DOI: 10.1542/peds.2013-1451; originally published online September 2, 2013; 2013;132;730Pediatrics

Meghan E. McGrady and Kevin A. HommelA Systematic Review

Medication Adherence and Health Care Utilization in Pediatric Chronic Illness:  

  http://pediatrics.aappublications.org/content/132/4/730.full.html

located on the World Wide Web at: The online version of this article, along with updated information and services, is

 

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2013 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from

DOI: 10.1542/peds.2013-1451; originally published online September 2, 2013; 2013;132;730Pediatrics

Meghan E. McGrady and Kevin A. HommelA Systematic Review

Medication Adherence and Health Care Utilization in Pediatric Chronic Illness:  

ServicesUpdated Information &

mlhttp://pediatrics.aappublications.org/content/132/4/730.full.htincluding high resolution figures, can be found at:

Supplementary Material

8/peds.2013-1451.DCSupplemental.htmlhttp://pediatrics.aappublications.org/content/suppl/2013/08/2Supplementary material can be found at:

References

ml#ref-list-1http://pediatrics.aappublications.org/content/132/4/730.full.htat:This article cites 40 articles, 13 of which can be accessed free

Permissions & Licensing

tmlhttp://pediatrics.aappublications.org/site/misc/Permissions.xhtables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2013 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by Gerri Mattson on October 1, 2013pediatrics.aappublications.orgDownloaded from

Medication Adherence and Health Care Utilization inPediatric Chronic Illness: A Systematic Review

abstractBACKGROUND AND OBJECTIVE: Advanced understanding of modifiablepredictors of health care use in pediatric chronic illness is critical toreducing health care costs. We examined the relationship betweenmedication non-adherence and health care use in children andadolescents who have a chronic medical condition.

METHODS: A systematic review of articles by using PubMed, PsycINFO,and CINAHL was conducted. Additional studies were identified bysearching reference sections of relevant manuscripts. Studies thattested the relationship between medication non-adherence andhealth care use (ie, hospitalizations, emergency department visits,outpatient visits) or cost in children and adolescents (mean age #18years) who have a chronic medical condition were included. Extraction ofarticles was completed by using predefined data fields.

RESULTS: Ten studies met our inclusion criteria. Nine of the 10 studiesreviewed (90%) demonstrated a relationship between medication non-adherence and increased health care use. The directionality of thisrelationship varied depending on the outcome variable of interest.

CONCLUSIONS: Medication non-adherence is related to increasedhealth care use in children and adolescents who have a chronicmedical condition and should be addressed in clinical care. Futurestudies should include randomized controlled trials examining theimpact of adherence promotion efforts on health care use andcosts. Pediatrics 2013;132:730–740

AUTHORS: Meghan E. McGrady, PhD and Kevin A. Hommel,PhD

Division of Behavioral Medicine and Clinical Psychology, Centerfor Adherence and Self-Management, Cincinnati Children’sHospital Medical Center, Cincinnati, Ohio

KEY WORDSmedication adherence, health care costs, health care quality,access, and evaluation, chronic disease

ABBREVIATIONSED—emergency departmentMPR—medication possession ratio

Dr McGrady contributed to the manuscript by aiding inconception and design, acquisition of data, analysis andinterpretation of data, drafting of manuscript, and statisticalanalysis; Dr Hommel contributed to the manuscript by aiding inconception and design, analysis and interpretation of data,critical revision of the manuscript for important intellectualcontent, and supervision; and both authors approved the finalmanuscript as submitted.

www.pediatrics.org/cgi/doi/10.1542/peds.2013-1451

doi:10.1542/peds.2013-1451

Accepted for publication Jul 23, 2013

Address correspondence to Meghan E. McGrady, PhD, Center forAdherence and Self-Management, Cincinnati Children’s HospitalMedical, Center MLC 7039, 3333 Burnet Ave, Cincinnati, OH 45229-3026. E-mail: meghan.mcgrady@cchmc.org

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2013 by the American Academy of Pediatrics

FINANCIAL DISCLOSURE: The authors have indicated they haveno financial relationships relevant to this article to disclose.

FUNDING: The design and conduct of the study; collection,management, analysis, and interpretation of the data; andpreparation of the manuscript were supported in part by grantT32HD068223 for Dr McGrady. Dr Hommel is funded in part byR01HD067174. Funded by the National Institutes of Health (NIH).

POTENTIAL CONFLICT OF INTEREST: The authors have indicatedthey have no potential conflicts of interest to disclose.

730 McGRADY and HOMMEL

In 2010, the United States spent $2.6trillion on health care, an increase of3.9% from 2009, and 17.9% of its grossdomestic product.1 Hospital care andphysician and clinical services (ie,emergency department [ED] visits,hospitalizations, and office visits, here-after defined as “health care use”) arethe 2 largest components of health carespending,2,3 and accounted for over 50%of the growth in health care spendingfrom 2009 to 2010.1 Continued risinghealth care costs despite the recession,the increasing US national deficit, andthe increasing percentage of US healthspending financed by the government($1.2 trillion, 45% of all US healthspending) have prompted investigationof modifiable factors to reduce healthcare use and associated costs.4 Al-though there is no single strategy foreffectively controlling costs, research-ers have begun to focus on the 83% ofhealth care resources consumed byindividuals who have a chronic medicalcondition.4,5

The number of children and adoles-cents diagnosedwith a chronicmedicalcondition has been steadily increasingover the past 20 years,6 driven in partby increases in the prevalence of obe-sity7 and asthma8 as well as advancesin medical care that increase survival(eg, cystic fibrosis, kidney transplant).9,10

Increases in the prevalence of chronicmedical conditions have only increasedthe already disproportionate healthcare expenses accounted for by childrenand adolescents who have a chronicillness.5,11,12 In 2000, children andadolescents with a special health careneed made up 16% of youth butaccounted for 53% of hospital days.13

Consequently, targeting factors thatinfluence health care use in this pop-ulation may provide 1 of the greatestopportunities to reduce pediatrichealth care spending.

Non-adherence refers to a lack ofcorrespondence between patient

self-managementbehaviorandmedicalor health advice14 and significantlycontributes to health care use in adultswho have a chronic illness,15–18 ac-counting for an estimated 33% to 69%of adult hospital admissions and $100to $300 billion in annual health carecosts.19,20 Because much of the healthcare use attributable to non-adherenceincludes excess use of urgent care andhospitalizations for preventable com-plications,16 it represents avoidablecosts, an ideal target outcome forinterventions aimed at reducing healthcare spending. As non-adherence ismodifiable with intervention,21 adultadherence promotion efforts have be-gun to focus on both improving healthstatus and reducing health care use,resulting in programs that effectivelydecrease health care costs.15

Efforts to contain pediatric health carecosts could prove to be equally as ef-fective as adult-focused interventions,but have not yet been examined despitethe widespread nature of non-adherencein pediatrics. Approximately 63% ofchildren and adolescents who havea chronic illness are prescribed medi-cation,22 but 50% to 88% of childrenand adolescents are non-adherent totheir prescribed regimens.23–25 As inadult samples, pediatric non-adherencemay be a modifiable predictor of healthcare use and resulting health carecosts. For example, an adolescent whohas asthma may require hospitalizationafter failing to take his controller med-ication for several days and suffering anexacerbation. If his medical care in-cluded regular assessments of his ad-herence and adherence promotionefforts as necessary, this hospitaliza-tion and its associated costs may havebeen avoided.16,26

Understanding the impactof adherencepromotion interventions on health carecosts specific to pediatric populationsis necessary given the numerous de-velopmental factors thatmake pediatric

adherence unique (ie, influence of adultcaregivers and systems on adherence,evolution of illness burden with age,increasing autonomy for disease man-agement) and the variations in healthcare systems and outcomes across thelifespan. For example, health care fi-nancing for over one-third of childrenand adolescents is provided by Med-icaid and Children’s Health InsurancePrograms.27 With inconsistencies inbenefits and reimbursement ratesbetween states, it is difficult to com-pare health care costs across pediat-ric and adult samples. Furthermore,given the more advanced diseasecourse and higher likelihood of com-plications experienced by adults whohave a chronic illness, the benefits inquality of life and cost savings result-ing from pediatric prevention effortsare likely to be less visible in short-term evaluations typically used withthese populations.28

An increased understanding of the re-lationship between non-adherence andhealth careuse inpediatric populationspresents an opportunity to examinea highly prevalent and modifiable(amenable to behavioral intervention)potential contributor to pediatrichealth care costs. Unlike other stablepredictors of health care use (ie,gender, socioeconomic status), non-adherence can be modified with be-havioral intervention, and thus is ofclinical significance. If non-adherencecontributes to pediatric health careuse, there is the potential to have evengreater impact on health care usethan adult efforts, because long-termself-management behaviors are oftendeveloped in childhood and adoles-cence.23 Specifically, promoting adher-ence in pediatric populations may havethe potential to reduce short-termhealth care use as well as long-termhealth care use that may result fromthe increased morbidity associatedwith non-adherence.23

REVIEW ARTICLE

PEDIATRICS Volume 132, Number 4, October 2013 731

Despite the potential significance, toour knowledge no studies have re-viewed the extant literature with a tar-geted examination of the relationshipbetweenmedicationnon-adherenceandhealth care use in children and ado-lescentswho have a chronic illness. Thisissue is timely as health care reformefforts and discussions regarding cov-ered services and service deliverymodels (ie, fee for service versus per-member, per month) move forward.We need to better understand howmodifiable factors contribute to healthcare costs and value to identify relevantand significant priorities of health carereform.Applicationoffindingswithadultpopulations would exclude many of theunique modifiable variables that impactpediatric adherence (ie, family func-tioning, caregiver involvement).14 Thepurpose of this systematic review is toprovide a comprehensive summary ofthe empirical studies examining therelationship between non-adherenceand health care use in childrenand adolescents who have a chronicmedical condition. Recommendationsfor advancing research in this fieldand policy implications are alsodiscussed.

METHODS

The systematic review was conductedin accordance with the publishedguidelines presented in the PreferredReporting Items for Systematic Reviewsand Meta-Analyses Statement.29

Data Search

PubMed, PsycINFO, and CINAHL weresearched in April 2013 for peer-reviewed original research articlespublished in English before April 2013.The search strategy included a combi-nation of Medical Subject Headingsterms (eg, chronic disease, patientcompliance,medicationadherence,healthcare costs, hospitalization, emergencyservice, ambulatory care) and key

words (eg, adherence, non-adherence,self-management, compliance, treat-ment concordance, health care utiliza-tion, outpatient visit, emergencydepartment, hospitalization, healthcare cost, service utilization, serviceuse, health care use, cost, chronic ill-ness, chronic care, chronic) (see Sup-plemental Information). To ensure allrelevant articles were captured, anadditional search was done replacingterms describing “chronic disease”with “diabetes,” “asthma,” or “HIV,” ef-fectively searching some of the mostdeveloped adherence literature basesin pediatrics (see Supplemental In-formation). The search did not includerestrictions on publication date orstudy design.

Study Selection and Screening

The Preferred Reporting Items forSystematic Reviews and Meta-Analyses4-phase flow diagram detailing studyselection is depicted in Fig 1.29 Theinitial search resulted in 4519 records.After the removal of duplicates, wescreened the abstracts of 4346 manu-scripts. Of these, 73 manuscripts metthe following inclusion criteria: (1)pediatric sample with a chronic illness;(2) original research article; and (3)mention of adherence and health careuse or cost. We then reviewed full-textversions of these 73 manuscripts. Weexcluded studies that did not assessnon-adherence and health care use,did not examine the relationship be-tween these variables, or includedparticipants with a mean age .18years.

Data Extraction

One reviewer (Dr McGrady) completeddata extraction from included articlesby using a standardized data collectionform. A second reviewer (Dr Hommel)checked the abstracted data for ac-curacy. Data retrieved from articlesincluded: (1) identification of the

manuscript (authors, year); (2) char-acteristics of participants (child age,child chronic illness, caregiver par-ticipating); (3) study design in accor-dance with published guidelines30

(cohort, cross-sectional, case-control,or randomized clinical trial); (4) mea-sure of non-adherence; (5) measure ofhealth care use; and (6) relationshipbetween non-adherence and healthcare use. In cases in which multiplemeasures of non-adherence or out-comes were used, all variables wereincluded. Questions regarding in-clusion or relevant data were resolvedvia discussion between authors.

Definitions of Variables

Measures of MedicationNon-Adherence

Medication non-adherence was as-sessed using measures of varyingvalidity and reliability,31 including self-report questionnaires, self-reportstructured interviews, electronic mon-itoring devices, prescription refill his-tories, and biochemical assays (seeTable 1). The time period during whichnon-adherence was assessed rangedfrom 24 hours to 365 days. Of note,somemeasures included non-adherencethreshold measurements with cut-points. As these cut-points differedacross studies, definitions of “good,”“high,” or “optimal” adherence pre-sented in this manuscript are thoseprovided by the authors of reviewedstudies.

Measures of Health Care Use

To facilitate comparison across stud-ies, we classified health care use dataas ED visits, outpatient visits, hospi-talizations, or “other” measures (eg,combined number of ED visits, out-patient visits, and hospitalizations) ofhealth care use. Measures of healthcare use were obtained via caregiverself-report, review of the electronic

732 McGRADY and HOMMEL

medical record, or review of insurancedatabases.

Risk of Bias

To ascertain the reliability of the re-sults of the included studies, the riskfor bias was assessed by using theitems from the Newcastle-OttawaQuality Assessment Scale.32,33 TheNewcastle-Ottawa Quality AssessmentScale is a valid and reliable measuredeveloped to assess the quality ofnonrandomized studies included insystematic reviews. High-quality stud-ies are identified using a “star system,”with more stars representing higherquality. Two of the 8 domains of studyquality (representativeness of the co-hort and outcome assessment method)were relevant to the included studies.We rated each study in accordancewith published guidelines on these 2domains.

RESULTS

Ten studies met all inclusion criteriaand were included in this review (see

Fig 1). Nine of the studies includedchildren and/or adolescents who hadasthma, and 1 study included a sampleof adolescents and young adults whohad type 1 diabetes. Mean age of studyparticipants ranged from 6.2 to 16years. Study designs included pro-spective observational cohort (n = 4),cross-sectional analyses (n = 3), ret-rospective observational cohort (n = 2),and randomized-controlled trials (n = 1).Three studies used managed care or-ganizations, Medicaid, or Children’sHealth Insurance Programs to accesspopulation-level data (range of sam-ple sizes, 1474–18 456). The otherstudies included data from single(n = 4; range of sample sizes, 63–150)or multiple hospitals/clinics (n = 3;range of sample sizes, 158–2960).Measures of adherence varied acrossstudies and included pharmacy refillrecords (n = 5), self-report measures(n = 3; eg, “In the past 2 weeks, howmany days would you guess that yourchild has forgotten to take his or hermedicine?”), and electronic monitors(n = 2).

Medication Non-Adherence and EDVisits

Six of the 7 studies examining the re-lationship between medication non-adherence and ED visits in childrenand adolescents who had asthmafound that greater non-adherenceusing multiple measures was relatedto more ED visits. Children and ado-lescents who had asthma who didnot fill any prescriptions for anti-inflammatory medications had a higherrisk for an ED visit than those whofilled at least 1 prescription.34,35 Fur-thermore, lower percentage of timeduring which children and adoles-cents possessed anti-inflammatorymedications (medication possessionratio [MPR] = [number of days coveredby dispensed medication/365] 3 100)was related to higher odds of an EDvisit.36 Consistent with these findings,a lower percentage of doses taken, asmeasured by electronic monitoringdevices, was associated with increasedED visits.37

In a study comparing groups of chil-dren who had asthma seen in the ED

FIGURE 1Study selection flow diagram.

REVIEW ARTICLE

PEDIATRICS Volume 132, Number 4, October 2013 733

TABLE1

StudiesExam

iningtheRelationshipBetweenAdherenceandHealth

Care

Usein

PediatricPopulations

Author

(Year)

StudyDesign

SampleSize

and

StudyPopulation

Measure

ofAdherence

ControlledFor

RelationshipBetweenAdherenceandHealth

Care

Use

EDVisits

Outpatient

Visits

Hospitalizations

OtherMeasuresof

Health

Care

Use

Adam

setal

(2001)

34Cohort

11195children

(ages3–15

yr)

with

asthmaa

Numberofcanisters

orcontainers

ofdrug

dispenseda

Age,gender,

site,reliever

medication

dispensing

Anydispensing

ofan

anti-inflam

matory

medicationwas

associated

with

adecreasedrisk

forED

visits.AdjustedRR

=0.4

(95%

CI:0.3–0.5),

P,

.01

XAnydispensing

ofan

anti-inflam

matory

medicationwas

associated

with

adecreasedrisk

forhospitalization.

Adjusted

RR=0.4(95%

CI:0.3–0.6),P

,.01

Ashkenazi

etal

(1993)

38

Cross-

sectional

100children(ages

2–14

y)with

asthmareferred

toapediatricED;

50childrenwith

asthmaattending

anHAC

Self-reportmeasure

including2

dichotom

ousitems

assessingwhether

medications

were

takeninthedose

and

frequencyprescribed

Compliancetoprescribed

dose

(P,

.001)and

frequency(P

,.001)

was

higher

inparticipantsin

theHACgroupthan

intheED

group

XX

STLa

STLwerehigher

inparticipantsinthe

HACgroupthan

inthe

EDgroup,P,

.001

XX

Bartlettetal

(2004)

40Cohort

158children

with

asthma

“How

oftendoes

your

child

forgettotake

his/herasthma

medication?”

Child

age,family

income,asthma

morbidity,m

aternal

depressive

symptom

s

NS,valuesnot

reported

XX

“Inthepast2weeks,how

manydays

wouldyou

guessthatyour

child

hasforgottentotake

hisor

hermedicine?”

Child

age,family

income,asthma

morbidity,

maternaldepressive

symptom

s

NS,valuesnot

reported

Baum

anetal

(2002)

41Cohort

1199

children

(ages4–9y)

with

asthmaa

Admitted

non-adherence:

numberoftim

esthat

caregivers

admitted

noncom

pliancewith

aphysician

recommendationfor

asthmamanagem

ent

XX

Adherencewas

not

associated

with

whether

thechild

was

hospitalized

inthepast

9months,

P=.059

Increasedadherence

was

associated

with

fewer

unscheduled

asthmavisits

(providerand

EDvisits),

P,

.001

Boylston

Herndon

etal

(2012)

36

Cohort

18456children

(ages2–18

y)with

asthmaa

MPR

forICS:the

percentage

ofdays

withina365-dperiod

thatan

ICSwas

supplied;Groupedinto:

0%to19%MPR;20%

to49%MPR,and

$50%

MPR

a

Demographics,

health

characteristics

$50%MPR

was

associated

with

lower

odds

ofan

EDvisitthan0%

to19%MPR,OR=0.56

(95%

CI:0.43–0.73),

P,

.001

Greatermedication

adherencewas

associated

with

morefrequent

asthma-related

office

visits,

P,

.01

20%to49%MPR

was

associated

with

increasedodds

ofahospitaladm

ission

than

0%to19%MPR,

OR=1.27

(95%

CI:1.04–

1.55),P,

.01

Higher

adherence

was

associated

with

higher

per-

mem

ber

per-month

paym

ents

734 McGRADY and HOMMEL

TABLE1

Continued

Author

(Year)

StudyDesign

SampleSize

and

StudyPopulation

Measure

ofAdherence

ControlledFor

RelationshipBetweenAdherenceandHealth

Care

Use

EDVisits

Outpatient

Visits

Hospitalizations

OtherMeasuresof

Health

Care

Use

MPR

forLI:the

percentage

ofdays

withina365-

dperiod

thatan

LIwas

supplied;Groupedinto:

0%to19%MPR;20%

to49%MPR,and

$50%MPR

a

Demographics,

health

characteristics

$50%MPR

was

associated

with

lower

odds

ofan

EDvisitthan0%

to19%MPR,OR=0.68

(95%

CI:0.53–0.86),

P=.002

Greatermedication

adherencewas

associated

with

morefrequent

asthma-related

office

visits,

P,

.01

NS,P

=.52

Higher

adherence

was

associated

with

higher

per-mem

ber

per-month

paym

ents

McNallyetal

(2009)

43Random

ized

controlled

trial

63children

(ages5–17

y)with

persistent

moderateor

severe

asthma

Percentofprescribed

oralmontelukast

dosesreceived

each

dayas

recorded

byan

electronicmonitoring

device;Grouped

into:

lowandhigh

adherencegroups

basedon

theupperand

lower

quartiles

ofadherencea

Demographicand

clinicalcovariates

exam

ined

butn

otincluded

inmodel

owingto

NS

XX

XLowadherencegroup

demonstratedan

increase

inhealth

careuse(num

berofED

visits,hospitalizations,

andclinicvisits

attributableto

asthma)

over

the

course

ofthe

study,P,

.05;High

adherencegroup

demonstratedno

significant

change

inhealth

care

use,P,

.05

Percentofprescribed

inhaledfluticasone

dosesreceived

each

dayas

recorded

byan

electronicmonitoring

device;Grouped

into:

lowandhigh

adherencegroups

basedon

theupperand

lower

quartiles

ofadherencea

Demographicand

clinicalcovariates

exam

ined

butn

otincluded

inmodel

owingto

NS

XX

XLowadherencegroup

demonstratedan

increase

inhealth

care

use(num

ber

ofED

visits,

hospitalizations,

andclinicvisits

attributableto

asthma)

over

the

course

ofthe

study,P,

.05;

High

adherence

groupdemonstrated

nosignificant

change

inhealth

care

use,

P,

.05

REVIEW ARTICLE

PEDIATRICS Volume 132, Number 4, October 2013 735

TABLE1

Continued

Author

(Year)

StudyDesign

SampleSize

and

StudyPopulation

Measure

ofAdherence

ControlledFor

RelationshipBetweenAdherenceandHealth

Care

Use

EDVisits

Outpatient

Visits

Hospitalizations

OtherMeasuresof

Health

Care

Use

Morrisetal

(1997)

42Cohort

89adolescentsandyoung

adults(age

,30

y)with

type

1diabetes

a

Numberofdays

(out

of365)

with

maximum

possibleinsulin

coverage

(as

determ

ined

bycomparing

the

medically

recommendedinsulin

dose

with

the

cumulativevolumeof

insulin

prescriptions

supplied);Grouped

intoquartiles

a

XX

Increasedadherence

was

associated

with

decreasedodds

ofadmission

fordiabetic

ketoacidosis(P,.001)

Increasedadherencewas

associated

with

decreasedodds

ofhospitaladm

issions

relatedto

complications

ofdiabetes

(P=.008)

Smith

etal

(2007)

35Cohort

1474

children

(ages2–17

y)with

persistent

asthmaa

Numberoffilled

prescriptions

foran

inhaled

corticosteroida

1–2filledprescriptions

comparedwith

0filledprescriptions

decreasedodds

ofan

EDvisitfor

asthma,Adj

OR=0.08

(95%

CI:

0.05–0.15),P,

.001

XX

Numberoffilled

prescriptions

for

bronchodilators

a

Age,gender,PCP

visits,asthm

aPCPvisits

NSinmultivariate

model

XX

Dichotom

ousindicator

ofwhether

aprescriptionfor

anycontroller

medicationwas

filled

withinthepast3moa

Thosewith

afilled

prescriptionwereless

likelytohavean

EDvisit

forasthma(P

,.001)

than

thosewithout

afilledprescription

Thosewith

afilled

prescriptionwere

morelikelytohave

anasthmaPCPvisit(P,

.001)than

those

withouta

filled

prescription

X

Waldersetal

(2004)

37Cross-

sectional

75children

(ages8–16

y)with

persistent

asthma

Totaldoses

takenperday

dividedby

prescribed

dosesperdaya

Increasedadherencewas

associated

with

decreasedED

visits

(r=20.25,P

,.05)

XX

Zhao

etal

(2012)

39Cross-

sectional

2960

children

(age

#14

y)with

asthmaa

Self-reported

months

(out

of12)patient

adheredtoprescribed

corticosteroiduse

Increasedadherencewas

associated

with

fewer

EDvisits,P

=.00

XIncreasedadherence

was

associated

with

fewer

hospitalizations,

P=.00

Self-reported

months

(out

of12)patient

adheredtoprescribed

leukotrienereceptor

modulator

use

NS,P

..05

XNS,P

..05

CI,confidenceinterval;HAC,hospitalasthm

aclinic;ICS,inhaled

corticosteroids;LI,leukotriene

inhibitors;NS,notsignificant;OR,odds

ratio,PCP,primarycare

physician;STL,serumtheophyllinelevels.

aRepresentativenessofexposedcohortor

assessmentofoutcom

ejudged

tobe

ofhigh

quality

perNewcastle-Ottaw

aQuality

AssessmentScale.

736 McGRADY and HOMMEL

to those seen in asthma clinic, self-reported non-adherence was higherin those seen in the ED.38 Similarly,caregiver report of more non-adherencewas associated with more ED visits forchildren who had asthma.39 In con-trast, a third study using a self-reportmeasure of adherence found no re-lationship between non-adherenceand ED visits when included in a multi-variate model.40

Medication Non-Adherence andOutpatient Visits

Both of the studies examining medica-tion non-adherence and outpatientvisits found that non-adherence wasassociated with fewer asthma-relatedoffice visits. In a sample of childrenwho had asthma, the fewer daysa child possessed inhaled cortico-steroids or leukotriene inhibitors, theless often asthma-related office visitsoccurred.36 Similarly, children who hadnot filled a prescription for a controllermedication in the past 3 months wereless likely than children who had filleda prescription to have an asthma pri-mary care physician visit.35

Medication Non-Adherence andHospitalizations

Findings of the 5 studies examiningmedication non-adherence and hospi-talizations were mixed and variedbased on method of adherence mea-surement. Four of these studies in-cluded children who had asthma and1 included adolescents and youngadults who had type 1 diabetes. Forchildren who had asthma, failing tofillanyprescriptionsforanti-inflammatorymedication was associated with anincreased risk for hospitalization.34

However, when MPRs for inhaled cor-ticosteroids were calculated andgrouped, those with 20% to 49% MPRhad increased odds of a hospital ad-mission as compared with those withan MPR of 0% to 19%.36 MPRs for

leukotriene inhibitors were not asso-ciated with hospitalizations.36

Caregiver self-reports of non-adherenceto corticosteroids were associatedwith more hospitalizations in a sampleof children who had asthma.39 In an-other sample, however, a compositescore comprised of asking caregiversyes/no questions related to their ad-herence to filling prescriptions, givingmore/less medication than prescribed,and obtaining recommended devicesyielded no relation to whether the childwas hospitalized in the past 9 months.41

In a sample of adolescents and youngadults who had type 1 diabetes, non-adherence was associated with in-creased odds for admission for diabeticketoacidosis and other diabetes-relatedcomplications.42

Medication Non-Adherence andOther Measures of Health Care Use

Similar to the aforementioned resultslinking non-adherence with more fre-quent ED visits, caregivers of childrenwith higher levels of self-reportednon-adherence also reported morefrequent unscheduled asthma visits(defined as provider and ED visits).41

Results of growth curve modelinganalyses suggest that medication non-adherence may be associated not onlywith rates of health care use, but alsochanges in health care use patterns.43

In a sample of African American chil-dren who had asthma, non-adherencewas associated with greater increasesin health care use over time than thosedemonstrated by groups of partic-ipants with better adherence.43

The only study to directly examinecosts related to non-adherence in-dicated that non-adherence wasassociated with lower per-member,per-month public insurance paymentsin a sample of children who hadasthma.36 As adherence to medicationrequires payment for prescriptions,Boylston Herndon et al found that

non-adherent children had lower drug-related costs than children with higheradherence levels. In this study, thecosts saved by not filling prescriptionswere more than the increased costs ofvisiting the ED or being hospitalizedincurred by these participants.

Risk for Bias

Risk for bias as identified by the itemsfrom the Newcastle-Ottawa QualityAssessment Scale and the inclusion ofconfounding variables is presented inTable 1. Three of the 10 studies (33%)included a cohort likely representa-tive of the relevant pediatric pop-ulation. Over three-quarters of studies(n = 8, 80%) used objective outcomemeasures.

DISCUSSION

Nine of the 10 studies reviewed (90%)demonstrated a relationship betweenmedication non-adherence and increa-ses in at least 1 health care use vari-able. Of note, the directionality of therelationship between medication non-adherence and health care use depen-ded on the outcome of interest. Themajority of studies examined reportedthat medication non-adherence wasassociated with more ED visits andhospitalizations.

The relationship between medicationnon-adherence and ED visits and hos-pitalizations is consistent with findingsin adult samples18,44,45 and suggeststhat there is potential to reduce the$1800 per child spent on hospital carein 2010.46 Children and adolescentswho were non-adherent suffered frommore exacerbations and complicationsrequiring medication attention andthus required more ED visits or hospi-talizations.

Although non-adherence was linkedwith more ED visits and hospital-izations, it was also related to feweroutpatient visits.35,36 These studies did

REVIEW ARTICLE

PEDIATRICS Volume 132, Number 4, October 2013 737

not differentiate between recom-mended outpatient visits (eg, regularlyscheduled planned clinic visits) andexcessive outpatient visits (eg, sickvisits that may have been preventedwith optimal adherence), and are likelyillustrating the relationship betweenmultiple aspects of treatment adher-ence. Specifically, patients who arenon-adherent to 1 aspect of theirmedical regimen (taking medication)are also more likely to be non-adherentto other parts of their medical regimen(attending regular clinic appoint-ments). Thus, this relationship mayrepresent non-adherence to recom-mended necessary treatment (bothmedications and clinic attendance) asopposed to less excessive health careuse.

In all, thefindings of this review suggestthatmedication non-adherence is likelyan important contributor to health careuse in pediatric chronic illness pop-ulations. Clinically, these results sug-gest that interventionefforts to improveadherencemayresult in reducedhealthcare use and ultimately reduced healthcarecosts.Currentlyavailablepediatricadherence interventions can effec-tively improve adherence and healthoutcomes.21,47 To our knowledge, nostudies have examined the cost-effectiveness of adherence inter-ventions with pediatric populations.Emerging adult literature, however,suggests that the reduced healthcare and indirect (eg, productivityloss, absenteeism) costs significantlyoutweigh the costs of implementing

behavioral adherence-promoting inter-ventions.48–51

Several limitations and resultingimplications for future research exist.First, the relatively small number ofstudies that met inclusion criteria andthe varied risk for bias across studiesindicate that this is a nascent area ofresearch that requires further in-vestigation. To broaden the literaturebase in this important field, future re-search examining adherence shouldinclude health care use and cost anal-yses and be conducted with additionalillness populations. Including addi-tional illness populations is of partic-ular importance, as it will allow for animproved understanding of how ad-herence relates to health care costsacross a variety of illnesses with vary-ing medical regimens and health careusepatterns. Second, the observationalstudy design of the reviewed manu-scripts negates the possibilities of de-termining the implicationsof improvingadherenceonhealthcareuseandcosts.Future research should examine thisquestion in the context of randomizedcontrolled trials using longitudinaldata. Third, as the method of assessingadherence varied significantly betweenstudies, it is not possible to directlycompare results. Given the significantimplications of this research, futureefforts should use evidence-basedmethods of adherence assessment.31

Fourth, currently available methodscan result in misleading findings, suchas cost-savings associated with lowerdrug-related costs attributable to

non-adherence. Researchers usingrelatively brief follow-up periods cap-ture only the short-term drug costsavings realized from non-adherenceand not the long-term, more costlyconsequences (eg, more frequenthospitalizations) of non-adherence.Thus, future research should addressthis methodologic issue by longitudinaldesigns that account for expectedcosts (eg, medication refills) versusunexpected costs (eg, ED visits) andfollow participants for periods of timethat allow for long-term consequencesof non-adherence to be realized. Fifth,the generalizability of our findings tonon-Western populations may be lim-ited by our exclusion of non-Englishmanuscripts. As the prevalence ofnon-adherence is a global concern,15

more inclusive reviews should be con-sidered.

Resources allocated to improving ad-herence are often fairly limited. In thecurrent economic and political cli-mate, however, increasing attentionhas been paid to novel approaches toimproving health outcomes while re-ducing spending. Results of this reviewsuggest that targeting non-adherencein children and adolescents who havea chronic illness may provide a uniqueopportunity to reduce health care usein a continually growing population.Consistent with the call for actionproposed by the World Health Organi-zation,15 multidisciplinary efforts topromote adherence should be a pri-ority for all medical providers andpolicy makers.

REFERENCES

1. Martin AB, Lassman D, Washington B,Catlin A; National Health ExpenditureAccounts Team. Growth in US healthspending remained slow in 2010; healthshare of gross domestic product wasunchanged from 2009. Health Aff (Millwood).2012;31(1):208–219

2. Schoenman J, Chockley N. UnderstandingU.S. Health Care Spending. Washington, DC:NIHCM Data Brief; 2011:1–13

3. Haber S, McCall N, Cromwell G. ChildrenWith High Health Care Utilization and Spe-cial Health Care Needs: Draft Manuscript.Washington, DC: RTI International; 2007

4. Emanuel EJ. Where are the health care costsavings? JAMA. 2012;307(1):39–40

5. Partnership for Solutions. Chronic Conditions:Making the Case for Ongoing Care. Baltimore,MD: The Johns Hopkins University; 2002

6. Van Cleave J, Gortmaker SL, Perrin JM.Dynamics of obesity and chronic health

738 McGRADY and HOMMEL

conditions among children and youth.JAMA. 2010;303(7):623–630

7. Wang G, Dietz WH. Economic burden ofobesity in youths aged 6 to 17 years: 1979-1999. Pediatrics. 2002;109(5). Available at:www.pediatrics.org/cgi/content/full/109/5/e81

8. Weiss KB, Sullivan SD, Lyttle CS. Trends inthe cost of illness for asthma in the UnitedStates, 1985-1994. J Allergy Clin Immunol.2000;106(3):493–499

9. Cystic Fibrosis Foundation. Cystic FibrosisFoundation Patient Registry 2010 AnnualData Report. Bethesda, MD: Cystic FibrosisFoundation; 2010

10. Smith JM, Martz K, Blydt-Hansen TD. Pedi-atric kidney transplant practice patternsand outcome benchmarks, 1987-2010: Areport of the North American PediatricRenal Trials and Collaborative Studies.Pediatr Transplant. 2013;17(2):149–157

11. Berry JG, Hall M, Hall DE, et al. Inpatientgrowth and resource use in 28 children’shospitals: a longitudinal, multi-institutionalstudy. JAMA Pediatr. 2013;167(2):170–177

12. Chevarley FM. Utilization and Expendituresfor Children With Special Health CareNeeds. Rockville, MD: Agency for HealthcareResearch and Quality; 2006

13. Newacheck PW, Kim SE. A national profile ofhealth care utilization and expenditures forchildren with special health care needs.Arch Pediatr Adolesc Med. 2005;159(1):10–17

14. Modi AC, Pai AL, Hommel KA, et al. Pediatricself-management: A framework for research,practice, and policy. Pediatrics. 2012;129(2).Available at: www.pediatrics.org/cgi/con-tent/full/129/2/e473

15. Sabate E. Adherence to Long-term Therapies:Evidence for Action. Geneva, Switzerland:World Health Organization; 2003

16. Bender BG, Rand C. Medication non-adherence and asthma treatment cost.Curr Opin Allergy Clin Immunol. 2004;4(3):191–195

17. Muenchberger H, Kendall E. Predictors ofpreventable hospitalization in chronic dis-ease: priorities for change. J Public HealthPolicy. 2010;31(2):150–163

18. Roebuck MC, Liberman JN, Gemmill-ToyamaM, Brennan TA. Medication adherence leads tolower health care use and costs despite in-creased drug spending. Health Aff (Millwood).2011;30(1):91–99

19. Berg JS, Dischler J, Wagner DJ, Raia JJ,Palmer-Shevlin N. Medication compliance:a healthcare problem. Ann Pharmacother.1993;27(9 suppl):S1–S24

20. DiMatteo MR. Variations in patients’ ad-herence to medical recommendations:

a quantitative review of 50 years of re-search. Med Care. 2004;42(3):200–209

21. Kahana S, Drotar D, Frazier T. Meta-analysisof psychological interventions to promoteadherence to treatment in pediatricchronic health conditions. J Pediatr Psy-chol. 2008;33(6):590–611

22. Newacheck PW, Taylor WR. Childhoodchronic illness: prevalence, severity, andimpact. Am J Public Health. 1992;82(3):364–371

23. Rapoff MA. Adherence to Pediatric MedicalRegimens. New York, NY: Springer; 2010

24. Logan D, Zelikovsky N, Labay L, Spergel J.The Illness Management Survey: identifyingadolescents’ perceptions of barriers toadherence. J Pediatr Psychol. 2003;28(6):383–392

25. Hommel KA, Davis CM, Baldassano RN. Ob-jective versus subjective assessment oforal medication adherence in pediatric in-flammatory bowel disease. Inflamm BowelDis. 2009;15(4):589–593

26. de Brantes F, Rosenthal MB, Painter M.Building a bridge from fragmentation toaccountability—the Prometheus Paymentmodel. N Engl J Med. 2009;361(11):1033–1036

27. Federal Interagency Forum on Child andFamily Statistics. America’s Children inBrief: Key National Indicators of Well-Being,2012. Washington, DC: U.S. GovernmentPrinting Office; 2012

28. Stille C, Turchi RM, Antonelli R, et al; Aca-demic Pediatric Association Task Force onFamily-Centered Medical Home. The family-centered medical home: specific consid-erations for child health research andpolicy. Acad Pediatr. 2010;10(4):211–217

29. Moher D, Liberati A, Tetzlaff J, Altman DG;PRISMA Group. Preferred reporting itemsfor systematic reviews and meta-analyses:the PRISMA statement. BMJ. 2009;339:b2535

30. Vandenbroucke JP, von Elm E, Altman DG,et al; STROBE Initiative. Strengthening thereporting of observational studies in epi-demiology (STROBE): explanation and elab-oration. PLoS Med. 2007;4(10):e297

31. Quittner AL, Modi AC, Lemanek KL, Ievers-Landis CE, Rapoff MA. Evidence-basedassessment of adherence to medicaltreatments in pediatric psychology. J PediatrPsychol. 2008;33(9):916–936, discussion 937–938

32. Wells G, Shea B, O’connell D, et al. TheNewcastle-Ottawa Scale (NOS) for assess-ing the quality of nonrandomised studies inmeta-analyses. Paper presented at 3rdSymposium on Systematic Reviews: Beyondthe Basics; 2000

33. Higgins JPT, Altman DG. Assessing Risk ofBias in Included Studies. Cochrane Hand-book for Systematic Reviews of Inter-ventions. West Sussex, England: John Wiley& Sons, Ltd; 2008:187–241

34. Adams RJ, Fuhlbrigge A, Finkelstein JA,et al. Impact of inhaled antiinflammatorytherapy on hospitalization and emergencydepartment visits for children with asthma.Pediatrics. 2001;107(4):706–711

35. Smith SR, Wakefield DB, Cloutier MM. Re-lationship between pediatric primary pro-vider visits and acute asthma ED visits.Pediatr Pulmonol. 2007;42(11):1041–1047

36. Herndon JB, Mattke S, Evans Cuellar A,Hong SY, Shenkman EA. Anti-inflammatorymedication adherence, healthcare utiliza-tion and expenditures among Medicaid andchildren’s health insurance program en-rollees with asthma. Pharmacoeconomics.2012;30(5):397–412

37. Walders N, Kopel SJ, Koinis-Mitchell D,McQuaid EL. Patterns of quick-relief andlong-term controller medication use inpediatric asthma. J Pediatr. 2005;146(2):177–182

38. Ashkenazi S, Amir J, Volovitz B, Varsano I.Why do asthmatic children need referral toan emergency room? Pediatr AllergyImmunol. 1993;4(2):93–96

39. Zhao J, He Q, Zhang G, et al. Status ofasthma control in children and the effect ofparents’ knowledge, attitude, and practice(KAP) in China: a multicenter study. AnnAllergy Asthma Immunol. 2012;109(3):190–194

40. Bartlett SJ, Krishnan JA, Riekert KA, ButzAM, Malveaux FJ, Rand CS. Maternal de-pressive symptoms and adherence totherapy in inner-city children with asthma.Pediatrics. 2004;113(2):229–237

41. Bauman LJ, Wright E, Leickly FE, et al. Re-lationship of adherence to pediatricasthma morbidity among inner-city chil-dren. Pediatrics. 2002;110(1 pt 1). Availableat: www.pediatrics.org/cgi/content/full/110/1/e6

42. Morris AD, Boyle DI, McMahon AD, GreeneSA, MacDonald TM, Newton RW. Adherenceto insulin treatment, glycaemic control, andketoacidosis in insulin-dependent diabetesmellitus. The DARTS/MEMO Collaboration.Diabetes Audit and Research in TaysideScotland. Medicines Monitoring Unit. Lan-cet. 1997;350(9090):1505–1510

43. McNally KA, Rohan J, Schluchter M, et al.Adherence to combined montelukast andfluticasone treatment in economically dis-advantaged African American youth withasthma. J Asthma. 2009;46(9):921–927

REVIEW ARTICLE

PEDIATRICS Volume 132, Number 4, October 2013 739

44. Encinosa WE, Bernard D, Dor A. Does pre-scription drug adherence reduce hospital-izations and costs? The case of diabetes.Adv Health Econ Health Serv Res. 2010;22:151–173

45. Valenti WM. Treatment adherence improvesoutcomes and manages costs. AIDS Read.2001;11(2):77–80

46. Agency for Healthcare Research and Qual-ity. Total health services—mean and me-dian expenses per person with expenseand distribution of expenses by source ofpayment: United States, 2010. Medical Ex-penditure Panel Survey Household Component

Data. Available at: http://meps.ahrq.gov/mepsweb/. Accessed July 8, 2013

47. Graves MM, Roberts MC, Rapoff M, Boyer A.The efficacy of adherence interventions forchronically ill children: a meta-analytic re-view. J Pediatr Psychol. 2010;35(4):368–382

48. Chapman RH, Ferrufino CP, Kowal SL, ClassiP, Roberts CS. The cost and effectiveness ofadherence-improving interventions for an-tihypertensive and lipid-lowering drugs. IntJ Clin Pract. 2010;64(2):169–181

49. Chapman RH, Kowal SL, Cherry SB, Ferru-fino CP, Roberts CS, Chen L. The modeledlifetime cost-effectiveness of published

adherence-improving interventions for an-tihypertensive and lipid-lowering medi-cations. Value Health. 2010;13(6):685–694

50. Desborough JA, Sach T, Bhattacharya D,Holland RC, Wright DJ. A cost-consequencesanalysis of an adherence focused pharmacist-led medication review service. Int J PharmPract. 2012;20(1):41–49

51. Patrick AR, Schousboe JT, Losina E, SolomonDH. The economics of improving medica-tion adherence in osteoporosis: validationand application of a simulation model.J Clin Endocrinol Metab. 2011;96(9):2762–2770

WHAT’S IN A NUMBER: While I have not thought too much about my three sons’ability to father children, researchers are worried about the fertility of thecurrent generation of men between the ages of 18 and 25 years. As reported inThe Wall Street Journal (In the Lab: July 15, 2013), many in the reproductivescience world believe that we are entering a “sperm crisis” due to falling spermcounts. In general, men have approximately 60 million sperm in each milliliter ofsemen. As long as a man has approximately 40 million in each milliliter, he isconsidered fertile. As the number falls beneath this threshold (and particularlybelow 20 million), either the ability to conceive begins to drop or the time toconception increases. Several European studies have demonstrated that menhave much lower sperm counts now than previously. A study of nearly 26,000healthy French men followed at fertility clinics because of partner infertilitydemonstrated a 30% decline in sperm count from 1989 to 2005. Another largestudy showed that sperm counts had declined by almost 50% in the past 50 years.While on average French men still have healthy sperm counts (49 million permilliliter of semen), approximately 1 in 5 northern European men have spermcounts low enough to affect fertility. Why sperm counts have declined is notknown. Possible explanations include in utero such as maternal smoking, earlylife exposures to phthalates, and sedentary jobs, frequent hot baths, fatty foods,and marijuana use. Even central obesity has been implicated. Scientists areworried about the implications beyond fertility as sperm count may be an easyand sensitive measurement of other effects lifestyle may have on other bodyfunctions or processes. Others are not so convinced that a “crisis” is at hand.While the study of French men only included presumably healthy men, manystudies have been in men presenting to clinics because of infertility and thus maynot be representative of the general population. I am not sure how to counsel mysons on this topic. Other than to wait until they are sure they are absolutely readyto start a family, I generally simply recommend a healthy lifestyle and appro-priate protection should fertility be something they find themselves needing toconsider.

Noted by WVR, MD

740 McGRADY and HOMMEL