ORIGINAL INVESTIGATIONS

Briefly Trained Medical Students Can EffectivelyIdentify Rheumatic Mitral Valve Injury Using aHand-Carried Ultrasound

Hezzy Shmueli, M.D.,* Yuval Burstein, M.D.,* Iftach Sagy, M.D.,* Zvi H. Perry, M.D., M.A.,* Ruben Ilia, M.D.,†Yaakov Henkin, M.D.,† Tali Shafat, B.Sc.,‡ Noah Liel-Cohen, M.D.,† and Sergio L. Kobal, M.D.†

*Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer Sheva, Israel; †Cardiology Department,Soroka University Medical Center, Beer Sheva, Israel; and ‡Clinical Research Center, Soroka UniversityMedical Center, Beer Sheva, Israel

Background: Rheumatic heart disease (RHD) is common and remains a major cause of morbidity,particularly in developing countries. Its diagnosis relies on expertise-dependent echocardiographicstudies. We evaluated the accuracy of briefly trained examiners in identifying RHD utilizing a hand-carried cardiac ultrasound (HCU) device. Methods: Three medical students received 8 hours of trainingin cardiac ultrasound, focused on assessment of rheumatic valve injury and its complications, using aprototype of HCU device, OptiGo. The students, blinded to the patients’ medical condition, performedan auscultation-based physical examination and a focused HCU study on volunteers and patients withknown RHD. A standard echocardiography study was used to validate the results. Results: Each studentperformed a physical examination followed by an HCU study on 45 subjects (mean age 57 � 14 years,52% men), 14 of whom (31%) had rheumatic mitral valve injury. The students’ averaged sensitivity fordiagnosing RHD by HCU examination was 81%, while specificity was 95%. The interrater agreement(kappa) of the 3 students’ HCU study and the standard echocardiography examination were between0.55 and 0.88 (P < 0.01), and among the students themselves between 0.57 and 0.74 (P < 0.01), asstudents 1 and 2 had better results than student 3. Auscultation-based physical examination renderedlow sensitivity (16%) for diagnosing rheumatic valve complications, namely mitral regurgitation andstenosis; however, it improved by 26% when students based their diagnosis on an HCU study. Conclu-sions: The ability to detect rheumatic valve injury using a portable ultrasound device by operators whoonly received brief echocardiographic training is remarkably high. However, the diagnosis of RHDcomplications is only modest. This result highlights the utility of portable cardiac ultrasound devicesoperated by basically trained personnel as a valuable diagnostic tool for RHD. (Echocardiography2013;30:621-626)

Key words: echocardiography, hand-carried cardiac ultrasound, physical examination, rheumatic heartdisease, screening

Rheumatic heart disease (RHD) is a commondisease that remains a major cause of morbidityand premature death, particularly in developingcountries.1–7 Physical examination alone is notsufficient to efficiently identify patients with RHD.Previous studies have demonstrated the low sen-sitivity of auscultation-based physical examina-tion as a screening tool in diagnosing RHDamong school-aged children.8,9 The most sensi-tive method for diagnosing RHD and its compli-cations is echocardiography, which isnoninvasive, free of radiation, and has no side

effects. However, traditional echocardiographydevices are expensive, cumbersome, and requireskilled personnel for the acquisition of the imagesand interpretation of the results. For the lastdecade small, mobile, user-friendly, and relativelyinexpensive echocardiography devices known ashand-held cardiac ultrasound (HCU), and morerecently the latest generation of pocket devicesare available for clinical use.10–12 Previous studieshave shown that HCU devices are accurate inassessing certain cardiac parameters, includingventricular function,13–15 ventricular hypertro-phy,16–18 valvular disease,19–21 pericardial effu-sion,22 left atrial size,23 patient’s central venouspressure,24 and fluid responsiveness in ventilatedpatients25,26 in the hands of both expert andnonexpert operators.27,28 The HCU, which is

Address for correspondence and reprints request: Sergio L.Kobal, M.D., Soroka University Medical Center, Beer-Sheva,Israel. Fax:+972-8-6400896;E-mail: kobal@bgu.ac.il

621

© 2013, Wiley Periodicals, Inc.DOI: 10.1111/echo.12122 Echocardiography

battery-operated, is ideal for RHD screening,especially in rural areas and developing countries,lacking infrastructures and experts.29–31

We hypothesized that medical students couldbe trained in a short period of time to performfocused echocardiography studies with an HCUdevice to identify features of rheumatic valveinjury. We also postulated that their diagnosis ofvalvular dysfunction, namely regurgitation and/or stenosis resulting from the rheumatic lesionwould be more accurate based on HCU resultsthan on auscultation-based physical examinationalone.

Methods and Materials:Three final-year medical students who weretrained to operate a mobile echocardiographydevice, OptiGo (Philips Medical Systems, Ando-ver, MA, USA), with the purpose of diagnosingRHD, participated in the study. The ultrasounddevice, a first generation of portable cardiacultrasound, weighs 2.9 kg and is equipped witha 2.5 MHz phased-array transducer, providestwo-dimensional imaging and conventional colorflow Doppler. It is equipped with controls tofreeze and scroll the image and calipers for linearmeasurements.

Medical Students Training:The students received 8 hours of training thatwere divided into 2 parts: the first 3 hours con-sisted of frontal lecture with an hour on echocar-diographic anatomy, examination views, anddiagnosing rheumatic valve injury and its compli-cations, and an additional 2 hours of observingnormal and pathological echocardiographicexaminations from the echocardiographic lablibrary. During the second part of the training, thestudents carried out 5 hours of practice with asenior sonographer, performing examinationswith HCU on random patients and healthy volun-teers. The students learned to identify rheumatic

valve injury of the mitral valve in a two-dimensional echocardiography examination bydemonstrating restricted movement of at least 1valve leaflet and/or thickening of the valve leaflets,especially at their edges, which causes a typicaldome-like movement during the opening of thevalve (Fig. 1). They also used the color Dopplertechnique for assessing blood flow stream and tur-bulence, which allows identification of rheumaticinjury complications, such as mitral stenosis (MS)and mitral regurgitation (MR). Stenosis was diag-nosed by apparent rapid diastolic turbulent flowcausing aliasing through the open valve (Fig. 2),while regurgitation was diagnosed according to asystolic backflow jet inside the closed valve’spreliminary cardiac chamber (Fig. 3). Finally, forthis study, the medical students also received a2-hour booster session of cardiac physical exami-nation by using an audiovisual web site on cardiacauscultation.

Diagnosis of RHD and its Complications:Each student examined 45 people, divided into 2groups of examinees: known RHD patients and acontrol group with normal echocardiograms,according to previous echocardiography studiesthat had been conducted for different reasons.Each student was double-blinded to the exam-inee’s medical situation, and a 2-step examina-tion consisting of a physical examination thatincluded cardiac auscultation was performed, toidentify possible complications of RHD, such asMS and/or MR, and a focused HCU study wascarried out to identify rheumatic mitral valveinjury and its possible complications. Both ofthese examinations were done in quiet rooms.Each student filled in a standard form about hisfindings on each examination. Students’ resultswere compared with and validated accordingto a standard echocardiography examinationperformed by a qualified sonographer andinterpreted by an echocardiographer.

Figure 1. Two-dimensional parasternal long-axis (on the left) and four-chamber view (on the right) showing rheumatic injury ofthe mitral valve as demonstrated by limitation in motion of the edge of the valves with the characteristic doming.

622

Shmueli, et al.

Statistical Analysis:Sample size calculation: Previous studies haveshown that the sensitivity of the physical exam is10 times lower than echocardiography regardingthe diagnosis of RHD.32,33

Under these assumptions, the rate of patientsaffected by RHD was considered 3%, with anexpected error of 5% and a = 5%. This led us toan estimated sample size of 45 patients.

The results are presented as the mean (SD) forcontinuous variables and as the total patients(percentage of total patients) for categoricaldata. T-test was used for comparison of the con-tinuous variables or chi-square test for categoricaldata with the use of Fisher’s exact test if needed.The kappa score was used to measure agreementbetween student’s examination and standardechocardiography examination, and among the3 students (j > 0.8 was considered good agree-ment). A 2-tailed P-value of �0.05 was consid-ered significant. The statistical analysis was doneusing SPSS version 17 (IBM, Armonk, NY, USA).

The Helsinki Committee of our institutionapproved the study protocol (number 4817),and the patients gave informed consent beforetheir participation in the study.

Results:Forty-five subjects were examined 3 times, in thesame session, by the 3 medical students (a totalof 135 examinations). There were 14 (31%)patients with rheumatic heart injury and 31nonrheumatic subjects as a control group. Thecharacteristics of these 2 subgroups were similaras shown in Table I.

In the rheumatic patients’ group, the rheu-matic injury was limited to the mitral valve.Among these 14 patients, 8 had both MS andMR, 5 had only MS, and 1 patient had neither.Of these overall 21 cases of MS and MR, 13 ofthem were of at least moderate severity. Theaverage examination time for a focused HCUstudy performed by the students was4.5 � 1.5 minutes. Image quality of the HCUstudies was not assessed but was satisfactory inall the subjects recruited in the study allowingthe students to assess mitral valve anatomy in allof the scanned patients.

Diagnosing Structural Features of RheumaticMitral Injury:Among the patients with RHD, an average of81% were successfully diagnosed as rheumaticmitral injury patients by the 3 students based onHCU results, and an average of 95% weresuccessfully diagnosed as nonrheumatic patientsamong those in the control group. Both sensitiv-ity and specificity of the 3 students’ HCU studyare described in Table II. Students’ diagnosticaccuracy was inconsistent, and one of them wassignificantly less successful than the others.

Figure 2. A high-velocities, turbulent jet seen in the left ven-tricle during the diastolic filling period as consequence of theblood crossing a stenotic valve, on a four-chamber view.

Figure 3. A regurgitant jet seen by color Doppler in the leftatrium during the ventricular systolic period due to incompe-tence of the valve, on a parasternal long-axis view.

TABLE I

General Characteristics of RHD Patients and a Control Group

CharacteristicsRHD Group(N = 14)

Control Group(N = 31)

P-Value

Male gender (%) 5 (38) 18 (58) 0.17Age, years(mean � SD)

59.5 � 11.8 55.8 � 15.2 0.43

Height, m(mean � SD)

1.67 � 0.06 1.68 � 0.07 0.66

Weight, kg(mean � SD)

77.3 � 11.8 75.5 � 10.8 0.61

Body mass index(mean � SD)

27.6 � 4.3 26.7 � 3.5 0.42

Heart rate, bpm(mean � SD)

71 � 11 68 � 9 0.24

RHD = rheumatic heart disease.

623

After Brief Training, Students Can Identify Rheumatic MS

No significant differences between misdiag-nosed and successful examinations were found inany parameters, including gender, body massindex (BMI), heart rhythm and rate.

The interrater agreement (Kappa) of the 3 stu-dents’ HCU study and the standard echocardiog-raphy examination were between 0.55 and 0.88(P < 0.01), and among the students themselvesbetween 0.57 and 0.74 (P < 0.01), as students 1and 2 had better results than student 3.

Diagnosing Functional Complications ofRheumatic Mitral Injury:The students attempted to determine mitralvalve dysfunction, specifically MR and/or MS, inthe injured valve, first by conducting an ausculta-tion-based physical examination and then byperforming a focused HCU study. The students’ability to identify MS (13 lesions) and MR (8lesions) by cardiac auscultation was poor andimproved only slightly based on the HCU results(Fig. 4). Also in the diagnostic rate of MS andMR, significant differences were observed amongthe 3 medical students (Table III). The diagnostic

value of cardiac auscultation was also poor whenonly moderate and severe MS and MR wereexamined. Of a total of 42 examinations of casesof moderate or severe MS (30 exams) and mod-erate or severe MR (12 exams), the studentsdetected 9 cases by cardiac auscultation (22%)and 16 (38%) cases based on HCU study.

Discussion:The main finding of this study is that medical stu-dents, after brief training in cardiac ultrasound,can successfully diagnose rheumatic mitral valveinjury with a portable echocardiography device.Although significant differences were observedamong the 3 students, even the least successfulstudent detected 65% of the cases. It is plausiblethat a somewhat longer training period mighthave decreased the differences betweenstudents. In an analysis, no HCU-biased factorswere found. It might have been possible toassume that high BMI would be found in thosepatients who were misdiagnosed (false negative),but the results suggest otherwise.

Another important finding of this study is that8 hours of training in cardiac ultrasound are notsufficient to significantly improve medical stu-dents’ diagnostic rate of valvular dysfunctionwhen a brief HCU study is added to auscultation-based physical examination. In our study, weshowed that the capability of medical students todiagnose MR and MS by cardiac auscultation ispoor, even among patients with at least moder-ate MS and/or MR (sensitivity of 20% and 25%for MS and MR, respectively). These results con-firm data from previous studies: Mangioneet al.34 demonstrated that only 56% of expertcardiologists and 39% of residents were suffi-ciently proficient in cardiac auscultation. Further-more, Kobal et al.21 have shown wide variabilitybetween rates of diagnosis of different heartmurmurs in physical examination carried out bysenior cardiologists, who were able to diagnose

TABLE II

Students’ Diagnosis of Rheumatic Valve Injury and NormalValve by Hand-Carried Cardiac Ultrasound

Parameter Student 1 Student 2 Student 3 Average

Sensitivity(%)

86 93 64 81

Specificity(%)

100 95 90 95

PPV (%) 100 93 82 92NPV (%) 94 95 77 89Accuracy(%)

96 93 82 90

j (P-value) 0.87(P < 0.01)

0.88(P < 0.01

0.55(P < 0.01)

0.77

PPV = positive predictive value; NPV = negative predictivevalue.

TABLE III

Students’ Diagnosis of Mitral Stenosis (13 Cases) and MitralRegurgitation (8 Cases) among Rheumatic Heart Disease

Patients by Hand-Carried Cardiac Ultrasound Study

Parameter Student 1 Student 2 Student 3 Average

Mitral stenosisSensitivity (%) 23 8 31 21Specificity (%) 100 100 100 100Accuracy (%) 29 14 36 26

Mitral regurgitationSensitivity (%) 50 25 25 33Specificity (%) 83 67 67 72Accuracy (%) 64 43 43 50

Figure 4. In gray bars the students’ performance in diag-nosing mitral regurgitation and/or stenosis by physical exami-nation and the improved diagnosis rates when based on ahand-carried ultrasound study (dotted bars).

624

Shmueli, et al.

62% of valvular pathology causing systolic mur-murs (including MR) but only 16% of the valvularlesions responsible for diastolic murmurs (includ-ing MS). The poor diagnostic rate of the studentsin identifying valvular dysfunction comparedwith their high capability to detect rheumaticinjury is related, most probably, to the trainingthey received. Our study, and therefore the train-ing sessions, focused on diagnosing rheumaticvalvular injury rather than on diagnosing andquantifying its complications. In addition, theHCU device used in our study lacked spectralDoppler, making diagnosis of MS more difficult.On the other hand, MR that is detected by colorDoppler (available in the prototype of HCUdevice used by our students) was morefrequently identified by the students. To date, avariety of portable ultrasound devices are avail-able, including devices with a spectral Doppleroption, so it is possible that with these newerdevices, the diagnostic rate of MS will be evenhigher. However, for screening large populationsit is sufficient for the operator to diagnose thepresence of rheumatic valvular changes, as a con-sequence of which the patient can be referredfor a more detailed evaluation in a specializedcenter.

Although in developed countries the extent ofthe RHD has decreased significantly, in poornations its incidence is increasing.4–7 Previousstudies have demonstrated the low sensitivity ofauscultation-based physical examination as ascreening tool in diagnosing RHD among school-aged children.8,9 Marijon et al.8 showed thatbased on echocardiography studies, experiencedphysicians can detect RHD with a sensitivity of94%. The diagnostic criteria for diagnosing RHDare controversial.5,35 These criteria take intoaccount both functional findings assessed bycolor-Doppler and structural abnormalities: limi-tation in valve leaflet movement, subvalvularthickening, and valvular thickening. In this study,we have shown that even briefly trained cardiacultrasound handlers can effectively diagnoserheumatic valve injury based only on structuralfindings. However, our study population and thescreening population examined by Marijon werequite different. Our study population consisted ofadult patients while his screening target popula-tion consisted of children, who have a much ear-lier, less obvious rheumatic mitral valve injurycompared with adult patients who have chronicinjuries. It is possible that the diagnosis of RHDwould be more difficult due to the nature of therheumatic injury in these cases of young patients.However — based on our experience — it wouldmost probably be possible to prepare the person-nel to screen the young RHD patients in a rela-tively short period of training. Our experience

demonstrates the diagnostic utility of HCU toidentify cardiovascular diseases at the pointof patient contact in developing countries.29

Providing the HCU operators with the skills toassess ventricular morphology and function aswell as other pathologic conditions of the valvesduring a screening program for RHD can be ofmajor importance because it is precisely in thisunderserved population that the brief physician–patient contact is a unique opportunity to iden-tify diseases. Misdiagnosis generally means thatthe disease will not be recognized for a longtime, if ever. However, the ability to train stu-dents for these tasks remains to be established.

Limitations:In the studied population, there was isolatedrheumatic injury of the mitral valve. This factraises the question of whether the high diagnos-tic rate of the students for detecting RHD wouldbe similar with rheumatic injury located on othervalves. Since the mitral valve is the valve mostcommonly affected by rheumatic injury, andsince even if there is rheumatic injury in anothervalve, in the majority of cases the mitral valve isinvolved as well,36 it stands to reason that assess-ment of the mitral valve only would be efficientas a screening method for RHD.

The HCU studies were performed in “labora-tory” conditions on adult patients. A study of ayounger population in endemic areas of RHDperformed by briefly trained HCU operators iswarranted to assess the accuracy, cost-benefitratio of this modality as a screening tool for RHDin its early stage.

The prototype of HCU device used in ourstudy belongs to the first generation of portablecardiac ultrasound and was the device used atour institution at the time the research protocolwas planned. A second generation of portabledevices with technologic advances, includingcapacity to store video clips for offline consulta-tion are currently available at our institution. It isplausible to anticipate that the better image reso-lution of new generation HCU devices mightimprove student’s ability to diagnose RHD and itscomplications.

Conclusions:The ability to detect the presence of rheumaticvalve injury using a portable ultrasound device bybriefly trained operators is remarkably high. It isplausible that a longer training period wouldimprove the diagnostic accuracy for RHD compli-cations. This result highlights the utility of porta-ble cardiac ultrasound devices operated bybasically trained personnel as a valuable diagnos-tic tool for RHD.

625

After Brief Training, Students Can Identify Rheumatic MS

References1. Steer A, Carapetis J, Nolan T, et al: Systemic review of

rheumatic heart disease prevalence in children in devel-oping countries: The role of environmental factors. J Pae-diatr Child Health 2002;38:229–234.

2. Carapetis J, Steer A, Mulholland E, et al: The global bur-den of group A streptococcal diseases. Lancet Infect Dis2005;5:685–694.

3. Gerber M, Baltimore R, Eaton C, et al: Prevention of rheu-matic fever and diagnosis and treatment of acute strepto-coccal pharyngitis. Circulation 2009;119:1541–1551.

4. Essop MR, Nkomo VT: Rheumatic and nonrheumaticvalvular heart disease: Epidemiology, management andprevention in Africa. Circulation 2005;112:3584–3591.

5. WHO Rheumatic fever and rheumatic heart disease:report of a WHO expert consultation, Geneva, 29 Octo-ber–1 November 2001. Geneva, Switzerland: WorldHealth Organization, 2004.

6. Lee G, Wessels M: Changing epidemiology of acute rheu-matic fever in the United States. Clin Infect Dis2006;42:448–450.

7. Miyake CY, Gauvreau K, Tani LY, et al: Characteristics ofchildren discharged from hospitals in the United States in2000 with the diagnosis of acute rheumatic fever. Pediat-rics 2007;120:503–508.

8. Marijon E, Ou P, Celermajer DS, et al: Prevalence of rheu-matic heart disease detected by echocardiographicscreening. N Engl J Med 2007;357:470–476.

9. Bhaya M, Panwar S, Beniwal R, et al: High prevalence ofrheumatic heart disease detected by echocardiographyin school children. Echocardiography 2010;27:448–453.

10. DeCara JM, Lang RM, Spencer KT: The hand-carriedechocardiographic device as an aid to the physical exam-ination. Echocardiography 2003;20:477–485.

11. Kobal SL, Atar S, Siegel RJ: Hand-carried ultrasoundimproves the bedside cardiovascular diagnosis. Chest2004;126:693–701.

12. Kimura BJ, DeMaria AN: Technology insight: Hand-car-ried ultrasound cardiac assessment-evolution, not revolu-tion. Nat Clin Pract Cardiovasc Med 2005;2:217–223.

13. Kimura BJ, Amundson SA, Willis CL, et al: Usefulness of ahand-held ultrasound device for bedside examination ofleft ventricular function. Am J Cardiol 2002;90:1038–1039.

14. Lemola K, Yamada E, Jagasia D, et al: A hand-carried per-sonal ultrasound device for rapid evaluation of left ven-tricular function: Use after limited echo training.Echocardiography. 2003:20:309–312.

15. Lipczy�nska M, Szyma�nski P, Klisiewicz A, et al: Hand-carried echocardiography in heart failure and heart fail-ure risk population: A community based prospectivestudy. J Am Soc Echocardiogr 2011;24:125–131.

16. Vourvouri EC, Poldermans D, Schinkel AF, et al: Left ven-tricular hypertrophy screening using a hand-held ultra-sound device. Eur Heart J 2002;23:1516–1521.

17. Senior R, Galasko G, Hickman M, et al: Communityscreening for left ventricular hypertrophy in patients withhypertension using hand-held echocardiography. J AmSoc Echocardiogr 2004;17:56–61.

18. Perez-Avraham G, Kobal SL, Etzion O, et al: Left ventricu-lar geometric abnormality screening in hypertensivepatients using a hand-carried ultrasound device. J ClinHypertens 2010;12:181–186.

19. Spencer K, Anderson A, Bhargava A, et al: Physician-per-formed point-of-care echocardiography using a laptopplatform compared with physical examination in the car-diovascular patient. J Am Coll Cardiol 2001;37:2013–2018.

20. Xie T, Chamoun AJ, McCulloch M, et al: Rapid screeningof cardiac patients with a miniaturized hand-held ultra-sound imager-comparisons with physical examinationand conventional two-dimensional echocardiography.Clin Cardiol. 2004:27:241–245.

21. Kobal SL, Trento L, Baharami S, et al: Comparison ofeffectiveness of hand-carried ultrasound to bedside car-diovascular physical examination. Am J Cardiol2005;96:1002–1006.

22. David LD, Howell EE, Ziegelstein EE, et al: Hand-carriedutrasound performed by hospitalists: Does it improve thecardiac physical examination?. Am J Med 2009;122:35–41.

23. Kimura BJ, Fowler SJ, Fergus TS, et al: Detection of leftatrial enlargement using hand-carried ultrasound devicesto screen for cardiac abnormalities. Am J Med2005;118:912–916.

24. Brennan JM, Blair JE, Goonewardena S, et al: A compari-son by medicine residents of physical examination versushand-carried ultrasound for estimation of right atrialpressure. Am J Cardiol 2007:99:1614–1616.

25. Brennan JM, Blair JE, Hampole C, et al: Radial artery pulsepressure variation correlates with brachial artery peakvelocity variation in ventilated subjects when measuredby internal medicine residents using hand-carried ultra-sound devices. Chest 2007;131:1301–1307.

26. Sicari R, Galderisi M, Voigt JU, et al: The use of pocket-size imaging devices: A position statement of the Euro-pean Association of Echocardiography. Eur J Echocardiogr2011;12:85–7.

27. Giusca S, Jurcut R, Ticulescu R, et al: Accuracy of hand-held echocardiography for bedside diagnostic evaluationin tertiary cardiology center: Comparison with standardechocardiography. Echocardiography 2011;28:136–41.

28. Croft LB, Duvall WL, Goldman ME: A pilot study of theclinical impact of hand-carried cardiac ultrasound in themedical clinic. Echocardiography 2006;23:439–46.

29. Kobal SL, Lee S, Willner R, et al: Hand-carried cardiacultrasound enhances healthcare delivery in developingcountries. Am J Cardiol 2004:94:539–541.

30. Steer AC, Kado J, Wilson N, et al: High prevalence ofrheumatic heart disease by clinical and echocardio-graphic screening among children in Fiji. J Heart Valve Dis2009:18:327–335.

31. Wilson N: Rheumatic heart disease in indigenous popula-tions – New Zealand experience. Heart Lung Circ2010;19:282–288.

32. Saxena A, Ramakrishnan S, Roy A, et al: Prevalence andoutcome of subclinical rheumatic heart disease in India:The RHEUMATIC (Rheumatic Heart Echo Utilisation andMonitoring Actuarial Trends in Indian Children) study.Heart 2011;97:2018–22.

33. Webb RH, Wilson NJ, Lennon DR, et al: Optimizing echo-cardiographic screening for rheumatic heart disease inNew Zealand: Not all valve disease is rheumatic. CardiolYoung 2011;21:436–43.

34. Mangione S, Nieman L, Gracely E, et al: The teachingand practice of cardiac auscultation during internal medi-cine and cardiology training: A nationwide survey. AnnIntern Med 1993;119:47–54.

35. Marijon E, Celermajer DS, Tafflet M, et al: Rheumaticheart disease screening by echocardiography: The inade-quacy of World Health Organization criteria for optimiz-ing the diagnosis of subclinical disease. Circulation2009;120:663–668.

36. Choopra P, Gulwahi H: Pathology and pathogenesis ofrheumatic heart disease. Indian J Pathol Microbiol2007;50:685–697.

626

Shmueli, et al.