Journal of Pediatric Psychology, Vol. 15, No. 4, 1990, pp.

423-436

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
Issues in Adherence With Pediatric Regimens
Annette M. La Greca1
University of Miami

Received February 9, 1990; accepted February 9, 1990

Adherence to recommendations for health care is an area of serious concern

for pediatric psychologists. The present article highlights several key issues
affecting research and clinical work in this area. They include: difficulties
defining and measuring adherence; developmental aspects of adherence; spe-
cial concerns with low income populations; and limitations of intervention
research. In addition, several directions for future investigation in the area
of pediatric adherence are addressed.
KEY WORDS: pediatric adherence; developmental factors; assessment of adherence; low in-
come populations; intervention.

Adherence with pediatric regimens is a critical area of concern for child

health. Empirical reports indicate that substantial percentages of youngsters
and families do not adequately adhere to prescribed medical regimens for
a wide range of pediatric problems. Significant nonadherence has been
documented for acute illnesses, such as otitis media and streptococcal pharyn-
gitis (Arnhold et al., 1970; Becker, Drachman, & Kirscht, 1972; Bergman
& Werner, 1963; Colcher & Bass, 1972; Mattar, Markello, & Yaffe, 1975),
as well as for chronic conditions, such as asthma (Eney & Goldstein, 1976;
Radius et al., 1978; Sublett, Pollard, Kadlec, & Karibo, 1979), insulin-
dependent diabetes mellitus (Johnson, Silverstein, Rosenbloom, Carter, &
Cunningham, 1986), and rheumatoid arthritis (Litt & Cuskey, 1980; Varni
& Jay, 1984), among others. Most remarkably, problems with treatment com-
pliance have been observed even when nonadherence is potentially life-
1
All correspondence should be sent to Annette M. La Greca, Department of Psychology, Univer-
sity of Miami, P. O. Box 248185, Coral Gables, Florida 33124.
423
0146-8693/90/0800-O423J06.00/0 © 1990 Plenum Publishing Corporation
424 La Greca

threatening, as is the case with medication regimens for renal transplant pa-
tients (Korsch, Fine, & Negrete, 1978).
Although Litt and Cuskey (1980) estimated an overall adherence rate
of 50% for pediatric conditions, general adherence rates may be misleading.
Problems with adherence tend to be considerably higher for long-tern ver-
sus short-term regimens (Sackett & Snow, 1979), and may vary as a function
of the complexity of the disease and treatment regimen, the type of patient

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
population, the age of the patient, and the method for assessing adherence
(Epstein & Cluss, 1982). Moreover, since most studies only include patients
who are willing to participate in research projects, and who are actively in-
volved in medical care (at least to the extent of seeking medical treatment),
rates of adherence may grossly underestimate the extent of the problem (Ep-
stein & Cluss, 1982; La Greca, 1988a).
Concerns about treatment adherence are likely to continue as advance-
ments in medical science lead to newer and more complex treatments for many
pediatric conditions. With childhood diabetes, for instance, new technologies
for home blood glucose monitoring have lead to more complex, intensified
regimens (e.g., Schade, Santiago, Skyler, & Rizza, 1983). Although young-
sters with diabetes can achieve much better metabolic control with intensi-
fied regimens, they require multiple daily blood glucose tests, skills at using
reflectance meters to obtain accurate test results, and daily reevaluation and
readjustment of insulin needs and/or food intake, among other tasks. This
is considerably more demanding for the youngster and family than conven-
tional treatments which typically involve twice-daily glucose tests, and insu-
lin adjustments every 3 to 4 months under medical supervision. As another
example, progress in pediatric oncology has generated new life-saving treat-
ments that require cooperation with aversive medical procedures, such as
chemotherapy and medications for bone marrow transplantation (Phipps &
DeCuir-Whalley, this issue). As we approach the 21st century, issues regarding
adherence with pediatric regimens will continue to demand the attention of
pediatric psychologists.
It is for this reason that the present issue of the Journal is devoted to
adherence with pediatric regimens. The articles herein represent a range of
pediatric conditions, and examine varied facets of this behavioral-medical
problem. Together they provide a tantalizing view of the kinds of research
problems and issues that investigators confront in this challenging field of
study.
In this opening editorial, I highlight several issues raised by the arti-
cles. I also consider several conceptual and methodological advances that
are needed to further our understanding of why adherence represents such
a significant problem and how we may better prepare youngsters and fami-
lies to cope with the demands of complex and challenging treatment regimens.
Issues in Adherence 425

DEFINITIONS OF ADHERENCE: WHAT ARE WE REALLY

STUDYING?

The terms adherence and compliance often have been used in the pedi-
atric literature. According to the American Heritage Dictionary of the En-
glish Language (Morris, 1983), adhere means "to follow without deviation."
Among health psychologists, compliance has been defined as, "the extent

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
to which a person's behavior . . . conincides with medical advice" (Haynes,
1979, pp. 2-3). Such wording implies that there are criteria against which
to compare the individual's actions and behavior. In medical settings, the
criteria would be the patient's prescribed regimen. However, one apparent
contradiction in this field is that most measures of adherence or compliance
do not actually measure the individual's behavior in relation to a prescribed
medical regimen. Moreover, different regimens may be prescribed for pa-
tients with the same disease, further obscuring the meaning of the term ad-
herence. To cope with these difficulties, some investigators examine
individuals' behaviors in comparison with an (implicit or explicit) ideal regi-
men or simply measure the frequency of health-related behaviors, without
making any comparisons to standards or prescriptions.
Although these are practical solutions, they raise the question, "What
are we actually measuring?" Perhaps self-care or disease management would
be more accurate ways of describing the construct. Note that none of the
investigations in this special issue explicitly compares youngsters' behaviors
with their doctors' advice. Although some use an ideal regimen as a stan-
dard for comparison, it is not clear that everyone was prescribed this regi-
men by their doctor. The advantage, however, of using an ideal regimen is
that we can at least begin to understand the extent to which following an
ideal regimen (or not) is related to treatment outcome—an important con-
sideration for medical management of a given disease. However, it compli-
cates our study of factors related to self-care as we must then consider whether
proper advice was given, and whether this was adequately communicated
to the youngsters and families. We cannot assume that failures to follow the
ideal are a function of the patient and family, as it is also plausible that devi-
ations from the ideal are a function of problems with the medical provider's
treatment recommendations and/or patient-provider communication, among
other possibilities.
Regardless of the terminology, it should be apparent throughout this
special issue that investigators use very different definitions of adherence.
Two general definitional approaches are represented in this issue. One ap-
proach, which appears to be most common in initial efforts to study treat-
ment adherence, involves establishing categories of "adherent" and
426 La Greca

"nonadherent" patients (e.g., Mattar et al., 1975; Phipps & DeCuir-Whalley,

this issue). These patient groups are subsequently compared on other varia-
bles of interest. Although this represents a reasonable starting point for in-
vestigations of adherence, the arbitrary nature of such classifications is a
serious drawback. From a methodological standpoint, the use of nonstan-
dardized cutoffs for classifying patients as adherent limits our ability to make
comparisons across studies, or even across different aspects of a particular

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
treatment regimen (Epstein & Cluss, 1982; La Greca, 1988a). A further com-
plication is that for most illnesses it is not known what level of treatment
adherence is essential for achieving the desired therapeutic result (Epstein
& Cluss, 1982). Our lack of precision in understanding the relationships be-
tween medical prescriptions and therapeutic outcomes represents a serious
limitation for the categorical approach. In this issue, Lemanek discusses sever-
al problems with the categorical approach as it applies to medication regi-
mens for asthma.
Rather than identifying patients as adherent or not, a second defini-
tional approach concentrates on developing adherence rates for various health
care behaviors. In this issue, work by Johnson and colleagues perhaps best
illustrates this approach (Reynolds, Johnson, & Silverstein, this issue). The
advantage of this perspective is that it places adherence on a continuum,
avoiding the use of arbitrary classifications. Furthermore, many investiga-
tors who use this approach calculate adherence rates separately for various
aspects of a regimen, so that the complexity of treatment adherence can be
appreciated.
Separate adherence indices are particularly advantageous for studying
chronic diseases with multicomponent regimens, such as diabetes, as it ap-
pears that adherence to one aspect of treatment is not necessarily predictive
of adherence to other management tasks (Glasgow, McCaul, & Schafer, 1987;
Johnson et al., 1986; La Greca, 1988b). With diabetes, adherence tends to
be much higher for insulin administration than for tasks that require greater
life-style alterations, such as following a specific meal plan or exercise pro-
gram (Glasgow et al., 1987; Hauser et al., this issue; Johnson et al., 1986;
La Greca, 1988b). One would expect similar independence of regimen tasks
for diseases such as asthma or cystic fibrosis. Moreover, problems and bar-
riers to care may vary as a function of the specific health-care task (La Gre-
ca & Hanna, 1983; La Greca, 1988b).
There is considerable debate regarding whether adherence should be
investigated as a general construct or as a group of distinct behaviors. Propo-
nents of the global approach argue that the effects of the entire treatment
program may be missed when discrete components of a regimen are examined
separately. In many cases, to achieve good therapeutic outcome, multiple
aspects of the regimen must be followed. On the other hand, a global ap-
Issues in Adherence 427

proach masks the differential significance of individual adherence behaviors,

which may vary in their therapeutic impact. A further concern with the global
approach is that different investigators use different criteria to derive an over-
all adherence index (cf. Anderson, Auslander, Jung, Miller, & Santiago, this
issue; Jacobson et al., this issue), making cross-study comparisons very
difficult.
Clearly, no one definition of adherence or disease management is satis-

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
factory in all cases. Investigators must carefully consider their objectives in
order to choose wisely.

MEASURING ADHERENCE

As we move beyond clinical description of adherent and nonadherent

patients (e.g., Phipps & DeCuir-Whalley, this issue) perhaps the single most
difficult question confronting investigators is how to measure adherence.
Some areas of pediatric adherence are further along than others in terms of
the extent of measurement development. Yet, even for a given illness or treat-
ment, measures of adherence vary widely from one study to the next. In the
present issue, Lemanek reviews several measures for assessing compliance
among youngsters with asthma. It is also instructive to review the articles
on youngsters with diabetes; measures of adherence have included very differ-
ent kinds of parent and patient reports (see Anderson et al. and Reynolds
et al., this issue) as well as physicians' ratings (see Hauser et al. and Jacob-
son et al., this issue). The studies in this special issue also suggest that rat-
ings completed by different informants, such as the parent and child
(Anderson et al.) or parent and physician (Hazzard, Hutchinson, & Krawiecki,
this issue), do not always correspond very highly.
Measures of adherence could be viewed as ranging on a continuum from
direct to indirect. The most direct measures include drug assays (blood, se-
rum, or urine tests) and behavioral observations. Pill counts and indices of
therapeutic outcome are less direct, but typically are not as subjective as phy-
sician ratings and parent or patient reports. Each type of measure has its
assets and limitations and should be used carefully (for further discussion,
see Cluss & Epstein, 1985; La Greca, 1988a; Mathews & Christophersen,
1988).
With medication regimens, drugs assays represent one of the most com-
mon methods for evaluating adherence. In this issue, Hazzard et al. based
their adherence index on information obtained from drug assays for anticon-
vulsive medications. Assays are appealing, as they represent an objective,
reliable, and quantifiable method for indexing adherence, as long as the medi-
cation can be traced by an assay and the cost is not too prohibitive. However,
428 La Greca

in many instances, drug assays reflect medication-taking over very short in-
tervals Oust prior to the testing), and may not be representative of adher-
ence over longer periods. This clearly represents a drawback with chronic
conditions, such as epilepsy and asthma, as patients who are generally nonad-
herent, but who take medications just prior to medical appointments, may
appear more adherent than actually is the case. Other drawbacks to this type
of measure include individual variability in drug absorption rates (Lemanek,

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
this issue), and the complexity of tracking multiple medication regimens (Haz-
zard et al., this issue), among others.
For many pediatric regimens, a common approach has been to obtain
reports of adherence from youngsters and/or their parents. Such assessments
are relatively easy to obtain and are cost-effective (La Greca, 1988a). On
the negative side, patient and parent reports appear to be biased in the direc-
tion of overreporting adherence (Epstein & Cluss, 1982; La Greca, 1988a;
Mathews & Christophersen, 1988), although patients who admit to poor ad-
herence also appear to be nonadherent based on more objective measures
(Epstein & Cluss, 1982).
Generalizations regarding patient and parent reports must be made very
cautiously, as these reports can vary tremendously in their detail and the extent
to which the treatment regimen is comprehensively assessed. In this issue,
for example, both Hazzard et al. and Anderson et al. use single items to evalu-
ate parents' and youngsters' reports of adherence. This constrasts sharply
with the approach illustrated by Reynolds et al., where detailed 24-hour recall
interviews were conducted to determine adherence to multiple aspects of a
diabetes regimen. It is of interest to note that Reynolds et al. systematically
compared youngsters' reports of their diabetes care with direct observations
of their behavior. For the most part, youngsters' reports corresponded well
with observers' ratings, with the exception of dietary and exercise behaviors.
A third measurement strategy represented in this special issue involves
physician's ratings of adherence (see Hauser et al., and Jacobson et al., this
issue). Although ratings by health care providers have been viewed as an im-
provement over patient reports by some investigators (Mathews &
Christophersen, 1988), such ratings typically are based on information provid-
ed by patients. As such, physician ratings may be subject to the same biases
inherent in patient reports (La Greca, 1988a). In addition, physicians have
access to other information about the patient (e.g., history of disease con-
trol and management) that could potentially bias their ratings. On the other
hand, physicians have experience with a wide range of patients, and may
be able to make meaningful distinctions in adherence behaviors.
Regardless of the type of measure, it is important to document a rela-
tionship between adherence and therapeutic outcome (e.g., Anderson et al.
and Hazzard et al., this issue). If high rates of adherence are not linked with
Issues in Adherence 429

better therapeutic outcomes, one might question the adequacy of the meas-
ures used to evaluate these constructs, or the efficacy of the prescribed treat-
ment regimen.
No one measure of adherence or disease management is satisfactory
in all cases; each has its merits and drawbacks. However, moving toward
the development of objective measures that can be replicated reliably across
studies, and that demonstrate consistent relationships with treatment out-

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
come, is essential for the progress of the field.

DEVELOPMENTAL ISSUES

Developmental status is a critical variable for understanding youngsters'

reactions to illness and the extent of their involvement in disease manage-
ment (La Greca, 1988a). Throughout this issue, investigators demonstrate
considerable sensivity to developmental aspects of treatment adherence. Sever-
al authors note that adolescents appear to be less adherent than younger chil-
dren (Anderson et al. and Lemanek, this issue). This pattern is consonant
with other research findings pertaining to youngsters with chronic illness (e.g.,
Johnson et al., 1986; Litt & Cuskey, 1980; Smith, Rosen, Trueworthy, &
Lowman, 1979). Of further interest is the finding that age near the time of
disease onset is also an important correlate of disease management. In this
issue, the studies by Jacobson, Hauser, and colleagues suggest that young-
sters with diabetes who are older (i.e., adolescents) at the time of disease
onset have more problems with short-term and long-term adherence than
younger children. Despite consistent findings of this nature, we actually know
very little about why developmental status might affect adherence, or the
specific mechanisms that underlie such a relationship.
A common interpretation is that the turmoil of adolescence contrib-
utes to rebellion and problems with parental and medical authority; this in
turn is presumed to affect treatment adherence. Although theoretically and
intuitively appealing, studies have not actually documented that adolescents
who experience problems with adherence are in fact high on measures of re-
bellion or have authority problems. Thus, these theoretical linkages have not
been empirically validated and remain speculative at this point. At the same
time, a number of competing hypotheses have been ignored.
For one, some diseases present a greater challenge for adolescents to
control, due to the marked physical growth and metabolic fluctuations that
accompany the onset and course of puberty (La Greca, 1988; La Greca &
Skyler, in press). For instance, puberty is associated with a marked decrease
in insulin sensitivity (Bloch, Clemmons, & Sperling, 1987) which, in the case
of diabetes, necessitates a careful revaluation of insulin needs and changes
430 La Greca

in youngsters' treatment regimens (La Greca & Skyler, in press). If disease

management at this developmental stage does not produce positive therapeutic
results, it is possible that adolescents will disengage from intensive self-care
efforts. In this scenario, physical changes associated with adolescence may
lead to problems with adherence.
Another factor that may mediate a relationship between age and ad-
herence is the degree of responsibility that youngsters and family members

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
assume for treatment management. As they become older, youngsters as-
sume greater responsibility for their medical care (Anderson et al., this is-
sue; La Greca, 1988b). However, Anderson and colleagues note that in some
families communication regarding responsibilities for treatment management
may not be clear. This, in turn, could contribute to problems with adher-
ence and lead to poor therapeutic outcome.
Yet another wrinkle in this picture is youngsters' level of knowledge
and skills for implementing treatment regimens. Studies suggest that although
adolescents are more knowledgeable about their disease than younger chil-
dren (Johnson et al., 1982; La Greca, 1988b; Susman et al., 1982), mothers
are the most knowledgeable group (Johnson et al., 1982). The team of mother
(or parent) and child working together may be more effective for disease
management than the youngster alone; the latter situation is much more com-
mon with adolescents.
Although these hypotheses suggest a few alternate ways of viewing com-
monly reported associations between age and treatment adherence, answers
can only be obtained by empirical investigations that carefully examine poten-
tial mediators. To accomplish this, it is important to recognize the inade-
quacies of using youngsters' chronological age as the sole index of
development. At best, age provides a very gross summary of the many cog-
nitive, social, emotional, behavioral, and physical changes that occur through-
out childhood and adolescence (La Greca, 1988a; also see Achenbach, 1978).
A more informative approach would be to consider specific developmental
parameters associated with chronological age that may help to explain and
mediate age/adherence relationships.
One additional point is worth noting. Age-adherence relationships may
vary as a function of the demands of certain diseases or treatment regimens.
The most consistent links between adolescence and poor adherence have been
obtained for chronic diseases with complex treatment regimens, such as dia-
betes and asthma (see Lemanek, this issue), or for regimens that affect phys-
ical appearance (Korsch et al., 1978; also see La Greca, 1990). In contrast,
Phipps and DeCuir-Whalley (this issue) found adolescents to be more ad-
herent than younger children with a regimen for bone marrow transplanta-
tion that involved aversive medications. In the case of bone marrow
transplantation, it is possible that the greater cognitive maturity associated
with adolescence facilitated youngsters' endurance of aversive treatments.
Issues in Adherence 431

Further research is important to determine the specific factors that mediate

age/adherence relationships for various types of medical regimens.

LOW-INCOME POPULATIONS

Little systematic attention has been devoted to differences among vari-

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
ous pediatric populations. Yet, relative to middle-income patients, different
concerns may arise with low-income pediatric populations. Articles by Haz-
zard, Saylor, and Finney, and their colleagues (this issue) address issues as-
sociated with treatment adherence with primarily low-income clinic
populations. In this context, additional efforts to engage parents and chil-
dren in medical and psychological services appear to be desirable.
One prerequisite for participation in medical care entails keeping ap-
pointments with health care professionals. Saylor, Elksnin, Farah, and Pope
(this issue) and Finney, Lemanek, Brophy, and Cataldo (this issue) focus
on ways to promote better appointment keeping as a first step toward better
treatment adherence. Saylor et al.'s surveys of mothers and professionals rev-
ealed a number of logistical and tangible strategies that might facilitiate bet-
ter appointment keeping. Their findings dovetail with other research
identifying financial and transportation problems as barriers to appointment
keeping in clinic populations (Francis, Korsch, & Morris, 1969). Among a
number of suggestions, Saylor et al. recommend that professionals consider
using tangible rewards to promote better clinic attendance for early inter-
vention programs.
Finney et al. take this concept further by systematically evaluating the
efficacy of incentives and telephone reminders for improving appointment
keeping among youngsters and parents in an allergy clinic. Although these
authors found that tangible rewards promoted better appointment keeping,
they cautioned that some individuals may not be able to maintain these
changes when such interventions are withdrawn. Investigation of strategies
that promote utilization of medical services and improve treatment manage-
ment among low-income populations is an important direction for future
research.
Organizational difficulties may also contribute to problems with ad-
herence in clinic settings. In this issue, the article by Hazzard et al. high-
lights organizational factors that might affect adherence among patients
attending a pediatric neurology clinic. In their study, parents' satisfaction
with medical care was related to medication adherence among youngsters
with seizure disorders. The main reasons for parents' dissatisfaction were
long waiting periods, having a number of different doctors providing medi-
cal care, and unclear communication by doctors—problems that often charac-
432 La Greca

terize clinic settings. Other research has associated poor compliance with long
waiting periods (Haynes, 1979) and good compliance with continuity of health
care professionals (Mattar & Yaffee, 1974). Further study of organizational
strategies that enhance medical care and parent satisfaction may lead to im-
proved treatment adherence.

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
INTERVENTION RESEARCH

Judging by the articles that appear herein, considerably more energy

is devoted to investigations of factors related to treatment adherence than
to methods for facilitating adherence among children, adolescents, and fa-
milies. Yet, well-controlled intervention studies can be very revealing with
respect to factors that influence adherence in a causal manner.
To date, much of the literature examining interventions to improve ad-
herence with pediatric regimens has focused on youngsters with chronic con-
ditions, such as asthma and hemophilia (see Lemanek, this issue; Varni &
Wallander, 1984). This emphasis most likely reflects recognition of the ex-
treme difficulties inherent in managing chronic conditions, as well as the seri-
ous health consequences associated with persistent nonadherence. In fact,
Varni and Wallander (1984) noted that we should expect nonadherence with
long-term regimens, in the absence of behavioral interventions or incentives
to support positive health-care behaviors.
For youngsters with chronic pediatric regimens, the predominant fo-
cus in the intervention literature has been on single-subject research designs
that evaluate behavioral strategies (see La Greca, 1988a; Varni & Wallander,
1984, for reviews). In the present issue, the article by Finney et al. provides
a good illustration of this research approach. These authors found that in-
centives and reminders improved appointment-keeping behaviors in young-
sters attending an allergy clinic. Noteworthy features of this study were the
involvement of both high-adherent and low-adherent subjects, and the ex-
tensive follow-up period for monitoring youngsters' appointment-keeping be-
havior. It is interesting to observe that some, but not all, of the youngsters
and families experienced a deterioration in clinic attendance once the incen-
tive program was withdrawn — even though the telephone reminders con-
tinued. This finding highlights the difficulty of programming treatment
maintenance as well as the individual variability in youngsters' and families'
responses to intervention. These two issues (treatment maintenance; varia-
bility in response to intervention) deserve continued empirical attention.
In a review of literature on interventions for promoting pediatric ad-
herence (La Greca, 1988a), several general conclusions emerged which con-
tinue to be relevant. First, with short-term regimens for acute illnesses, several
Issues in Adherence 433

strategies may aid adherence, such as providing the youngster and family
with verbal and written instructions or visual cues and reminders, and in-
creasing medical supervision and support. Second, with chronic pediatric con-
ditions that have complex regimens, the most successful interventions have
been multifaceted, combing intensive education, parental involvement, self-
monitoring, and/or reinforcement strategies.
Despite these encouraging findings, intervention studies have been

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
plagued by a variety of methodological difficulties, such as reliance on small
samples, failure to establish a relationship between adherence and therapeu-
tic outcome, focus on short-term interventions with limited follow-up peri-
ods, or use of subject selection procedures that exclude resistant or
low-adherent youngsters and families (La Greca, 1988a; also see Lemanek,
this issue). Clearly, there is a need for well-controlled research in this area,
using methodologically sound instruments and better experimental designs.
It is also noteworthy that few intervention studies focus on factors that
have been identified as correlates of adherence and nonadherence in descrip-
tive research. Furthermore, given the multitude of factors related to problems
with disease management, one type of intervention strategy is not likely to
be effective for all or even most youngsters with a given illness. Efforts to
identify factors contributing to problems with adherence, and then match-
ing interventions to the needs of youngsters and families, are important con-
siderations for intervention research.

DIRECTIONS FOR THE FUTURE

Several directions for future research in pediatric adherence merit at-

tention. First, greater emphasis on longitudinal and prospective research de-
signs is essential for furthering our understanding of adherence to long-term
treatment regimens. With few exceptions (Finney et al., Hauser et al., Jacob-
son et al., this issue), the studies in this special issue report correlational find-
ings; this emphasis reflects trends that are apparent in the broader adherence
literature. We know that a multitude of individual differences (e.g., coping
strategies, psychological adjustment) and family factors (e.g., cohesion, con-
flict, maternal anxiety) are associated with adherence, yet we know consider-
ably less about the ways in which these variables influence and are influenced
by treatment management.
Especially with chronic conditions, adherence is not static and unvary-
ing; adherence rates may change considerably over time for the same young-
ster and family. This point is illustrated by the work of Hauser, Jacobson,
and their colleagues in this issue. Among youngsters with diabetes, initial
adherence rates were significantly related to adherence over a subsequent
434 La Greca

4-year period; however, the magnitude of these relationships suggested that

fluctuations in adherence also occurred. Moreover, overall ratings of adher-
ence declined each year after disease onset. At the present time we know very
little about factors that predict changes in youngsters' disease management.
How do youngsters' attempts to manage their disease affect subsequent ad-
herence? When efforts to manage a disease do not yield positive outcomes,
is treatment management adversely affected? Questions such as these can

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
only be answered with longitudinal and prospective research designs.
Along with a shift away from correlational methods, it may be fruitful
to delineate subgroups of nonadherers. Variables associated with adherence
are complex and multifaceted; they include disease knowledge and daily
management skills, individual psychological variables, family environment,
physical factors, and financial barriers, among others. Yet, for any given
youngster and family, problems in one or more of these areas may impede
disease management. By identifying appropriate subgroups, interventions
could be directly tailored to the needs of youngsters and their families.
Finally, the development of conceptual models for guiding adherence
research is a critical concern for the future. Descriptive research examining
correlates of pediatric adherence will not likely advance our understanding
of youngsters' disease management unless it is guided by well-reasoned, mul-
tivariate, conceptual models. Several investigators in the area of childhood
diabetes have developed models for examining factors related to treatment
adherence and metabolic control (e.g., Hanson, Henggeler, & Burgen, 1987;
Johnson, 1989; La Greca, 1988b; La Greca & Skyler, in press). Further ef-
forts in this direction are important and desirable.
In summary, this editorial highlighted a number of conceptual and
methodological issues confronting researchers in the area of pediatric ad-
herence. Several directions for future research were discussed that should
be critical for the progress of the field. Of the suggestions offered, perhaps
the greatest need is in the area of measurement development. The availabili-
ty of psychometrically sound measures to assess children's health care is es-
sential for advancing our understanding of how youngsters and families cope
with the challenge of pediatric illness.

REFERENCES

Achenbach, T. M. (1978). Psychopathology of childhood: Research problems and issues. Journal

of Consulting and Clinical Psychology, 46, 759-776.
Arnhold, R. G., Adebonojo, F. O., Callas, E. R., Callas, J., Carte, E., & Stein, R. C. (1970).
Patients and prescriptions: Comprehension and compliance with medical instructions
in a suburban pediatric practice. Clinical Pediatrics, 9, 648-651.
Becker, M. H., Drachman, R. H., & Kirscht, J. P. (1972). Predicting mothers' compliance with
pediatric medical regimens. Journal of Pediatrics, 18, 843-854.
Issues in Adherence 435

Bergman, A. B., & Werner, R. J. (1963). Failure of children to receive penicillin by mouth.
New England Journal of Medicine, 268, 1334-1338.
Bloch, C. A., demons, P., & Sperling, M. A. (1987). Puberty decreases insulin sensitivity.
Journal of Pediatrics, 110, 481-487.
Cluss, P. A., & Epstein, L. (1985). The measurement of medical compliance in the treatment
of disease. In P. Karoly (Ed.), Measurement strategies in health psychology (pp. 403-432).
New York: John Wiley & Sons.
Colcher, I. S., & Bass, J. W. (1972). Penicillin treatment of streptococcal pharyngitis: A com-
parison of schedules and the role of specific counseling. Journal of the American Medi-
cal Association, 222, 657-659.

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
Eney, R. D., & Goldstein, E. O. (1976). Compliance of chronic asthmatics with oral adminis-
tration of theophylline as measured by serum and salivary levels. Pediatrics, 57, 513-517.
Epstein, L. M., & Cluss, P. A. (1982). A behavioral medicine perspective on adherence to long-
term medical regimens. Journal of Counsulting and Clinical Psychology, SO, 950-971.
Francis, V., Korsch, B. M., & Morris, M. J. (1969). Gaps in doctor-patient communication:
Patients' response to medical advice. New England Journal of Medicine, 280, 535-540.
Glasgow, R. E., McCaul, D. K., & Schafer, L. C. (1987). Self-care behaviors and glycemic
control in type I diabetes. Journal of Chronic Diseases, 40, 399-412.
Haynes, R. B. (1979). Introduction. In R. B. Haynes, D. W. Taylor, & D. L. Sackett (Eds.),
Compliance in health care (pp. 1-10). Baltimore: Johns Hopkins University Press.
Heisler, A. B., & Friedman, S. B. (1981). Social and psychological considerations in chronic
disease: With particular reference to the management of seizure disorders. Journal of
Pediatric Psychology, 6, 239-250.
Johnson, S. B. Pollak, T. Silverstein, J. H., Rosenbloom, A. L., Spillar, R., McCallum, M.,
& Harkavy, J. (1982). Cognitive and behavioral knowledge about insulin-dependent di-
abetes among children and parents. Pediatrics, 69, 708-713.
Johnson, S. B., Silverstein, J., Rosenbloom, A., Carter, R., & Cunningham, W. (1986). As-
sessing daily management in childhood diabetes. Health Psychology, 5, 545-564.
Korsch, B. M., Fine, R. N., & Negrete, V. F. (1978). Noncompliance in children with renal
transplants. Pediatrics, 61. 872-876.
La Greca, A. M. (1988a). Adherence to prescribed medical regimens. In D. K. Routh (Ed.),
Handbook of pediatric psychology (pp. 299-320). New York: The Guilford Press.
La Greca, A. M. (1988b). Children with diabetes and their families: Coping and disease manage-
ment. In T. M. Field, P. M. McCabe, & N. Schneiderman, Stress and coping across
development (pp. 139-159). Hillsdale, NJ: Erlbaum.
La Greca, A. M. (1990). Social consequences of pediatric conditions: Fertile area for future
investigation and intervention? Journal of Pediatric Psychology, 15, 285-308.
La Greca, A. M., & Skyler, J. S. (in press). Psychosocial issues in IDDM: A multivariate frame-
work. In P. McCabe, T. Field, J. Skyler, & N. Schneiderman (Eds.), Stress, coping and
disease. Hillsdale, NJ: Erlbaum.
La Greca, A. M., & Hanna, N. C. (1983). Health beliefs of children and their mothers: Impli-
cations for treatment. Diabetes, 32 (Suppl. 1), 66 (Abstract).
Litt, I. F., & Cuskey, W. R. (1980). Compliance with medical regimens during adolescence.
Pediatric Clinics of North America, 27, 3-15.
Mattar, M. E., Markello, J., & Yaffe, S. J. (1975). Pharmaceutic factors affecting pediatric
compliance. Pediatrics, 55, 101-108.
Mattar, M. E., & Yaffe, S. J. (1974). Compliance of pediatric patients with therapeutic regi-
mens. Postgraduate Medicine, 55, 181-188.
Matthews, J. R., & Christopherson, E. R. (1988). Measuring and preventing noncompliance
in pediatric health care. In P. Karoly (Ed.), Handbook of child health assessment, (pp.
519-557). New York: John Wiley & Sons.
Morris, W. (1983). American Heritage dictionary. Boston: Houghton Mifflin Co.
Radius, S. M., Becker, M. H., 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 of Asthma Research, 15, 133-149.
Sackett, D., & Snow, J. (1979). The magnitude of compliance and noncompliance. In R. Haynes,
D. Taylor, & D. Sackett (Eds.), Compliance in health care (pp. 11-22). Baltimore: Johns
Hopkins Press.
436 La Greca

Schade, D. S., Santiago, J. V., Skyler, J. S., & Rizza (1983). Intensive insulin therapy. Prince-
ton, NJ: Excerpta Medica.
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.
Sublett, J. L., Pollard, S. J., Kadlec, G., & Karibo, J. M. (1979). Non-compliance in asthmatic
children: A study of theophylline levels in a pediatric emergency room population. An-
nals of Allergy, 43, 95-97.
Susman, E. J., Hersh, S. P., Nannis, E. D., Strope, B. E., Woodruff, P. J., Pizzo, P. A.,
& Levine, A. S. (1982). Conceptions of cancer: The perspectives of child and adolescent

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Fransisco on April 21, 2015
patients and their families. Journal of Pediatric Psychology, 7, 253-261.
Varni, J. W., & Jay, S. M. (1984). Biobehavioral factors in juvenile rheumatoid arthritis: Im-
plications for research and practice. Clinical Psychology Review, 4, 543-560.
Varni, J. W., & Wallander, J. L. (1984). Adherence to health-related regimens in pediatric chronic
disorders. Clinical Psychology Review, 4, 585-596.