chttcrtcpimtgotpro

mga

0d

Electrocardiographic Screening in Children With Attention-DeficitHyperactivity Disorder

William T. Mahle, MDa,b,*, Camden Hebson, MDa,b, and Margaret J. Strieper, DOa,b

Some investigators have suggested that children receiving stimulant medications to man-age attention-deficit hyperactivity disorder should undergo screening electrocardiographyto identify asymptomatic cardiac disease. However, no study to date has examined theefficacy and costs of this strategy. In the present study we sought to determine the utilityof electrocardiographic screening in children with attention-deficit hyperactivity disorder.We reviewed the clinical experience of electrocardiographic screening of subjects withattention-deficit hyperactivity disorder <21 years of age from April to September 2008.Additional cardiac care and testing that resulted from an abnormal initial electrocardio-gram were recorded. Screening electrocardiograms were obtained in 1,470 children withattention-deficit hyperactivity disorder and were interpreted as abnormal in 119 subjects(8.1%). Further evaluation of these 119 subjects included 63 transthoracic echocardio-grams, 5 stress tests, and 9 Holter monitor studies. Cardiac disease was identified in 5subjects (0.3% of entire cohort), yielding a positive predictive value of 4.2%. Cardiacdiagnoses included ventricular pre-excitation syndrome (n � 2), bicuspid aortic valve (n �2), and moderate secundum atrial septal defect (n � 1). The mean cost of electrocardio-graphic screening including further testing for subjects with abnormal initial screen resultswas $58 per child. The mean cost to identify a true-positive result was $17,162. Inconclusion, electrocardiographic screening for children with attention-deficit hyperac-tivity disorder can successfully identify cardiac disease in otherwise asymptomaticsubjects, although the positive predictive value is low. Ongoing studies are needed toknow what role electrocardiographic screening should play in the management ofchildren with attention-deficit hyperactivity disorder. © 2009 Elsevier Inc. All rights

reserved. (Am J Cardiol 2009;104:1296 –1299)

M

Caesiessermspr

nlcwTdc

iF

A recent American Heart Association (AHA) statementoncluded that screening with electrocardiogram may beelpful to assess for cardiac disease in children with atten-ion-deficit disorder who are treated with stimulant medica-ion.1 This statement suggested that, “It is reasonable toonsider adding an ECG [electrocardiogram], which is ofeasonable cost, to the history and physical examination inhe cardiovascular evaluation of children who need to re-eive treatment with drugs for ADHD [attention-deficit hy-eractivity disorder].” Although these recommendations arentriguing, to date it is unknown whether such a strategyight be successful in detecting occult cardiac disease. In

he present study we analyzed the efficacy of electrocardio-raphic screening in an attention-deficit hyperactivity dis-rder population that resulted from changes in clinical prac-ice because of the AHA recommendations. In particular weaid attention to the impact of false-positive studies thatesult because of this strategy and report the economic costsf such an approach.

aSibley Heart Center, Children’s Healthcare of Atlanta, and bDepart-ent of Pediatrics, Emory University School of Medicine, Atlanta, Geor-

ia. Manuscript received April 29, 2009; revised manuscript received andccepted June 19, 2009.

*Corresponding author: Tel: 404-315-2672; fax: 404-325-6021.

“E-mail address: mahlew@kidsheart.com (W.T. Mahle).

002-9149/09/$ – see front matter © 2009 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2009.06.052

ethods

With the approval of the institutional review board ofhildren’s Healthcare of Atlanta (Atlanta, Georgia), wenalyzed data from a cohort of children who underwentlectrocardiography performed for the purpose of cardiaccreening for stimulant medication. The study populationncluded children 5 to 21 years of age who obtained anlectrocardiogram from April 21 to September 21, 2008. Atart date of April 21 was chosen because the AHA consen-us guideline was published at that time. We assumed thatlectrocardiograms ordered after that date would represent aesponse to the AHA guidelines. Although the AHA state-ent did not explicitly recommend electrocardiographic

creening, many clinicians (psychiatrists and primary careroviders) began to refer children receiving or likely toeceive stimulants for this test.

Electrocardiograms from 3 major sources were included,amely outpatient cardiology practices, a hospital-basedaboratory performing outpatient electrocardiography, andommunity hospitals that perform the studies and then for-ard them to the Sibley Heart Center (Atlanta, Georgia).hose subjects who were identified as having attention-eficit hyperactivity disorder screening or stimulant medi-ation screening were included in the analysis.

Those subjects who were identified as having an abnormalnitial electrocardiogram were defined as positive screens.indings that were considered variants of normal such as

sinus arrhythmia” or “early repolarization” were consid-

www.AJConline.org

efivtd

ttefinw

sdSemPbeTm

sTwtcg

R

efmubs5

(dmnhcitcctt

jchm1pt

oicphimste3vIiuw

tiTnfTiwp

d

Fe

TM

A

LRPRLRLPPV

1297Miscellaneous/Electrocardiographic Screening for Stimulant Medications

red negative screens. Based on the AHA statement, thesendings would be classified as category 1 (normal or normalariants). The outcome for those with positive initial elec-rocardiograms (AHA category II and III readings) wasetermined.

For those subjects with a positive initial screening elec-rocardiogram who presented for further cardiology consul-ation, records were reviewed to determine results of repeatlectrocardiogram if performed and physical examinationndings. Ancillary testing that was performed in the diag-ostic work-up was also recorded. Final cardiac diagnosisas reported.To analyze the added cost of an electrocardiographic

creening program for attention-deficit hyperactivity disor-er, we included the cost of the initial electrocardiogram.ubsequent evaluation such as cardiology consultation,chocardiography, and stress testing were included in theodel. Costs for all tests were determined based on Currentrocedural Terminology codes and 2008 Medicare reim-ursement rates.2 The cost model assumed that the firstlectrocardiogram did not result in any lost parental wages.he model did assume that follow-up cardiac consultationight result in lost wages.3

The proportion of positive initial electrocardiographiccreens was reported, as were 95% confidence intervals.he positive predictive value of electrocardiographic screenas reported. Because the number of true-positive results in

he population could not be determined, we were unable toalculate the sensitivity and specificity of electrocardio-

Entire CohortN=1470

1st ECG NormalN=1351

1st ECG AbnormalN=119

2nd ECG PerformedN=97

2nd ECG Not PerformedN=22

2nd ECG AbnormalN=61

2nd ECG NormalN=36

igure 1. Results of electrocardiographic testing in study cohort. ECG �lectrocardiogram.

able 1ost common abnormalities on initial electrocardiogram

bnormality Subjects (%)

eft ventricular hypertrophy 34 (28.2%)ight ventricular hypertrophy 24 (20.5%)rolonged QT interval 22 (18.5%)ight bundle branch block 17 (14.2%)eft-axis deviation 7 (5.9%)ight-axis deviation 7 (5.9%)eft atrial enlargement 3 (2.5%)remature ventricular complexes 2 (1.7%)remature atrial complexes 2 (1.7%)entricular pre-excitation 2 (1.7%)

raphic screening. c

esults

During the 5-month study period 1,470 children hadlectrocardiography performed for purposes of screeningor attention-deficit hyperactivity disorder medications. Theean age of this population was 9.8 � 2.9 years. Those

ndergoing electrocardiographic screening were primarilyoys (1,237 of 1,470, 84%). The primary payers for childrencreened with electrocardiogram were private payers in5%, governmental in 44%, and self pay in 1%.

Of the screened cohort of 1,470, there were 119 subjects8.1%) who were found to have an “abnormal” electrocar-iogram. Electrocardiographic abnormalities were foundore commonly in children undergoing testing at commu-

ity hospitals than in pediatric cardiology offices or childrenospitals (11.2% vs 7.5%, respectively, p � 0.04). The mostommon abnormalities are listed in Table 1. The mediannterval from first electrocardiogram to cardiology consul-ation was 11 days (range 1 to 49). Most subjects referred toardiology evaluation (78 of 119, 65%) were already re-eiving stimulant medications at the time of screening elec-rocardiogram; the remaining subjects were referred in an-icipation of starting stimulant medications.

Of the 119 with abnormal electrocardiograms, 97 sub-ects had a second electrocardiogram obtained at the time ofardiology consultation. More than 35% of subjects whoad repeat electrocardiography were thought to have a nor-al electrocardiogram when the study was repeated (Figure

). The most common original diagnoses that were reinter-reted as normal were prolonged QT interval and left ven-ricular hypertrophy, respectively.

Physical examination performed by a cardiologist dem-nstrated abnormalities in 14 subjects. These abnormalitiesncluded systolic ejection murmur (n � 7), systolic ejectionlick (n � 3), irregular heart rhythm (n � 5), and prominentrecordial impulse (n � 1). Seven subjects were thought toave an innocent heart murmur. Additional ancillary testingncluded complete echocardiography in 63 subjects, Holteronitoring in 9 subjects, and exercise stress testing in 5

ubjects. Indications for echocardiograms were early sys-olic ejection click in 2, heart murmur in 5, and abnormallectrocardiograms with a variety of findings (n � 56), the

most common being left ventricular hypertrophy, rightentricular hypertrophy, and right bundle branch block.ndications for Holter monitors were prolonged QT intervaln 5, ventricular pre-excitation in 2, and premature ventric-lar complexes in 2. Indication for exercise stress testingas prolonged QT interval in all 5 patients.Final diagnoses included no cardiac disease in 114, ven-

ricular pre-excitation syndrome in 2, bicuspid aortic valven 2, and moderate secundum atrial septal defect in 1.herapeutic interventions undertaken because of new diag-oses included transcatheter closure of an atrial septal de-ect (n � 1) and ablation of an accessory pathway (n � 2).herefore, the proportion of children with cardiac disease

dentified with an electrocardiographic screening strategyas 0.003 (95% confidence interval 0.001 to 0.008). Theositive predictive value was 4.2%.

Total cost for testing in the entire cohort of 1,470 chil-ren was $85,868. Therefore, the cost to screen a single

hild is $58. The calculated cost to identify a child with

sanacl

D

rghaectcns

dPbf[tertrpcimcfi

fcaarpsPsicd

dhtcsdaIcn

disn

ddAshsbasmd

sptcWtannw

pbstrccatscaeSmcs

ttOamfcmuwtpt

1298 The American Journal of Cardiology (www.AJConline.org)

ome form of heart disease is $17,162. If one were tossume that the additive risk of stimulant medication isegligible in mild congenital heart lesions, such as bicuspidortic valve and atrial septal defect, the cost to identify ahild with complex congenital heart disease or potentialife-threatening arrhythmias is $42,904.

iscussion

This study in a small, contemporary cohort of childreneceiving stimulant medications demonstrates electrocardio-raphic screening can identify otherwise asymptomaticeart disease. Electrocardiographic screening can promptdditional testing and ultimately diagnosis of cardiac dis-ase that might be life-threatening and that may represent aontraindication to the administration of stimulant medica-ions. Conversely, the positive predictive value of electro-ardiographic screening is low, meaning that a significantumber of children without heart disease will likely beubjected to additional testing.

Recent recommendations from the AHA have engen-ered considerable controversy. The American Academy ofediatrics writing group stated that, “the harm outweighs theenefit of recommending routine ECGs [electrocardiograms]or healthy children starting stimulant medication for ADHDattention-deficit hyperactivity disorder].”4 These investiga-ors raised concerns about the sensitivity and specificity of thelectrocardiogram as a general screening test. Much of theesistance to routine electrocardiographic screening relates tohe concern that this strategy will produce many false-positiveesults. Studies of electrocardiographic screening in otheropulations have generally reported that 5% to 10% ofhildren will have an abnormal electrocardiogram onnitial testing.5,6 Findings from our analysis are veryuch in agreement with these previous publications be-

ause we reported that 8.1% had an abnormal screen onrst examination.

Many quality initiatives could decrease the number ofalse-positive results. For example, we found that electro-ardiograms obtained in community hospitals generally hadhigher rate of abnormalities than those obtained in pedi-

tric cardiology clinics and children’s hospitals. This mayesult from improper lead placement in centers that care forrimarily for adults.7 Also, there may be significant incon-istencies among interpretations by pediatric cardiologists.8

revious investigators have suggested that there is con-iderable variability in the interpretation of QT intervaln pediatric cardiologists and that reinterpretation by spe-ialists at large referral centers can demonstrate �30%iscordance.9

A key assumption of any screening strategy is that earlyetection of the disorder might improve outcome. To date,owever, the link between stimulant medications used toreat attention-deficit disorder and an increased cardiovas-ular risk is tenuous. The original Health Canada warninguggested a possible association between congenital heartefects—even mild defects such as bicuspid aortic valve—nd sudden death while receiving Adderall XR (Shire US,nc., Florence, Kentucky).10 Therefore, the discovery ofongenital heart disease with minimal or mild hemody-

amic significance may be a valuable finding of electrocar- b

iographic screening. However, it should be recognized thatn the 2 subjects with bicuspid aortic valve identified in thiseries, the minor electrocardiographic abnormalities wereot thought to be related to valvar disease.

The balance of risks and benefits of stimulants in chil-ren with heart disease and attention-deficit hyperactivityisorder continues to be evaluated by the Food and Drugdministration and other organizations. However, use of

timulant medications in children with known structuraleart disease is controversial. Even with more complextructural heart lesions, practitioners may believe that theenefits of medical management of attention-deficit hyper-ctivity disorder may outweigh any cardiovascular risks. Asuch, the advantage of identification of these predominantlyild forms of congenital heart disease through electrocar-

iographic screening remains speculative.In contrast, many cardiologists would likely agree that

timulant medications may pose a risk in children withotentially life-threatening arrhythmias such as supraven-ricular tachycardia, long QT syndrome, and hypertrophicardiomyopathy.11,12 Importantly, with diseases such as

olff-Parkinson-White syndrome (or ventricular pre-exci-ation), ablation procedures can effectively treat the disordernd decrease any potential risk. For other entities that can-ot be “cured,” such as long QT syndrome, further diag-ostic testing and risk stratification might allow clinicians toeigh the risks and benefits of stimulant medications.With any screening strategy cost-effectiveness is an im-

ortant issue. To calculate cost-effectiveness one needs toe able to quantify the economic benefits of early detection,uch quantifiable decrease in morbidity or mortality. Givenhe very limited—and controversial—data regarding theisk of sudden death related to the use of stimulant medi-ations, one cannot say with certainty whether an electro-ardiographic screening program has the potential of savingny lives. The cost of a single electrocardiogram is rela-ively low—approximately $22. However, one must con-ider the additional costs incurred from added testing. Theost per subject increased approximately threefold due to thedded diagnostic work-up that is required when the initiallectrocardiogram is abnormal. Assuming that in the Unitedtates approximately 2.5 million children receive stimulantedication for attention-deficit hyperactivity disorder, the total

ost of a screening electrocardiographic program for a singlecreen per child would be $146 million.13

In addition to financial costs associated with false-posi-ive electrocardiograms, concerns have been raised abouthe adverse psychological impact of screening programs.ne of the factors in weighing the adverse consequences offalse-positive test result is the length of time to takes toake a final diagnosis. In our series, the median interval

rom initial electrocardiogram to evaluation by a pediatricardiologist was relatively short—�2 weeks. However, oneight imagine that wait times could be potentially longer in

nderserved areas. Although most of the time a single visitith a pediatric cardiologist allowed a definitive diagnosis,

here were 13 subjects who required additional testing, thusrolonging the diagnostic process and potentially adding tohe anxiety of a false-positive screen.

There are several limitations to the present study. First,

ecause this is not a population-based study, it is unknown

wigrppva

1

1

1

1

1299Miscellaneous/Electrocardiographic Screening for Stimulant Medications

hat proportion of children receiving stimulant medicationsn metropolitan Atlanta were referred for electrocardio-raphic screening. Primary physicians may have selectivelyeferred for electrocardiographic testing. Second, the sam-le is relatively small. The 95% confidence interval for theositive detection rates varies from 0.001 to 0.008. Suchariability would dramatically alter the utility of screeningnd our cost estimates.

1. Vetter VL, Elia J, Erickson C, Berger S, Blum N, Uzark K, Webb CL.Cardiovascular monitoring of children and adolescents with heartdisease receiving stimulant drugs: a scientific statement from theAmerican Heart Association Council on Cardiovascular Disease in theYoung Congenital Cardiac Defects Committee and the Council onCardiovascular Nursing. Circulation 2008;117:2407–2423.

2. 2008 Medicare physician fee schedule. Available at: www.cms.hhs.gov/center/physician.asp. Accessed October 10, 2008.

3. U.S. Department of Labor, Bureau of Labor Statistics. Employmentand earnings, 2007 annual averages and the monthly labor review.Available at: www.dol.gov/wb/stats/main.htm. Accessed November 1,2007.

4. Perrin JM, Friedman RA, Knilans TK. Cardiovascular monitoring andstimulant drugs for attention-deficit/hyperactivity disorder. Pediatrics2008;122:451–453.

5. Pelliccia A, Culasso F, Di Paolo FM, Accettura D, Cantore R, Cast-agna W, Ciacciarelli A, Costini G, Cuffari B, Drago E, Federici V,Gribaudo CG, Iacovelli G, Landolfi L, Menichetti G, Atzeni UO,Parisi A, Pizzi AR, Rosa M, Santelli F, Santilio F, Vagnini A, CasascoM, Di LL. Prevalence of abnormal electrocardiograms in a large,unselected population undergoing pre-participation cardiovascular

screening. Eur Heart J 2007;28:2006–2010.

6. Haneda N, Mori C, Nishio T, Saito M, Kajino Y, Watanabe K, KijimaY, Yamada K. Heart diseases discovered by mass screening in theschools of Shimane Prefecture over a period of 5 years. Jpn Circ J1986;50:1325–1329.

7. Garson AJ, Bricker J, Fisher D, Neish S, eds. Electrocardiography inthe Science and Practice of Pediatric Cardiology. Baltimore: Williamsand Wilkins, 1998.

8. Chiu CC, Hamilton RM, Gow RM, Kirsh JA, McCrindle BW. Eval-uation of computerized interpretation of the pediatric electrocardio-gram. J Electrocardiol 2007;40:139–143.

9. Taggart NW, Haglund CM, Tester DJ, Ackerman MJ. Diagnosticmiscues in congenital long-QT syndrome. Circulation 2007;115:2613–2620.

0. Health Canada suspends the market authorization of ADDERALLXR®, a drug prescribed for attention deficit hyperactivity disorder(ADHD) in children. Available at: http://www.hc-sc.gc.ca/ahc-asc/media/advisories-avis/_2005/2005_01-eng.php. Accessed February 9,2005.

1. Gracious BL. Atrioventricular nodal re-entrant tachycardia associatedwith stimulant treatment. J Child Adolesc Psychopharmacol 1999;9:125–128.

2. Lehnart SE, Ackerman MJ, Benson DW Jr, Brugada R, Clancy CE,Donahue JK, George AL Jr, Grant AO, Groft SC, January CT, LathropDA, Lederer WJ, Makielski JC, Mohler PJ, Moss A, Nerbonne JM,Olson TM, Przywara DA, Towbin JA, Wang LH, Marks AR. Inheritedarrhythmias: a National Heart, Lung, and Blood Institute and Office ofRare Diseases workshop consensus report about the diagnosis, pheno-typing, molecular mechanisms, and therapeutic approaches for primarycardiomyopathies of gene mutations affecting ion channel function.Circulation 2007;116:2325–2345.

3. Mayes R, Bagwell C, Erkulwater J. ADHD and the rise in stimulant

use among children. Harv Rev Psychiatry 2008;16:151–166.