Predicting maternal adherence and child health status in childhood epilepsy

MISSING IMAGE

Material Information

Title:
Predicting maternal adherence and child health status in childhood epilepsy
Physical Description:
Book
Creator:
Fritz, Sobha P., 1975-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 30435808
oclc - 54703266
System ID:
AA00022328:00001

Table of Contents
    Title Page
        Page i
    Dedication
        Page ii
    Acknowledgement
        Page iii
    Table of Contents
        Page iv
        Page v
        Page vi
    Abstract
        Page vii
        Page viii
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
    Chapter 2. Aims and study justifications
        Page 21
        Page 22
        Page 23
        Page 24
    Chapter 3. Methods
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
    Chapter 4. Results
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
    Chapter 5. Discussion
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
    Appendix A. Background information form - Time 1
        Page 71
        Page 72
        Page 73
        Page 74
    Appendix B. Background information form - Time 2
        Page 75
        Page 76
    Appendix C. Medical chart information form - Time 1
        Page 77
    Appendix D. Medical chart information form - Time 2
        Page 78
    List of references
        Page 79
        Page 80
        Page 81
        Page 82
    Biographical sketch
        Page 83
        Page 84
        Page 85
        Page 86
Full Text










PREDICTING MATERNAL ADHERENCE AND CHILD HEALTH
STATUS IN CHILDHOOD EPILEPSY: AN EXPLORATORY MODEL














By

SOBHA P. FRITZ


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

UNIVERSITY OF FLORIDA


2003


























This document is dedicated to my husband, parents, and sisters for their support and
unconditional love. My family's emphasis regarding the importance of education and
helping others has guided me into a career for which I am well suited. I would like to
especially thank my husband, Jason Jon Fritz, for always believing in me and providing a
secure foundation of warmth and encouragement.














ACKNOWLEDGMENTS

This investigation could not have been carried out without the mentorship and

direction of Robert Glueckauf, Ph.D., the valuable contributions ofNicole Jagusztyn, the

dedication of Paul Carney, M.D., and the staff at the University of Florida Pediatric

Neurology Clinic. I would like to acknowledge my committee members for their

scholarly assistance and guidance. I would additionally like to thank my family for their

support and encouragement throughout this project.














TABLE OF CONTENTS
Page

ACKN O W LED GM EN TS ............................................................................................ iii

ABSTRA CT...................................................................................................................... vii

CHAPTER

1 IN TRODUCTION ........................................................................................................ 1

Epilepsy: An Overview ................................................................................................. I
Etiology.........................................................................................................................2
D iagnosing Epilepsy..................................................................................................... 3
Classification of the Epilepsies..................................................................................... 4
Partial Seizures ...................................................................................................... 4
Generalized Seizures ............................................................................................. 6
Treatm ent......................................................................................................................8
Psychological Im pact of Epilepsy ................................................................................ 9
A dherence to the Regim en: An Overview .................................................................. 10
Adherence to the Regim en: Epilepsy ......................................................................... 14
M aternal Em otional Functioning................................................................................ 18

2 A IM S AN D STUD Y JU STIFICATION S .................................................................. 21

Purpose .......................................................................................................................21
Study M odel................................................................................................................22
Hypotheses..................................................................................................................23
Maternal Emotional Distress, Adherence, Child Health Status .......................... 23
Maternal Perceptions of the Treatment Regimen, Adherence, Child Health
Status................................................................................................................ 23
Condition and Regimen-related Variables, Adherence, Child Health Status......23
Demographic Characteristics, Adherence, Child Health Status.......................... 24

3 M ETH OD S ................................................................................................................ 25

Study Design...............................................................................................................25
Participants .................................................................................................................26
M measures .....................................................................................................................29
Participant Background Inform ation................................................................... 29
M medical Chart Inform ation.................................................................................. 30








Child Health Status and Adherence..................................................................... 31
Maternal Emotional Distress............................................................................... 32

4 RESULTS ................................................................................................................... 35

Preliminary Analyses.................................................................................................. 35
Primary Analyses........................................................................................................ 38
Correlations among Background Variables, Predictors, Criterion Variables......38
Hypothesis 1: Maternal Emotional Distress, Adherence, Child Health Status ...40
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence,
Child Health Status .......................................................................................... 42
Hypothesis 3: Condition and Regimen-related Variables, Adherence, Child
Health Status .................................................................................................... 46
Hypothesis 4: Demographic Characteristics, Adherence, Child Health Status...49
Model Testing...................................................................................................... 51
Prediction of adherence................................................................................ 51
Prediction of health status ............................................................................ 52
Summary of Findings ................................................................................................. 55

5 DISCUSSION............................................................................................................. 57

Sample Characteristics of Demographics and Maternal Distress............................... 57
Hypotheses.................................................................................................................. 58
Hypothesis 1: Maternal Emotional Distress, Adherence, Child Health Status ...58
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence,
Child Health Status .......................................................................................... 60
Hypothesis 3: Condition and Regimen-related Variables, Adherence, Child
Health Status .................................................................................................... 61
Hypothesis 4: Demographic Characteristics, Adherence, Child Health Status...63
Model Testing............................................................................................................. 64
Study Strengths and Limitations................................................................................. 65
Future Directions ........................................................................................................ 67
Clinical Implications................................................................................................... 69








APPENDIX

A BACKGROUND INFORMATION FORM TIME 1 .............................................. 71

B BACKGROUND INFORMATION FORM TIME 2 .............................................. 75

C MEDICAL CHART INFORMATION FORM TIME 1 ......................................... 77

D MEDICAL CHART INFORMATION FORM TIME 2 ......................................... 78

LIST O F REFEREN CES................................................................................................... 79

BIOGRAPHICAL SKETCH............................................................................................. 83














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

PREDICTING MATERNAL ADHERENCE AND CHILD HEALTH STATUS IN
CHILDHOOD EPILEPSY: AN EXPLORATORY MODEL

By

Sobha P. Fritz

December 2003

Chair: Robert L. Glueckauf
Major Department: Clinical and Health Psychology

Epilepsy is a chronic illness that typically begins in early childhood. Poor

adherence to the medication regimen can result in increased seizure frequency and

severity. Young children are reliant on their parents for their daily routine and

functioning. This issue is particularly salient for children with seizure disorders, who

rely on their parents to take anti-convulsant medicationss. Depressive symptomatology

and maternal stress have been associated with negative child status. Maternal emotional

factors such as depression and stress may play an important role in adherence to

medication for younger children with epilepsy. In the current study, we examined the

relationship between maternal adherence to medication (pill count percentage) and child

health status (seizure frequency, seizure-related emergency room visits) for children with

epilepsy and the following factors: maternal emotional distress (maternal depression,

maternal stress), demographics (family annual income, race, maternal educational level),

condition-related and regimen-related variables (disease type, disease duration, type and








complexity of the medication regimen, concordance between prescribed regimen and

reported regimen), and maternal perceptions of the treatment regimen (perceived

efficacy, perceived knowledge, perceived difficulty).

Research participants included children, ages one through eleven, diagnosed with

epilepsy and were approached for study participation at the University of Florida

Pediatric Neurology Clinic. Eighty-two mother-child dyads completed the study.

Results indicated that as maternal stress, parent-child dysfunction, and maternal

perceived regimen difficulty increased, child seizure frequency significantly increased.

As maternal perceived regimen efficacy levels increased, child seizure frequency

significantly decreased. Results revealed that as maternal perceived regimen difficulty

levels increased, seizure-related emergency room (ER) visits significantly increased. As

regimen concordance rates increased, seizure-related ER visits significantly decreased.

Results demonstrated that as concordance rates increased, pill count increased.

Final hierarchical regression analyses suggested that parent-child dysfunction was

predictive of seizure frequency. Maternal perceived regimen difficulty ratings and

concordance ratings between the prescribed and reported regimen were predictive of

seizure-related ER visits. Concordance ratings between the prescribed and reported

regimen were predictive of pill count percentage. These findings suggest that the

identification of specific variables that influence maternal adherence and child health

status may be particularly helpful for children with epilepsy.













CHAPTER 1
INTRODUCTION

Epilepsy: An Overview

Epilepsy is a chronic illness that typically begins in early childhood in which the

patient is prone to experience recurrent epileptic seizures, where two or more attacks

have occurred. Epilepsy is defined as a manifestation of an abnormal and excessive

synchronized discharge of a set of cerebral neurons (Shorvon, 2000). Clinical

manifestations are sudden and transient and can include a wide variety of motor, psychic,

and sensory phenomena with or without alteration in consciousness or awareness. The

specific symptoms that occur depend on the part of the brain involved. Status epilepticus

is defined as a condition in which epileptic seizures continue, or are repeated without

consciousness, for a period of 30 minutes or more and this occurrence often requires a

visit to the emergency room. Untreated status epilepticus can be fatal.

Provoked seizures have an obvious and immediate preceding cause, such as an

acute systemic, metabolic, or toxic insult or an acute cerebral event (stroke, trauma, and

infection). Provoked seizures are often isolated events and do not tend to recur once the

cause is removed. A person is described as having active epilepsy when at least one

seizure has occurred in the preceding 2-5 years. Epilepsy is said to be in remission when

no seizures have occurred in the preceding 2-5 years (Appleton & Gibbs, 1998). There

are approximately 80 cases of epilepsy per 100,000 people per year, with a range of 50-

120 cases. Nearly 75 per 100,000 adolescents per year go on to develop some form of

epilepsy. The cumulative incidence, risk of an individual developing epilepsy in his or








her lifetime, is between 3% and 5%. The highest incidence rates are observed in

neonates and young children with a second peak in old age. Greater than 50% of seizures

start in childhood and begin before age 20 (Browne & Holmes, 2000).

The prevalence of active epilepsy is 5-10 cases per 1,000 people per year. Isolated

seizures (first and only) occur in approximately 20 people per 100,000 each year. The

prevalence of epilepsy in school age children (aged 5-17 years) is 0.7%-0.8%. The

prevalence of epilepsy in adolescents is 6-7 per 1,000 adolescents. One-third of all

epilepsies that begin in childhood will have spontaneously remitted by puberty and this

change is usually sustained throughout adult life. Epilepsy is relatively static after early

childhood but shows a tendency to rise in old age. Epilepsy is more common in

developing countries and is thought to be associated with poorer nutrition, poorer

hygiene, higher prevalence of infectious diseases, and a higher proportion of children in

the population (Appleton & Gibbs, 1998).

Etiology

Congenital and genetic conditions are most commonly related to early childhood

epilepsy. Genetics are thought to play an important role in seizure disorders. Hereditary

and genetic factors are related to idiopathic or primary epilepsy syndromes and possibly

in some secondary epilepsies, but known single-gene disorders are related to epilepsy in

only 1-2% of cases. In older children and young adults, common causes of epilepsy

include inherited predisposition, hippocampal sclerosis, alcohol or drug abuse, and

trauma. Vascular disease is a predominant cause of epilepsy in the elderly. Tumors and

infections can occur at any age but malignant tumors are more likely to occur after age 30

(Shorvon, 2000). In 70-75% of childhood cases of epilepsy, a specific cause may not be








noted, but the etiology of the disease may be apparent from clinical information and test

results (Appleton & Gibbs, 1998).

Patients with specific syndromes such as Fragile X syndrome, Sturge-Weber

syndrome, Huntington's, Niemann-Pick disease type C, tuberous sclerosis (Boumrneville's

syndrome), neurofibromatosis (type 1 and type 2), and Down's syndrome may develop

epilepsy. Infectious diseases that involve the brain such as viral and bacterial meningitis

also lead to epilepsy. Cerebral tuberculosis frequently causes epilepsy. Malaria, strep

infections, and encephalitis are all thought to be related epilepsy. Cerebral tumors are

related to approximately 6% of newly diagnosed cases of epilepsy. In children, tumors

are not typical causes of epilepsy. Head injuries account for approximately 2-12% of all

cases of epilepsy (traffic accidents and falls). Drugs, alcohol, toxins, and metabolic

disorders have been associated with seizure disorders. Additional causes of epilepsy

include brain malformations, unruptured aneurysms, intracranial hemorrhages, lesions,

perinatal injury, and hippocampal sclerosis. Hippocampal sclerosis is the most common

cause of temporal lobe epilepsy (Shorvon, 2000).

Diagnosing Epilepsy

A physician needs detailed information to render an epilepsy diagnosis and a

diagnosis of epilepsy is dependent almost entirely on the history given. During the

assessment the doctor will typically gain the following information: a very detailed

history of the seizure(s), an eyewitness account of the seizure(s), a history of previous

illnesses or traumatic events, family history, and history of any co-occurring diseases)

and treatmentss. Noteworthy is the fact that an accurate eyewitness account is essential

and may be the only history available for young children. A physical examination will be

conducted. Video recordings, sleep studies, and EEGs of suspected seizures are useful








aids for diagnosis. Some characteristics that may mimic epilepsy include headaches,

jitters, tic disorder, shuddering, shivering, repetitive ritualistic movements, vacant stares,

daydreaming, inattention, psychological disturbances, and excessive sleep (Browne &

Holmes, 2000).

Classification of the Epilepsies

Seizures are classified into two main categories: partial and generalized. These two

categories are further classified. Partial seizures include simple partial, complex partial,

and partial seizures evolving to secondary generalized. Generalized seizures are

classified as typical absence (petit mal), atypical absence, myoclonic, clonic, tonic, tonic-

clonic (grand mal), atonic, and unclassifiable (Appleton & Gibbs, 1998).

Partial Seizures

Partial seizures are further classified as "simple," in which consciousness is

retained, or "complex," in which consciousness is impaired or lost. Simple partial

seizures (beginning locally) can occur with motor symptoms, sensory symptoms,

autonomic symptoms, and psychic symptoms. Simple partial seizures with sensory,

autonomic, or psychic symptoms may be easily overlooked in younger children, who may

be unable to describe such symptoms. Partial seizures may become secondary

generalized, resulting in tonic-clonic (grand mal) convulsions (Browne & Holmes, 2000).

Simple partial seizures last only a few seconds and the symptoms that occur depend

on the part of the cortex involved. Simple partial seizures can include motor

manifestations such as jerking or spasms, which usually occur from frontal or central

region epilepsies. Simple partial seizures can include somatosensory manifestations such

as simple hallucinations that can take the form of tingling or numbness, an electrical

shock-like feeling, burning, pain, or a feeling of heat. The epileptic focus is usually in








the central or parietal regions. Simple visual phenomena such as flashing lights and

colors usually occur if the calcarine cortex is affected. A rising epigastric sensation is the

most common manifestation reported, which typically occurs when the mesial temporal

lobe is affected. Simple partial seizures can also include autonomic manifestations such

as changes in skin color, blood pressure, heart rate, and pupil size and these changes

usually are related to generalized or complex partial seizures of frontal or temporal

origins. Psychic manifestations are common as well. Psychic auras can take various

forms and are more common in complex partial seizures. They usually occur from

epilepsy arising from a temporal, frontal, or parietal focus. There are six principal

categories of auras: dysphasic symptoms (cortical or speech areas are effected),

dysmnestic symptoms (disturbance of memory), cognitive symptoms (dreamy states and

depersonalization), affective symptoms (fear, depression, anger, irritability, and elation),

illusions of size, shape, weight, distance or sound, or structural hallucinations (visual,

auditory, gustatory, or olfactory forms) (Appleton & Gibbs, 1998).

Complex partial seizures have three main components: auras, altered

consciousness, and automatisms. The auras are equivalent to those of simple partial

seizures and usually last a few seconds, but in rare cases a prolonged aura persists for

minutes, hours, or even days. Altered consciousness may follow the aura and can include

motor arrest, during which the patient is motionless. The patient may be described as

vacant or glazed and often associated is a spasm, posturing, or mild jerking.

Automatisms are involuntary motor actions that occur during or in the aftermath of

epileptic seizures, in a state of unconsciousness where there is total amnesia for the

events.








Partial seizures (simple, complex, or simple evolving to complex) may spread to

become generalized. The partial seizure is often experienced as an aura in the seconds

before the generalized seizure. The generalized seizure is usually tonic-clonic, tonic, or

atonic (Browne & Holmes, 2000).

Generalized Seizures

There are two types of absence seizures: typical absence (petit mal seizures) and

atypical absence. In typical absence seizures there is an abrupt and sudden loss of

consciousness, cessation of all motor activity, tone is usually preserved, there is no fall,

the patient is unaware and inaccessible, the patient often appears glazed or vacant, the

attack ends as abruptly as it started, and previous activity is resumed as if nothing

happened. In atypical absence seizures, the duration is longer, loss of awareness is often

incomplete, associated tone changes are more severe, and the onset and cessation of

attacks are not as abrupt (Appleton & Gibbs, 1998).

With myoclonic seizures, there is a brief contraction of a muscle, muscle group, or

several muscle groups that can be single or repetitive, varying in severity from a nearly

imperceptible twitch to a severe jerking which can result in a sudden fall or the

propulsion of handheld objects (flying saucer syndrome). Clonic seizures consist of

asymmetrical and irregular jerking. Clonic seizures are most frequent in neonates,

infants, and young children. Tonic seizures are associated with muscle contractions with

altered consciousness. In tonic seizures, the contraction causes extension of the neck,

contraction of facial muscles, widening of the eyes, up-turning of the eyeballs,

contraction of respiratory muscles, and spasms of the upper limb muscles (Browne &

Holmes, 2000).








Tonic-clonic seizures are most associated with the type of epilepsy that the public

imagines. The tonic phase is sometimes preceded by a period in which an attack is

anticipated (prodromal period) defined by a vague feeling or by the occurrence of

increasing myoclonic jerking. If an aura occurs, this indicates that the tonic-clonic

seizure is secondary generalized. The seizure is initiated by loss of consciousness where

the patient will fall if standing and there is a brief period of tonic flexing and then a

longer phase of rigidity, with the eyes rolled up, the jaw clamped shut, the limbs stiff, and

the fists clenched, then respiration ceases. The clonic phase of tonic-clonic seizures has a

tonic stage that lasts 10-30 seconds and is followed by a clonic phase, during which

convulsive movements occur, usually of all four limbs and the jaw and facial muscles.

Breathing may be labored and saliva may froth from the mouth containing blood from

tongue-biting. The convulsive movements decrease in frequency, typically to about 4

clonic jerks per second, and increase in amplitude as the attack progresses. Autonomic

features (flushing, changes in BP, changes in pulse, and increased salivation) are

common. The clonic phase lasts 30-60 seconds and is followed by further brief tonic

contraction of the muscles, sometimes accompanied by incontinence. The final phase

lasts between 2-30 minutes and is characterized by flaccidity of muscles and then

consciousness is slowly regained (Browne & Holmes, 2000).

The most severe form of the atonic seizure is the classic drop attack in which all

postural tone is suddenly lost, causing the person to collapse. In atonic seizures, tone can

be more restricted, resulting in bobbing of the head, a bowing movement, or sagging at

the knee. These seizures are short and are usually followed by immediate recovery.








At least 1/3 of all seizure disorders are not classifiable and fall into the

unclassifiable category. Seizures are described as unclassifiable because typical clinical

and EEG patterns do not conform to the defined categories of seizures (Browne &

Holmes, 2000).

Treatment

Various drug treatments for epilepsy are available. One of the most difficult

aspects of drug treatment is adherence to medication. Some issues that arise are when to

start drug therapy, which drug to use, what dose to give, when to change the drug and/or

add another drug, when to stop a drug, and when to measure blood levels of the drug.

Drug treatments can vary from dosing once or twice a day to several times a day and drug

types include pills, liquids, sprinkles, chewables, suppositories, and inhalers (Appleton &

Gibbs, 1998).

Most physicians would recommend beginning drug treatment after a cluster of

seizures. Once a drug regimen is started, the goal is to achieve seizure control without

extreme side effects. Before adding additional medications, the dose of the initial drug

should be increased to the maximum tolerated level. Currently, the recommended first-

line of drugs in treating childhood epilepsy is sodium valproate for generalized

epilepsies, and carbamazepine for partial seizures. The second line of defense includes

carbamazepine, gabapentin, lamotrigine, and topiramate for generalized seizures, and

vigabatrin, gabapentin, lamotrigine, and topiramate for partial seizures. The third line of

defense includes phenytoin, ethosuximide, clonazepam, phenobarbitone, clobazam,

carbamazepine, acetazolamide, and topiramate for generalized seizures, and sodium

valproate, clobazam, and phenytoin for partial seizures. Many physicians take a polydrug








treatment approach of epilepsy treatment, specifically for those patients who do not

respond well to first, second, and third line treatment approaches (Shorvon, 2000).

Surgical treatment is an option for patients with severe intractable epilepsy that

fails to respond to medication and other forms of treatment. Alternative methods of

treatment include the ketogenic diet, steroids, intravenous immunoglobulins, behavior

therapy, acupuncture, herbal therapy, vagus nerve stimulation, and seizure alert dogs.

Psychological Impact of Epilepsy

Psychiatric disorders occur more frequently with epilepsy than with any other

chronic disorder (Appleton & Gibbs, 1998). Children and adolescents with poorly

controlled epilepsy have more restrictions imposed upon them to protect them from

getting injured during a seizure. They may be unable to participate in activities with their

peers such as swimming, driving, cycling, climbing, and sports. These restrictions can

result in fewer contacts with peers. Social stigma concerns such as teasing, isolation, and

poor social contact have been associated with increased restrictions. Restrictions,

depression, and isolation have been associated with poorer adherence in children and

adolescents with epilepsy (Kyngas, 1990). Poorer adherence to medication is associated

with increased seizure frequency, resulting in a larger number of epilepsy-related injuries

and deaths. One preventable cause of increased seizure frequency and its associated

injuries is taking the anti-epileptic medication as prescribed. Adherence to the medical

regimen becomes an important factor in decreasing seizure frequency.

Epilepsy is a disorder that is associated with a high degree of uncertainty. This

uncertainty and sudden loss of control have been associated with anxiety and depression

in patients with epilepsy including adults, children, and caregivers (Kyngas, 2000).

Increased stress levels and anxiety have been demonstrated as related to an increase in








seizure frequency and seizure severity (Spector et al., 1999). The effects of maternal

emotional adjustment and adherence with young children with epilepsy have received

limited attention.

Adherence to the Regimen: An Overview

One-third of all patients and 50-55% of chronically ill patients are non-adherent to

their prescribed medical regimens (Rapoff, 1999). Non-adherence can negatively impact

health status and health-related quality of life and can lead to increased risks of morbidity

and mortality. Non-adherence is also related to increased health care costs, where the

cost of non-adherence in the United States is nearly $100 billion each year (Rapoff,

1999). Additionally, non-adherence may inadvertently negatively impact clinical

decisions made by physicians. Physicians who attribute disease-related problems to

ineffective or incorrect prescriptions and not to the patient's adherence problems may

change the patient's medical regimen unnecessarily or inappropriately.

Identifying correlates of adherence is important because they facilitate the

development of risk profiles, medical regimens, and future research using matched

samples. Patient and family factors, disease-related factors, and treatment regimen

factors have been identified as correlates of adherence (Rapoff, 1999). The majority of

studies have focused exclusively on patient and family factors and their relationship to

adherence.

Patient/family correlates include demographics (child age, parental education level,

work status, number of children in the home, and socioeconomic status), knowledge

about disease course and treatment, as well as adjustment and coping. Previous research

has shown that patient and family demographic variables have a significant influence on

child and maternal adherence. Younger children with cancer, cystic fibrosis, diabetes,








and renal disease are more likely to be adherent to their medication regimen compared to

adolescents (Anderson, Auslander, Jung, Miller, & Santiago, 1990; Beck et al., 1980;

Bond, Aiken, & Somerville, 1992; Brownbridge & Fielding, 1994; Gudas, Koocher, &

Wyplj, 1991; Jacobsen et al., 1987; LaGreca, Follansbee, & Skyler, 1990; Patterson,

1985; Smith, Rosen, Truworthy, & Lowman, 1979; Tebbi et al., 1986). Boys tend to be

less adherent than girls to regimens for cystic fibrosis and diabetes (Lorenz, Christensin,

& Pichert, 1985; Patterson, 1985). Lower socioeconomic status (SES) and lower parental

education levels have been related to non-adherence to regimens for asthma, cystic

fibrosis, diabetes, and renal disease. Patients in larger families or whose mothers work

outside of the home tend to be less adherent to cancer and cystic fibrosis regimens

(Patterson, 1985; Tebbi et al., 1986).

Knowledge about the treatment regimen is an interesting and important factor in

predicting adherence to treatment. However, studies that have examined this factor have

shown an inconsistent pattern of results. Several investigators have reported that patients

who are less knowledgeable about their specific condition and treatment regimen tend to

be less adherent to regimens for cancer, cystic fibrosis, and diabetes (Gudas et al., 1991;

LaGreca et al., 1990; Tebbi et al., 1986) whereas knowledge appears to be unrelated to

adherence for patients with renal disease (Beck et al., 1980). Maternal knowledge for

children with diabetes was positively related to adherence for preadolescents and

unrelated to adherence for adolescents (LaGreca et al., 1990).

The relationship of adjustment and coping to adherence has been examined across a

variety of childhood disorders. Poor parental coping has been associated with non-

adherence to medical regimens for children with juvenile rheumatoid arthritis and renal








disease. Increased parental depression has been related to poor adherence to renal disease

regimens (Brownbridge & Fielding, 1994; Wynn & Eckel, 1986). Greater parental

anxiety and perceived stress have been associated with lower adherence to anti-epileptic

medications for children with epilepsy (Hazzard, 1990). Parental compliance behavior is

related to perceptions of the medications' efficacy as well as its side effects (Shope,

1988). Austin (1989) suggests that attitudes toward giving anti-epileptic medications and

behavioral intentions are important variables in the prediction of parental compliance

behavior. Parental knowledge about the efficacy and side effects of medications

influences parental attitudes towards medications and improves their medication-giving

behavior (Austin, 1989).

Patients who lack parental monitoring of regimen procedures or have families

where the role of the parental monitor is ambiguous tend to be less adherent to their

prescribed medical regimen. Parental monitoring has been found to be a key factor in

improving medication adherence for adolescents with chronic illness (Rapoff, 1999).

Condition-specific correlates include duration, course, symptoms, and perceived

severity of the condition. Diseases with a longer duration are related to poorer adherence.

Poorer adherence to regimens for diabetes, juvenile rheumatoid arthritis, and renal

disease has been related to longer illness duration. The relationship of disease course and

adherence has not been specifically examined for children with chronic illness (Rapoff,

1999). Condition-specific symptoms are an important factor to examine in their

relationship to adherence. Increased health problems and hospitalizations are related to

poorer adherence for patients with renal disease (Beck et al., 1980). Greater seizure

activity has been related to decreased adherence to anticonvulsant medications for








children with epilepsy (Hazzard, 1990). Maternal perceptions of higher disease severity

have been associated with better adherence to asthma medications (Radius et al., 1978).

Medication regimen-related correlates include type and complexity, costs, side

effects, and perceived efficacy of the medication regimen. Patients with more complex

regimens tend to be less adherent to their prescribed regimen in that more complex

regimens are more often associated with polydrug treatments, with higher numbers of

medication types and dosages. This is particularly true for patients with cystic fibrosis,

diabetes, and juvenile rheumatoid arthritis. Treatment costs can be related to non-

adherence especially for those patients and families who are economically disadvantaged.

Those medical regimens that produce a greater number of negative side effects tend to be

related to poorer adherence to the regimen. Patient and parental perceptions of

medication efficacy are especially related to adherence. Higher levels of perceived

benefits rated by patients and parents are associated with better adherence to regimens for

asthma and diabetes (Bobrow, Avruskin, & Siller, 1985; Bond et al., 1992; McCaul,

Glasgow, & Schafer, 1987; Radius et al., 1978).

In summary, important correlates of adherence for children and adolescents with

chronic illness include patient/family factors (demographics, knowledge, adjustment and

coping, parental monitoring), disease factors (duration, course, symptoms, perceived

severity), and regimen factors (type and complexity, costs, side effects, efficacy). These

summarized findings demonstrate that correlates of adherence to medical regimens are

critical in improving adherence. The majority of studies conducted to identify correlates

of adherence to medical regimens for children with chronic disease have focused on the

following: asthma, cancer, cystic fibrosis, diabetes, and renal disease. Studies examining








the specific correlates of adherence for children and adolescents with epilepsy are limited

in number and require further examination.

Adherence to the Regimen: Epilepsy

The vast majority of people with epilepsy begin having their seizures before the age

of 20, and more than 50% of epilepsy cases begin in childhood. Thus, most people

experience their first seizure at a time that is critical to the acquisition and development

of basic cognitive and social competencies, which are crucial for long-term academic,

interpersonal, and vocational adjustment. Due to the long-term nature of epilepsy and its

related medication regimen, medication adherence to anti-epileptic drugs (AEDs) is

extremely important from the time of diagnosis and throughout the course of treatment.

Although correlates of adherence have been identified in a wide range of pediatric

chronic illnesses, only a few studies have been conducted to date on adherence to

treatment among children and adolescents with seizure disorders. We continue to lack

basic information about the psychological, environmental, and familial factors that

influence adherence in children with epilepsy. Furthermore, there is only a modicum of

outcome research on intervention approaches to improve adherence in this at-risk

population. A few factors have been identified that are specifically related to individuals

with seizure disorders.

Well-controlled seizure patients are asymptomatic, which may gradually contribute

to decreased adherence to the medication regimen. Furthermore, nonadherence does not

necessarily lead to immediate symptomatology for patients with epilepsy. Pediatric

seizure patients are typically on long-term medication regimens, for which adherence is

worse than for short-term regimens (Becker & Maiman, 1975; Davis, 1966; Rapoff&

Christopherson, 1982).








A demographic factor generally associated with parental non-adherence is lower

socioeconomic status. Additionally, ethnic minority patients have been found to be more

likely to be nonadherent to their child's medical regimen (Becker & Maiman, 1975;

Rapoff & Christopherson, 1982).

Conducting research on adherence with pediatric seizure patients is made difficult

by the methodological problems involved in assessing patient's adherence. Although

blood levels of anticonvulsant medications can be obtained, it is difficult to translate

blood levels into comparable measures of compliance for patients on different

medications and doses. Therapeutic ranges vary for different medications and vary

according to dosage, time since the medication was ingested, and individual metabolic

rates, and hormonal changes (which are especially salient for adolescent females).

Furthermore, for patients taking more than one medication, there are medication

interactions, with some medications altering the blood level concentrations of other

medications. Because of these complications, studies examining pediatric patient

adherence to anticonvulsant medication regimens have tended to assess only one

medication level or have used group differences in blood levels to assess the validity of

other compliance measures (self report, pill counts). With these limitations, estimates of

parental adherence to anticonvulsant medication regimens for pediatric patients ranges

from 25% to 54% (Zysset, et al., 1981). These estimates are generally consistent with

adherence rates from other pediatric populations where a long-term medication has been

prescribed (Sackett & Haynes, 1976).

Only three studies have assessed factors related to parental medication adherence

with pediatric seizure patients (Freidman et al., 1986; Peterson et al., 1982). The studies








assessed at least some factors from the demographic, illness, treatment regimen, and

attitudinal domains. However, the Peterson et al. study focused primarily on adults, with

only 18 adolescents in the sample. Demographic variables were not demonstrated to be

associated to adherence in the pediatric seizure disorder populations in the two studies

(Freidman et al., 1986; Peterson et al., 1982). In terms of illness variables, Peterson et al.

(1982) found that patients experiencing more seizures in the past year, more recent

seizures, or having tonic-clonic seizures were more likely to be adherent. The only

treatment-related variable related to adherence for children who had an active role in

taking their medication, particularly adolescents, was perceived personal independence

(Freidman et al., 1986). Research in children with other medical conditions has also

shown that children with greater personal freedom (or fewer behavioral restrictions) were

more likely to be adherent with their medication regimen. Peterson et al. (1982) found

that satisfaction with the physician's explanation was unrelated to adherence. However,

in studies with other populations, patient satisfaction and good patient-doctor

communication have been related to good adherence. Peterson et al. (1982) also found

that worry about one's own health was related to self-reported adherence.

Shope reviewed a study that closely followed the Health Belief Model (HBM)

framework for pediatric epilepsy patients (Shope, 1992). Parents of a random sample of

patients drawn from an urban medical center pediatric seizure clinic were seen at their

regular appointment and interviewed from January of 1975 through May of 1975.

Questions asked included variables from the HBM as well as a wide range of medical,

psychosocial, and demographic variables. Serum antiepileptic levels were taken at the

clinic visit on 65 of the 90 study participants. Data presented were based on 65 patients.








Patients were judged to be compliant or non-compliant by an experienced clinician's

relating the prescribed dosage of medication to the obtained blood level. This process

was used, rather than therapeutic blood levels, because some patients had less than

therapeutic dosages prescribed.

The patient sample was 55% female and 78% African American with a mean age of

10 years. For 55% of the children, seizure disorder was the only medical problem

identified. Age of onset ranged from birth to 11.5 years and duration of the seizure

disorder ranged from 6 months to 19 years. Forty-eight percent of the children were

prescribed one medication, 39% were prescribed two, and 13% were prescribed three.

Phenobarbital was prescribed for most of the patients. The assessment of compliance, by

looking at blood levels, revealed that 57% of the parents gave their children their

medication as prescribed, 25% were giving less than the prescribed dose, and 18% were

giving no medication. Forty-three percent of the patients were not receiving their

medication as prescribed.

Results indicated that parents who reported at least one occurrence of running out

of their child's medication for several days were more likely to have children whose

blood levels revealed a discrepancy between the obtained and prescribed dosage (p < .05)

and parents who reported that they missed giving their child a medication more than once

a week were less likely to be compliant, as measured by blood levels (p < .001)

(motivational factors of the HBM). Recency of seizures was significantly positively

correlated with compliance (p < .05) (value of illness threat reduction of the HBM).

Child's enrollment in school was significantly and negatively correlated with compliance








(p < .01). Parents who reported that their child did not have any additional medical

problems were more compliant (p < .05).

In summary, factors that appear to influence parental non-adherence to anti-

epileptic medications for children with epilepsy include lower socioeconomic status and

ethnic minority group membership. However, the criteria for assessing socioeconomic

status have not been well defined nor have they been applied consistently across studies,

thus rendering comparability of socioeconomic findings somewhat suspect. Research

findings indicate that parents whose children had increased seizure frequency, more

recent seizure activity, and tonic-clonic seizures were more adherent to their child's anti-

epileptic medication regimen. Parents who perceived an increased level of disease

severity (more severe seizure events with greater frequency and intensity) were more

adherent to their child's medication regimen. The current study examined factors that are

related to medication adherence in pediatric epilepsy populations, specifically with

younger children with epilepsy.

Maternal Emotional Functioning

Depression is a highly prevalent mental disorder, especially among those women of

childbearing age. Downey and Coyne (1990) found that there is a high rate of serious

psychological problems in children of depressed mothers and that depressed mothers

have a host of difficulties as parents. Murray, Sinclair, Cooper, Ducoumau, Turner, and

Stein (1999) reported that exposure to maternal depression in the early postpartum

months may have an enduring influence on childhood psychological adjustment.

Cummings and Davies (1994) indicated that children of depressed parents are at

increased risk for the development of psychopathology and that they are two to five times

more likely to develop behavior problems than are children of normal parents. Depressed








mothers view their parenting role less positively than non-depressed mothers, experience

negative feelings towards parenting roles, and have feelings of rejection and hostility

towards their child (Downey & Coyne, 1990). Depressed mothers also tend to

demonstrate difficulties interacting with their children and displaying positive emotions

towards their children.

Hodes and Garralda (1999) examined the effects of maternal expressed emotion on

the psychological adjustment of children with epilepsy. Maternal expressed emotion and

risk for psychiatric disorders (for children and mothers) were assessed. The researchers

found that mothers showed significantly more emotional overinvolvement and a trend for

more hostility towards their children with epilepsy than they displayed towards sibling

controls. Maternal emotional overinvolvement was unrelated to child behavior problems.

High levels of criticism and hostility also were shown to covary with child behavior

problems (Hodes & Garralda, 1999). The effects of maternal expressed emotion on

adherence to medication were not assessed. However, the results suggest that maternal

expressed emotion is related to negative psychological functioning in children with

epilepsy. Thus, it seems highly plausible that maternal expressed emotion may have a

substantial impact on adherence to anti-epileptic medications for children with epilepsy.

Otero and Hodes (2000) assessed the relationship between maternal expressed

emotion and adherence to treatment in children with epilepsy. The researchers reported

that good treatment compliance was related to less maternal hostility and criticism for

epileptic children and their mothers. Children and mothers in the good compliance group

had fewer reported psychiatric symptoms. Poor treatment compliance for children with

epilepsy was associated with greater maternal psychological disturbance. Children with a








greater number of behavioral problems were less compliant. A relationship between

maternal depression and poorer compliance was reported (Otero & Hodes, 2000).

The relationship between maternal depression and behavior problems has been

investigated intensively across a variety of pediatric medical populations. Poor maternal

emotional functioning appears to exert a consistent negative impact on children with

disabilities and their healthy siblings. Research findings demonstrate a clear relationship

between maternal emotional functioning and childhood behavior problems.

Spector et al. (1999) found that increased maternal stress and anxiety levels were

related to an increase in seizure frequency and severity. Hazzard (1990) reported that

increased levels of parental anxiety and perceived stress were related to lower adherence

to anti-epileptic medications for children with a seizure disorder.

It is feasible that a true relationship exists between poor maternal adjustment and

difficulty with adherence to medication regimens for younger children. We examined the

relationship between maternal emotional distress variables (maternal depression and

maternal stress) and maternal adherence to anti-epileptic drugs (pill count percentage)

and child health status variables (seizure frequency, seizure-related emergency room

visits) for children diagnosed with epilepsy.













CHAPTER 2
AIMS AND STUDY JUSTIFICATIONS

Purpose

In summary, epilepsy is a chronic illness that typically begins in early childhood.

Poor adherence to the medication regimen can result in increased seizure frequency and

severity, and possibly, death. Young children are reliant on their parents, usually their

mothers, for their daily routine and functioning. This issue is particularly salient for

children with seizure disorders, who rely on their mothers to take their anti-convulsant

medicationss. It is well known that depression has a negative effect on daily

functioning. Additionally, depressive symptomatology and stress in mothers have been

associated with negative child health status and maternal adherence outcomes. Maternal

emotional factors such as depression and maternal stress may play an important role in

adherence to medication for younger children with epilepsy. To our knowledge, the

influence of maternal emotional distress on adherence to medication regimens of children

with epilepsy has not yet been examined.

An improved understanding of the impact of maternal emotional functioning on the

health status of children with epilepsy would enhance our ability to effectively intervene

with distressed mothers, and aid in developing interventions to prevent the possible

complications associated with poorer adherence, and improve overall psychological

funimctioning in both mother and child. In the current study, we examined the relationship

between maternal adherence to medication (pill count percentage) and child health status

(seizure frequency and seizure-related emergency room visits) for children with epilepsy








and the following factors: maternal emotional distress (maternal depression, maternal

stress), demographics (family annual income, race, maternal educational level),

condition-related and regimen-related variables (disease type, disease duration, type and

complexity of the medication regimen, concordance between prescribed regimen and

reported regimen), and maternal perceptions of the treatment regimen (perceived

efficacy, perceived knowledge, perceived difficulty).

Study Model

The purpose of the current study was to test a one-month predictive model of

maternal adherence and child health status to prescribed anti-epileptic medication in a

sample of 1-11 year old children diagnosed with epilepsy, focusing on the influence of

four predictor factors. The four predictor factors are defined as (1) maternal emotional

distress, (2) maternal perceptions of the treatment regimen, (3) condition-related and

regimen-related variables, and (4) demographic variables. The three outcome variables

are defined as (1) seizure frequency, (2) seizure-related emergency room visits, and (3)

pill count percentage, where (1) and (2) assess child health status and (3) assesses

maternal adherence. The current study extends earlier research in several ways. First, we

will integrate the four key sets of predictor factors identified in previous research in a

unifying predictive framework. Moreover, we strengthen the model by examining the

roles of the four, predictor factors at Time 1 (assessment) in predicting subsequent Time

2 (1-month follow-up) adherence and health status, controlling for Time 2 (1-month

follow-up) predictor factors (see Figure 1). This controls for the possibility that the

predictive role of adherence may be affected by maternal emotional distress, maternal

perceptions of the treatment regimen, condition-related and regimen-related variables,








and demographic variables at Time 2. The present study examined the following specific

hypotheses.

Hypotheses

Maternal Emotional Distress, Adherence, and Child Health Status

1. It was hypothesized that Time 1 maternal emotional distress would be negatively
related to Time 2 adherence to the child's medication regimen and positively with
child health status.
a. Time 1 maternal depression scores would be negatively associated with Time
2 adherence to the child's medication regimen and positively with health
status.
b. Time 1 maternal stress scores would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.

Maternal Perceptions of the Treatment Regimen, Adherence, and Child Health
Status

2. We predicted that Time 1 ratings of maternal perceptions of the treatment regimen
(perceived knowledge, perceived difficulty, and perceived efficacy), and Time 1
maternal distress and Time 1 maternal perceptions of the treatment regimen would be
associated with Time 2 adherence to the child's medication regimen and child health
status.
a. Maternal ratings of perceived knowledge regarding the child's disease and
medication regimen at Time 1 would be positively associated with Time 2
adherence to the child's medication regimen and negatively with health status.
b. Maternal ratings of perceived difficulty in giving the child the prescribed
medication at Time 1 would be negatively associated with Time 2 adherence
to the child's medication regimen and positively with health status.
c. Maternal ratings of perceived efficacy regarding the child's medication
regimen at Time 1 would be positively associated with Time 2 adherence to
the child's medication regimen and negatively with health status.
d. Maternal ratings of maternal distress and maternal perceptions of the
treatment regimen at Time 1 would be associated with Time 2 adherence to
the child's medication regimen and health status.

Condition and Regimen-related Variables, Adherence, and Child Health Status

3. We predicted that Time 1 condition and regimen-related variables would be
associated with Time 2 adherence to the child's medication regimen and child health
status.
a. A tonic-clonic seizure diagnosis would be positively related to adherence to
the child's medication regimen and health status.
b. Time since diagnosis at Time 1 would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.








c. Complexity of the medication regimen (higher doses, multiple medications,
more frequent doses) at Time 1 would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.
d. Percentage of concordance between the physician-prescribed medication
regimen and self-reported regimen at Time 1 would be positively associated
with Time 2 adherence to the child's medication regimen and negatively with
health status.

Demographic Characteristics, Adherence, and Child Health Status

4. We predicted that Time 1 demographic characteristics would be positively related to
Time 2 adherence to the child's medication regimen and negatively with child health
status.
a. Total household income at Time 1 would be positively associated with Time 2
adherence to the child's medication regimen and negatively with health status.
b. Maternal educational level at Time 1 would be positively associated with
Time 2 adherence to the child's medication regimen and negatively with
health status.

















CHAPTER 3
METHODS


Study Design

The current study is a prospective study using a time lag design, in which we


examined the influence of predictor variables at Time 1 on outcome variables at Time 2,


while controlling for predictor variables at Time 2 (see Figure 1).


Time 1


Time 2


Depression

Maternal Stress




Efficacy
K now ledge
Difficulty


Com p lexity

Time Since Dx -
Regimen Type* -
Disease Type -


M aternal Perceptions
of the )
Treatment Regime n No s----.Seizure Frequency
---Health Sa ER Visits

---o--- Pill Counts
Adherence


Education Level
Annual Income
Race


Figure 1: Study Model








After obtaining informed consent, mothers were given a background questionnaire

(which included questions about condition and regimen-related variables), a

questionnaire assessing maternal stress, and a questionnaire assessing depressive

symptomatology to complete. The questionnaires were completed during a regularly

scheduled pediatric neurology clinic visit. Participants were asked to participate in a one-

month follow-up phone call. Approximately one month later, the investigator contacted

all participants at home by phone to re-assess condition and regimen specific-variables,

maternal perceptions of the treatment regimen, and maternal emotional distress. The one-

month delay in the assessment was designed to minimize sampling of temporary behavior

change that may have occurred immediately after a clinic visit and to re-assess maternal

ratings of emotional distress, condition and regimen-related measures, and maternal

perceptions regarding the treatment regimen, and their relationship to maternal adherence

to the child's medication regimen and child health outcomes.

Participants

Participants with a diagnosed seizure disorder were recruited from the Pediatric

Neurology clinic at Shands Hospital at the University of Florida's Health Science Center.

Based on data obtained from the pediatric neurologist, there are over 300 children with a

diagnosed seizure disorder between the ages of 1 and 11 who were eligible to participate

in the study. Young children are the focus of this study because: 1) epilepsy typically

begins to develop in early childhood, 2) mothers of younger children are more likely to

be highly involved in their child's medication regimen, and 3) one-third of epilepsies

spontaneously remit by adolescence.

Inclusionary criteria included: (1) An epilepsy diagnosis from the University of

Florida's pediatric neurologist which was noted in the patient's medical chart; (2) 1-11








years of age; (3) English-speaking; (4) Ability to read and write; (5) Telephone available

in the home. Exclusionary criteria included: (1) < 1 years of age or > 11 years of age; (2)

Non-English speaking; (3) Inability to read and write; (4) No telephone in the home.

A total of 82 participants completed all study measures. Data indicated the sample

consisted of 36 male and 46 female patients, aged 1-11 years (M = 5.76; SD = 2.93). The

mean education of mothers in the sample was 13.18 years (SD = 2.69). Mothers were

married (67.1%) and were homemakers (29.3%) or working full-time (35.4%). Average

family income was $44,450 (SD = 41,993). The ethnic makeup of the maternal sample

was 78.1% White, 11% African American, 6.1% Hispanic American, 2.4% Native

American, and 2.4% Other. The ethnic makeup of the child sample was 75.6% White,

9.8% African American, 7.3% Hispanic American, 2.4% Native American, and 4.9%

Biracial/Other. Approximately 93% of children in the sample were taking medication in

pill form, with seven percent taking medication in liquid form. Approximately 32% of

children in the study were diagnosed with partial seizures, 66% with generalized seizures,

and 2% with a diagnosis of other type. Study sample demographic characteristics are

presented in Table 1.

Table 1

Demographic and Medical Characteristics

Mean age of children in years (SD) 5.76 (2.93)

Sex of the child
Boys 36 (43.9%)
Girls 46(56.1%)

Mean education of mothers in years (SD) 13.18 (2.69)








Table 1 Continued


Mother Marital Status
Single/ Never Married
Married
Separated/Divorced

Mother Employment Status
Unemployed
Volunteer
Part-time
Full-time
Homemaker
Student
Other


Mean annual family income in dollars (SD)


Child race
White
African American
Hispanic American
Native American
Biracial/Other

Mom race
White
African American
Hispanic American
Native American
Biracial/Other


44,450 (41,993)


62 (75.6%)
8 (9.8%)
6 (7.3%)
5 (2.4%)
3 (4.9%)


64(78.1%)
9(11%)
5 (6.1%)
2 (2.4%)
2 (2.4%)


76 (92.7%)
6 (7.3%)


Regimen Type
Pills
Liquid


Seizure Type
Partial
Generalized
Other
Note. (n = 82).


26(31.7%)
54.9 (65.9%)
2 (2.4%)


Participants were recruited from the University of Florida's Pediatric Neurology

Outpatient Clinic at Shands Hospital during the patient's usual clinic visit. This clinic


15(18.3%)
55(67.1%)
12(14.6%)


13(15.9%)
1(1.2%)
8 (9.8%)
29 (35.4%)
24 (29.3%)
5 (6.0%)
2 (2.4%)








treats over 300 children a year diagnosed with epilepsy. We offered the study to all

patients who met the study's eligibility criteria.

The researcher and/or research assistants approached participants individually and

presented the rationale for the study. Parental informed consent and child assent was

obtained for children (if possible). Questions concerning subject obligation and

confidentiality were addressed at this time. After obtaining informed consent/assent,

participants completed a brief background questionnaire including questions regarding

maternal perceptions of the treatment regimen, the CES-D, and the PSI-SF. The entire

process took approximately 30-40 minutes.

The investigator recorded the following information from the participant's clinic

visit or from the medical chart review conducted to gather other pertinent information:

disease-specific information including date of diagnosis, time since diagnosis, type of

diagnosis, type of medication, dose of medication, seizure frequency, blood serum levels,

pill counts, and emergency room visits. Approximately one month after the clinic visit,

the mother was contacted by phone to re-assess condition-related information, maternal

perceptions regarding the treatment regimen, and maternal emotional distress.

Measures

Participant Background Information

A background questionnaire was completed at the time of initial assessment in

order to gather the following variables: age of child, age of mother, disease type, age of

onset, time since initial diagnosis, race, mother's level of education, family income, and

marital status. Time since initial diagnosis was divided into two categories: (1) those

who were newly diagnosed, and (2) those with a diagnosis of epilepsy greater or equal to

one year prior to the study. Mothers were asked to describe their child's seizure








diagnosis, seizure frequency, and medication regimen. The type(s) of seizure disorders)

the mothers reported were noted. Maternal report of the child's monthly seizure

frequency for the past month was also assessed. Furthermore, complexity of the

medication regimen reported by mothers was evaluated based on 4 indicators (i.e., type of

medication, number of medications per day, dosing, and timing). A complexity rating for

the medication regimen was calculated by multiplying the number of anti-epileptic

medications by the total number of times per day each medication is prescribed.

Maternal ratings of perceived difficulty (0 not difficult to 10 extremely difficult) and

perceived efficacy (0 not good to 10 extremely good) in carrying out the child's

medication regimen were assessed. Perceived difficulty ratings included items about

maternal perceptions of difficulty of medication type, medication dosage, and medication

timings. Perceived efficacy ratings included items about maternal perceptions of efficacy

of the child's medications and efficacy of maternal management of the child's medication

regimen. Mothers were asked to rate how they perceived their level of knowledge about

their child's medical condition. The knowledge ratings also ranged from 0 to 10 (0 not

knowledgeable to 10 extremely knowledgeable). Knowledge ratings included items

about maternal perceptions of knowledge of the child's seizure disorder and medications

(type, timing, number, and dosage) (see Appendix A & Appendix B).

Medical Chart Information

Condition-specific information including date of diagnosis, time since diagnosis,

type of diagnosis, type of medication, dose of medication, seizure frequency, blood

levels, pill counts, and emergency room (ER) visits were collected by a review of the

medical chart. The child's medication regimen was noted from the chart (type of

medication, number of medications per day, dosage, and timing) and compared to the








mothers' report of the child's medication regimen. Concordance ratings were calculated

between physician prescribed anti-epileptic medications) and maternal reports of

prescribed medications) by determining the percentage correct of maternal reports

compared to physician reports (see Appendix C & Appendix D). Specifically, points for

concordance were assigned for a correct match between the prescribed and reported

regimen as follows: 50 (name), 30 (dose), and 20 (timing). More weight was given to the

name of the medication and dosage of the medication, with the rationale that if the name

and dosage of the medication is incorrect, the mother might not be appropriately

administering the correct medication and/or dosage.

Child Health Status and Adherence

Child health status included two components: current seizure frequency and

seizure-related ER visits. Current monthly seizure frequency was assessed by maternal

self-report and by chart review over the last month. Seizure-related ER visits were

identified by conducting a chart review. Chart reviews consisted of a systematic

examination of the child's medical chart for the most recent data regarding child health

status. Written prescriptions and medication refill information were not noted

systematically in each medical chart and thus, this information was not available to be

collected.

Pill counts were assessed during the clinic visit if medication bottles were brought

to the clinic appointment. If mothers did not have the child's medication bottle at the

clinic visit, they were contacted at home on the same day of the appointment in order to

report a current pill count. Mothers were telephoned at one-month follow-up and asked

to report a pill count. Maternal adherence to the child's medical regimen was assessed by

percentage of pills used after one month [(Time 1-Time 2) / (Time 1)] x 100%.








Adherence was defined as an 80% pill count percentage. Blood serum levels were not

included as a measure of adherence because these values were not comparable across

different medication types; additionally, many participants had no blood serum levels

drawn.

Maternal Emotional Distress

The Center for Epidemiologic Studies Depression (CES-D; Radloff, 1977) is a

self-report measure developed to be used in epidemiological studies of depression and

was designed to measure current levels of symptoms which accompany depression in a

general adult population. The CES-D is a 20-item self-report measure of depressive

symptoms. Each item provides a statement representing a symptom characteristic of

depression (e.g. "I had crying spells"), followed by a 4-point Likert-type response scale

ranging from "rarely or none of the time" (less than 1 day) to "most all of the time" (5-7

days). Participants were instructed to bubble in the number of each statement which best

described how often they felt or behaved that way during the past week. Sixteen of the

scales range from 1 to 4 and the remaining four scales range from 4 to 1.

Four major factors have been derived based on a principal components factor

analysis of data from the general population. These factors have been described as

depressed affect, positive affect, somatic and retarded activity, and an interpersonal

factor. Total scores were used for the purpose of the current study. The CES-D is scored

by summing the ratings of the 20 items and scores range from 0 to 60. Total severity is

calculated by reversing scores for items 4, 8, 12, and 16 (the items that control for

response bias). Higher scores indicate higher levels of depressive symptoms experienced

during the past week. Investigators have designated a cutoff score of 16 as a suitable

indicator to determine depressed from nondepressed patients (Nezu, 2000).








Regarding internal consistency, Cronbach's alpha is high across a variety of

populations and ranges from 0.85 for the general population and 0.90 for the clinical

population. Split-half reliability is also high, ranging from 0.77 to 0.92. Test-retest

reliability studies ranging over 2-8 weeks show moderate correlations (r = 0.51-0.67).

Studies examining effects between African Americans, Caucasians, and Mexican

Americans found no differences in measures of internal consistency reliability.

In samples of outpatients with depression, alcoholism, drug addiction, or

schizophrenia, correlation coefficients between CES-D scores and Symptom Checklist-90

Depression Subscale Scores were high, ranging from 0.73-0.89. Correlations with the

Hamilton Rating Scale for Depression and Raskin Scale ranged from 0.45 and 0.28 for

patients with acute depression to 0.85 and 0.79 for patients with schizophrenia,

respectively. Scores on the CES-D have been shown to decrease significantly following

treatment for depression. For major depression, sensitivity was 79.5% and specificity

was 71.1%. The positive predictive value was 27.9%. (American Psychiatric

Association, 2000).

The Parenting Stress Index Short Form (PSI-SF; Abidin, 1995) is a 36-item short

form directly derived from the long form of the PSI. The PSI-SF was developed for use

in a variety of primary health care settings where it could be administered in less than 10

minutes. Participants were asked to respond to questions based on the five-point-scale.

The PSI-SF parent self-report scale contains three factor-analytically-derived

subscales (Parental Distress, Parent-Child Dysfunctional Interaction, and Difficult Child)

and a Total Stress Score. The Parental Distress (PD) scale determines the distress a

parent is experiencing in their role as a parent as a function of personal factors that are








directly related to parenting. The Parent-Child Dysfunctional Interaction (P-CDI)

subscale focuses on the parent's perception that their child does not meet their

expectations, and the interactions with their child are not reinforcing to them as a parent.

The Difficult Child (DC) subscale focuses on some of the basic behavioral characteristics

of children that make them either easy or difficult to manage. These characteristics can

include defiant, noncompliant, and demanding behavior. The Total Stress score is

designed to provide an indication of the overall level of parenting stress an individual is

experiencing. Parents who obtain a Total Stress score above a raw score of 90 are

experiencing clinically significant levels of parenting stress.

The three PSI-SF subscales differ from that of the longer PSI, but correlate in the

expected direction with the domains of the longer form. The PSI and PSI-SF total scores

are highly correlated with one another (0.94). The short form subscales have shown

Cronbach's alphas of 0.80 to 0.91 and 6-month test-retest reliabilities of 0.68 to 0.85.

Positive predictive values of the PSI-SF range from 25% to 45%.













CHAPTER 4
RESULTS

The overall findings of the data analyses are presented as follows: 1) descriptive

statistics of the predictor and criterion variables included in the conceptual model, 2)

internal consistency reliabilities of the questionnaire measures and the distributional

properties of key predictor and criterion variables, 3) correlations between the

background variables not included in the conceptual model and those between key

predictor variables and change scores on the criterion variables, 4) regression analyses

that control for the effects of Time 2 predictors on Time 2 criterion variables, and 5)

exploratory model testing to identify goodness of fit of the specific regression models.

Due to small sample size, structural equation modeling (SEM) was not conducted

as anticipated. The statistical approach shifted from SEM to variable-level model testing,

in which the goodness of fit of variables located under the four primary factors of the

conceptual framework was assessed (see Figure 1).

Preliminary Analyses

Medical characteristics of the child with epilepsy were variables collected at Time

1, except for seizure frequency. The latter was also collected at Time 2, since this time

point was of main interest in our primary analyses. Study sample medical characteristics

are displayed in Table 2.








Table 2

Mean Scores and Standard Deviations for Child Medical Characteristics


Measures M (SD)

Complexity of the regimen (uncorrected) 8.7(13.1)
Complexity of the regimen (corrected) 1.13 (0.5)
Time Since Diagnosis in years 4.0(2.8)
Concordance between reported and prescribed regimen-Time 2 72.8 (32.7)
Seizure Frequency-Time 2 (uncorrected) 50.0 (87.2)
Seizure Frequency-Time 2 (corrected) 2.5 (1.1)
Note. (n = 82). Corrected scores are corrected for skewness to ensure analyses integrity.

Means and standard deviations for maternal self-reports of maternal stress

(measured by the PSI-SF), depression (measured by the CES-D), and maternal

perceptions of the treatment regimen (measured by maternal reports of difficulty,

efficacy, and knowledge) for Time 1 and Time 2 are presented in Table 3 below.

Table 3

Mean Scores and Standard Deviations for Study Measures at Time 1 and Time 2

Measures M(SD) M (SD)
Time I Time 2
PSI-SF, Total score 85.2 (28.5) 65.5 (8.9)
PSI-SF, Parental distress 27.5(10.5) 24.5(3.5)
PSI-SF, Parent-child dysfunction 25.9(9.9) 21.6(5.1)
PSI-SF, Difficult child 31.8(11.5) 19.4 (3.7)
CES-D, Severity score 7.8(3.3) 12.4(2.2)
Difficulty, Total score uncorrected 6.0 (5.8) 12.9 (4.5)
Difficulty, corrected 2.1 (1.5) 1.5 (0.8)
Efficacy, Total score 14.3 (4.5) 11.0 (3.5)
Knowledge, Total score 39.4 (13.84) 39.5 (13.82)
Note. PSI-SF = Parenting Stress Index-Short Form; CES-D = Center for Epidemiologic
Studies-Depression. Corrected scores are corrected for skewness to ensure analyses
integrity.

Cronbach's alphas were calculated to assess the reliability (i.e. internal consistency)

of mother-completed self-report measures. Coefficients ranged from .68 to .96 for the

total sample at Time 1, suggesting adequate internal consistency reliability for all








measures used. Coefficients ranged from -.002 to .58 for the total sample at Time 2,

suggesting lower levels of internal consistency reliability for all measures used compared

to Time 1. Time 1 and Time 2 reliability statistics are presented in Table 4 below.

Table 4

Alpha Coefficients of Parent-completed Measures for Total Sample at Time 1 and Time 2

Measure N a a
Time 1 Time 2

PSI-SF, Total score 82 .96 .56
CES-D, Total score 82 .94 .35
Difficulty, Total score 82 .86 -.18
Efficacy, Total score 82 .68 -.002
Knowledge, Total score 82 .92 .08
Note. PSI-SF = Parenting Stress Index-Short Form; CES-D = Center for Epidemiologic
Studies-Depression.

We also analyzed the distribution properties of demographic, medical, predictor,

and outcome variables. Skewed distributions were corrected with the creation of new

variables. Annual family income was highly positively skewed and ranged from $1500

to $250,000 and was recorded into three income categories with a value of one

representing poverty level of $1500 to $18,000, a value of two representing middle level

of $18,001 to $116,000, and a value of three representing an upper middle level of

$116,001 to $250,000. The poverty level cutoff was determined using the United States

Census Bureau poverty guidelines (United States Census Bureau, 2001). The means and

standard deviations for uncorrected and corrected annual income were $44,450 (41,993)

and 1.2 (.62), respectively. The original scores for complexity of the regimen, seizure

frequency, and maternal perceptions of regimen difficulty were also highly positively

skewed. In order to correct for data skewness, the data was collapsed into groups that

were equally distributed, creating new complexity, seizure frequency, and difficulty








variables. Original complexity scores ranged from 1 to 80 and were recorded into three

categories, with one representing a complexity score of 1 to 26.7, two representing a

complexity score of 26.8 to 53.4, and three representing a complexity score of 53.5 to 80.

Means and standard deviations for uncorrected and corrected complexity scores were 8.7

(13.1) and 1.13 (0.5), respectively. Original seizure frequency scores ranged from 0 to

320 and were recorded into four categories with equal distributions of the sample in each

group, with one representing 0 seizures, two representing 1 to 7 seizures, three

representing 8 to 60 seizures, and four representing 61 to 320 seizures. The means and

standard deviations for uncorrected and corrected seizure frequency scores were 50.0

(87.2) and 2.5 (1.1), respectively. Difficulty scores were positively skewed, ranged from

0 to 24, and were recorded into four categories, with one representing a score of 0 to 6,

two representing a score of 6.1 to 12, three representing a score of 12.1 to 18, and four

representing a score of 18.1 to 24. The means and standard deviations for uncorrected

and corrected difficulty scores were 6.0 (5.8) and 2.1 (1.5) for Time 1 and 12.9 (4.5) and

1.5 (0.8) for Time 2, respectively.

Primary Analyses

Correlations among Background Variables, Predictors, and Criterion Variables

Correlational analyses were conducted between background variables not included

in the model (child sex, marital status, number of years married, employment status,

number of children in the home) and Time 1 key predictor variables of the following

factors: maternal emotional distress, maternal perceptions of the regimen, condition and

regimen-related variables, and demographic variables. No significant associations were

found between the background variables not included in the model and Time 1 predictors








(p > .05). No significant associations were found between the background variables not

included in the model and criterion variable change scores (see Table 5 below).

Table 5

Correlations between Background Variables not in the Model and Specific Time 2
Outcomes Change Scores

Background Factors Seizure Seizure-related ER
Frequency Visits
Child sex .160' -.006+
Marital status -.027+ .083+
Years married .085+ .060+
Employment status -.098+ -.072+
Number of children in home .062+ -.015+
Note. +p > .05.

Correlational and hierarchical regression analyses were conducted to examine the

relationships between Time 1 predictors (maternal emotional distress, maternal

perceptions of the treatment regimen, condition and regimen-related factors, and

demographic factors), and Time 2 outcome measures (seizure frequency, emergency

room visits, and pill count percentages). Note that the effects of Time 2 predictor

measures were partialed out in regression analyses to control for Time 2 predictor

variable effects on the criterion measures. The order of entry into the regression analyses

were as follows: (1) Time 2 predictors were entered in a block in a forced-entry fashion,

(2) Time 1 predictors were entered in blocks in a stepwise fashion. Note that Time 2

predictor measures were not included in correlational analyses due to low internal

consistency reliability. New variables that were created as corrections for data skewness

were used in all subsequent analyses; analyses conducted with skewed data did not reveal

any significant relationships. Partial correlation and regressions analyses were used in

determining which predictors to enter into the final hierarchical regression models.








Hypothesis 1: Maternal Emotional Distress, Adherence, and Child Health Status

We hypothesized that Time 1 maternal emotional distress (measured by the PSI-SF

and CES-D) would be negatively related to Time 2 adherence to the child's medication

regimen (pill count percentage) and positively associated with child health status (seizure

frequency and seizure-related emergency room visits). Specifically, it was predicted that

Time 1 maternal stress and depression scores would be negatively associated with Time 2

adherence to the child's medication regimen and that Time 1 ratings of maternal stress

and depressive affect would be positively associated with the child's health status. These

hypotheses were partially supported. Hierarchical regression analyses were conducted to

assess the association between Time 2 outcomes and Time 1 maternal emotional distress

variables, while controlling for the effects of Time 2 maternal emotional distress

variables. Time 1 PSI-SF overall mean scores and PSI-SF parent-child dysfunction

subscale scores significantly predicted Time 2 seizure frequency, while controlling for

Time 2 predictor variables. No substantial associations were found between Time 1

ratings of maternal stress and Time 2 seizure-related emergency room visits or Time 2

maternal adherence. No significant relationships were found between Time 1 ratings of

maternal depression and child health status or maternal adherence. Next, Time 1

maternal emotional distress variables did not predict Time 2 seizure-related emergency

room visits or pill count percentages, while controlling for Time 2 maternal emotional

distress. The results of the two sets of regression analyses for Time 1 maternal emotional

distress variables on Time 2 outcomes are presented in Tables 6 and 7.








Table 6

Summary of Hierarchical Regression Analyses for Time 1 PSI-SF Overall Score
Predicting Child Seizure Frequency (N = 82)


Variables Included P3 t R R2 Adj.R2 AR2 F AF
Controlling for T2 .091 .008 -.017 .008 .332 .332
CES-D mean T2 .055 .466
PSI-SF mean T2 -.093 -.784

Model 1 .253 .064 .028 .056 1.78 4.64*
CES-D mean T2 .090 .773
PSI-SF mean T2 -.121 -1.04
PSI-SF mean T1 .239 2.16*
Note. CES-D = Center for Epidemiologic Studies-Depression; PSI-SF = Parenting
Stress Index-Short Form; T1 = Time 1; T2 = Time 2.
*E < .05

Table 7

Summary of Hierarchical Regression Analyses for Time 1 PSI-SF Subscale Variables
Predicting Child Seizure Frequency (N = 82)


Variables Included t R R2 Adj. R2 AR2 F AF
Controlling for T2 .084 .007 -.031 .007 .183 .183
PD mean T2 -.048 .466
P-C dys T2 -.007 -.784
DC T2 -.059

Model 1 .278 .077 .029 .070 1.62 5.88*
PD mean T2 -.060 -.535
P-C dys T2 -.015 -.124
DC T2 -.078 -.629
P-C dys T1 .267 2.42*
Note. PSI-SF = Parenting Stress Index-Short Form; PD = Parental Distress; DC =
Difficult Child; P-C dys. = Parent-Child Dysfunction; T1 = Time 1; T2 = Time 2.
*p < .05

The data revealed significant relationships (partial correlations) between Time 1

PSI-SF overall mean scores and Time 2 seizure frequency (r = .24, p < .05), as well as

between Time 1 PSI-SF parent-child dysfunction subscale scores and Time 2 seizure








frequency (r = .27, p < .05). As hypothesized, overall Time 1 maternal stress levels were

positively related to Time 2 child seizure. Further analysis of the PSI-SF subscales

indicated that Time 1 PSI-SF parent-child dysfunction was positively related to Time 2

child seizure frequency. As maternal stress levels and parent-child dysfunction levels

increased, seizure frequency concomitantly increased. Contrary to prediction, no other

significant correlational relationships were found with Time 1 maternal emotional

distress and Time 2 outcomes. Correlational data between Time 1 maternal emotional

distress, Time 2 maternal adherence, and Time 2 child health status are presented in

Table 8 below.

Table 8

Partial Correlations between Time 1 Maternal Emotional Distress and Time 2 Outcomes

Health Status Adherence

Maternal Emotional Distress Seizure Seizure-related ER Pill Count
Frequency Visits Percentage
PSI-SF Total .237* .061 -.104
PSI-SF, Parental Distress .014 -.026 -.045
PSI-SF, Parent-child dysfunction .266* .056 -.165
PSI-SF, Difficult child -.096 .025 -.078
CES-D Total .045 .101 -.099
Note. PSI-SF = Parenting Stress Index; CES-D = Center for Epidemiologic Studies-
Depression.
*p < .05.

Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence and
Child Health Status

We predicted that Time 1 ratings of maternal perceptions of the treatment regimen

(measured by perceived knowledge, perceived difficulty, and perceived efficacy), and

Time 1 maternal distress (measured by the PSI-SF) and Time 1 maternal perceptions of

the treatment regimen would be associated with Time 2 adherence to the child's








medication regimen (pill count percentage) and child health status (seizure frequency and

seizure-related emergency room visits). Specifically, maternal ratings of perceived

knowledge and efficacy regarding their child's medical condition and medication

regimen at Time 1 would be positively associated with Time 2 adherence to the child's

medication regimen and negatively with health status. In addition, maternal ratings of

perceived difficulty in giving the child the prescribed medication at Time 1 would be

negatively associated with Time 2 adherence to the child's medication regimen and

positively associated with child health status, and last, maternal ratings of maternal

distress and maternal perceptions of the treatment regimen at Time 1 would be associated

with Time 2 adherence to the child's medication regimen and health status. These

hypotheses received mixed support.

Time 2 outcomes of child health status and maternal adherence were regressed on

Time 1 maternal perceptions of the treatment regimen variables, while controlling for the

effects of Time 2 maternal perceptions of the treatment regimen predictor variables.

Time 1 difficulty and efficacy scores successfully predicted Time 2 seizure frequency,

while controlling for Time 2 difficulty and efficacy predictor variables. Time 1 difficulty

scores also successfully predicted Time 2 seizure-related emergency room visits, while

controlling for Time 2 difficulty predictor variables. No other significant relationships

were found between Time 1 ratings of difficulty, efficacy, and knowledge and Time 2

outcomes of child health status and maternal adherence. Next, Time 1 maternal

perceptions of the treatment regimen variables were not successful in predicting Time 2

pill count percentages, while controlling for Time 2 maternal perceptions of the treatment

regimen. The findings of the two sets of regression analyses for Time 1 maternal








perceptions of the treatment regimen variables on Time 2 outcomes are presented in

Tables 9 and 10.

Table 9

Summary of Hierarchical Regression Analyses for Time 1 Maternal Perceptions of the
Treatment Regimen Predicting Child Seizure Frequency (N = 82)


Variables Included 3 t R R2 Adj.R2 AR2 F AF
Controlling for T2 .222 .049 .013 .049 1.35 1.35
Difficulty T2 -.025 -.229
Knowledge T2 .226 2.01*
Efficacy T2 -.046 -.406

Model 1 .334 .111 .065 .062 2.41 5.38*
Difficulty T2 -.034 -.312
Knowledge T2 .212 1.93
Efficacy T2 -.022 -.196
Difficulty T1 .250 2.32*

.459 .211 .159 .100 4.07 9.62**
Model 2
Difficulty T2 -.023 -.226
Knowledge T2 .406 3.35**
Efficacy T2 -.043 -.411
Difficulty T1 .172 1.63
Efficacy T1 -.376 -3.10**
Note. T1 = Time 1; T2 = Time 2.
*P <.05 **p<.l01

Table 10

Summary of Hierarchical Regression Analyses for Time 1 Maternal Perceptions of the
Treatment Regimen Predicting Seizure-related Emergency Room Visits (N = 82)


Variables Included 03 t R R2 Adj. R2 AR2 F AF
Controlling for T2 .227 .052 .015 .052 1.41 1.41
Difficulty T2 .060 .541
Knowledge T2 .211 1.87
Efficacy T2 -.110 -.972








Table 10- Continued

Model 1 .365 .134 .089 .082 2.97 7.29**
Difficulty T2 .051 .472
Knowledge T2 .195 1.80
Efficacy T2 -.082 -.750
Difficulty T1 .288 2.70**

Model 2 .377 .142 .085 .008 2.51 .729
Difficulty T2 .054 .449
Knowledge T2 .251 1.98
Efficacy T2 -.088 -.804
Difficulty T1 .265 2.41*
Efficacy T1 -.108 -.854
Note. T 1 = Time 1; T2 = Time 2.
*<.05 ** <.01

The data revealed significant relationships (partial correlations) between Time 1

difficulty scores and Time 2 seizure frequency (r = .26, p < .05), Time 1 efficacy scores

and Time 2 seizure frequency (r = -.34, p < .01), and Time 1 difficulty scores and Time 2

seizure-related emergency room visits (r = .29, p < .01). As hypothesized, Time 1

maternal perceptions of regimen difficulty were positively related to Time 2 child seizure

frequency and emergency room visits. As maternal perceptions of regimen difficulty

significantly increased, seizure frequency and seizure-related emergency room visits also

significantly increased. Time 1 maternal perceptions of regimen efficacy were negatively

related to Time 2 seizure frequency in that as efficacy significantly increased, seizure

frequency significantly decreased. Contrary to study hypotheses, no other significant

correlational relationships were found with Time 1 maternal perceptions of the treatment

regimen and Time 2 outcomes. Correlational data between Time 1 maternal perceptions

of the treatment regimen, Time 2 maternal adherence, and Time 2 child health status are

presented in Table 11.








Table 11

Partial Correlations between Time I Maternal Perceptions of the Treatment Regimen
and Time 2 Outcomes

Health Status Adherence

Maternal Perceptions of the Seizure Seizure-related ER Pill Count
Treatment Regimen Frequency Visits Percentage

Difficulty .255* .294** .037
Efficacy -.335** -.097 -.024
Knowledge .215 .194 .074
Note. *p <.05 **p <.01

Hypothesis 3: Condition and Regimen-related Variables, Adherence, and Child
Health Status

We predicted that condition and regimen-related variables (complexity, time since

diagnosis, regimen type, disease type, and concordance between the prescribed and

reported regimen) would be associated with Time 2 adherence to the child's medication

regimen (pill count percentage) and child health status (seizure frequency and seizure-

related emergency room visits. First, we hypothesized that a tonic-clonic seizure

diagnosis would be positively related to adherence to the child's medication regimen.

Second, we predicted that child health status and time since diagnosis and complexity of

the medication regimen (higher doses, multiple medications, more frequent doses) would

be negatively associated withTime 2 adherence to the child's medication regimen and

positively with health status. Third, we predicted that percentage of concordance

between the physician-prescribed medication regimen and self-reported regimen would

be positively associated with Time 2 adherence to the child's medication regimen and

negatively with health status. These hypotheses were partially confirmed. Hierarchical

regression analyses were conducted to predict Time 2 outcomes with condition and








regimen-related variables. Concordance scores successfully predicted Time 2 seizure-

related emergency room visits and Time 2 pill count percentage. No other significant

predictive relationships were obtained. Furthermore, condition and regimen-related

factors were not successful in predicting Time 2 seizure frequency. Regression data for

condition and regimen-related variables on Time 2 outcomes are presented in Tables 12

and 13.

Table 12

Summary of Hierarchical Regression Analyses for Condition and Regimen-related
Variables Predicting Seizure-related Emergency Room Visits (N = 82)


Variables Included 13 t R R2 Adj. R2 AR2 F AF
Model 1 .076 .006 -.020 .006 .222 .222
Seizure type -.074 -.653
Regimen type .019 .163

Model 2 .408 .166 .110 .161 2.96 4.76**
Seizure type .015 .140
Regimen type -.035 -.324
Concordance -.219 -1.97*
Complexity .210 1.80
Time since dx. .177 1.57
Note. dx = diagnosis.
* <.05 **'1<.01

Table 13

Summary of Hierarchical Regression Analyses for Condition and Regimen-related
Variables Predicting Pill Count Percentage (N = 82)


Variables Included 13 t R R2 Adj. R2 AR2 F AF

Model 1 .122 .015 -.011 .015 .586 .586
Seizure type -.074 -.653
Regimen type .019 .163








Table 13 Continued

Model 2 .755 .570 .541 .555 19.66 31.9**
Seizure type .015 .140
Regimen type -.035 -.324
Concordance -.219 -1.97*
Complexity .210 1.80
Time since dx. .177 1.57
Note. dx = diagnosis.
*E2<.05 **E<.01

The data revealed significant relationships (partial correlations) between

concordance scores and Time 2 seizure-related emergency room visits (r = -.22, p < .05),

as well as concordance scores and Time 2 pill count percentage (r = .69, p <.01). As

hypothesized, concordance scores were negatively related to Time 2 seizure-related

emergency room visits and positively related to Time 2 pill count percentage. As

concordance scores significantly increased, emergency room visits significantly

decreased and pill count percentages significantly increased. Contrary to study

hypotheses, a tonic-clonic seizure diagnosis was not associated with better adherence,

increased seizure frequency, or emergency room visits. No other significant relationships

were found with condition and regimen-related variables and Time 2 outcomes.

Correlational data between condition and regimen-related variables, Time 2 maternal

adherence, and Time 2 child health status are presented in Table 14.

Table 14

Partial Correlations between Time 1 Condition and Regimen-related Variables and Time
2 Outcomes

Health Status Adherence

Condition and Regimenb Seizure Seizure-related ER Pill Count
Variables Frequency Visits Percentage
Complexity .098 .205 -.108








Table 14 Continued

Time Since Diagnosisa .157 .180 -.076
Regimen Type" .089 -.038 .027
Disease Typea .114 -.016 -.083
Concordance"b -.160 -.223* .691**
Note. <.05 ** < .01

Hypothesis 4: Demographic Characteristics, Adherence, and Child Health Status

We predicted that demographic characteristics (total annual income, maternal

education level) would be positively related to Time 2 adherence to the child's medication

regimen (pill count percentage) and negatively associated with child health status (seizure

frequency and seizure-related emergency room visits). Specifically, total household

annual income and maternal education level would be positively associated with Time 2

adherence to the child's medication regimen and negatively associated with child health

status. These hypotheses were partially supported. Hierarchical regression analyses were

conducted to predict Time 2 outcomes with demographic variables. Family income

approached significance (i.e., trend in the predicted direction) for predicting Time 2 pill

count percentage. No other significant associations were found. Demographic variables

were not successful in predicting Time 2 seizure frequency or seizure-related emergency

room visits. Regression data for demographic variables on Time 2 outcomes is presented

in Table 15.








Table 15

Summary of Hierarchical Regression Analyses for Demographic Variables Predicting
Pill Count Percentage (N = 82)


Variables 13 t R R2 Adj. R2 AR2 F AF p
Included
Model 1 .343 .118 .064 .118 2.21 2.21 .078+
Family income .237 2.04 .058+
Mom education .132 1.13
Child race -.283 -1.17
Mom race .093 .388
Note. + > .05

The data revealed a significant relationship (partial correlation) between annual

family income and Time 2 pill count percentage (r = .24, p < .05). As hypothesized,

annual family income was positively related to Time 2 pill count percentage in that as

family income increased, pill count percentages increased. Contrary to study hypotheses,

no other significant correlational relationships were found between demographic

variables and Time 2 child health status or maternal adherence outcomes. Regression

analyses did not identify any additional variables that significantly predicted Time 2

outcomes. Correlational data between demographic variables, Time 2 maternal

adherence, and Time 2 child health status is presented in Table 16 below.

Table 16

Partial Correlations between Demographic Variables and Time 2 Outcomes

Health Status Adherence

Demographic Factors Seizure Seizure-related ER Pill Count
Frequency Visits Percentage
Total Annual Income .088 .089 .244*
Mother's Education -.057 -.101 .138
Child Race .037 -.130 -.143
Mother Race -.025 -.100 -.048
Note. *P < .05








Model Testing

Predictor variables for Time 2 outcomes were included based on the analyses

described above by selecting variables that showed both significant (3 weights and

accounted for significant R2 change. These variables were subsequently included in tests

of the overall goodness of fit to identify promising alternative regression models for use

in follow-up research (see Table 17). The statistical approach included two steps: (1)

Time 2 measures were entered in step one of the hierarchical regression analyses in a

forced-entry fashion and (2) Time 1 variables were entered in a stepwise fashion with

variables included in the model if r < .05.

Table 17

Identified Predictors of Time 2 Outcomes

Seizure Frequency Seizure-related ER Visits Pill Count Percentage
PSI-SF overall mean Difficulty Concordance
PSI-SF parent-child dysfunction Concordance Annual family income
Difficulty
Efficacy

Prediction of adherence

Hierarchical regression analyses were performed to test the hypotheses that the

variables of concordance and annual family income predicted Time 2 maternal adherence

(measured by pill count percentage). Concordance and annual family income were

entered in a stepwise manner with Time 2 pill count percentage as the outcome variable.

Only one model was generated, with the concordance variable retained as a significant

predictor. The concordance variable accounted for a significant proportion of the

variance in the Time 2 pill count percentage outcome measure (Adjusted R2 = .44; tp <

.001). The results suggested that the best predictor of maternal adherence was

concordance between mother and physician on the child's drug regimen. This measure








accounted for approximately 44% of the variance in the Time 2 pill count percentage

measure. In examining the significance of P3 weights for all predictor variables in the

model, only the concordance variable was significant (13 = .67; t = 7.4; p = .000) (see

Table 18).

Table 18

Summary of Regression Model for Predicting Maternal Adherence (N = 82)


Variables Included 13 t R R2 Adj. R2 AR2 F AF
Model l** .665 .443 .435 .443 54.8 54.8*
Concordance .665 7.4*
Note. *p < .01 **Best prediction model

Prediction of health status

Hierarchical regression analyses were performed to test the hypotheses that the

variables of maternal stress (measured by the PSI-SF), parent-child dysfunction

(measured by the subscale of the PSI-SF), and maternal perceptions of difficulty and

efficacy regarding the child's medication regimen successfully predicted Time 2 seizure

frequency. The PSI-SF overall score, parent-child dysfunction score, difficulty score,

and efficacy score for Time 2 were entered as a block in a forced-entry fashion to partial

out Time 2 effects on seizure frequency. The PSI-SF overall score, parent-child

dysfunction score, difficulty score, and efficacy score for Time 1 were then entered in a

stepwise fashion with Time 2 seizure frequency as the outcome variable. One model was

generated in which the Time 1 parent-child dysfunction subscale was a predictor. Parent-

child dysfunction accounted for a small but significant proportion of the variance in the

Time 2 seizure frequency outcome measure (Adjusted R2 = .022; p = .015). The results

suggested that the best predictor of Time 2 seizure frequency was parent-child








dysfunction, which accounted for approximately 2.2% of the variance in the Time 2

seizure frequency measure. In examining the significance of 13 weights for all predictor

variables in the model, only the parent-child dysfunction variable was significant (13P =

.29, t = 2.5, p = .015) (see Table 19).

Table 19

Summary of Regression Models for Predicting Seizure Frequency (N = 82)


Variables Included 3 t R R2 Adj. R2 AR2 F AF
Controlling for T2 .084 .007 -.045 .007 .135 .135
PSI-SFmeanT2 -.130 -.610
P-C dys. T2 .065 .306
Difficulty T2 -.001 -.010
Efficacy T2 -.006 -.049

Model 1** .287 .082 .022 .075 1.37 6.25*
PSI-SF mean T2 -.166 -.802
P-C dys. T2 .087 .421
Difficulty T2 -.057 -.497
Efficacy T2 .056 -.489
P-C dys. TI .287 2.50*
Note. PSI-SF = Parenting Stress Index; P-C dys. = Parent-Child Dysfunction; T1 =
Time 1; T2 = Time 2.
*" <.05 **Best prediction model

Hierarchical regression analyses were performed to test the hypotheses that the

variables of difficulty regarding the child's medication regimen and concordance between

the prescribed and reported regimen successfully predicted seizure-related emergency

room visits. The difficulty score for Time 2 was entered first to partial out Time 2 effects

on seizure-related emergency room visits. The difficulty score for Time 1 and

concordance were then entered as a block in a stepwise fashion with seizure-related

emergency room visits as the outcome variable. This resulted in the generation of two

models, the first of which included the difficulty variable as a predictor and the second








added concordance as a predictor. The first model accounted for a significant proportion

of the variance in the emergency room visits outcome measure (Adjusted R2 = .072; =

.006), with 7.2% of the variance accounted for by this block. The second model

accounted for a significant proportion of the variance in the emergency room visits

beyond model one (Adjusted R2 = .146; p = .007), with an additional 7.4% of the

variance accounted for by this model. The results suggested that the best model for

predicting seizure-related emergency room visits was Model 2, accounting for

approximately 15% of the variance in seizure-related emergency room visits. In

examining the significance of 13 weights for all factors in Model 2, two betas, difficulty (3

= .32, t = 3.1, p = .002) and concordance (13 = -.293, t = -2.8, p = .007), were statistically

significant (see Table 20).

Table 20

Summary of Regression Models for Predicting Seizure-related Emergency Room Visits (N
= 82)


Variables 13 t R R2 Adj.R2 AR2 F AF
Included
Controlling for T2 .062 .004 -.009 .004 .307 .307
Difficulty T2 .062 .554

Model 1 .308 .095 .072 .091 4.15 7.96*
Difficulty T2 .054 .505
Difficulty T1 .302 2.82*

Model 2** .421 .177 .146 .082 5.60 7.80*
Difficulty T2 -.003 -.026
Difficulty T1 .322 3.13*
Concordance -.293 -2.79*
Note. T1 = Time 1; T2 = Time 2.
*r < .01 **Best prediction model








Summary of Findings

Overall, we found mixed confirmation for the hypothesized relationships among

the predictor variables, health status, and adherence criterion variables.

First, the maternal stress measure of maternal emotional distress was found to be

significantly related to the child health status variable of seizure frequency. Maternal

stress and the subscale of parent-child dysfunction were significantly related to seizure

frequency and were significant predictors of seizure frequency in hypothesis testing.

Parent-child dysfunction was identified as a significant predictor of seizure frequency in

exploratory model testing. Note also that maternal emotional distress on measures of the

traditional psychological constructs of depression and maternal stress were not clinically

elevated in the study sample.

Second, the difficulty and efficacy variables of maternal perceptions of the

treatment regimen were identified as being significantly related to child health status

variables. Difficulty was significantly related to seizure frequency and seizure-related

emergency room visits. Difficulty was found to be a significant predictor of both seizure

frequency and seizure-related emergency room visits in hypothesis testing, while efficacy

was only a significant predictor of seizure frequency in hypothesis testing. Difficulty was

also identified as a significant predictor of seizure-related emergency room visits in

exploratory model testing.

Third, the concordance measure of condition and regimen-related variables were

identified as significantly related to the child health status variable of seizure-related

emergency room visits and maternal adherence. Concordance was found to be a

significant predictor of maternal adherence and seizure-related emergency room visits in

hypothesis testing, as well as in exploratory model testing.








Fourth, the demographic factor of annual family income was identified as being

related to maternal adherence. Annual family income was identified as a significant

predictor of maternal adherence in hypothesis testing. However, no significant predictive

relationships were identified between annual family income and outcomes in exploratory

model testing.

Finally, findings related to the predictive validity of Time 1 measures of maternal

emotional distress, maternal perceptions of the treatment regimen, condition and

regimen-related factors, and demographics on Time 2 child health status (seizure

frequency and seizure-related emergency room visits) and maternal adherence (pill count

percentage) outcomes suggest that several measures (PSI-SF parent-child dysfunction,

difficulty, concordance, and annual family income) play an important role in child health

status and maternal adherence outcomes.













CHAPTER 5
DISCUSSION

Sample Characteristics of Demographics and Maternal Distress

In regards to the study sample, mothers were highly educated, having attended an

average of one year of college, with their annual family income falling in the middle

income range ($18,001 to $116,000). The racial make-up of mothers completing this

study was predominately White (78.1%), with traditionally labeled minorities (e.g.,

African American, Hispanic, etc.) showing a fairly uniform distribution pattern. Two-

thirds of the mothers were married and nearly half worked outside the home at least part-

time. The racial profile of patients in this study is representative of the reported

distribution of children with epilepsy living in the United States (Adams, Hendershot, &

Marano, 1996). Patients were predominately White (75.6%), girls (56.1%), and between

five and six years of age. The racial heterogeneity observed in the current study suggests

that its findings are generalizable to the overall epilepsy population within the United

States.

Maternal stress and depression scores were below the cut-off for clinically

significant levels of 90 and 16, respectively, indicating that mothers were not reporting

heightened levels of maternal stress or depressive symptoms at Time 1 or Time 2. It was

surprising that the study sample did not meet criteria for significant levels of maternal

stress or depressive symptomatology since mothers in previous studies reported higher

levels of depressive symptomatology (Kyngas, 2000) and maternal stress (Spector et al.,

1999). One possible explanation for this finding may be that mothers of children with








epilepsy in this study may have experienced more stigma and emotional strain, as well as

daily pressures (e.g., frequent doctor's appointments) when the child was much younger

during the initial months following the child's diagnosis. Perhaps maternal stress and

depression scores were not significant because, compared to when the child was first

diagnosed, the current situation is a calmer and less stressful time for mothers. Studies

indicate that the initial phase following the child's diagnosis leads to negative emotional

consequences (e.g., increased levels of stress and depressive symptomatology) for

mothers of children with chronic illnesses (Brownbridge & Fielding, 1994; Austin, 1989).

It is also possible that only mothers of children who experienced low levels of stress or

depressive symptomatology opted to participate in the study.

Hypotheses

Hypothesis 1: Maternal Emotional Distress, Adherence, and Child Health Status

It was hypothesized that Time 1 maternal emotional distress variables would be

negatively related to Time 2 adherence to the child's medication regimen and positively

with child health status outcome measures. Overall, mixed support was obtained for the

study hypotheses. Hypothesis 1 was partially supported in that mothers who reported

high levels of maternal stress at Time 1, also reported higher seizure frequency levels for

their children at Time 2. No significant relationships were found between maternal

distress measures and seizure-related emergency room visits. Maternal emotional

distress variables were not significantly related to maternal adherence.

The maternal stress measure and parent-child dysfunction subscale scores on the

PSI-SF were positively related to seizure frequency of the child health status factor.

Mothers who reported higher stress levels and more dysfunctional interactions with their

children also reported increased seizure frequency at Time 2 for their children. The








results are consistent with previous findings that increased parental stress levels are

related to an increase in child seizure frequency and severity (Spector et al., 1999).

Spector et al. (1999) investigated the efficacy of a range of psychological interventions

and found that short-term group interventions combining supportive and cognitive-

behavioral components to improve coping strategies and decrease stress may be effective

for improving seizure control,

These findings may be attributable to the idea that mothers who have children with

increased levels of seizure frequency may be more stressed about their child's health and

well-being, which may carry over into increased levels of overall maternal stress. Of

note is that increased patient stress levels have been associated with increased seizure

frequency (Kyngas, 1990). Kyngas (1990) studied the compliance of adolescents with

epilepsy and found that good motivation, a strong sense of normality, subjective

outcome, energy and will-power, support from parents, physicians and nurses, a positive

attitude towards the disease and its treatment, no threat to social and emotional well-

being, and no fears of complications or seizures were related to good compliance and

decreased seizure frequency.

It is possible that dysfunctional mother-child relationships are related to increased

stress levels in children. Perhaps children involved in a dysfunctional parent-child

relationship are more likely to have increased stress levels and in turn, an increase in

seizure frequency as is suggested by Kyngas in 1990. Increased parent-child conflict has

been associated with poorer coping and increased behavioral and emotional problems in

children (Burt, Krueger, McGue, & lacono, 2003). Hodes and Garralda (1999) found that

mothers were more hostile towards their child with epilepsy than towards healthy siblings








and high levels of hostility and criticism were related to child behavior problems.

Additionally, mothers of children with higher levels of seizure frequency may have

reduced opportunities for positive interactions with their children and therefore, may

view their interactions with their child as more dysfunctional in nature.

Previous research has shown that greater parental stress and depression levels have

been associated with lower adherence to medications for children with epilepsy (Hazzard,

1990; Otero & Hodes, 2000). Hence, it was surprising that maternal emotional distress

variables were not significantly related to seizure-related emergency room visits or

maternal adherence. Note, however, that in these prior aforementioned studies, mothers

reported moderately high levels of maternal stress and depressive affect while in the

current study, mothers reported low levels of maternal stress and maternal depression

with the average number of depressive symptoms endorsed by mothers in our study

reported as two. This discrepancy suggests that low levels of maternal stress and/or

depressive affect may have accounted for the lack of a relationship between the maternal

emotional distress factor and the criterion variables.

Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence, and
Child Health Status

We predicted that Time 1 ratings of maternal perceptions of the treatment regimen

(perceived knowledge, perceived difficulty, and perceived efficacy), and Time 1 maternal

distress and Time 1 maternal perceptions of the treatment regimen would be associated

with Time 2 adherence to the child's medication regimen and child health status. Similar

to Hypothesis 1, a mixed pattern of results was obtained. Mothers who rated higher

levels of difficulty with their child's medical regimen at Time 1 concomitantly reported

increased seizure frequency levels and seizure-related emergency room visits for their








children at Time 2. Maternal perceptions regarding difficulty of the child's medical

regimen were positively related to child health status outcome variables. The relationship

between regimen complexity and maternal perceptions of regimen difficulty approached

significance (r = .205, p = .06). These findings may be related to the premise that

children with poorer health outcomes (e.g., increased levels of seizure frequency and

seizure-related emergency room visits) may have more difficult and complex regimens in

general, due to the nature of their illness. Furthermore, mothers who reported higher

levels of regimen efficacy had children with lower levels of seizure frequency.

No significant relationships were found between maternal perceptions of the

treatment regimen variables (e.g., regimen difficulty, efficacy, and knowledge) and

adherence. Overall findings suggest that maternal perceptions of the treatment regimen

are significantly related to poorer health outcomes for children with epilepsy but not

significantly related to maternal adherence. A possible explanation for this null finding is

that mothers of children with more difficult regimens and poorer health outcomes are

more adherent to their child's condition in an effort to minimize negative health outcomes

(e.g., increased seizure frequency and seizure-related emergency visits). Additionally,

relationships between maternal perceptions of the treatment regimen and adherence may

not have been identified because the questions comprising regimen perceptions are

preliminary in nature and are in need of further exploration, modification, and ultimately,

validation.

Hypothesis 3: Condition and Regimen-related Variables, Adherence, and Child
Health Status

We predicted that Time 1 condition and regimen-related variables would be

associated with Time 2 adherence to the child's medication regimen and child health








status. Hypothesis 3 received mixed confirmation. While no significant relationships

were found between condition and regimen-related variables and seizure frequency, one

condition and regimen-related variable was associated with adherence and child health

status outcome variables, partially supporting our hypothesis.

Relationships between condition and regimen variables and seizure frequency were

not related. Condition and regimen-related variables are more directly related to seizure-

related emergency room visits and maternal adherence to the medical regimen.

The concordance measure was related to the seizure-related emergency room visits

variable of child health status and pill count percentage indicator of adherence.

Concordance between the prescribed regimen and reported regimen was negatively

associated with seizure-related emergency room visits and positively associated with pill

count percentage. Mothers who were unable to consistently report their child's

medication regimen with the prescribed regimen may not have administered the child's

medication as prescribed, resulting in negative child health outcomes. The results also

demonstrate that mothers who were able to report their child's medication regimen were

more adherent to their child's medication regimen. It follows from these findings that

mothers who were unable to accurately report their child's medication regimen may not

have followed the prescribed regimen and in turn, reported increased levels of emergency

room visits related to their child's seizures. High concordance ratings may be

representative of good parent-physician communication. Peterson et al. (1982) found that

good patient-doctor communication has been related to better adherence. The significant

predictive relationship demonstrated between concordance ratings and outcomes (e.g.,

seizure frequency and maternal adherence) are of theoretical and practical importance in








that concordance can be easily assessed during each child's clinic visit and further

examined for its relationship to child health status outcomes and maternal adherence.

Hypothesis 4: Demographic Characteristics, Adherence, and Child Health Status

We predicted that Time 1 demographic characteristics would be positively related

to Time 2 adherence to the child's medication regimen and negatively with child health

status. Hypothesis 4 was only partially supported in that one demographic variable was

associated with adherence; family income was related to the pill count percentage

measure of adherence. Families with greater financial resources show better maternal

adherence to the child's medication regimen than those with less income. No significant

relationships were found between race, maternal education and Time 2 outcomes of child

health status and maternal adherence. A possible explanation for the lack of a significant

relationship between demographic variables and outcome measures is that the families in

the sample were not truly representative of families with children diagnosed with a

seizure disorder. Families in the present sample were wealthier than the average family

of children with epilepsy and mothers were more educated than average participants in

previous studies (Friedman et al., 1996). However, earlier research indicated that lower

socioeconomic status and lower parental education levels were correlated significantly

with non-adherence for asthma, cystic fibrosis, and renal disease (Patterson, 1985; Tebbi

et al., 1986) and lower socioeconomic status was related to parental non-adherence

(Becker & Maiman, 1975; Rapoff& Christopherson, 1982). One explanation for these

findings is that the population at the University of Florida's Health Science Center (UF

HSC) may not be representative of the larger demographic distribution. It is feasible that

wealthier and more educated families, because of their financial stability, are better

equipped to meet the demands of their child's medical appointments, including time off








work in order to travel to and from the UF HSC. Less wealthy families without such

resources may only be able to be seen closer to home in order to regularly attend medical

appointments. Second, it is plausible that wealthy and educated parents are more aware

of cutting-edge research and treatment available at the UF HSC and are more willing to

participate in clinical research. Finally, the small sample was from one study site and

may not have included sufficient participants to be representative of the true demographic

distribution.

Model Testing

Sets of predictors were identified as being related to the child health status

variables of seizure frequency and seizure-related emergency room visits and the

adherence measure of pill count percentage. Parent-child dysfunction appears to play a

major role in predicting seizure frequency. Improving the dysfunctional relationship

between parent and child may be beneficial in that a stronger, more positive parent-child

relationship may be developed that is related to decreased seizure frequency. However,

another interpretation of this finding is that parents of children with increased seizure

frequencies may have limited opportunities for positive interactions with their child.

Children with greater levels of seizure frequency may have less time to spend engaging in

non-medical interactions with their parents such as playtime and open communication.

In predicting seizure-related emergency room visits, the regimen-related variable of

concordance and maternal perceptions regarding difficulty of the child's medication

regimen emerge as important variables. These measures may be better predictors of

seizure-related emergency room visits compared to maternal emotional distress measures

because they specifically assess maternal understanding of the child's regimen and its

difficulty. It is suggested that more difficult regimens or poor concordance are more








closely related to seizure-related emergency room visits than maternal emotional distress

variables.

In predicting adherence, maternal emotional distress does not play a key role, but

the regimen-related variable of concordance indeed does. Concordance is likely to be

more predictive of maternal adherence than maternal emotional distress variables,

because concordance taps specifically into the mother's understanding of the child's

regimen. Our findings strongly indicate that incorrect reporting of the child's medication

regimen has a more negative effect on maternal adherence than maternal mood. In that

outcome measures are not solely related to maternal mood, it should be noted that these

variables warrant further investigation.

Study Strengths and Limitations

Studies identifying correlates and predictors of adherence have typically focused on

the following disease groups: diabetes, cystic fibrosis, juvenile rheumatoid arthritis,

asthma, and renal disease. A major strength of the current study was that it assessed

several variables in their relationship to child health status and maternal adherence for

children ages one to eleven diagnosed with epilepsy. Due to this study's multi-variable

approach, it is superior to research in earlier factor-based studies because it includes

individual variable relationships with outcome measures that may have been excluded or

overlooked by factor-based approaches, allowing for the reporting of new information in

the area of pediatric epilepsy research.

A significant methodological strength of the current study was the time-lag design.

This study design provided us with a better predictive model for Time 1 measures on

Time 2 outcomes, in that Time 2 measures were partialed out of Time 2 outcomes.








However, the time lag was relatively short (i.e., one month) and needs to be extended in

future studies to test the reliability of the results over a longer time interval.

The study also allowed us to tap into maternal perceptions of the treatment regimen

with the creation of questions to assess difficulty, efficacy, and knowledge regarding the

child's medical regimen. These types of questions are new to the area of epilepsy

research and can provide researchers with a great deal of information about maternal

perceptions of the treatment regimen and are important in their relationship to child

health status and maternal adherence outcomes.

However, several study limitations should be noted. Although attempts were made

to obtain a larger sample size, the final sample included 82 mother-child dyads. Small

sample size may have contributed to difficulties in detecting significant differences

between predictor and outcome variables. With this small sample size, statistical power

and prediction stability may have been methodological issues. The prediction models

may have been accurate in accounting for variance in the outcome measures but one

needs to examine the current results with caution due to this small sample size. The

correlations between some predictor and outcome variables were quite strong, and led us

to believe that a significant relationship does indeed exist between these measures.

However, a larger sample size would have resulted in greater power, which would have

enabled us to detect differences in other variables that may exist but were not apparent

due to low power leading to poor detectability. For example, difficulty scores

approached significance for inclusion in the final regression model predicting seizure

frequency (p = .056) as did the PSI-SF overall mean (p = .058). Having a larger sample

would also increase the power of prediction because variance on the measures would








increase, resulting in stronger predictive power. The generalizability of the results is

limited due to small sample size, higher than average family income, and higher than

average maternal education level.

Study limitations also include a lack of more concrete adherence measures such as

blood serum levels or prescription refill information. In the current study, we were

unable to report on blood serum levels because more than half of the patients did not have

these levels drawn consistently and some medications do not have normative data on

blood serum levels. Reporting of blood serum levels would have most likely resulted in

skewed findings related to this measure. Also, blood serum levels are not comparable

across different types of medications; thus, a coding system indicating therapeutic and

non-therapeutic blood level ranges suggesting adherence would be valuable in future

studies. Prescription refill information was also unavailable in the medical chart and was

not able to be obtained prospectively due to financial constraints of the current study.

Such information (e.g., blood levels and prescription refill information) is extremely

important to adherence research and should be included in future research.

Additional study limitations involve maternal perceptions of the treatment regimen.

The questions comprising the measures of difficulty, efficacy, and knowledge were

preliminary in nature and are not validated measures. In light of its strengths, the present

study can be used as a catalyst for future research in this area and serve as a model for

related fields.

Future Directions

In order for future studies to better assess predictors of maternal adherence and

child health status for young children with epilepsy, as well as outcome measures, the

following criterion should be met: (1) larger sample size; (2) assessment of child stress








levels; (3) validation of maternal perceptions of the treatment regimen questions; (4)

better measures of adherence; (5) inclusion of multiple study sites; (6) father data; and (7)

control group. With a larger sample size, structural equation modeling (SEM) procedures

would permit specific factor testing rather than the current variable-level testing model.

Multiple regression and other multivariate techniques can address a number of statistical

questions, but have a common limitation in that each technique can examine only a single

relationship at a time. SEM combats this problem and examines a series of predictive

relationships simultaneously. SEM techniques are distinguished by two characteristics:

(1) estimation of multiple and interrelated dependence relationships and (2) the ability to

represent unobserved concepts in these relationships and account for measurement error

in the estimation process. Relationships among predictor variables in each factor seem

promising for the SEM model given that there are significant relationships between the

specific variables comprising each factor of maternal emotional distress, maternal

perceptions of the treatment regimen, condition and regimen-related variables, and

demographic variables.

Assessment of child stress levels can assist in demonstrating a relationship between

maternal stress, child stress, and increased seizure frequency. It would be valuable to

empirically validate the specific questions assessing maternal perceptions of the treatment

regimen in the epilepsy population. Additional questions can then be created through this

validation process. This validation could result in the creation of an epilepsy-specific

measure that may have very important implications in child health status and maternal

adherence outcomes.








With proper funding, future studies could include blood levels, pharmacy refill

information, and electronic medication lid devices. Blood levels can be drawn as a part

of the study protocol, in order to have more consistent data regarding levels of

medication in the bloodstream for each study participant. The physician can aid in

identifying if the blood level falls into the therapeutic or non-therapeutic category. Study

protocol can also include permission to track the participant's pharmacy refill

information at their local pharmacy in order to have a more solid measure of adherence.

Electronic counts in which a device counts each time a medication bottle is opened and

closed, may be helpful in gaining additional adherence information.

The current study was exploratory in nature. Multi-site studies may serve in

gaining larger sample sizes and an overall distribution more representative of the general

population, especially in terms of annual family income and maternal education level.

Greater than 80% of patients were accompanied to the clinic visit by their mother.

However, attempts should also be made to include fathers in future studies by utilizing

mail-outs and/or semi-structured telephone interviews. A control group of children with

controlled seizures may be useful to compare to children that are currently experiencing

seizures. The consideration of additional variables and attempts at intervention studies in

this population is suggested for future research programs in order to gain more

information in this newly growing area of research.

Clinical Implications

The results of this study identified variables that were predictive of child health

status and maternal adherence to the child's medical regimen. An important clinical goal

is to assess maternal emotional distress, maternal perceptions of the treatment regimen,








condition and regimen-related, and demographic variables that may impact maternal

adherence to the child's medication regimen and child health status.

The PSI-SF could be utilized in pediatric clinical settings to identify mothers of

children with epilepsy who are in need of immediate interventions to assist them in

dealing with parenting stress and more specifically, parent-child dysfunction.

Therapeutic interventions could include stress management training, parent training,

progressive muscle relaxation training, and support groups for mothers of children with

seizure disorder. If maternal perceptions of the treatment regimen are identified as

particularly problematic, medical staff can assist with clarifying maternal concerns

regarding the regimen. Reported regimens can be compared with the prescribed regimen

to aid in identifying mothers who are in need of further explanation regarding their child's

medical regimen. Lastly, those mothers with lower education levels may need additional

assistance and additional time in order to properly understand the full nature of their

child's medication regimen. In conclusion, maternal stress, perceptions of the treatment

regimen, report of the regimen, and education level can be quickly assessed in the clinical

setting in order to have a beneficial effect on child health status and adherence outcomes.















APPENDIX A
BACKGROUND INFORMATION FORM TIME 1

Date_______


BACKGROUND INFORMATION

Child' s Name:

Mother's Name:

Address:


Phone #:( )


Child's Sex:


Male


Female


Child's Age: ____


Mother's Age: ____


Child's Birthdate:
m

Mother's Birthdate:


/ /
onth day year

/ /
month day year


Mother's Race:


White
African American
Hispanic American
Asian American
Native American
Biracial
Other:


1. White
2. African American
3. Hispanic American
4. Asian American
5. Native American
6. Biracial
7. Other:


Mother's Marital Status:

1. Single/Never married
2. Married
3. Separated
4. Divorced
5. Widowed


Number of Years Married (if applicable):

Child's Native Language:

Mother's Native Language:


Child's Race:






72



Mother's Occupation: _________

Current Employment Status:
(please circle all that apply)

1. Unemployed
2. Volunteer
3. Employed part-time
4. Employed full-time
5. Homemaker
6. Student
7. Retired
8. Other- Please describe ______

Years of mother's total formal education: (please circle number of years completed)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20+

Post-secondary education: If attended college, number of years completed: ______

Degrees received:

Attended trade or technical school? Number of years completed: ______

How many children are in the family?______

What is the child's current grade in school?

Has this child ever had to repeat a grade in school?______ If so when?______

Please list the name(s) of the anti-seizure medications) your child is currently taking and the dose (how much
and how many times per day):


Please rate how difficult it is to manage which type of medication to give your child:

0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage how much medication to give your child:

0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage when to give medication to give your child:

0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult


9 10
Extremely
Difficult



9 10
Extremely
Difficult



9 10
Extremely
Difficult









Please name your child's seizure diagnosis:___________

Over the past month, how many seizures did your child have per week?______
per month?______

Please rate how much you know about your child's seizure disorder:


0 1 2 3 4 5
Not Somewhat
Knowledgeable Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how much you know about your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5 6 7 8 9 10
Somewhat Extremely
Knowledgeable Knowledgeable


Please rate how much you know about the type of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how much you know about the timing of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5 6 7 8 9 10
Somewhat Extremely
Knowledgeable Knowledgeable


If applicable, please rate how much you know about the number of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely N/A
Knowledgeable


Please rate how much you know about the dosin2 of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how good you feel your child's seizure medications are:


1 2 3 4 5
Somewhat
Good


6 7 8 9 10
Extremely
Good


Please rate how good you feel you are in managing your child's seizure medications:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Good Good Good


0
Not
Good









Please describe what type of information would be helpful to further your understanding of seizure
disorders:






Financial Status:

It is important for this study that we know something about your financial circumstances. We realize these are
extremely personal matters and we wish to assure you that your responses will be kept strictly confidential. Would
you please put an "X" next to the number that gives the best estimate of your total household income?


0. __ prefer not to answer

S1. ___ Under $10,000

2. ___ $10,000 to $20,000

3.___ $20,001 to 30,000


4. ___ $30,001 to $40,000

5. __ $40,001 to $50,000

6. ___ $50,001 to $60,000

7. ___ Over $60,000


Please do not answer the following, the researchers will fill out this portion.

Family ID: __________


Notes:















APPENDIX B
BACKGROUND INFORMATION FORM TIME 2

Date_______


BACKGROUND INFORMATION

Child' s Name: _________________

Mother's Name: ________________


Please list the name(s) of the anti-seizure medications) your child is currently taking and the dose (how much
and how many times per day):


Please rate how difficult it is to manage which type of medication to give your child:
0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage how much medication to give your child:
0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage when to give medication to give your child:
0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult
Please name your child's seizure diagnosis:___________


9 10
Extremely
Difficult


9 10
Extremely
Difficult


9 10
Extremely
Difficult


Over the past month, how many seizures did your child have per week?______
per month?_____

Please rate how much you know about your child's seizure disorder:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Knowledgeable Knowledgeable Knowledgeable

Please rate much you know about your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5 6 7 8 9 10
Somewhat Extremely
Knowledgeable Knowledgeable










Please rate how much you know about the type of your child's seizure medicationss:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Knowledgeable Knowledgeable Knowledgeable


Please rate how much you know about the timing of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


If applicable, please rate how much you know about the number of your child's seizure medicationss:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely N/A
Knowledgeable Knowledgeable Knowledgeable

Please rate how much you know about the dosing of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how good you feel your child's seizure medications are:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Good Good Good

Please rate how good you feel you are in managing your child's seizure medications:


0 1 2 3 4 5
Not Somewhat
Good Good


6 7 8 9 10
Extremely
Good


Please describe what type of information would be helpful to further your understanding of seizure
disorders:


Please do not answer the following, the researchers will fill out this portion.

Family ID: __________


Notes:















APPENDIX C
MEDICAL CHART INFORMATION FORM TIME 1

MEDICAL CHART INFORMATION

Patient's Name:

Patient's DOB: ______ Age_____

Date of Diagnosis: _______

Newly Diagnosed: yes_____ no____
If no, time since diagnosis (months):_______

Seizure Frequency (over the past month):per week _____ per month ____

Seizure Type(s) {the type(s) of seizures your child is diagnosed with}:


Medication(s) and Dose Info:


ER Visits Dates:

Blood Levels:


Pharmacy Refill Information:


% Concordance between prescribed medication regimen and reported medication
regimen_______


Complexity_____















APPENDIX D
MEDICAL CHART INFORMATION FORM TIME 2

MEDICAL CHART INFORMATION

Patient's Name:

Patient's DOB: _____ Age_____

Date of Diagnosis: _______

Newly Diagnosed: yes_____ no____
If no, time since diagnosis (months):_______

Seizure Frequency (over the past month):per week ______ per month ______

Seizure Type(s) {the type(s) of seizures your child is diagnosed with}:





Medication(s) and Dose Info:______________


ER Visits Dates:

Blood Levels:


Pharmacy Refill Information:


% Concordance between prescribed medication regimen and reported medication
regimen______


Complexity______














LIST OF REFERENCES


Abidin, R.R. (1995). Parenting Stress Index.: Professional Manual (3rd ed). Odessa:
Psychological Assessment Resources.

Adams, P.F., Hendershot, G.E., & Marano, M.A. (1996). Current Estimates from the
National Health Interview Survey: National Center for Health Statistics. Vital
Health Statistics, 10(200), 85-86.

Anderson, B.J., Auslander, W.F., Jung, D.C., Miller, J.P., & Santiago, J.V. (1990).
Assessing family sharing of diabetes responsibilities. Journal ofPediatric
Psychology, 15, 477-492.

Appleton, R. & Gibbs, J. (1998). Epilepsy in Childhood and Adolescence. Malden:
Blackwell Science Incorporated.

Austin, J.K. (1989). Predicting parental anticonvulsant medication compliance using the
theory of reasoned action. Journal ofPediatric Nursing, 4(2), 88-95.

Beck, D.E., Fennell, R.S., Yost, R.L., Robinson, J.D., Geary, D., & Richards, G.A.
(1980). Evaluation of an educational program on compliance with medication
regimens in pediatric patients with renal transplants. The Journal of Pediatrics,
96, 1094-1097.

Bobrow, E.S., Avruskin, T.W., & Siller, J. (1985). Mother-daughter interaction and
adherence to diabetes regimens. Diabetes Care, 8, 146-151.

Bond, G.G., Aiken, L.S., & Somerville, S.C. (1992). The health belief model and
adolescents with insulin-dependent diabetes mellitus. Health Psychology, 11,
90-198.

Brownbridge,G. & Fielding, D.M. (1994). Psychosocial adjustment and adherence to
dialysis treatment regimens. Pediatric Nephrology, 8, 744-749.

Browne, T.R. & Holmes, G.L. (2000). Handbook of epilepsy (2nd edition).
Philadelphia: Lippincott, Williams, & Wilkins.

Burt, S.A., Krueger, R.F., McGue, M., & lacono, W. (2003). Archives of General
Psychiatry, 60(5), 505-513.








Cummings, E.M. & Davies, P.T. (1994). Maternal depression and child development.
Journal of Child Psychology and Psychiatry, 35(1), 73-112.

Downey, G. & Coyne, J. (1990). Children of depressed parents: An integrative review.
Psychological Bulletin, 108(1), 50-76.

Friedman, I.M., Litt, I.F., King, D., Henson, R., Holtzman, D., Halverson, D., &
Kraemer, H. (1986). Compliance with anticonvulsant therapy by epileptic youth.
Journal ofAdolescent Health Care, 7, 12-17.

Gudas, L.J., Kroocher, G.P., & Wyplj, D. (1991). Perceptions of medical compliance in
children and adolescents with cystic fibrosis. Developmental & Behavioral
Pediatrics, 12, 236-247.

Hazzard, A., Hutchinson, S.J., & Krawiecki, N. (1990). Factors related to adherence to
medication regimens in pediatric seizure patients. Journal ofPediatric
Psychology, 15(4), 543-555.

Hodes, M. & Garralda, M.E. (1999). Maternal expressed emotion and adjustment in
children with epilepsy. Journal of Child Psychology and Psychiatry, 7, 1083-
1093.

Jacobsen, A.M., Hauser, S.T., Wolfsdorf, J.I., Houlihan, J., Milley, J.E., Herskowitz,
R.D., Wertlieb, D., & Watt, E. (1987). Psychologic predictors of compliance in
children with recent onset diabetes mellitus. Journal of Pediatrics, 110, 805-811.

Kyngas, H. (2000). Compliance with health regimens of adolescents with epilepsy.
Seizure, 9, 598-604.

LaGreca, A.M., Follansbee, D., & Skyler, J.S. (1990). Developmental and behavioral
aspects of diabetes management in youngsters. Children's Health Care, 19, 132-
139.

Lorenz, R.A., Christensen, N.K., & Pichert, J.W. (1985). Diet-related knowledge, skill,
and adherence among children with insulin-dependent diabetes mellitus.
Pediatrics, 75, 872-876.

McCaul, K.D., Glasgow, R.E., & Schafer, L.C. (1987). Diabetes regimen behaviors:
Predicting adherence. Medical Care, 25, 868-881.

Murray, L., Sinclair, D., Cooper, P., Ducournau, P., Turner, P., & Stein, A. (1999). The
socioemotional development of 5-year-old children ofpostnatally depressed
mothers. Journal of Child Psychology and Psychiatry, 40(8), 1259-1271.

Myers, L.B. & Midence, K. (Eds.). (1998). Adherence to treatment in medical
conditions. New Delhi: Harwood Academic Publishers.









Nezu, A.M. (Ed.). (2000). Practitioner's guide to empirically based measures of
depression. New York: Kluwer Academic, Plenum Publishers.

Otero, S. & Hodes, M. (2000). Maternal expressed emotion and treatment compliance of
children with epilepsy. Developmental Medicine and Child Neurology, 42(9),
604-608.

Patterson, J.M. (1985). Critical factors affecting family compliance with home treatment
for children with cystic fibrosis. Family Relations, 34, 79-89.

Peterson, G.M., McLean, S., & Millingen, K.S. (1982). Determinants of patient
compliance with anticonvulsant therapy. Epilepsia, 23, 607-613.

Radius, S.M., Marshall, H.B., Rosenstock, I.M., Drachman, R.H., Schuberth, K.C., &
Teets, K.C. (1978). Factors influencing mothers' compliance with a medication
regimen for asthmatic children. Journal ofAsthma Research, 15, 133-149.

Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the
general population. Applied Psychological Measurement, 1, 385-401.

Rapoff, M.A. (1999). Adherence to pediatric medical regimens. New York: Kluwer
Academic & Plenum Publishers.

Rapoff, M.A. & Christophersen, E.R. (1982). Improving compliance in pediatric
practice. Pediatric Clinics of North America, 29(2), 339-357.

Shorvon, S. (2000). Handbook of epilepsy treatment. Oxford: Alden Press Limited.

Shope, J.T. (1988). Compliance in children and adults: review of studies. Epilepsy Res.
Suppl., 1, 23-47.

Smith, S.D., Rosen, D., Trueworthy, R.C., & Lowman, J.T. (1979). A reliable method
for evaluating drug compliance in children with cancer. Cancer, 43, 169-173.

Spector, S., Tranah, A., Cull, C., & Goldstein, L.H. (1999). Reduction in seizure
frequency following a short-term group intervention for adults with epilepsy.
Seizure, 8, 297-303.

Tebbi, C.K., Richards, M.E., Cummings, K.M., Zevon, M.A., & Mallon, J.C. (1986).
The role of parent-adolescent concordance in compliance with cancer
chemotherapy. Adolescence, 28, 599-611.

United States Census Bureau. (2001). Poverty in the United States: 2000. (U.S. Census
Publication No. P 60-214). Washington, DC: U.S. Government Printing Office.





82


Wynn, K.S. & Eckel, E.M. (1986). Juvenile rheumatoid arthritis and home physical
therapy program compliance. Physical & Occupational Therapy in Pediatrics, 6,
55-63.

Zysset, T., Redeberg, A., Vasella, F., Kupfer, A., & Bircher, J. (1981). Phenytoin
therapy for epileptic children: Evaluation of salivary and plasma concentrations
and of methods of assessing compliance. Developmental Medicine and Child
Neurology, 23, 66-75.














BIOGRAPHICAL SKETCH

Sobha P. Fritz was bom in St. Louis, Missouri in 1975 and raised in Tampa,

Florida. Sobha received a B.A. in psychology from the University of South Florida in

1997. Before coming to the University of Florida in 1998, she worked as a research

assistant for the Psychosocial Oncology Program at the H. Lee Moffitt Cancer Center.

Sobha married her high school sweetheart, Jason Jon Fritz, in December 1998, who is

completing his doctorate in molecular genetics and microbiology. Sobha received an

M.S. in clinical psychology from the University of Florida in 2000. She completed her

doctoral and internship training in the University of Florida's Department of Clinical and

Health Psychology, with a specialization in pediatric psychology, in 2003. Sobha will

begin a postdoctoral fellowship at Emory University and the Children's Healthcare of

Atlanta hospitals in Atlanta, Georgia in January 2004.








I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy. /


Robert Glueckatf, ClIaf.
Professor of Clinical and Health Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor ofPhiloso .


Pssell Bauer
Professor of Clinical and Health Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.


Paul Camrney
Professor of Neuroscience

I certify that I have read this study and that in my opinion it conforms to ac ptable
standards of scholarly presentation and is fully adequate, in scope and quality, as
dissertation for the degree of Doctor of Philosophy.
/ /

Gary Gef en
Associate Professor of Clinical and Health
Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
," _, -_ .... .__ ,. "--_.__..__
Claydell Home
Associate Professor of Nursing







This dissertation was submitted to the Graduate Faculty of the College of Health
Professions and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.

December 2003- _____' ....-
Dean, College of Health Professions


Dean, Graduate School








































UNIVERSITY OF FLORIDA
III1 1 IIIIIIIMt11 tl1 IIIIII11111
3 1262 08555 2866