Ketamine has also been found to produce

hallucinations (Meer, Downing, & Colman, 1973). General

anesthesia is used today mainly for cesarean deliveries. A

combination of an intravenous drug with an inhalant is often

used so that the required dose of each can be reduced. The

use of this type of anesthesia in vaginal deliveries is

reserved for emergencies and for women who desire to be

unconscious during the delivery.


Local anesthetics are being used more and more

frequently in obstetric practices. They work by blocking

the transmission of sensory impulses along nerve fibers.

Local anesthetic agents include the ester compounds such as

procaine, tetracaine and chloroprocaine and the amide

compounds such as bupivacaine, lidocaine, mepivacaine and

etidocaine. These drugs are administered via different

routes to provide either numbness at the site of injection

or regional analgesia. For local infiltration, the agent is

injected directly into the area to be anesthetized. During

vaginal deliveries, this area is the perineum. A local is

injected into this site immediately before delivery in

preparation for an episiotomy or immediately after for

repair of a tear or of the episiotomy.

Local anesthetic agents are also administered

regionally to block nerve impulses leading to the lower part

of the body. They are used in this way for both vaginal and

cesarean deliveries. Regional anesthesia is classified

according to the space in the woman's back into which the

agent is injected and at which point the nerve impulses are

blocked. This includes epidural, or caudal, spinal, or

saddle, pudendal and paracervical blocks. These injection

sites differ in the amount and concentration of the local

anesthetic agent which is needed to provide effective

analgesia, in the amount of time it takes for the agent to

be effective and in the region of the body which becomes

anesthetized. Therefore, the route chosen for the


administration of the drug depends in part on the time

during labor at which pain relief is desired and how quickly

it is needed.

Medication used during parturition for purposes other

than pain relief includes oxytoxics, to induce or augment

labor, vasopressors and vasoconstrictors, to retard

absorption of local anesthetics and prevent hypotension, and

narcotic antagonists, to avoid respiratory depression

associated with narcotics. These are typically given in

addition to analgesics and/or anesthetics.

Perinatal Pharmacology

The dose and route of administration of a drug, the

time it is given and the properties of the agent all affect

the amount of drug that circulates in the mother and crosses

the placenta, the duration of fetal exposure to the drug and

its concentration in the newborn at delivery. Any drug

administered to the mother by any route eventually enters

her bloodstream. The rate at which it enters specific

tissues depends on the rate of blood flowing through the

tissues and the ease at which the drug molecules pass

through the capillary membranes (Julien, 1978). The brain

receives a greater amount of blood flow than any other

organ. Thus, drugs which can bypass the blood brain barrier

have a sufficient chance to exert their effect on the CNS.

Any substance which is given to affect the CNS will

readily cross the placenta. The placental barrier is

similar to the blood brain barrier. Passage through these


membranes is accessible only to the free unbound form of the

drug. The concentration of free drug in the maternal plasma

is affected by the rate of absorption from the site of

administration, the protein binding capacity of the drug and

its distribution through various tissues. It is also

affected by the rate of uptake by the liver, subsequent

biotransformation and, finally, excretion from the kidneys.

Most drugs enter the placenta passively by simple

diffusion. In other words, they flow from an area of high

concentration to one of low concentration. The rate at

which they pass depends on several properties of the

specific agent: its molecular weight, spatial

configuration, lipid solubility, protein binding capacity

and degree of ionization. Small nonionized molecules which

are fat soluble cross placental membranes most quickly.

Theoretically, increasing the protein binding capacities of

molecules should decrease the amount of free drug available

for placental transport. However, drugs with higher protein

binding capacities are also more lipid soluble. The

solubility counteracts the binding capacity making much of

the drug readily available to pass directly into the

placenta (Ostheimer, 1977).

Another variable which affects the rate of drug

transfer is the difference between maternal and fetal acid-

base statuses. According to the law of mass action, ionized

molecules accumulate on the side with the lower pH (Brown,

Bell, & Alper, 1976). Normally fetal blood is slightly more


alkaline than maternal blood. However, local anesthetics

have a depressant effect on the heart and blood vessels and

may cause maternal hypotension (de Jong, 1970). This drop

in maternal blood pressure reduces the oxygen content of the

fetal blood which lowers the pH and causes acidosis. The

alkaline drugs such as local anesthetics become trapped in

fetal circulation after crossing the placental barrier. The

increased acidity of the fetal blood leads to increased

ionic dissociation which reduces the drug's diffusibility

through the placenta and back into maternal circulation.

Since local anesthetics are myocardial depressants, a

vicious cycle is established with the increased drug

concentration causing further acidosis and the acidosis

causing a further increase of the drug in the fetal blood

stream, etc. (Brown et al., 1976).

Several other maternal conditions may influence the

rate of transfer of drugs to the baby. Hypoproteinemia

allows a greater amount of drugs which normally bind to

maternal protein to cross the placenta. Liver diseases

reduce the capacity of the mother to metabolize drugs. This

results in an increased concentration of the drug in the

maternal bloodstream and, thus, the transfer of a greater

amount of the drug to the fetus. Obesity creates a

condition in which lipid soluble drugs which cross into the

fetal bloodstream quickly return to the mother. This occurs

because fat molecules absorb these lipid soluble drugs

leaving the concentration of free drug circulating in the


mother lower than that in the fetus (Cohen & Olsen, 1970;

Ralston & Shnider, 1978).

The placenta itself may have a role in altering the

rate of drug transfer. Although little is known about

changes which take place within this organ, it is possible

that some compounds are metabolized there. This would cause

a change in the structure of the drug molecule and,

consequently, in its rate of entry into fetal circulation

(Cohen & Olsen, 1970). The placental vasculature and

metabolic capacities may also be directly altered by the

drug. This could, in turn, further affect the rate of drug

transfer. It could also affect fetal functions, thereby

lowering the nutritional status of the neonate (Dubowitz,

1975; Scanlon, 1974).

Maternal diseases which may affect the rate of drug

transfer by altering the placental vascular bed are toxemia,

diabetes, eclampsia and chronic hypertension. The increased

permeability of the placenta caused by these illnesses may

speed up the exchange of drugs with the fetus.

The amniotic fluid has been suggested to be another

likely route of drug transfer (Cohen & Olsen, 1970). Since

the pH of the amniotic fluid is significantly lower than the

pH of maternal or fetal blood, high concentrations of some

drugs may pass into the fluid and subsequently be ingested

by the fetus. However, very little is known about this

process as a possible means of drug exchange.


Once the drug has entered the fetal bloodstream, the

inherent vulnerability of the immature organism makes both

the fetus and the neonate more susceptible than the adult to

the effects of the drug. Their immature livers lack many of

the enzymes which are necessary for the metabolism of drug

molecules. Also, their kidneys are not fully functional so

that their ability to excrete drugs is limited. The

elimination of drugs in the newborn is further slowed down

by hypothermia. The common difficulty neonates have with

temperature regulation is exacerbated in the presence of

acidosis. Hypothermia is also pronounced in premature and

malnourished infants (Morishima, Mueller-Heubach, & Shnider,


Fetuses and neonates also have different membrane

permeabilities, regional circulation, tissue affinities and

plasma protein binding capacities than adults (Alper, 1979).

The brain is the area most vulnerable to these factors. The

blood brain barrier of the young organism is poorly

developed allowing for greater diffusion of drugs into the

CNS. Additionally, there is more blood flow to the brain

during labor and delivery than at other times because of the

decrease in fetal P02 and the increase in PC02 (Julien,

1978). The use of anesthesia has been related to an

increase in the length of labor (Berges, 1971). Increasing

the length of labor to the point where it becomes

dysfunctional can lead to fetal asphyxia (Aleksandrowicz,

1974). Fetal asphyxia produces a release of catecholomine


which causes a shunting of blood to the brain in an effort

to increase its oxygenation. Asphyxia also aggravates

acidosis thereby causing a further accumulation of the

anesthetic in the bloodstream. The protein binding capacity

of the drug is disrupted by the asphyxia and the vascular

permeability of the brain is increased.

In summary, the immaturity of the brain at birth makes

the neonate especially vulnerable to the effects of any

substance entering its bloodstream inutero. Vulnerability

is increased by an insufficient ability to metabolize and

excrete drug molecules, by a more permeable blood brain

barrier and by a greater blood flood to the brain during

gestation and parturition. Susceptibility to obstetric

medication is further increased by common complications of

labor and delivery such as asphyxia, acidosis and


Empirical Studies of the Behavioral Effects of Obstetric

Awareness of the morphological and functional

immaturity of the neonate as well as of the remarkable

capabilities of the neonate has led to research focusing on

the behavioral effects of obstetric medication. Over the

past ten years, there have been fifty studies published in

the child development, pediatric and anesthesiology

literature relating the effects of drugs used during labor

and delivery to infant behavior. Although medication has

been used to relieve the pain of childbirth since the mid


1800's, there are only nine studies about its effect on

infant behavior reported prior to 1971 and none before 1948.

The literature on the effects of maternal medication

has varied in both the independent and dependent parameters

under investigation. The drugs, doses and routes of

administration which are popular to use in obstetric

practices today have changed over the years as a result of

dispelling the myth that drugs do not cross the placenta and

the increasing focus on neonatal outcome. New technology to

measure the concentration of drugs in the bloodstream and

the development of different drugs have also contributed to

the changes. In general, there has been a decrease in the

use of centrally depressant narcotics, sedatives and general

anesthesia and an increased use of local or regional

anesthesia so that the mother can remain alert and aware

during labor and delivery (American Academy of Pediatrics

Committee on Drugs, 1978). However, overall, the use of

medication to relieve anxiety and pain during labor is on

the increase (Brackbill, 1979).

Some of the earliest literature investigated the

behavioral effects of heavier doses of sedatives than are

typically used today (Brazelton, 1961; Hughes, Ehemann, &

Brown, 1948; Hughes, Hill, Green, & Davis, 1950). More

recently, varying agents frequently used for epidural

anesthesia such as mepivacaine, lidocaine and bupivacaine

have been studied. As the rate of cesarean section

deliveries has increased, so have investigations of


different agents and routes of administration used to

alleviate pain during surgery. Premedication has continued

to be used and studied in both vaginal and cesarean

deliveries over the years, especially meperidine. Another

common parameter in the behavioral studies of obstetric

medication has been the potency score. Many studies have

not differentiated routes and agents of premedication or of

anesthesia in their analyses but have lumped them together

and given them a clinical rating from high to low depending

on the dose and/or characteristics of the drugs and routes

of administration.

The literature also varies in the measures of behavior

used to investigate the effects of obstetric medication and

in the age of testing. Most studies have examined behavior

over the first three days of life. Many of them have

followed subjects to seven or ten days of life but only a

few have gone beyond this age. The standardized exams which

have been used most frequently as the dependent parameters

in these studies are the Brazelton Neonatal Behavioral

Assessment Scale (Brazelton, 1973) and the Scanlon Early

Neonatal Neurobehavioral Scale (Scanlon, 1974). The

Brazelton Scale was designed to measure the capabilities the

infant has for interacting with a caregiver over the first

month of life. It consists of 27 behavioral items and 17

reflexive items administered in an order determined by the

individual infant's state patterns. Special care is taken

by the examiner to elicit the best performance the baby is


capable of (for a more complete description of the Brazelton

Scale, see the Methodology chapter). The literature on

obstetric medication has summarized Brazelton items in

several different ways for the purpose of data analyses.

These include using item-by-item analysis, factor analysis,

"marker" variables, elicited and emitted item summary

scores, the four a priori summary scores of Als, Tronick,

Lester, and Brazelton (1977) and the seven conceptual

clusters of Lester, Als and Brazelton (1982). Some studies

have included modifications of the Brazelton Scale, e.g.,

the Kansas version (Horowitz, Sullivan, & Linn, 1978) which

scores both the infant's best and average performance on

several items. It also includes six additional summary

scores. The Mother's Assessment of the Behavior of her

Infant (Field, Dempsey, Hallock, & Shuman 1978) which

includes the mother's scores of her baby's performance

during the administration of the Brazelton exam has also

been used in some studies.

The Scanlon Scale has been the exam used most

frequently in studies of obstetric medication, particularly

among those appearing in the journals of anesthesiology. It

consists of ten items taken from the Brazelton Scale but

administered in a predetermined order. The Scanlon Scale

was designed to be a short and easily administered exam

which could be used by physicians in the first few hours of

life. The items were chosen to screen specifically for the

effects of obstetric medication on higher CNS functioning.


Standardized assessment scales which have been used

less frequently are the Graham (Graham, Matarazzo, &

Caldwell, 1956) and the Graham/Rosenblith (Rosenblith, 1961;

1975) Behavioral Examination of the Neonate and Prechtl and

Beintema's (1964) and Parmalee's (1974) neurological exams.

Most recently, the Neurologic and Adaptive Capacity Score

(Amiel-Tison, Barrier, Shnider, Levinson, Hughes, & Stefani,

1982) has been designed as a screening exam to look for the

effects of obstetric medication. It is quicker to

administer than the Scanlon exam, taking 4 minutes to

administer as opposed to 7 minutes for the Scanlon exam, and

focuses on the muscle tone of the newborn.

After the neonatal period, the Bayley Scales of Infant

Development (Bayley, 1969) and the Denver Developmental

Screening Test (Frankenburg & Dodds, 1967) have been used to

evaluate mental and motor development during the first year.

The Carey Infant Temperament Questionnaire (Carey, 1970) has

also been used during this time period to assess maternal

perception of infant temperament. At one year of age, the

Catell Infant Intelligence Scale (Catell, 1960) has been


Other measures of behavioral outcome which have been

investigated include habituation to visual and auditory

stimuli, visual fixation, sucking, electroencephalographic

patterns (EEGs), organization of sleep states and measures

of mother-infant and father-infant interaction.


In order to summarize the findings of the literature

and examine what aspects of a baby's behavior may be

affected by obstetric medication, it is possible to group

most of the behavioral measures studied into the seven

conceptual clusters described by Lester, Als and Brazelton

(1982). These clusters were designed to summarize the

Brazelton Scale items. They include habituation,

orientation, motor performance, range of state, regulation

of state, autonomic regulation and reflexes. Although the

behavioral measures described previously often differ from

each other in their administration and scoring of particular

items, they can generally be grouped into these seven areas

of behavioral functioning. It is important to remember that

not every study has assessed behaviors which fall into all

seven clusters. For instance, the Scanlon exam does not

assess autonomic functioning or state dimensions. Even

studies using the Brazelton exam frequently do not assess

habituation due to babies beginning the exam in an

inappropriate state for the administration of the

habituation items. Also, the reflex scores are frequently

not reported because they are not included in the Als et al.

(1977) four a priori dimensions.

A summary of the studies which have found significant

effects of obstetric medication on behaviors which fit into

the seven clusters can be seen in Table 1.

Table 1. Summary of Studies Finding Significant Effects of

Obstetric Medication
Seven Clusters.

on Behaviors Fitting into



Aleksandrowicz &
Aleksandrowicz 1974
Belsey et al. 1981
Bonta et al. 1979
Borgstedt & Rosen 1968
Brackbill et al. 1974
Brackbill 1976
Brazelton et al. 1979
Brown et al. 1975
Clark et al. 1976
Conway & Brackbill 1970
Corke 1977
Friedman et al. 1978
Hodgkinson et al. 1976
Hodgkinson et al. 1977
Hodgkinson et al. 1978a
Hodgkinson et al. 1978b
Hodgkinson et al. 1978c
Hollmen et al. 1978
Horowitz et al. 1977
McGuinness et al. 1978
Meis et al. 1978
Merkow et al. 1980
Moreau & Birch 1974
Murray et al. 1981
Nesheim et al. 1979
Palahniuk et al. 1977
Richards & Bernal 1972
Rosenblatt et al. 1981
Scanlon et al. 1974
Scanlon et al. 1976
Standley et al. 1974
Stechler 1964
Tronick et al. 1976
VanderMaelen et al. 1975
Wiener et al. 1979
Woodson &
DaCosta-Woodson 1980

d+ d

a a

a a,d

b b

a a

a a a a

d d

b b b

b b b


* Clusters: III=Motor; I=Habituation; II=Orientation;
VII=Reflexes; VI=Autonomic Regulation; IV=State Range;
V=State Regulation.
+ Obstetric Medication: a=premedication;
b=local anesthesia; c=general anesthesia;
d=mixed analgesia and anesthesia.


Motoric Functioning

Significant effects of maternal medication have been

found most frequently on neonatal motoric functioning. In

general, babies of medicated deliveries seen to be more

hypotonic, less active, less able to lift their heads and

have less control over their movements. Motoric behaviors

have been found to be depressed up to 6 weeks of age.

Brackbill, Kane, Manniello and Abramson (1974a; 1974b)

found that infants whose mothers had received meperidine

were less able to swipe at a cloth placed over their eyes on

the second day of life than a group whose mothers had not

had meperidine. Also using the Brazelton scale, Belsey,

Rosenblatt, Lieberman, Redshaw, Caldwell, Notariani, Smith

and Beard (1981) found that the greater the exposure of the

infant to meperidine as determined by the umbilical cord

blood concentration and the time before delivery that the

drug was administered, the more depressed motoric behaviors

were at 3 days and 1, 3 and 6 weeks of life. Spontaneous

motor activity during the Prechtl exam was also depressed on

the first day. Decreased motor tone shortly after birth was

observed by Richards and Bernal (1972) in babies whose

mothers received a compound which included both meperidine

and an antihypotensive and by Corke (1977) in babies of

meperidine and promethazine deliveries.

Effects of meperidine on motor functioning have also

been found by examining the reversal of effects when

naloxone is administered to the baby or mother. Naloxone is

a narcotic antagonist given to reverse depression caused by

meperidine or other narcotics. Using the Scanlon exam at

varying times over the first day of life Clark, Beard,

Griefenstein and Barclay (1976) and Bonta, Gagliardi,

Williams and Warshaw (1979) found a meperidine group of

babies without naloxone to have lower tone than a group

receiving naloxone. However, both groups had lower tone

than a group without meperidine. Hodgkinson, Bhatt, Grewal

and Marx (1978a) found that the administration of naloxone

improved motor scores at 2 hours after birth but not at 4 or

24 hours, due to the fact that length of time the narcotics

were active exceeded the effectiveness of the short acting


Local anesthetics have also been found to depress motor

behavior. This may be dueto the direct action of the drug

on the neuromuscular function itself. Several experimental

studies have shown that local anesthetics influence

neuromuscular transmission and function (Bianchi & Bolton,

1967; Usubiaga & Standaert, 1968).

Effects of bupivacaine epidurals were found on motoric

items of the Brazelton exam by Murray, Dolby, Nation and

Thomas (1981) on day 1, Wiener, Hogg and Rosen (1979) on

days 1 and 2 and Rosenblatt, Belsey, Lieberman, Redshaw,

Caldwell, Notarianni, Smith and Beard (1981) up to 6 weeks.

Tronick, Wise, Als, Adamson, Scanlon and Brazelton (1976)

and Scanlon, Brown, Weiss and Alper (1974) found that motor

performance on the Scanlon exam was depressed more during


the first half day of life by mepivacaine and lidocaine

epidurals than by spinals, locals or no medication. Tronick

et al. also found decreased activity during the Brazelton

exam on day 10.

Nesheim, Lindbaek, Storm-Mathisen and Jenssen (1979)

found low tonus over the first three days in lidocaine and

bupivacaine paracervical block groups and in a lidocaine

local group. Both groups had also received diazepam and

nitrous oxide.

General anesthetics used during vaginal deliveries or

cesarian sections have also been found to depress motor

performance on the Scanlon exam. In separate studies,

Hodgkinson, Wang, and Marx (1976), Hodgkinson, Marx, Kim and

Miclat (1977) and Hodgkinson, Bhatt, Kim, Grewal and Marx

(1978b) found that infants of mothers anesthesized with

thiopental had lower scores on days 1 and 2 than those of

deliveries using ketamine. Both drugs depressed motoric

scores more than chloroprocaine epidurals or tetracaine

spinals. Palahniuk, Scatcliff, Biehl, Wiebe and Sankaran

(1977) compared thiopental generals combined with either

methoxyflurane or nitrous oxide to lidocaine epidurals and

found that tone was depressed in both the nitrous oxide and

lidocaine epidural groups. However, Hollmen, Jouppila,

Koivisto, Maatta, Pihlaianiemi, Puukka and Rantakyla (1978)

found hypotonia on the Prechtl-Beintema exam with lidocaine

epidurals only when hypotension occurred following the

initiation of anesthesia. They also found low tone up to


one week among babies whose mothers received intravenous

thiopentol and had hypertension during labor. Bupivacaine

epidurals and tetracaine spinals have also been related to

depressed tonus at four hours among cesarian section infants

(McGuinness, Merkow, & Kennedy, 1978). Conway and Brackbill

(1970) found a decrease in muscle tension with increasing

potency of medication on the Graham scale on days 2 and 5

and on the Bayley Motor Index at one month. Standley,

Soule, Copans, and Duchowny (1974) found that poorer motor

performance on the Brazelton exam on day 3 was related to a

greater potency score. The Brazelton motor cluster was

found to be lower up to day 10 among a group of infants of

Greek middle-class mothers who had received varying

obstetric medication than among a lower-class group that had

received no medication (Brazelton, Tryphonopoulou, & Lester,

1979). In another cross-cultural study of the effects of

obstetric medication, Horowitz, Ashton, Culp, Gaddis, Levin,

and Reichmann (1977) found babies of light to moderately

medicated mothers in Israel and Uruguay to be poorer on

defensive performance and lower on activity level on day 3

than nonmedicated babies. An American sample which was more

heavily medicated did poorly on several motor items up to

one month of age.


Significant effects of obstetric medication have been

found almost as frequently on the infant's rate of

habituation as on motoric functioning. Dependent measures


which have been used to determine habituation are, most

commonly, a decrement in motoric responses to repeated

presentations of auditory, visual or tactile stimuli, a

decrement of eye blinks, and a change in heart rate

responses. Both the Brazelton and the Scanlon exams include

items which assess the infant's ability to habituate to

repetitive stimuli.

Brackbill et al. (1974a;1974b) premedicationduring

labor was more closely related to neonates' ability to

inhibit responding than to their ability to respond to

stimuli. The authors investigated the effects of meperidine

on the infant's ability to inhibit motoric responses to

white noise on days 2 and 3. They found that infants whose

mothers did not take meperidine during labor habituated

twice as fast as those who had taken the medication. The

dose and timing of the administration of the drug and the

birthweight of the baby were found to be related to the rate

of habituation. Wiener et al. (1979) found that slower

inhibition of responses to repeated presentations of a

rattle and bell on the Brazelton exam on days 1 and 2 was

related to meperidine usage. Also, Corke (1977) found that

both meperidine and promazine affected rate of habituation

to a pinprick in the the first four hours on the Scanlon

exam. In a sample of infants of cesarean section deliveries

with general anesthesia, Hodgkinson et al. (1978a) found

that meperidine was the variable related to a slower rate of

habituation on the first day. As with the motor items,

naloxine improved habituation at two hours but not later in

the day.

A higher portion of infants in a group with an

obstetric history of maternal epidural anesthesia with

mepivacaine or lidocaine could not inhibit their responses

to a pinprick over the first six hours of life than a mixed

group with either alphaprodine, secobarbital, local or

spinal anesthesia or no medication at all (Scanlon et al.,

1974). However, Tronick et al. (1976) found that groups of

infants with histories of lidocaine locals or alphaprodine

and/or promazine were slower to habituate at 3 days to a

pinprick than a mepivacaine or lidocaine epidural group.

The mean umbilical cord blood concentration of mepivacaine

reported in the Scanlon et al. study was slightly higher

than that in the Tronick et al. study (1.68 vs. 1.58 ug/ml)

but mean concentration of lidocaine in the epidural group

was slightly lower (.48 vs. .55 ug/ml). The actual doses of

maternal drug intake were reported in neither study.

Habituation to items on the Scanlon exam have been

examined with respect to the effects of paracervical blocks.

Nesheim et al. (1979) found decrement to the light, pinprick

and Moro was slower up to 3 days with both bupivacaine and

lidocaine paracervicals than with lidocaine locals.

Randomly assigning mothers who desired paracervicals to

differing agents, Merkow, McGuinness, Erenberg and Kennedy

(1980) found that infants from bupivacaine or chloroprocaine

groups habituated slower at four hours than those from the


mepivacaine group. However, Meis, Reisner, Payne and Hobel

(1978) found decrement to presentations of the light was

faster at 2 hours in a bupivacaine paracervical block group

than in a group with locals, spinals or pudendal blocks.

Infants delivered with general anesthesia were found to

habituate more slowly on day 2 to white noise and to a

somesthetic stimulus than infants delivered without general

anesthesia regardless of what other medication was used

(Moreau and Birch, 1974). Hodgkinson et al. (1976; 1977;

1978b) found that general anesthesia with thiopental among

cesarean sectioned babies was related to slower habituation

on the Scanlon exam on the first two days of life than was

general anesthesia with ketamine. Both agents were related

to worse performance than chloroprocaine epidurals or

tetracaine spinals. The authors found similar results among

infants delivered vaginally, with meperidine further

lowering the scores of babies delivered with general

anesthesia (Hodgkinson et al., 1978c).

Assigning potency scores to varying combinations of

obstetric medication, Conway and Brackbill (1970) found that

habituation of the orienting reflex to white noise was

correlated .66 with potency at 2 days, .64 at 5 days and .61

at one month. They also found that babies delivered with

general anesthesia took over two and one-half times as long

to habituate as babies of nonanesthetized deliveries. Using

the same paradigm, VanderMaelen, Strauss and Starr (1975)

also found that a slower rate of habituation at 3 days was


related to an increased level of obstetric medication.

In a study by Aleksandrowicz and Aleksandrowicz (1974),

obstetric medication accounted for 9 to 21 percent of the

variance among scores on response decrement to the pinprick

on the Brazelton exam over the first month of life.

Horowitz et al. (1977) found that a group of light to

moderately medicated babies in their Uruguyan sample were

slower to habituate to the light on the Brazelton exam at 3

days than a group of nonmedicated babies.

Friedman, Brackbill, Caron and Caron (1978)

investigated longer term effects of obstetric medication on

the habituation process by having 4 and 5 month olds view

slides of regular or distorted faces and measuring the

amount of time they fixated on the stimuli. They found that

premedication, oxytocin and general anesthesia were related

to slower rates of habituation. At eight months of age,

Brackbill (1976) found that the potency of anesthesia

continued to be correlated with the rate of habituation as

defined by a shift in heart rate response to white noise

from deceleration to acceleration.


Performance on the orientation cluster has also been

frequently related to the effects of obstetric medication.

This cluster assesses the infant's ability to orient to

visual and auditory stimuli and his or her overall quality

of alertness. This dimension of behavior is especially

important to the parent-infant relationship as it reflects


some of the infant's capabilities for interacting with his

or her parents.

The dose of meperidine has been correlated with

orientation and alertness scores on the Brazelton exam on

day 2 (Brackbill et al., 1974a,b). Hodgkinson et al.

(1978a, c) also found decreased alertness and responsiveness

to sound on the Scanlon exam on days 1 and 2 in a group of

babies exposed to meperidine inutero. They (1978a) found

that naloxone improved orientation scores at 2 hours of age,

but that the improvement was temporary. However, Bonta et

al. (1979) found that naloxone improved alertness at 1 and 4

hours, and responsiveness to sound at 24 hours. Although

meperidine doses were similar in the two studies, .4 mg. of

naloxone was administered intraveneously to the mother 15

minutes prior to delivery in the Hodgkinson et al. study and

in the Banta et al. study, 20 mcg. per kilogram of body

weight was injected intramuscularly into the baby a few

minutes after birth. The brief reversal of behavioral

depression in the Hodgkinson et al. study may be due to the

small amount of naloxone given to the mother and to its

short acting duration.

Rosenblatt et al. (1981) found that the greater the

exposure to bupivacaine as epidural anesthesia, the poorer

the performance on Brazelton orientation items up to six

weeks of life. Comparing a group of babies from bupivacaine

epidural deliveries to a minimally or nonmedicated group,

Murray et al. (1981) found no differences over the first


month of life on the interactive Brazelton dimension which

assesses the orientation items. However, only 45% of the

mothers in the bupivacaine group reported that their babies'

interactive performance at one month was exceptional,

compared to 85% of the mothers in the control group.

Bupivacaine paracervical blocks have also been related to

decreased alertness on the Scanlon exam at 6 hours (Meis et

al., 1978).

Infants of cesarean deliveries with general anesthesia

have been found to be less alert on days 1 and 2 than those

of deliveries with regional anesthesia (Hodgkinson et al.,

1976; 1977; 1978b; 1978c). Again, thiopental has been

related to more depressive effects than ketamine.

Furthermore, Palahniuk et al. (1977) found that within a

thiopental group, nitrous oxide was related to less

alertness on day 1 than methoxyflurane.

In one of the earliest studies on the effects of

obstetric medication on infant behavior Stechler (1964)

examined the relationship between analgesics and attention.

He weighted varying agents according to their dose and time

of administration. Findings indicated that the greater the

dosage and the closer to delivery the drugs were

administered, the less attentive the baby was to different

visual stimuli. Brown, Bakeman, Snyder, Fredrickson, Morgan

and Hepler (1975) also investigated the effects of

analgesics and found that doses greater than 52 mg. were

associated with less responsiveness to auditory stimulation


and less overall alertness during a feeding observation than

were smaller doses.

In the Tronick et al. (1976) study, analgesics were

related to a decreased ability to orient to the voice during

the Brazelton exam on day 3. Poor performance on

orientation to inanimate auditory stimulation was associated

with local anesthesia. Aleksandrowicz and Aleksandrowicz

(1974) found that combination of analgesics and anesthetics

accounted for 24% of the variance among Brazelton

orientation scores on day 1, 13% on day 2, 11% at a week and

28% at a month. Also, the medicated Greek sample in the

Brazelton et al. (1979) study did poorer on Brazelton

orientation items over the first 10 days than the

nonmedicated group. In the Horowitz et al. study (1977),

medicated Israeli babies had difficulties with orientation

items during the first four days of life and the highly

medicated American group had difficulties up to 10 days.

Lastly, Conway and Brackbill (1970) found a correlation

of .51 between a potency of medication score and the vision

subscale of the Graham exam on day 5. They found that the

performance of babies from deliveries with general

anesthesia was worse than that of babies from deliveries

with local anesthesia which, in turn, was worse than

performance of babies from nonmedicated deliveries.


The most comprehensive assessments of infant reflexes

are found with neurological exams such as the Prechtl-

Beintema. Standardized behavioral and pediatric exams also

include the assessment of several reflexes, most notably,

sucking, rooting and the Moro. Reflexive responses have

been related to maternal obstetric medication, although not

as frequently as the three previously mentioned clusters.

Meperidine has been related to depressed reflexes more

than any other medication. Brackbill et al. (1974a; 1974b)

found that on day 2, babies exposed to meperidine inutero

scored less optimally on a cluster of reflex items from the

Brazelton exam than those who were not exposed. Wiener et

al. (1979) reported similar findings. They also found that

naloxone significantly improved reflex scores. Several

studies have found that the four reflexes assessed by the

Scanlon exam, i.e., rooting, sucking, placing and the Moro,

are lower among meperidine groups on days 1 and/or 2 than

among non-meperidine groups (Corke, 1977; Hodgkinson et al.,

1976; 1978a; 1978b; 1978c). Hodgkinson et al.(1978a) also

found that naloxone improved reflex scores, but only at 2

hours of age. Using a short screening exam at birth and the

Prechtl-Beintema exam at 8 days, Richards and Bernal (1972)

found that several reflexes were depressed among a group of

babies whose mothers received a compound including

meperidine and a hypotensive medication.


The Wiener et al. (1979) study found that reflex scores

on the Brazelton exam over the first 2 days were as low

among a group of bupivacaine epidural babies as they were

among a meperidine without naloxone group. The

intramuscular administration of 200 mcg. of naloxone at

delivery to the mother seemed to improve scores among a

meperidine group for the 2 days. Scanlon et al. (1974,

1976) found reflexes to be depressed during the first 8

hours of life among a mepivacaine or lidocaine epidural

group but not among a bupivacaine group. Reflex scores on

the Prechtl-Beintema exam were also found to be low among a

group of babies born by cesarean section with lidocaine

epidural anesthesia (Hollmen et al., 1978). Weak reflexes

were significantly correlated with maternal hypotension

within this group.

Babies who are the products of deliveries with

paracervical blocks with bupivacaine have been found to have

a weaker sucking reflex at 24 hours of age than babies of

local, pudendal or spinal deliveries (Meis et al., 1978).

Hodgkinson et al. (1978c) found that epidural babies had a

higher percentage of good reflex scores than pudendal block

babies. They also found along with Hollmen et al. (1978)

and Palahniuk et al. (1977) that babies born by cesarean

section with general anesthesia have lower reflex scores on

the first two days than cesarean babies delivered with

epidural anesthesia. Hodgkinson et al. (1976; 1977; 1978b;

1978c) found that thiopental was related to weaker reflexes


than ketamine. Palahniuk et al. found that within a

thiopental group, nitrous oxide seemed to be more depressive

than methoxyflurane.

Autonomic Regulation

The autonomic regulation cluster assesses the infant's

ability to control physiological responses to stress, i.e.,

startles, tremors and changes of skin color such as flushing

or cyanosis. Several investigators have found a

relationship between these responses as measured by the

Brazelton Scale and the use of obstetric medication.

Brackbill et al. (1974a; 1974b) found that infants

whose mothers took meperidine during labor startled more

frequently on the second day of life than infants whose

mothers took no meperidine. Using a screening tool which

they devised, Richards and Bernal (1972) found that use of

meperidine with an antihypotensive was related to poor skin

color shortly after birth.

Infants with more exposure to bupivacaine epidural

anesthesia had a greater number of tremors and startles over

the first 6 weeks of life in the Rosenblatt et al. study

(1981). Murray et al. (1981) also found poorer performance

on physiological response to stress among a bupivacaine

group, but only on day 1. On day 3, Standley et al. (1974)

found that an increase in tremulousness and startles was

related to use of anesthesia. Tronick et al. (1976) found

poorer skin color on day 10 among a group of infants

delivered with lidocaine or mepivacaine epidural anesthesia.


Horowitz et al. (1977) found poorer skin color over the

first month of life among their American sample of babies

from deliveries with general anesthesia than among those

from moderately medicated deliveries. A heavier amount of

obstetric medication was also related to a greater number of

tremors and startles at one month in the Standley et al.

(1974) study. Finally, Brazelton et al. (1979) found poor

performance on the autonomic regulation cluster on days 1,

5, and 10 among the Greek neonates whose mothers were

medicated during delivery than among those whose mothers

were unmedicated.

Range of State

Clusters of behavior assessing the infant's state

patterns have been related to obstetric medication less

frequently than the other clusters. As noted previously,

this may be because the Scanlon exam, which has been used

most frequently in these studies, omits state-related items.

Another reason for the lack of significant state-related

effects may be that different medications have opposite

effects on the infant's states so that they may tend to

cancel out the effects of each other. The range of state

cluster assesses the predominant states the baby is in

during an examination period, the number of state changes,

how irritable the baby is and the amount of stimulation it

takes to make the baby cry. The results of the literature

seem to indicate that analgesics are related to drowsiness

during the first few days of life and to irritability later


on while anesthetics are related to increased irritability

immediately after birth. Oxytocin may also be related to

drowsiness. Therefore, studies in which mothers have

received both premedication and anesthesia may find no

significant effects on items assessing to state patterns

(e.g., Brazelton et al., 1979).

Brackbill et al. (1974a; 1974b) found that babies whose

mothers took 50 to 100 mg. of meperidine during labor were

drowsier and had fewer state changes at 2 days of age on the

Brazelton exam than infants whose mothers had no meperidine.

Using the Brazelton on day 1 in Malaysia, Woodson and

DaCosta-Woodson (1980) found that when the effects of

parity, length of labor, maternal blood pressure and ethnic

group were controlled for, use of meperidine and promazine

was related to a decrease in an irritability factor. The

factor included the item assessing frequency of fussing,

time to buildup to the initial cry, general state of

arousal, number of state changes, ability to self-quiet or

be consoled and muscle tonus. Additionally, in an

investigation of the effects of meperidine and promethazine,

Borgstedt and Rosen (1968) had a pediatric neurologist rate

the baby's states and activity levels during the Prechtl-

Beintema exam on days 2 and 4 as "impaired" or not. The use

of the analgesics was related to drowsier babies with less

labile state patterns. Belsey et al. (1981) found that

during the first twenty minutes after birth, a greater

concentration of meperidine in cord blood was correlated


with a longer latency to cry and more intervals spent in a

drowsy state. A greater exposure to meperidine inutero was

correlated with less intervals in a crying state. The

authors also examined infants using the Brazelton exam on

days 1, 3, 7, 21 and 42. Although no significant

relationship between meperidine and range of state was found

during the first few days of life, a greater dose and/or

exposure level to meperidine was related to an increased

number of state changes and irritability from 1 through 6

weeks of age. Horowitz et al. (1977) also found that the

Israeli analgesic group was less irritable than a

nonmedicated group on the Brazelton over the first 3 days of

life and more irritable at one month. The analgesic group

also had fewer state changes and a lower peak of excitement

during the first few days and the opposite at one month.

Anesthesia seems to be more generally related to

increased irritability. Rosenblatt et al. (1981) found that

the greater the exposure to bupivacaine epidural anesthesia

inutero, the greater the irritability, peak of excitement

and number of state changes during the Brazelton exam on

days 1 and 3. The anesthesia was also related to increased

irritability at 6 weeks. Murray et al. (1981) found more

labile state patterns in a bupivacaine epidural group on

days 1 and 5 than in a minimally or nonmedicated group.

However, an epidural group whose mothers had also received

oxytocin for induction or augmentation of labor exhibited

the opposite state pattern, i.e., flat with few state


changes. Although no significant differences were found in

Brazelton performance at one month, mothers assessed their

infants at one month as continuing to exhibit the same state

patterns that were seen on days 1 and 5. Standley et al.

(1974) also found that the use of anesthesia for delivery

was related to increased irritability and state changes on

day 3. The agent and route used for the anesthesia were not

reported. Hollmen et al. (1978) assessed infants' states of

arousal during the Prechtl-Beintema exam in the first week

of life. They found no differences comparing groups of

cesarean section babies delivered with epidural or general

anesthesia. However, infants in the epidural group whose

mothers were hypotensive during labor and delivery were

found to be listless for the first week of life. Comparing

the state-related performance on the Brazelton exam of the

moderately and heavily medicated American infants, Horowitz

et al. (1977) related general anesthesia to an increased

peak of excitement and a greater number of state changes

over the first month of life. Infants delivered with

general anesthesia were also found to be less irritable on

the first day of life than other infants but more irritable

over the rest of the month.

Regulation of State

The regulation of state cluster is defined by items

assessing the infant's abilities to modulate his or her

states of consciousness. This includes the infant's

capabilities for self-quieting, sucking on his or her


fingers, quieting with help from the examiner or caregiver

and adapting to cuddling maneuvers. The Brazelton Scale is

the only standardized exam which assesses regulation of

state. Thus, all studies finding differences related to

obstetric medication have used this exam.

The dose of meperidine has been related to a decreased

ability to self-quiet at 3, 14 and 42 days, to decreased

hand-to-mouth ability at 3 and 42 days and to more

difficulty consoling at 42 days (Belsey at al., 1981).

Brackbill et al. (1974a; 1974b) found that infants whose

mothers had taken meperidine meperidine were more difficult

to console at 2 days and were less able to cuddle or to

relax and mould to the examiner's body when being held.

Tronick et al. (1976) also found more difficulty consoling

at 3 days among babies whose mothers took some type of


Exposure to bupivacaine was related to a decreased

ability to self-quiet on days 3 and 42 by Rosenblatt et al.

(1981). Murray et al. (1981) also found that babies in a

bupivacaine group had less control over their states on days

1 and 5 than a group of minimally or nonmedicated babies.

The mothers in the bupivacaine group also reported that

their babies had difficulties controlling their states at

one month of age. Tronick et al. (1976) reported decreased

hand-to-mouth ability on day 10 among a group of infants

delivered with lidocaine or mepivacaine epidural anesthesia.


Cuddliness has been found to be lower at one month of

age among groups of babies from highly medicated deliveries

(Aleksandrowicz and Aleksandrowicz, 1974). Horowitz et al.

(1977) also found that infants from the highly medicated

American sample scored lower than moderately medicated

Israeli or Uruguayan infants on day 3. Investigating the

behavior of Greek neonates, Brazelton et al. (1979) found

that babies of middle class mothers medicated with analgesia

and anesthesia performed worse on the state regulation

cluster over the first 10 days of life than babies of lower

class mothers who received no medication during labor and


Other Behaviors

Several behavioral measures which have been

investigated in relation to obstetric medication did not fit

into the clustering system previously described. These

behaviors are changes in electroencephalogram pattern, sleep

states, sucking, feeding and weight gain and parent-infant


Electroencephalogram patterns (EEGs)

In one of the earliest studies on the behavioral

effects of obstetric medication on the neonate, Hughes,

Hill, Green and Davis (1950) examined EEG patterns at 1 and

2 days in infants whose mothers received large doses of

meperidine (100-300 mg.), vinabarbitol (5-10 grains), or

morphine (10-15 mg.). They found that dose was correlated


with a clinical picture of drowsiness and that EEGs in these

babies were typically of low amplitude and high frequency,

consistent with pictures of cortical depression. In another

study, Hughes, Ehemann and Brown (1948) found that infants

whose mothers had taken varying doses of secobarbitol during

labor were drowsy for the first two days of life but that

their EEGs continued to be of low amplitude and high

frequency on the third day. Borgstedt and Rosen (1968)

related similar EEG patterns between 2 and 4 days to the

effects of obstetric medication. Most of the mothers in

their medicated group received meperidine and either

promethazine or phenobarbitol. Differences in EEGs during

auditory stimulation at 48 hours have also been related to

meperidine intake in a study by Brower, Crowell, Leung and

Cashman (1978).

Sleep states

Observing newborns for the first ten hours of life,

Emde, Swedberg and Suzuki (1975) found that a group of

infants born to mothers who had received low levels of

meperidine, diazepam or secobarbitol during labor had 400%

more quiet sleep in the first 2 hours and 50% more in the

next 8 hours than a group who had received no premedication.

A decrease in amount of time awake was correlated with an

increase in dose. The group born to mothers who had taken

meperidine had less than half the amount of wakefulness as

the unexposed group in the first 8 hours. Yang, Zweig,

Douthitt and Federman (1976) found that a greater amount of


quiet sleep and less active sleep during the third day of

life was correlated with an earlier administration and a

greater amount of analgesia and anesthesia during labor.

Feeding and interactive behaviors

Several studies have associated the effects of

obstetric medication to differences in neonatal feeding

behavior or sucking times and to differences in parental

responses to the infants. Studies relating the following

drugs to a weaker sucking reflex over the first 2 days of

life have been reviewed previously: meperidine, mepivacaine

and lidocaine epidurals, paracervical blocks with

bupivacaine and general anesthesia.

Dubignon, Campbell, Curtis and Partington (1969) found

that with length and type of labor covaried, nutritive

sucking was influenced by obstetric medication but

nonnutritive sucking was not. A decrease in sucking time

was related to general and epidural anesthesia despite

exposure to sedatives. Although sucking time was not found

to be influenced by sedatives, babies of mothers who had

been sedated were found to have a lower food intake when

calories and ounces were measured. Kron, Stein and Goddard

(1966) randomly assigned women to receive 200 mg. of

secobarbitol during labor or no medication at all. They

found that on days 2, 3 and 4, sucking rate, pressure and

amount consumed was impaired in the barbiturate group.

Observing feedings on day 2 to 10, Richards and Bernal

(1972) found that infants of mothers premedicated with the


compound of meperidine and an antihypotensive fed for

shorter periods of time and needed more stimulation from

their mothers to suck than a group that had no history of

premedication. They did not find statistically significant

differences between groups on a measure of nonnutritive

sucking but there was a trend toward a reduced amount of

sucking and less responsiveness to the removal of a nipple

among the meperidine group.

Brazelton (1961) assessed the effects of obstetric

medication on the infant's ability to establish a normal

breast feeding pattern as reported by the mother over the

first six days of life. Effective nursing was defined as no

more than 5 minutes of stimulation needed from the mother to

get the baby to feed and being able to establish a

continuous feeding period of at least'3 minutes on the first

2 days and 5 minutes on the next 4 days. High doses of bar-

biturates and general anesthesia were found to impair the

infants' ability to nurse effectively over the first 4 days.

There was also a 24 hour delay in effective weight gain

among infants in the highly medicated groups.

Kraemer, Korner, and Thoman (1972) studied a sample of

babies whose mothers had received any of 12 drugs during

labor in any of 41 combinations. These were classified into

four drug groups according to the type of anesthesia

administered and whether or not analgesia was administered

in addition to the anesthesia. They observed the first

feeding by a nurse at 12 hours of age. Results indicated


that before length of labor and parity were covaried, there

was a significant difference between the four groups in the

number of feed intervals. However, when the effects of

these two variables were removed, only exposure to analgesia

was related to a decrement in the number of feed intervals.

Brown et al. (1975) observed a feeding by mothers on

the third day of life. Their sample had received various

combinations of analgesics and tranquilizers not greater

than 150 mg. one to six hours prior to delivery. They

found that infants whose mothers received more than 52 mg.

of the drugs were more passive during the feeding. They

were held, fed and stimulated more by their mothers than the

group with less premedication.

Murray et al. (1981) also found that epidural

anesthesia was related to the number of times mothers had to

stimulate their babies to suck during a feed at 5 days and

one month. Mothers in bupivacaine groups with or without

oxytocin had to stimulate their infants more frequently than

those from the control group. There was also less

affectionate handling and less eye contact during the

feeding in these groups. According to diaries kept by the

mothers, infants in the bupivacaine with oxytocin group fed

less frequently than infants from the other two groups.

Observing mother-infant interactive behavior in the

first 20 minutes after birth, Lieberman, Rosenblatt, Belsey,

Packer, Mills, Caldwell, Notarianni, Smith, William and

Beard (1979) found that mothers in a bupivacaine epidural


group spent less time talking to their infants than mothers

in a control group without obstetric medication. Parke,

O'Leary and West (1972) found that when the effects of

length of labor were removed, a higher potency of obstetric

medication was correlated with a greater amount of maternal

stimulation in the form of rocking and vocalizations during

interactions between 6 and 42 hours after birth. There was

a trend toward a decreased amount of paternal stimulation

with increased medication.

Lack of Significant Behavioral Effects

The majority of the investigations into the effects of

obstetric medication have shown a substantial negative

relationship with the behavioral outcome for the infant, and

no study has shown a positive relationship. However,

several studies have found little or no effects of varying

drugs used during labor and delivery on infant behavior.

Ten of these studies have compared the effects of varying

agents on behavior and four have compared the effects of no

medication to varying agents. Four other studies for which

significant findings have been previously reported are

considered to have shown minimal effects of obstetric

medication when the total number of nonsignificant findings

are taken into account.

Most of the studies which have found no differences

between drugs have used the Scanlon exam as the measure of

behavioral outcome. Baraka, Noueihid and Haj] (1981)

investigated the effects of 1 vs. 2 mg. of intrathecal


morphine as parturitional medication and simply reported

that at 24 hours, the Scanlon scores in both groups were

normal. In another study, Baraka, Maktabi and Noueihid

(1982) compared 6 to 12 hour Scanlon scores between a group

of babies whose mothers received meperidine injected

epidurally followed by bupivacaine supplementation and a

group receiving bupivacaine only. Again, they merely

reported that all scores were normal. Writer, James and

Wheeler (1981) compared bupivacaine epidurals to morphine

epidurals. They found no significant differences between

the groups at 3 or 24 hours in Scanlon scores. However, at

three hours, three-quarters of the babies in the morphine

group were classified as "borderline" with respect to the

normalcy of their behavior.

In a recent study, Stefani, Hughes, Shnider, Levinson,

Abboud, Henriksen, William and Johnson (1982) randomly

assigned 61 women to receive either enflurane, nitrous oxide

or no inhalation agent during labor. Sixty-five percent of

the entire sample had also received local infiltration

and/or pudendal block and 40% had received meperidine or

alphaprodine. They compared group performance on the Amiel-

Tison exam at 15 minutes after birth and on both this exam

and the Scanlon exam at 2 and 24 hours. They reported

finding no significant differences in behavioral performance

between the drug groups on either of the exams. They

compared performance on the two exams by devising a summary

score for the Scanlon and comparing it to the Amiel-Tison,


which concludes with a summary score. No significant

differences were found. However, examining scores on the

individual Scanlon items which are presented in their paper,

it appears that infants from the control group did

significantly better than the inhalation analgesic groups on

head control during pull-to-sit and on alertness. It is

difficult to interpret these findings though since a varying

number of infants within each group received local

anesthetics and/or narcotics.

Shnider, Abboud, Levinson, Wright, Kim, Henrikson,

Hughes, Roizen and Johnson (1979) compared infants whose

mothers had received either halothane or .5 or 1% enflurane

in addition to thiopental induction for cesarean delivery.

They found no differences in Scanlon scores at 2 or 24 hours

between any of the groups. Also examining cesarean section

babies, Lund, Cwik, Gannon and Vassallo (1977) reported that

2 to 4 hour Scanlon scores were normal among babies whose

mothers received etidocaine epidurals with or without

adrenaline. No differences in performance were attributed

to the use of adrenaline. Datta, Corke, Alper, Brown,

Ostheimer and Weiss (1980) compared the use of etidocaine,

bupivacaine and chloroprocaine as epidural anesthesia for

elective cesarean sections. They found no differences in

Scanlon scores at 2 to 4 hours.

Scanlon, Ostheimer, Lurie, Brown, Weiss and Alper

(1976) studied the behavior of 20 infants delivered

vaginally with bupivacaine epidural anesthesia. They


reported that all 2 to 4 hour Scanlon scores were normal.

Comparing this group to a group of infants from a previous

study who had been exposed to mepivacaine or lidocaine

epidural anesthesia (Scanlon et al., 1974), they concluded

that there were no "appreciable" differences between the

groups but that the bupivacaine group did not have the

motoric depression or slow rate of habituation noted in the

previous sample.

In the study in which Corke (1977) compared Scanlon

performance between a meperidine-promethazine and an

unmedicated control group as described earlier, he also

included a group of bupivacaine epidural babies. No

significant differences were found in Scanlon scores at

approximately 4 hours of age between the bupivacaine and the

control groups. Abboud, Khoo, Miller, Doan and Henriksen

(1982) compared the use of varying agents for epidural

anesthesia during labor and delivery to the use of no

medication. They found no significant differences in 2 or

24 hour Scanlon scores between babies in a bupivacaine,

chlorcprocaine, lidocaine or control group. Investigating

weight gain over the first 5 days of life, Abouleish, Van

der Donck, Meeuwris, and Taylor (1978) found no significant

differences between babies exposed to spinal, epidural or

general anesthesia or to no obstetric medication. In a

study designed to compare different methods of induction

during labor, Ounsted, Boyd, Hendrick, Mutch, Simons and

Good (1978) also compared the results of babies delivered


spontaneously with epidural anesthesia and/or meperidine or

no medication at all. They found no significant differences

between these groups on a composite neurobehavioral score

given on the first day of life and at 2 months, or on the

Denver test given at 18 months. However, they did find that

babies from a group induced with oxytocin performed worse at

2 months and had lower fine motor scores at 18 months than

groups induced with prostaglandin E2.

A longitudinal study by Field and Widmayer (1980)

investigating the effects of cesarean section deliveries

with general anesthesia on infant development reported few

significant findings. Since this was a comprehensive study

and the results were unexpected, I will discuss it in more

detail than the other studies. The authors sampled a group

of infants of lower class black mothers, 20 who delivered by

emergency section with thiopental and nitrous oxide and 20

who delivered vaginally with lidocaine local, pudendal,

saddle back or without anesthesia. Each subject was given a

drug score according to the dose of obstetric medication

(anesthesia, narcotics, sedatives, tranquilizers and/or

oxytocin) multiplied by the time of administration. The

cesarean group had a significantly higher drug score than

the vaginal group. They also had a less optimal obstetric

complication score and higher pre- and postpartum maternal

blood pressure.

Infants were assessed at 2 days with the Brazelton

Scale, during the first month with the Mother's Assessment


of her Infant's Behavior, at 4 months with the Denver

Developmental Screening Test and the Carey Infant

Temperament Questionnaire and at 8 months with the Bayley

Scales of Infant Development and the Carey Questionnaire.

At 4 months, a feeding and play interaction with the mother

was also observed.

Very few significant differences were found. The only

measure on which the cesarean section infants performed less

optimally was on the Denver adaptability items (tracking,

reaching and grasping) at 4 months. Otherwise, cesarean

babies were rated more optimally on the temperament

questionnaires and both mothers and babies in this group

received better ratings on the interactional observations.

The authors offered several hypotheses for these

curious findings. They suggested that the stresses of

passage through the birth canal and the depressive effects

of local anesthesia may be greater in the case of the

vaginally delivered infant. Also, subjects in most studies

on the effects of general anesthesia have received

intravenous anesthetics as well as various inhalation

anesthetics. They suggested that the nitrous oxide may

provide an antagonistic effect against the effects of the

other medications given. However, as noted previously,

comparing groups receiving thiopental plus an inhalation

agent, Palahniuk et al. (1977) found that nitrous oxide

depressed scores on the Scanlon exams over the first day of

life more than methoxyflurane. Both groups performed worse

than an epidural group.

Field and Widmayer (1980) also suggested that the high

blood pressure among many of the cesarean mothers may have

decreased the rate of placental transfer of the medications

since hypertension has been found to alter the placental

vascular bed and possibly inhibit the passage of some drug

compounds (Cohen and Olsen, 1970). However, as previously

discussed, Hollmen et al. (1978) found that mothers who

received general anesthesia with thiopental for cesarean

section and who were hypertensive had infants who were

drowsy and hypotonic for the first week of life. Field and

Widmayer neglected to report if any of the mothers in the

vaginal group were hypotensive.

Lastly, the authors of this study suggest that the

special nature of the emergency cesarean section among a

lower class population may require more support from family

members than is typical for a natural delivery. Thus, the

mother may be especially attuned to her baby which would in

turn yield better temperament ratings and more optimal


One final comment on this study which the authors

failed to note is that an almost equal number of mothers in

both groups received meperidine (11 section and 8 vaginal),

promethazine (2 and 2) and oxytocin (20 section and 16

vaginal). One additional interpretation of the results

would be that the effects of cesarean section delivery were

masked by the more powerful depressive effects of these


drugs as reported in previous studies. Additionally,

varying extremes in the lengths of labor could have

potentially influenced the neonatal outcome. The authors

failed to report the lengths of labor in either group.

Thus, the results of this research are not easily

interpretable because of the nature of the sample and the

possibility of many confounding variables.

Several of the studies which have been reported

previously should be mentioned in this section because the

findings of significant effects of maternal medication on

infant behavior were relatively minimal when the total

number of dependent measures in these studies are taken into

consideration. For example, after covarying the length of

labor, Kraemer et al. (1972) found that drugs had no effect

on a visual pursuit score, irritability, state changes or

duration of sleep and wakefulness on the third day of life.

Although Brackbill (1976) found that a drug potency score

was related to habituation at 8 and 12 months as indicated

by heart rate responses to auditory stimulation, she found

no significant relationship at one and four months.

Brackbill postulated that the delayed effect of the

obstetric medication may be due to the immaturity of the

cardiovascular function in the first few months of life and

to the more stressful nature of the testing situation at the

later age. Thus, the effects of drugs may manifest

themselves only when the infant is mature enough to perceive

the event as being stressful.


In the Horowitz et al. (1977) study of Brazelton

performance among three populations, they found few

significant differences between the light to moderately

medicated and the unmedicated Israeli or Uruguyan samples.

However, comparing a moderately to highly medicated American

sample to the unmedicated samples of the Standley et al.

(1974) and Tronick et al. (1976) studies, they found many

behavioral differences. The authors suggested that a

combination of genetic, biological and attitudinal factors

may have contributed to the differences among the different


Finally, in the Tronick et al. (1976) study, only

thirteen significant differences were found comparing 54

subjects in 8 drug groups on 26 Brazelton items on each of 7

days. They found no significant differences between a group

of infants of epidural deliveries and a group with

premedication. This study has been criticized for excluding

babies with Apgar scores less than 7 and, therefore,

possibly excluding those babies who were most affected by

obstetric medication (Fanaroff, Kennell, McClelland and

Mortimer, 1977). Lester, Als and Brazelton (1982) recently

reanalyzed the Tronick et al. data and found a significant

relationship between medication and behavior in this sample

when the effects of other nonoptimal obstetric variables

were taken into account. This study will be reported later

in the next section.


Methodological Considerations

The literature on the behavioral effects of obstetric

medication on the infant has generated considerable

controversy within the scientific community. This is due to

the clinical importance of the area since approximately 95%

of the deliveries in this country are medicated (Brackbill,

1979), to the interdisciplinary nature of the subject and

frequent naivete of authors not thoroughly investigating

disciplines other than their own, to the diversity of

findings and to methodological problems within the

literature. Several of the studies which have been reported

in the previous section were accompanied by critical

editorials. Criticisms have covered study designs, subject

criteria, analyses of the drugs, appropriateness and

reliability of the dependent measures, use of blind

observers and the adequacy of data analyses. Psychologists

have been criticized for naivete regarding the principles of

perinatal pharmacology and anesthesiologists, obstetricians

and pediatricians, for naivete regarding the complexities of

behavior and statistical procedures.

The study of the effects of obstetric medication on

infants is severely limited by ethical considerations

regarding random assignment of women to drug groups.

Although the experimental method was used in an early study

by Kron et al. (1966) in which they randomly assigned

subjects to receive either 200 mg. secobarbitol or nothing

during labor, there has been enough evidence since then to


indicate that medication administered perinatally may not be

completely safe for the baby. Therefore, the experimental

design may no longer be used in this type of research which

necessitates the use of correlational designs. The only

exceptions have been studies comparing similar agents and

routes of administration. Merkow et al. (1980) randomly

assigned subjects who were to receive a pudendal block

during labor to receive bupivacaine, mepivacaine or

chloroprocaine as the agent. Nesheim et al. (1979)

randomly assigned women to a bupivacaine paracervical block

group or a lidocaine local group. Additionally, Hodgkinson

et al. (1978a) randomly assigned some women who had received

meperidine during labor to receive the antagonist naloxone

in addition to the meperidine. As more is known about the

differing effects of agents, this type of random assignment

will also become unfeasible.

Most of the literature on obstetric medication has used

two types of designs to correlate drugs with behavior. One

design is to choose subjects from the medical records who

fit the selection criteria, i.e., after receiving the

medication under investigation. The problem with this

design is that data such as blood samples normally collected

at the time of delivery cannot be obtained. Also, medical

records omit many variables crucial to studies of this

nature. The second design which is used frequently is to

recruit subjects who fit selection criteria prior to

delivery and to study the drugs which they end up receiving.


This method is fraught with problems in interpretation.

When a variety of medications have been administered

obstetrically, it is difficult to tease out the agents and

routes of administration which are having the major effect

on behavioral outcome. The dose and timing of the

medication have also been found to be important to consider.

Most studies have either ignored one or more of these

variables or have used a drug-weighting system which

attempts to combine some or all of these variables. This

latter approach may be inappropriate when assessing the

effects of more than one drug type and/or route because not

enough evidence exists to be able to rank-order different

drugs and routes of administration or to weight them

quantitatively. Most of the studies which have used this

approach ignore principles of perinatal pharmacology

(Scanlon and Hollenbeck, 1982). According to pharmaceutics,

it is unreasonable to equate the potential effects of

varying drugs or routes of administration on the baby

because of differences in the rate of placental transfer of

various agents, the time of maximum effectiveness as

determined by the agent, route and time of administration

and the varying doses used for different agents. One study

which has been highly criticized for using this approach is

that of Standley et al. (1974). For the purpose of

analyses, they lumped the use of narcotics and tranquilizers

together as analgesics and varying agents used as pudendal,

paracervical, spinal and epidural blocks as anesthesia.


Comparisons were made between groups receiving analgesia,

anesthesia and various combinations of the two. Hodgkinson,

Marx and Kaiser (1975), Ostheimer (1981) and Scanlon and

Sostek (1979) have argued that the groups are incomparable.

For example, spinals are put in the same group as epidurals

and compared to pudendal and paracervical blocks and locals.

The differing doses and metabolic rates of the various local

anesthetics make the comparisons meaningless.

Another deficiency of most obstetric medication studies

is that they lack samples of "clean" nonmedicated controls.

Studies in the anesthesiology literature typically compare

the effects of different drugs, often comparing newly

developed agents to older ones. When no differences are

found, they have concluded that the null hypothesis was

proven, i.e., that there are no differences between the

drugs. Use of the newly developed agents may therefore be

considered safe. In order to estimate potential effects of

drugs on behavior, it is important to examine the behavior

of infants delivered without medication. Since most births

are medicated, these subjects may be very difficult to find.

Murray et al. (1981) thought that they had recruited a

"clean" control group until analyzing cord blood samples and

discovering lidocaine in the plasma of over half of this

sample. The drug had been given for local infiltration of

the perineum but the hospital records had not reported

whether the drug had been administered before or after

delivery. Half of the unmedicatedd" group had also briefly

inhaled nitrous oxide.

Major criticisms of the sampling procedures may be made

of most studies. Selection criteria generally include only

the "clinically accepted ideal" or a population described as

healthy fullterms with no complications of pregnancy and

delivery. Optimal birthweights, maternal ages and other

variables are often specified. Infants with histories of

common nonoptimal pre- and perinatal conditions are not

included in an attempt to exclude variables other than drugs

which may affect behavior. However, the practice of using

strict subject selection criteria makes it difficult to

generalize from the "superbaby" samples which are recruited

to a typical population. Additionally, the exclusion of

subjects with conditions which may have been caused by the

drugs such as long labors, fetal heart rate variability,

acidosis and low Apgar scores make conclusions about the

safety of drugs negligible. Alper (1978) has noted that

complications arise in at least thirty percent of all births

which appear to be following normal progressions. It is,

therefore, very difficult, if not impossible, to control for

every nonoptimal condition that may occur during pregnancy

and delivery. These conditions may either mask drug effects

or interact with drugs to affect behavior. Most

investigators have ignored the variables which are not

controlled for in their studies. However, a few have

recognized the influence of variables other than drugs on

their findings and have either conducted post hoc analyses


on their cumulative effects, have excluded their effects or

have examined them individually. For example, Brackbill et

al. (1974a; 1974b) regressed scores on a habituation

paradigm and on a revised Brazelton exam using birthweight,

length of the first stage of labor, mother's age, parity,

low forceps and timing and dose of meperidine as predictors.

They found that birthweight and length of labor

significantly increased the predictability of the infants'

performance. The combination of these variables with the

dose of meperidine accounted for more of the variance in

scores than did dose alone.

Hollmen et al. (1978) investigated the effects of

maternal blood pressure within groups of infants delivered

by cesarean section with epidural or general anesthesia.

They found that the longer the hypotension within the

epidural group, the more depressed activity and neurologic

scores were. Also, hypertension had a cumulative effect on

behavior within the general anesthesia group. One problem

with this study is that their conclusions were based on

clinical findings among a small number of subjects within

their sample.

After finding significant differences with analysis of

variance between drug groups in feeding behaviors, Kraemer

et al. (1972) used analysis of covariance to eliminate the

effects of length of labor, parity, sex, birthweight and

maternal age from their results. They found fewer

differences in behavior between drug groups after covarying


length of labor. No further significant effects on behavior

were found with removal of the other variables although drug

use was found to be significantly related to both length of

labor and parity.

Woodson and DaCosta-Woodson (1980) also investigated

covariates of obstetric medication. They sampled a group of

babies in Malaysia and found that parity, ethnic group,

duration of the second stage of labor, gestational age and

maternal blood pressure during labor influenced the

relationship between analgesia and a Brazelton irritability

factor. When the effects of these variables were

controlled, infants from a group delivered with analgesia

were found to be less irritable than those from an

unmedicated group. This lends support to the Horowitz et

al. (1977) hypothesis that genetic, biological and

attitudinal factors may contribute to differences in drug

effects among different cultures.

Investigating the effects of several variables on

mother-infant interaction, Brown et al. (1975) found that

parity was the best predictor of behavior followed by dose

of analgesics and length of labor. However, more than two-

thirds of the variance associated with parity was confounded

with amount of analgesic taken and length of labor. In

separate analyses, the authors found that birthweight and

sex also had a significant effect on interactive behaviors.

These authors concluded their study with the following



In future studies, investigators might consider
selecting a large number of mothers and infants
at random and extracting the effects of background
variables through statistical analyses rather than
attempting to increase the homogeneity of their
sample through the imposition of stringent
selection criteria. At the present time not
enough is known about the importance of even the
most common background variables to enable
researchers to choose selection criteria without
introducing possible biases. (Brown et al., 1975,
p. 686)

With this recommendation in mind, if variables which

have not been controlled for by selection criteria act in

concert with obstetric medication to stress the infant, the

best way to examine the relationship is with multivariate

analysis as in the Brackbill et al. (1974a; 1974b) study.

This allows us to examine the predictive power of related

variables and to estimate the relative contribution of each

variable to the prediction of behavior. A recent reanalysis

of the Tronick et al. (1975) data by Lester, Als and

Brazelton (1982) using stepwise multiple regression serves

as a good illustration of the power of this method. This

analysis shows the cumulative effects of drug and nondrug

variables on behavior where few effects were found in the

Tronick et al. study as a function of drugs alone within a

sample of "superbabies" whose mothers received relatively

low doses of obstetric medication.

The sample was 54 infants more highly selected for

obstetric and pediatric well-being than in most studies.

Selection criteria are shown in Table 2. The Brazelton exam

was administered to each infant on days 1 through 5, 7, and


Table 2. Selection Criteria for Tronick et al. (1977) Study

Infant Was Excluded if Mother Had


Threatened Abortion

Diabetes mellitus

Chronic disease (renal, hyperthyroidism,

neurological disorders)

Age < 18 yr or > 35 yr

Prolonged first-stage labor > 24 hr

Prolonged second-stage labor > 6 hr

Rupture membranes > 24 hr

Hemorrhage or shock

Precipitous labor < 3 hr

Any abnormal delivery (breech, Caesarian section,

shoulder, brow, head presentation,

high, mid-forceps)

Infant Was Rejected at Delivery or After 24 Hours if


Gestation was > 41 wk

Birthweight < 2,712 gm or > 4,068 gm

Congenital anomaly

Apgar score < 7 at 1 min and 5 min

Cephalohematoma or other bruising

Development of severe illness during study

(sepsis, seizures, bleeding, etc.)


10. Each of the seven Brazelton clusters previously

mentioned on each of the seven testing days plus their slope

or rate of behavioral change over the ten days of life were

the outcome variables in the regressions. The independent

or predictor variables were (1) a drug group score based on

the type, route of administration and timing of the

medication (drug groups were ranked from 1 to 8 by an

anesthesiologist in increasing order according to the

magnitude of their expected effects on behavior); (2) a drug

factor score based on the amount of time from the first and

last drug administration to delivery, the number of

different drugs and the number of drug administrations; (3)

the ponderal index which is a weight-to-length ratio

assessing possible fetal malnutrition; (4) length of labor;

and (5) parity. There were few significant individual

correlations between any one of these predictor variables

and the outcome measures. However, 22 of the 49 multiple

correlations were significant. They ranged from .30 to .52,

indicating that the linear combination of the 5 predictor

variables accounted for 9 to 27% of the variation in the

cluster scores over the 10 days. Significant effects were

found on 5 of the 7 testing days for the autonomic

regulation cluster, 4 days for the range of state cluster

and 2 days for the orientation, regulation of state and

reflex clusters. The regressions on the rate of behavioral

change showed significant effects for the habituation,

orientation, motor, autonomic regulation and reflex

clusters. The individual beta weights indicated that worse

performance on all clusters was associated with higher drug

group and drug factor scores, a lower ponderal index, being

firstborn and having a long labor.

The results of this analysis suggest that even with a

sample as carefully selected to be low risk as this one was,

subtle indications of risk may be present which act in

concert with medication to affect behavior. A comparison of

the minimal effects found by Tronick et al. with the larger

effects found in this reanalysis suggests that the effects

of low levels of medication on behavior may be too subtle to

be identified when comparing performance between drug

groups. It suggests that investigating drugs in combination

with perinatal variables that may potentiate their effects

may expose the effects of even low levels of medication.

Although this reanalysis is important as an exploratory

study, the findings must be considered preliminary since

much of the medical information was retrieved

retrospectively for the post hoc analyses. The study has

been criticized for obtaining data on birthlengths from the

medical records (Scanlon, in press). Records of

birthlengths are generally considered to be inaccurate

because of the varying methods used by physicians to measure

babies. This study also suffers from rank-ordering

different types of drugs. This detracts from its clinical

value because it is impossible to delineate which types and

routes of medication are affecting which behaviors.


The Puroose of the Study

The purpose of the present study was to increase our

understanding of the effects of one commonly used type of

obstetric medication on early behavior by examining the

constellation of nonoptimal obstetric conditions found in a

relatively healthy sample of newborns with varying birth

histories. As in the Brackbill et al. and Lester et al.

studies, a multiple regression design was used to tease out

the interaction of the variables and to investigate the

variables which contributed to different behaviors. In

order to avoid the problems of weighting different drugs, no

medication for pain relief other than bupivacaine epidural

anesthesia was included. The dose and time of

administration were the only drug-related variables allowed

to vary in the study. This choice of obstetric medication

was made because bupivacaine epidurals are the drug and

route of administration most commonly used in the Boston

maternity hospitals. Statistics for the United States and

Great Britain have indicated that it is used in at least 40%

of all deliveries (Plumer, 1978; Reynolds, Hargrove & Wyman,


Bupivacaine has become the agent of choice for epidural

deliveries because it has a greater potency and a longer

duration of action than other local anesthetics. It also

has the greatest plasma protein binding capacity and is

highly lipid soluble (Covino, 1971). Bupivacaine has been

found to bind more readily to maternal than to fetal plasma

proteins and, therefore, to have a lower maternal/fetal

concentration ratio than other agents (Reynolds et al.,

1973; Reynolds & Taylor, 1970; Thomas, Long, Moore & Morgan,

1976; Tucker, Boyes, Bridenbaugh & Moore, 1970). The half-

life of bupivacaine is estimated to be eight hours and

approximately five to seven half-lives are needed to

eliminate the drug from the system (Belfrage, Berlin, Raabe

& Thalme, 1975). Although we were able to sample maternal

and umbilical blood for bupivacaine levels in this study, we

were unable to measure bupivacaine after birth because of

hospital restrictions on neonatal sampling.

The present study investigated drug-related variables

which have been found, in the past, to affect neonatal

behavior. They were dose, time of administration and number

of topup administrations of bupivacaine. Other obstetric

conditions were chosen as predictor variables because of

their role in perinatal pharmacology and because of previous

research findings. They were the acid-base status of the

umbilical arterial blood, neonatal ponderal index, maternal

weight-to-height ratio, duration of the administration of

oxytocin, length of the first and second stages of labor and

number of nonoptimal obstetric conditions.

The relevance of the acid-base status to the study of

obstetric medication was discussed previously. Datta,

Brown, Ostheimer, Weiss and Alper (1981) found that acidotic

infants had higher levels of bupivacaine in their umbilical

cord blood samples than infants with normal acid-base


statuses. They also found that acidosis prolonged the half-

life of the drug. Maternal blood pressure was found to

affect behavioral outcome in the Hollmen et al. (1978) and

Woodson and DaCosta-Woodson (1981) studies. In the present

study, the acid-base status or pH of the umbilical arterial

blood was used because its measurement is more reliable than

maternal blood pressure during labor and because the rate of

drug transfer is influenced by the fetal pH. It was

hypothesized that the combination of drug variables and pH

would predict behavior better than the drug alone.

The ponderal index was found to increase the

predictability of behavior in the Lester et al. (1982)

study. The authors suggested that thin babies respond

differently than heavier babies to the same level of

medication since the storage of drugs in fatty tissue

protects the baby from the influence of circulating drugs.

It has also been suggested that a low temperature which is

pronounced in thin malnourished babies slows down the

elimination of drugs (Aleksandrowicz, 1974; Morishima,

Mueller-Heubach and Shnider, 1974). Therefore, in the

present study, it was postulated that the addition of the

ponderal index to the medication variables would increase

the predictability of behavior.

The ratio of the mother's weight-to-height was also

investigated because of the role of fatty tissue in drug

distribution. The mother's height is typically the only

criteria used by physicians to determine the dosage of the


first administration of medication during labor. Dose is

not dependent on body mass. It may be expected, as with a

low ponderal index, that more drugs will circulate in the

baby of a thin mother than in the baby of a heavier mother

since drugs bind to fatty tissues. Therefore, use of this

variable should increase the predictability of behavior.

Oxytocin is given almost routinely with epidural

anesthesia to speed up labor that the anesthesia may slow

down especially when the block is given early in labor.

Most studies have ignored its use probably because the

amount of oxytocin received by the mother is very difficult

to estimate. It is typically administered by a manually

controlled intravenous drip or infusion pump. Any blockage

of the tube caused by a change in the maternal position

during labor may stop the flow of the oxytocin. Because the

use of this drug could not be omitted in the present study,

it was included as a predictor variable. The total amount

of time it was administered was used to estimate the

quantity of oxytocin. A few studies have related the use of

oxytocin to negative results in the baby. Schifrin (1972)

found an increase in late decelerations of fetal heart rate

with the use of oxytocin and/or epidural anesthesia with

lidocaine during labor. He suggested that the late

decelerations were a sign of fetal asphyxia. Investigating

performance on the Brazelton Scale, Murray et al. (1981)

found that a group of Australian babies delivered by

bupivacaine epidural anesthesia with oxytocin performed


worse than a group delivered with bupivacaine alone on the

motor cluster on day 1 and on state range on day 5.

However, Hodgkinson et al. (1977) found no effect of

oxytocin augmentation on Scanlon scores over the first 2

days of life. Ounsted et al. (1978) related lower

neurological scores at 2 months of age and poorer fine motor

performance at 18 months to the use of oxytocin during

parturition. Additionally, Friedman et al. (1978) found

that infants whose mothers received oxytocin during labor

habituated at a slower rate than those not exposed to

oxytocin. In the present study, it was hypothesized that

use of oxytocin would increase the depressing effects of

bupivacaine on behavior.

Length of labor has been shown to interact with drugs

in the Lester et al. (1982), Brackbill et al. (1974a;

1974b), Yang et al. (1976) and Kraemer et al. (1972)

studies. McGrade, Kessen and Leutzendorff (1965)

investigated the relationship between length of labor and

activity level, and found that babies of longer labors were

less active on days 3 and 4. The findings were interpreted

as representing the effects of fatigue after a long labor.

In the present study, it was expected that the use of length

of labor with the other independent variables would further

increase the predictability of behavior.

Previous research has indicated that the physical

status of both the mother and the fetus exerts an influence

on the pharmacological action of the drug (Covino, 1971;

Scanlon, 1974). Furthermore, numerous prenatal and

perinatal risk factors have been related to the behavioral

outcome of the infant (Lipsitt & Field, 1982). In order to

investigate the cumulative effect of obstetric variables

which may influence behavior, a modification of Prechtl's

(1968) scale of optimal obstetric conditions was used.

Prechtl's technique is a self-weighting one based on the

concept that a hazardous obstetric complication will not

occur by itself but exists in conjunction with other

complications and that multiple events place greater stress

on the central nervous system than single events. Parmalee,

Kopp and Sigman (1976) used this technique but revised the

scale to score the total number of obstetric risk factors.

Several studies have found the Parmalee et al. system to be

successful in differentiating high and low risk infants on

different behaviors (Kittner & Lipsitt, 1976; Field,

Dempsey, Hallock, & Shuman, 1978; Sepkoski, Coll, & Lester,

1982; Lester & Zeskind, 1978). In the present study, it was

proposed that the use of a cumulative risk score would add

to the drug variables in the prediction of behavior.

The Brazelton Neonatal Behavioral Assessment Scale was

used in the present study to assess neonatal behavior. This

scale was chosen because it is a comprehensive standardized

tool useful for assessing behavior over the first month of

life. As discussed previously, many studies have found it

to be sensitive to the effects of obstetric medication and

to perinatal risk. In the present study, it was


hypothesized that newborn behavior as measured by the

Brazelton Scale would be sensitive to the combination of

anesthesia and obstetric variables. It was further

hypothesized that the amount of variance accounted for by

these independent variables would differ from day to day and

from behavior to behavior.


Study Site

The study was begun at the largest maternity hospital

in the Boston area, the Boston Hospital for Women, Lying-In

Division (BLI). After approximately three-quarters of the

sample had been recruited, the BLI moved joining with other

hospitals in the area to become the Brigham and Women's (BW)

Hospital. The study was completed at this location.

The BLI/Brigham and Women's Hospital is a private

maternity hospital staffed by obstetricians and nurses from

both individual and group practices. The clientele come

from Boston and the surrounding vicinity and are from middle

and working class backgrounds. During the two year period

of data collection, 13,253 deliveries were performed at this

hospital. Seventy-seven percent (n=10,228) of these were

vaginal deliveries with 91% (n=9307) receiving some type of

anesthesia. Of the vaginal deliveries, 45% (n=4603)

received an epidural, 36% (n=3682) a local, 6% (n=614) a

spinal, and .3% (n=31) inhalation anesthesia.

Subject Recruitment

After receiving permission from the human subjects'

committee to proceed with the study, letters were sent

explaining the study to 55 obstetricians who deliver babies

at the BLI/BW Hospital. One obstetrician denied permission

for his patients' participation in the study; therefore, his


patients were not recruited. (See Appendix 1 for the

summary of obstetricians whose patients were recruited.)

Subjects were selected from mothers who delivered in-

house and lived within 35 miles of the hospital. Selection

criteria included fullterm (38 42 weeks gestation) vaginal

deliveries with bupivacaine epidural anesthesia or no

medication given for pain relief during labor and delivery.

Mothers who received any other medication to reduce pain or

relieve anxiety were excluded from the study. Infants were

selected if they were the product of a singleton birth, had

no congenital malformations or disorders and were admitted

to the regular newborn nursery. In order to reduce the

number of subject variables, only Caucasian infants of

English-speaking mothers were included in the study.

A research nurse at the BLI/BW Hospital recruited women

who fit the subject selection criteria and who would be

delivering during a time when she was available. A maximum

of two subjects per day or three per week were recruited.

Women were approached for consent while in labor because a

maternal blood sample at the time of delivery was requested.

The study was explained and written consent was obtained

(see Appendix 2). Women were recruited late in labor in

order to exclude those who night receive medication other

than epidural anesthesia.

A total of 100 women who were patients of 20 different

obstetricians were asked to participate in the study.

Fifty-six of the women delivered with epidural anesthesia


and 32 delivered without medication. Twelve of the women

received a narcotic or local infiltration late in labor.

They were, therefore, excluded from the study. Four of the

women (2 epidural, 2 nonmedicated) refused consent for the

study for the following reasons: being upset by frequent

requests during labor to participate in research projects,

being upset about receiving intravenous fluid when it was

not requested, wanting no intervention with the baby on the

first day and being too busy due to a disabled child at

home. Because consent was initially obtained during a

difficult time, the study was explained again to mothers

when they were first seen after delivery and a second verbal

consent was requested, One woman (epidural delivery)

withdrew consent at this time because her husband did not

want her to participate.

Twenty-three infants were excluded from the study for

various reasons other than receiving analgesics or local

infiltration. Five were excluded immediately after birth

because they were admitted to the special care nursery:

four (2 epidural, 2 nonmedicated) had respiratory

complications and one had shoulder dystocia (nonmedicated).

Four were excluded after the first Brazelton exam: one had

a club foot (nonmedicated), one had pneumonia (nonmedicated)

and two epiduralss) left the hospital immediately after the

first day and could not be contacted. Fourteen infants were

excluded after completing the study: the initial ten

(epidural) were considered pilot subjects, two subjects (one


epidural, one nonmedicated) received medication for thrush

during the first month, and two (nonmedicated) had surgery

for pyloric stenosis during the first month. (Both parents

of one of the babies with pyloric stenosis were

anesthesiologists. The mother of the other baby had a

history of infertility, miscarriages and a stillbirth.)


The final sample consisted of 60 infants, 38 who were

products of epidural deliveries and 22 of nonmedicated

deliveries. There were 27 girls and 33 boys. The mean

birthweight was 3451 grams (SD=388g.) and mean birthlength

was 50 cm. (SD=1.9 cm.). One minute Apgar scores ranged

from 5 to 9 (x=8, SD=1.0) and 5 minute Apgar scores ranged

from 8 to 10 (x=9, SD=.2).

The mothers in the sample ranged from 18 to 39 years of

age (x=31 years, SD= 3.9). Thirty-three percent of them

were having their first baby. All mothers except for one

received prenatal care for all three trimesters of

pregnancy. The reported number of prescription and over-

the-counter drugs taken during pregnancy ranged from none to

6 and averaged 2 drugs (SD=1.4). All mothers were married

and of intact families. All fathers were present at

delivery. The socioeconomic status of the families

according to the Hollingshead Four Factor Index

(Hollingshead, 1975) ranged from 20 to 66 indicating that

the sample was from middle and working class backgrounds.

All parents had completed high school. Sixty-three percent


of the mothers and 68% of the fathers had completed college.

Of the parents completing college, 45% of the mothers and

56% of the fathers also had graduate degrees.


Brazelton Neonatal Behavioral Assessment Scale

As previously discussed, the Brazelton Neonatal

Behavioral Assessment Scale (Brazelton, 1973) is an

assessment tool which was designed to measure the behavioral

capacities and individual differences of infants during the

first month of life. It is based on the concept that the

neonate is a complexly organized organism adapted to

interact with the environment in such a way to elicit

responses that are necessary for future growth and

development (Als, Tronick, Lester and Brazelton, 1979). For

example, the neonate should be able to elicit attention from

a caregiver, control movements and physiological responses

which might interfere with interaction with a caregiver, and

defend him/herself from negative stimuli in the environment.

Thus, the Brazelton Scale goes beyond traditional

examinations in its assessment of a range of neonatal

behavior which may affect the development of the infant-

caregiver relationship.

The Brazelton Scale consists of 27 behavioral and 17

reflexive items (see Appendix 3) and takes approximately 30

minutes to administer. The exam begins with a two minute

observation of the infant to assess his or her initial state

of consciousness and physical condition. If the baby is


sleeping, the examiner proceeds by repeating presentations

of a light, followed by a rattle, bell and tactile

stimulation of the heel to look for decrement of responses.

These items are excluded if the initial state is not one of

sleep or drowsiness. The rest of the exam consists of

manipulating the baby with progressively stronger tactile

stimulation to observe both elicited and spontaneously

emitted responses of the baby. The mild maneuvers include

elicitation of reflexes of the hands, feet, arms and legs,

the glabella, rooting and sucking. Moderate maneuvers

include undressing, pull-to-sit, cuddling, and standing,

walking, placing, incurvation, and crawling reflexes. The

strongest tactile maneuvers include rotation of the baby to

elicit tonic deviation of the head and eyes, elicitation of

a defensive reaction to a cloth placed over the eyes, the

tonic neck reflex and the Moro. At some point during the

exam when the baby is in an alert state, the infant's

ability to orient to the examiner's face and voice and to a

ball and rattle is observed.

There are several features which distinguish the

Brazelton Scale from other assessments of neonatal behavior.

The goal of the exam is to bring out the baby's best

possible performance, not just the average performance.

Thus, an attempt is made to elicit the best capabilities the

infant has for interacting with the environment. In order

to bring out best performance, it is necessary that the

order of administration of the Brazelton Scale be flexible,


or that items be administered when the examiner feels that

the best performance can be achieved. In order to do this,

it is necessary that the examiner continuously control and

monitor the baby's states of consciousness throughout the

exam. Prechtl and Beintema's (1964) categorization system

for six states is used during the exam (see Appendix 4).

The states range from deep sleep to alertness to robust

crying. The Brazelton Scale specifies the appropriate state

for the administration of each item. It is important that

the examiner be familiar with a range of behavior patterns

that may be expected during the Brazelton so that he or she

can interpret the baby's cues and order the exam to bring

out the baby's best performance. The examiner must go

through a training procedure of first examining and scoring

30 "pilot" babies to appreciate a range of behavior

patterns. He or she then participates in a two day workshop

with a qualified trainer. To achieve reliability on the

Brazelton Scale, the trainee must be judged reliable on its

administration by the trainer, and score within one point on

no less than 24 of the 26 behavioral items and within two

points on the remaining 2 items. The trainee must also

score the 17 reflex items correctly.

Another characteristic which distinguishes the

Brazelton Scale from most neonatal assessments is its

omission of a single index to summarize the infant's

behavior and/or to label the infant's progress. Instead, it

provides a profile of the functioning of several systems of


behavioral organization. To obtain this, after the

administration of the Brazelton exam, the examiner scores

the 17 reflex items on a 4 point scale from absent to

extreme. The 26 behavioral items are scored on scales from

one to nine. Some of these items are assessed from discreet

events during the exam as with the reflexes, i.e., response

decrement items, orientation items, pull-to-sit and the

defensive response, while other items are monitored

continuously and scored during the whole exam, i.e.,

measures of activity level, motor performance, general

tonus, physiological responses, state range and control of

states. These scores may then be reduced to the clustering

system developed by Lester, Als, and Brazelton (1982) which

was discussed in the previous chapter (see Appendix 5).

This system consists of one reflex cluster and six

behavioral clusters. The reflex cluster is the total number

of abnormal reflex scores. To derive the behavioral

clusters, the individual items that are curvilinear are

rescored so that higher scores in all items of the exam

indicate better performance. Six total mean scores are then

derived: 1) habituation, which includes items measuring

decrement of responses to the light, rattle, bell and touch

of the heel; 2) orientation, or responses to the animate and

inanimate auditory and visual stimuli, and overall

alertness; 3)motor organization, or the items assessing

motor maturity and overall muscle tonus; 4) range of state,

which examines the rapidity, peak, and liability of state


changes; 5) regulation of state, or the infant's efforts to

modulate state; and 6) autonomic regulation, which includes

signs of physiological stress, i.e., tremors, startles and

changes in skin color.


A maternal interview was designed to obtain additional

information about the mother and baby (see Appendix 6). It

was adapted in part from a pregnancy interview by Barnard

(1976) and a prenatal drug interview by Doering (1978).

Questions were asked concerning decisions made about labor

and delivery (e.g., use of anesthesia), the amount of time

the mother spent with the baby in the hospital, prenatal

drug intake and the mother's and father's occupation and

education. Also, information which is often missing in the

hospital medical records (e.g., length of labor) was


Obstetric Complicatons Scale (OCS)

The Obstetric Complication Scale designed by Parmalee,

Kopp and Sigman (1976) was used to assess a range of

nonoptimal obstetric conditions (see Appendix 7). The OCS

was derived from Prechtl's (1968) scale of optimal obstetric

conditions. Both scales are self-weighting based on the

concept that a hazardous obstetric condition will not occur

by itself but exists in conjunction with other complications

and that multiple events place greater stress on the central

nervous system than single events. Parmalee, Kopp and


Sigman revised Prechtl's scale to score nonoptimal rather

than optimal obstetric conditions. This scoring system

includes 41 items associated with maternal, parturitional

and fetal nonoptimal conditions. The items are scored as

present or absent. To obtain an OCS score, the total number

of conditions which are absent are divided by the total

number of items scored. This percentage of optimal scores

is converted to a standardized score based on a mean score

of 100 and a standard deviation of 20 (see Manual for

Obstetric Complications, Littman and Parmalee, 1974).

The OCS has been found to be a useful technique for

discriminating babies according to their risk status.

Infants classified as high or low risk according to this

system have been found to differ in cries (Lester & Zeskind,

1978) and heart rate (Kittner & Lipsitt, 1976).

Additionally, the number of maternal, parturitional and

fetal nonoptimal conditions on the OCS, the ponderal index,

gestational age and mother's age were found to significantly

predict six of the seven Brazelton clusters in a group of

Puerto Rican neonates (Sepkoski, Coll, & Lester, 1982). In

another study, it was found that a continuum of risk over

the first year of life could be predicted by both the OCS

and several Brazelton behavioral dimensions (Field, Hallock,

Demsey, Dabiri, & Shuman, 1978).


Blood Samples

Blood samples were taken at delivery with a hepernized

syringe from a vein in the mother's arm and from the vein

and artery of a doubly clamped segment of the umbilical

cord. Portions of the samples were immediately analyzed

using a Radiometer microelectrode to determine acid-base

status (pH). The remaining portions were frozen and later

analyzed using a gas chromatographic technique to measure

the whole blood concentrations of bupivacaine. The maternal

blood sample was not obtained if the woman refused consent

for it or if it was too difficult to obtain during delivery.

A total of 35 maternal blood samples were collected.

Neonatal Assessments

The Brazelton Neonatal Behavioral Assessment Scale was

administered by an examiner blind to the obstetric history

of the baby on days 1, 3, 7 and 28. The examiners were

trained to reliability in both the administration and

scoring of the scale according to the procedures outlined in

the manual by Brazelton (1973).

The first exam was administered between 3 and 6 hours

after birth in an empty postpartum room or in the nursery

examining room. The mother was not present for the initial

exam since she was generally still in the recovery room.

The examining room was darkened due to the sensitivity of

the neonate's eyes. (Silver nitrate is administered to the

baby's eyes immediately after delivery at the BLI/BW



Hospital). Following the exam the baby was measured as

recommended by Miller and Hassanein (1971) by putting

him/her in the tonic neck position in an anthropometer

(Infa-length, Olympic Surgical Co., Seattle).

The second exam was scheduled between 48 and 72 hours

after birth and no less than 4 hours after circumcision.

Exams were administered in the mother's hospital room with

the exception of one baby who went home early and was

examined at home on the fourth day because she left the

hospital prior to 48 hours. Mothers were present for these

exams. After the exam, the tester discussed with the mother

the baby's strongest points, e.g., alertness, activity,

ability to calm, etc. The second Brazelton exam was missed

for one infant who went home early.

Appointments were made by phone to examine the baby at

one week (7 + 1 day) and one month (28 + 2 days) of age in

the home. Often parents, siblings and grandparents were

present for these exams. An effort was made to explain the

exam to all of them (this lesson was learned during a visit

to one of the pilot subjects' homes when the examiner was

punched in the abdomen by a 5 year-old after eliciting a

Moro from his baby brother). Often siblings would attempt

to get the baby to orient to the rattle or their own voice

after the exam was over.

When the examination of the baby at one month was

finished, the mother was given the pregnancy and delivery

interview. After this home visit, additional medical


information was obtained from the mother's and baby's

medical records. Each family was sent a card expressing

gratitude for their participation in the study. When the

data analyses were completed, abstracts of the study were

sent to each family.


Characteristics of the Sample

Bupivacaine Variables

Thirty-eight of the women in the sample received a .5%

concentration of bupivacaine (without epinephrine) in an

isotonic solution of sodium chloride for relief of pain

during labor. The solution was injected at appropriate

intervals through a catheter which was inserted into the

mother's lumbar epidural space when her cervix was

appropriately 5 to 6 cm dilated. As illustrated in Table 3,

the first dose of the drug was administrated to these women

between one and ten hours prior to delivery (x = 3.5 hrs.,

SD = 2.1). One to six doses were administered (x = 2.5, SD

= 1.2) with the total amount of bupivacaine received by this

sample ranging from 50 to 290 mg. (x = 112.7 mg., SD =


Analyses of the whole blood concentrations of

bupivacaine showed that the amount of drug in both the

umbilical venous (UV) and umbilical arterial (UA) samples

ranged from .04 to .44 ug/ml. The mean UV concentration of

bupivacaine was .12 ug/ml (SD = .07) and the mean UA

concentration was .09 ug/ml (SD = .09). The concentration

of bupivacaine in the maternal venous (MV) blood samples

ranged from .04 to .88 ug/ml and averaged .28 ug/ml (SD =

.2). The ratio of the UV/MV concentrations ranged from .04

to 1.0 and averaged .33 (SD = .24).



Table 3. Descriptive Statistics of Bupivacaine Variables.

Mean Standard Range


Dose (mg) 112.7 45.6 50-290

Number of Administrations 2.5 1.2 1-6

Time of First 3.5 2.1 1-10
(hours prior to delivery)

UV concentration (ug/ml) 0.12 0.07 0.04-0.44

UA concentration (ug/ml) 0.09 0.09 0.04-0.44

MV concentration (ug/ml) 0.28 0.20 0.04-0.44

UV/MV concentration ratio 0.33 0.24 0.04-1.0

As seen in Table 4, Pearson product-moment correlations

were run between the following bupivacaine variables which

had been preselected as predictors for regressions: total

dose, number of administrations, time of administration to

delivery and umbilical venous concentration. The

intercorrelations were high, ranging from .54 to .91 (x =

.72). Because of the high correlations among these

variables, Fischer Z score transformations were also

performed to obtain a mean correlation. The mean result was

.746. Thus, because a high portion of the variance was

shared by these variables, only the dose of bupivacaine was

retained as a predictor variable for the regressions. Dose

was chosen over the other variables because it was the

single variable most highly correlated with the other three

and because it has been related to behavior in studies of

the effects of obstetric medication more frequently than the

other variables.

Data collected from the maternal interviews indicated

that half of the women receiving anesthesia decided on it

prior to labor and half decided during labor. Fifty-five

percent of these women indicated that they made the decision

to have the epidural themselves. Thirteen percent said that

their doctors made the decision and 32% said that it was a

joint decision between themselves and their doctors.

Complete analgesia was achieved in 74% of the medicated


Table 4. Correlations Between Bupivacaine Variables.

Dose Number Time

Dose (mg)

Number of Administrations

Time of First
(hours prior to delivery)

UV concentration (ug/ml)


.75 .66 .54 1.0

deliveries and partial analgesia in 18% of the deliveries.

Eight percent of the women said that they received no pain

relief from the epidural. Regarding plans for future

deliveries, 61% of these women said they would want the same

medication, 26% would want no medication, 5% would want less

medication and 3% would want more.

In contrast, 77% of the 22 women who received no

medication for pain relief said that they had decided on

this prior to the beginning of labor and 23% decided not to

have medication during labor. Seventy-seven percent

reported that they made this decision themselves, 5%

reported that it was made by their doctors and 18% made the

decision with their doctors. Only one of the nonmedicated

mothers said that she would like to have medication for the

relief of pain in future deliveries.

Predictor Variables

Table 5 illustrates the characteristics of the

preselected predictor variables for the entire sample of 60

babies. As determined by analyses of the umbilical arterial

blood samples, the neonatal pH ranged from 7.14 to 7.45 (x

7.28, SD = .07). Eight babies were mildly acidotic (pH =

7.2 7.24), and seven were severely acidotic (pH < 7.19).

One baby had a pH less than 7.14.

The ponderal index of this sample ranged from 2.2 to

3.48 with a mean of 2.72 (SD = .23) which falls at

approximately the 75th percentile according to Miller and

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs