The decline in physical examination skills during

Transcription

1 Use of Hand-carried Ultrasound Devices to Augment the Accuracy of Medical Student Bedside Cardiac Diagnoses Jeanne M. DeCara, MD, James N. Kirkpatrick, MD, Kirk T. Spencer, MD, R. Parker Ward, MD, Kristen Kasza, MS, Kathleen Furlong, RN, and Roberto M. Lang, MD, Chicago, Illinois Background: Hand-carried ultrasound (HCU) devices used by cardiologists as extensions of the physical examination have been shown to improve the accuracy of bedside diagnoses. We tested the feasibility of teaching medical students to use HCU devices to make bedside cardiac diagnoses and compared the accuracy of their HCU and physical examinations. Methods: In all, 10 fourth-year medical students enrolled in a 4-week medical school course on the cardiac examination. Students examined 12 standardized patients at 3 different time intervals: (1) on day 1 of the course; (2) on day 10 after review of cardiac physical examination using traditional teaching methods; and (3) after instruction on the use of HCU devices. Students were scored at each time interval for primary findings (most salient) and all findings, accounting for both errors of commission and omission. Scores could range from 12 to 12 for primary findings and from 22 to 22 for all findings. A perfect score was 12 for primary findings and 22 for all findings. Results: The average score for all students at baseline was and for primary and all findings, respectively. A significant improvement in the scores was noted with use of the HCU device ( and for primary and all findings, respectively) compared with the baseline and two subsequent physical examinations. Conclusion: Instruction of fourth-year medical students on the use of HCU device is feasible and results in significantly more accurate bedside diagnoses. (J Am Soc Echocardiogr 2005;18: ) The decline in physical examination skills during the past 2 decades is well documented and occurs at all levels of training. 1-9 Error rates for the auscultatory components of the cardiovascular examination have been particularly high, even when performed by cardiologists. 1,4,5,8-10 Moreover, the interobserver agreement of some auscultatory findings has been shown to be moderate to poor at best. 5 Several studies indicate that echocardiography provides a higher degree of diagnostic accuracy than physical examination using the stethoscope, especially in patients older than 50 years in whom cardiac pathology is more likely. 3,11-14 However, routine use of echocardiography at the bedside has From the Department of Medicine and Section of Cardiology and Department of Health Studies (K.K.), University of Chicago Hospitals. Supported in part by a Course Development Grant from the University of Chicago Pritzker School of Medicine. Dr DeCara is also funded by a Faculty Teaching Award from the University of Chicago, Pritzker School of Medicine. Reprint requests: Jeanne M. DeCara, MD, 5831 S Maryland Ave, MC 5084, Chicago, IL ( jdecara@medicine.bsd. uchicago.edu) /$30.00 Copyright 2005 by the American Society of Echocardiography. doi: /j.echo historically been cumbersome and expensive. Recently, miniature hand-carried ultrasound (HCU) devices have been developed that weigh less than 7 pounds and can easily be used at the bedside. These devices have been studied in a variety of clinical settings and demonstrate improved diagnostic accuracy compared with the physical examination The accuracy of HCU devices compared with standard echocardiography has been favorable even when used by noncardiologists. 18 The advent of truly portable ultrasound machines, at a time when physical examination skills acquired through traditional teaching modalities have been shown to be subadequate, presents an opportunity to use the latest technology in an innovative manner to augment traditional medical student instruction in cardiac examination and potentially improve the accuracy of their bedside diagnoses. Accordingly, our study had 3 aims. The first aim was to test the feasibility of teaching medical students to acquire and interpret limited echocardiographic studies using a HCU device within the confines of a 1-month medical school elective. Secondly, we wished to determine the additive value of HCU machines as teaching tools to augment the instruction of the cardiac examination in comparison with traditional teaching methods by testing the 257

2 258 DeCara et al March 2005 students at baseline; after a systematic review of anatomy, physiology, and physical examination skills using traditional teaching methods; and after learning to use a HCU device. Lastly, we sought to determine the diagnostic accuracy of student-performed HCU examinations in a group of standardized patients. METHODS Study Participants Data for this study were collected during a new medical school elective course on the cardiac examination offered as part of our curriculum. Eligible participants consisted of fourth-year medical students who had completed the mandatory second-year medical school physical examination course and had been participating in hospital ward and elective rotations for 16 months. No incentive was given for participation in the course or the study. Students were permitted to participate in the course without participating in the study. Standardized patients were incorporated into the course as examinees. These patients were selected from our cardiology outpatient clinic population to serve as typical examples of cardiac pathology based on abnormal physical examination findings associated with abnormal echocardiographic findings. These cardiac abnormalities are listed in Table 1. Some patients had more than one cardiac abnormality but, in all cases, there was one finding that was the most salient feature by examination, echocardiogram, or both (primary finding). During the course, students also examined inpatients and outpatients referred to the institutional noninvasive imaging laboratories for routine echocardiography. Neither standardized patients nor patients encountered during the echocardiographic training portion of the study were selected based on image quality. The protocol was reviewed and approved by the institutional review board. Written consent was obtained from all student participants and standardized patients. Oral consent was obtained from patients who were scanned during the training portions of the course. Study Design The medical school course had two educational objectives. The first was to review cardiac anatomy, physiology, and bedside examination skills. The second was to learn how to perform and interpret a HCU examination with the goal of being able to perform a brief and limited echocardiographic examination after the bedside physical examination. In this manner, HCU devices were used as visual teaching aids to reinforce the association between a given cardiac diagnosis and its physical examination correlate. The format for the course is outlined below and illustrated in Figure 1. Table 1 Diagnoses tested Valvular Cardiomyopathies Congenital Other AS Dilated ASD Pericardial Effusion AI Hypertrophic VSD Normal MS Right ventricular MR MVP TR AI, Aortic insufficiency; AS, aortic stenosis; ASD, atrial septal defect; MR, mitral regurgitation; MS, mitral stenosis; MVP, mitral valve prolapse; TR, tricuspid regurgitation; VSD, ventricular septal defect. Physical Examination Review All students underwent a 10-day review of cardiac anatomy, physiology, and cardiac bedside examination skills. This review was comprised of a 6-hour day consisting of lectures supported by CD-ROM recorded heart sounds for classroom and personal use. The intent of this portion of the course was to reiterate the material that the students were taught in their second-year physical examination course and had reinforced through their clinical rotations, using traditional teaching methods. Emphasis was placed on the cardiac abnormalities of each of the standardized patients who the students examined throughout the course. All instructors were from the section of cardiology. Echocardiographic Training After the 10-day review portion of the course, the students spent the remaining 10 days of the course learning to use the HCU device. This time included supervised scanning practice with a sonographer and individual unsupervised practice time. Each student was given a HCU device to use for the duration of the course (Optigo, Philips Medical Imaging, Andover, Mass). These devices were capable of 2-dimensional nonharmonic imaging with color Doppler, measurement calipers, and still-frame image storage. The students were only instructed on the use of the 2-dimensional imaging and color Doppler features. After learning the basic operational features of the HCU machine, students spent 5 8-hour days with a sonographer in the noninvasive imaging laboratories learning how to acquire images in the parasternal long-axis and apical 4-chamber views only. The students would perform a brief cardiac physical examination of referred unselected patients to arrive at a tentative diagnosis. After their physical examination, students immediately performed a HCU examination to confirm or revise their preliminary diagnosis. The adequacy of the student-performed scans and verification of the students echocardiographic diagnoses were assessed by the sonographer and reviewed with the student present using the standard echocardiogram for comparison. In addition, students rotated with a physician on the inpatient wards to perform bedside physical and HCU examinations on patients with noteworthy physical findings.

3 Volume 18 Number 3 DeCara et al 259 Figure 1 Course format for instruction and skills assessment. After baseline bedside examination skills were assessed (time 1) students underwent review of cardiac anatomy, physiology, and examination skills using traditional teaching methods typically used in standard medical curriculum, after which they were again tested on bedside examination skills (time 2). Instruction on acquisition and interpretation of hand-carried ultrasound (HCU) images then commenced. After using HCU device as instructional tool, students were tested on bedside examination using stethoscope alone (time 3A) and after using HCU as extension of physical examination (time 3B). To ensure that the students had ample opportunity to recognize abnormal pathology that is less frequently encountered in clinical practice (ie, mitral stenosis, atrial and ventricular septal defects), 1-hour audiovisual demonstrations were provided by physicians twice daily consisting of echocardiographic images of cardiac abnormalities relevant to the standardized patients who the students examined throughout the course. Data Acquisition Given only a 1-sentence history detailing age, symptomatology, and relevant comorbidities, each student examined 12 standardized patients with cardiac disease at 3 different time intervals: on the first day of the course (time 1), to establish each student s baseline competency in the cardiac examination; after a 10-day review of cardiac anatomy, pathophysiology, and auscultatory findings using traditional teaching methods consisting of lectures and CD-ROMs (time 2); and after 10 days of individualized and supervised instruction on the use of HCU devices (time 3). On this last encounter with the standardized patients, students were first asked to submit a diagnosis based on physical examination (time 3A) and then permitted to perform a HCU examination (time 3B). They were then allowed to resubmit a final diagnosis. All standardized patient examinations were performed in the outpatient cardiology clinic. Students were instructed not to ask the patient questions regarding symptomatology or diagnosis but were allowed to perform as complete a physical examination as they deemed necessary to make the diagnosis. Student findings were recorded on a forced-response checklist of potential diagnoses listed in Table 1. The selection of more than one diagnosis per patient was permitted if desired. Only one student was permitted in the examination room at any given time. Students were instructed not to discuss their findings with each other. An attending cardiologist supervised all visits in which data were acquired. At the completion of the course, students were given a questionnaire in which information was gathered regarding attendance at lecture, number of hours spent on independent study of the physical examination, number of HCU scans performed, confidence in their physical examination before and after the course, and assessment of difficulty in learning to use the HCU device to acquire and interpret limited echocardiographic images for the purpose of making a diagnosis. Data Analysis All data were analyzed by the primary investigator (J. M. D.) in conjunction with a statistician (K. K.). Student findings were entered into a database as a true-positive, false-positive, true-negative, or false-negative result. A scoring system was generated by assigning points for the various types of student responses. For instance, for each abnormal finding in a given standardized patient, 1 point was given for each true-positive finding, 0.5 points given for a true-positive finding with one or more false-positive results, and 0 points were given for a false-negative result (omission). A 1-point deduction was given for the combination of a false-negative result and one or more falsepositive results. In the case of a patient without cardiovascular abnormalities, one point was given if the student said the findings were normal and one point was deducted if the student incorrectly diagnosed the patient with a cardiac abnormality (false-positive result). A score for each student s performance was calculated for each standardized patient on each visit. There were 12 primary findings and 10 additional nonprimary findings for a total of 22 findings for all 12 patients examined. Thus, the overall score for the 22 findings could range from a perfect score of 22 to a completely inaccurate score of 22. Similarly, for the 12 primary findings, scores could range from 12 to 12. Scores for those students who did not examine every patient on any given encounter were adjusted by dividing their unadjusted score by the number of diagnoses they did make and then multiplying by the total number of diagnoses possible. Statistics Student scores for the primary findings and all findings were compared over all testing intervals using repeated measures analysis of variance with a Greenhouse-Geisser adjustment. The study was powered at 80% to detect a 20% difference in number of correct primary findings between times 2 and 3A at an alpha equal to Wilcoxon signed rank test was used to compare confidence levels in physical examination skills before and after the course. RESULTS Of the 10 students, 8 completed the examinations on every patient on each testing day. One student did not examine patients at time 1 and another student did not complete examinations on 3 patients

4 260 DeCara et al March 2005 Table 2 Mean score for primary and all findings Time N Primary findings, mean SD All findings, mean SD A B * * Times 1 3B as explained in Figure 1. *P.05 for all comparisons to Time 3B Figure 2 Trend in students scores for primary findings (top) and all findings (bottom) over time. Using handcarried ultrasound device after physical examination (time 3B) resulted in significantly higher score than all intervals in which students relied on physical examination alone (times 1, 2, and 3A). at times 3A and 3B. Scores were adjusted accordingly. Overall, repeated measures analysis of variance indicated the accuracy of the students bedside diagnoses improved by the end of the course for both primary findings (P.0045) and all findings (P.0012). In general, the majority of students showed a similar pattern of improvement during the 4 time intervals at which they were tested. Serial trends in scores for primary findings and all findings are shown in Figure 2. The average score for all students at baseline was and for primary and all findings, respectively. This score increased on each successive examination for all findings. Of note, the score at time 3B using the HCU device after physical examination was statistically higher than scores at times 1, 2, and 3A during which only physical examination was performed. There was no significant improvement in score between any of the visits in which physical examination was solely used to make the bedside diagnosis (Table 2). Table 3 indicates that with the use of a HCU device immediately after the physical examination, students failed to decrease errors of omission (falsenegative results) but made significantly more correct diagnoses, often in combination with at least one false-positive answer. Although the number of completely correct diagnoses (true-positive results only) using the HCU device at time 3B was significantly higher than at baseline, it was not significantly higher than at time 2 or 3A when the bedside diagnosis was made by physical examination alone. In contrast, the number of partially correct answers (a true-positive result in combination with one or more false-positive diagnoses) with the HCU device was significantly higher than that found using physical examination alone (times 2 and 3A) whereas the number of incorrect diagnoses (false-positive results) significantly decreased over time. These data indicate that it was the conversion of false-positive findings to partially correct responses (true-positive findings in combination with false-positive findings) that accounted for the increase in score seen at time 3B, because the scoring system assigned a greater penalty for a false-positive result than a partially correct result. For instance, in the case of mitral stenosis, a student may have misdiagnosed the patient with aortic insufficiency at baseline. In this case, not only did the student commit an error of omission but he also gave a false-positive diagnosis of aortic insufficiency, resulting in a 1-point deduction and net score of 1.0. At time 3B that same student may have correctly diagnosed the patient with mitral stenosis but also incorrectly given a diagnosis of mitral insufficiency. This partially correct answer would result in a score of 0.5 points. Therefore, the student would have a higher score at time 3B compared with baseline, primarily because of the conversion of the false-negative/false-positive response combination to a true-positive/false-positive response combination. Because each standardized patient was chosen for a unique salient cardiac abnormality, subgroup analysis on the diagnostic accuracy of HCU-enhanced bedside examination for any particular cardiac pathology could not be performed in this study. However, student performance in the detection of valvular abnormalities was compared with performance

5 Volume 18 Number 3 DeCara et al 261 Table 3 Distribution of findings and score for primary findings Time 1 Time 2 Time 3A Time 3B Omissions (FN) 12 (11.7%) 13 (10.8%) 8 (6.8%) 12 (10.3%) Correct (TP only) 22 (21.4%)* 39 (32.5%) 40 (34.2%) 49 (41.9%) Partially correct (TP FP) 14 (13.6%) 8 (6.7%) 9 (7.7%) 22 (18.8%) Incorrect (FP only) 55 (53.3%) 60 (50%) 60 (51.3%) 34 (29.0%)* Primary score * Times 1 3B as explained in Figure 1. FN, false negative result; FP, false positive result; TP, true positive result. *P.05 compared with all other times. P.05 compared with times 2 and 3A. in the detection of cardiomyopathies. The difference in mean percentage of true-positive responses among total responses offered by the students after examining standardized patients with valvular abnormalities between times 1 and 3B was 29.2% (mean % at time 1 compared with at time 3B, P.002). In comparison, the difference in mean percentage of true-positive responses among total responses offered after examining standardized patients with cardiomyopathies was 16.9% (mean at time 1 compared with at time 3B, P.09). Therefore, despite a trend toward improved detection of cardiomyopathies after a HCU-assisted bedside examination, only the detection of valvular abnormalities significantly improved from time 1 to 3B. Based on the results of the questionnaire at the conclusion of the course, 9 of 10 students (90%) indicated that learning to operate the HCU device was easy and that using it to acquire and interpret limited echocardiographic images for diagnostic purposes was convenient. The mean number of HCU examinations performed by the students was Students attended a mean of lectures and independently spent hours reviewing recorded heart sounds outside of class. Students rated their confidence in bedside examination before and after the course on a scale of 1 to 4, with 1 being extremely confident and 4 being not at all confident. The mean confidence rating before the course was compared with after the course (P.002). DISCUSSION Our study has 3 main findings. First, it demonstrates the feasibility of teaching fourth-year medical students how to acquire and interpret echocardiographic images, limited to two echocardiographic views, within the confines of a 1-month medical school elective. Secondly and most importantly, we demonstrated that the student-performed HCU examination, when used as an extension of the physical examination, yielded a significantly more accurate bedside diagnosis than physical examination alone, even when a focused effort to review traditional examination skills was undertaken. The benefit of HCU in improving the diagnostic accuracy of the bedside examination appeared to be greater for the detection of valve abnormalities than for cardiomyopathies, likely reflecting the relative ease of learning to interpret color flow Doppler data in comparison with interpretation of wall motion among novice HCU device users. Lastly, contrary to what one might expect, the HCU device did not appear to have a significant additive value in improving medical student physical examination skills. However, it should be noted that there was a 13% increase in number of correct primary findings from time 2 to 3A and, based on our power calculations, the number of students who participated in this study may have been insufficient to detect a difference in accuracy between these two time intervals. The data from our study reconfirmed the inadequate physical examination skills of this generation of new physicians who acquire these skills within the context of a traditional medical school curriculum. In this study, student performance on the physical examination using the stethoscope alone did not significantly improve over the course of the month. Notably, these students had not only undergone the training in the bedside examination offered in our standard medical school curriculum but also an organized review specifically targeted to the cardiac examination. These results imply that insufficient learning may occur despite an intensive conventional instruction effort. Recent efforts to use technology to improve physical examination skills, such as interactive computer simulation and even the Harvey mannequin, have resulted in modest success These techniques take place strictly in the classroom and do not involve actual patients in the learning process. In contrast, our study used the stethoscope in combination with the HCU device at the time of the patient encounter, thereby introducing a new paradigm of teaching that, like the stethoscope, involves direct patient interaction and results in improved accuracy of bedside diagnoses. The majority of students in our study believed that

6 262 DeCara et al March 2005 the learning process for HCU devices was easy and all reported greater confidence in their examination skills by the end of the course. Although ours is the first study to demonstrate the usefulness of HCU devices to augment instruction of the cardiac examination and to refine the accuracy of medical student bedside diagnoses, the use of these machines as instructional aids for gross anatomy has been reported with encouraging results. 26 Together, these studies suggest a potential role for the incorporation of HCU devices into the medical school curriculum. The HCU device used in this study costs approximately $15,870. In contrast, stethoscopes commonly used in practice can cost as little as $65. Although the expense of incorporating HCU devices into medical school curriculum would not be insignificant, one might speculate that a significantly more accurate bedside examination performed by these future physicians would result in more appropriate triage of patients and, consequently, more appropriate use of medical tests. Familiarity with this powerful diagnostic tool will prove invaluable as the costs of new technology decrease over time and use of HCU devices in the community becomes increasingly prevalent. Moreover, standardized and formalized instruction on the acquisition and interpretation of limited echocardiography during medical school education promotes responsible use of this technology and creates a context in which to assess competency. Our study has several limitations. The results reflect a single medical school s experience using a small student population. Results from a larger study in which multiple medical schools participate is needed to confirm our findings before widespread introduction of HCU devices into mainstream medical education can be considered. In that setting, it would also be of interest to re-evaluate the students learning process in a different patient subset at some remote point from the course to re-evaluate their continued ability to acquire and correctly interpret HCU images. It was not feasible to do so in the current study because the time that would need to be allocated to that aim conflicted with the time the students needed to meet their curriculum criteria before graduation. It is also important to note that the students in this study examined the same 3 patients on 3 separate occasions. Therefore, a potential bias resulting from a learning curve may have been introduced into the results. Although we considered this possibility, we chose to use the same standardized patients at each time interval to reduce the confounding influence of intersubject variability in physical examination findings, without which it would be unclear whether changes in student performance truly reflected a change in skill. Moreover, the lack of significant improvement in physical examination skills at times 2 and 3A suggests that this factor did not affect our results. Lastly, because the students were aware of the goals of the study, a pretest bias that would cause the students to be motivated to be more accurate when using HCU devices cannot be excluded nor differentiated from their perhaps generational enthusiasm for the use of new technology in a clinical setting. In summary, this study has shown that teaching medical students to perform and interpret a limited echocardiographic examination using a HCU device was feasible after a brief instruction period. Use of this new technology resulted in a significant improvement in the diagnostic accuracy of medical student bedside cardiac diagnoses. Further investigation will be required to determine the cost-effectiveness of widespread incorporation of these devices into mainstream medical education. The authors gratefully acknowledge Drs Ajoy Kapoor, Jeffrey Teuteberg, and Keith Miller for their assistance in teaching the medical school course and Philips Medical Imaging for loaning equipment for this project. REFERENCES 1. Raftery EB, Holland WW. Examination of the heart: an investigation into variation. Am J Epidemiol 1967;85: Aloia JF, Jonas E. Skills in history-taking and physical examination. J Med Educ 1976;51: Jaffe WM, Roche AH, Coverdale HA, McAlister HF, Ormiston JA, Greene ER. Clinical evaluation versus Doppler echocardiography in the quantitative assessment of valvular heart disease. Circulation 1988;78: Johnson JE, Carpenter JL. Medical house staff performance in physical examination. Arch Intern Med 1986;146: Lok CE, Morgan CD, Ranganathan N. The accuracy and interobserver agreement in detecting the gallop sounds by cardiac auscultation. Chest 1998;114: Mangione S, Nieman LZ, Gracely E, Kaye D. The teaching and practice of cardiac auscultation during internal medicine and cardiology training: a nationwide survey. Ann Intern Med 1993;119: Mangione S, Nieman LZ. Cardiac auscultatory skills of internal medicine and family practice trainees: a comparison of diagnostic proficiency. JAMA 1997;278: Mangione S. Cardiac auscultatory skills of physicians-in-training: a comparison of three English-speaking countries. Am J Med 2001;110: Wray NP, Friedland JA. Detection and correction of house staff error in physical diagnosis. JAMA 1983;249: St Clair EW, Oddone EZ, Waugh RA, Corey GR, Feussner JR. Assessing housestaff diagnostic skills using a cardiology patient simulator. Ann Intern Med 1992;117: Cheesman MG, Leech G, Chambers J, Monaghan MJ, Nihoyannopoulos P. Central role of echocardiography in the diagnosis and assessment of heart failure: British Society of Echocardiography. Heart 1998;80:S Rahko PS. Prevalence of regurgitant murmurs in patients with valvular regurgitation detected by Doppler echocardiography. Ann Intern Med 1989;111:

7 Volume 18 Number 3 DeCara et al Roldan CA, Shively BK, Crawford MH. Value of the cardiovascular physical examination for detecting valvular heart disease in asymptomatic subjects. Am J Cardiol 1996;77: Xu M, McHaffie DJ. Nonspecific systolic murmurs: an audit of the clinical value of echocardiography. N Z Med J 1993; 106: Rugolotto M, Hu BS, Liang DH, Schnittger I. Rapid assessment of cardiac anatomy and function with a new handcarried ultrasound device (OptiGo): a comparison with standard echocardiography. Eur J Echocardiogr 2001;2: Xie F, Breese MS, Nanna M, Lichtenberg GS, Allen MN, Meltzer R. Blinded comparison of an ultrasound stethoscope and standard echocardiographic instrument. Chest 1988;94: Spencer KT, Anderson AS, Bhargava A, Bales AC, Sorrentino M, Furlong K, et al. Physician-performed point-of-care echocardiography using a laptop platform compared with physical examination in the cardiovascular patient. J Am Coll Cardiol 2001;37: DeCara JM, Lang RM, Koch R, Bala R, Penzotti J, Spencer KT. The use of small personal ultrasound devices by internists without formal training in echocardiography. Eur J Echocardiogr 2003;4: Ewy GA, Felner JM, Juul D, Mayer JW, Sajid AW, Waugh RA. Test of a cardiology patient simulator with students in fourthyear electives. J Med Educ 1987;62: Harley A. Evaluation of a heart sound simulator in teaching cardiac auscultation. J Med Educ 1976;51: Horiszny JA. Teaching cardiac auscultation using simulated heart sounds and small-group discussion. Fam Med 2001;33: Lilienfield LS, Broering NC. Computers as teachers: learning from animations. Am J Physiol 1994;266:S Petrusa ER, Issenberg SB, Mayer JW, Felner JM, Brown DD, Waugh RA, et al. Implementation of a four-year multimedia computer curriculum in cardiology at six medical schools. Acad Med 1999;74: Sajid AW, Ewy GA, Felner JM, Gessner I, Gordon MS, Mayer JW, et al. Cardiology patient simulator and computer-assisted instruction technologies in bedside teaching. Med Educ 1990;24: Waugh RA, Mayer JW, Ewy GA, Felner JM, Issenberg BS, Gessner IH, et al. Multimedia computer-assisted instruction in cardiology. Arch Intern Med 1995;155: Wittich CM, Montgomery SC, Neben MA, Palmer BA, Callahan MJ, Seward JB, et al. Teaching cardiovascular anatomy to medical students by using a handheld ultrasound device. JAMA 2002;288: Correction In the article Subacute effusive constrictive pericarditis: Diagnosis by serial echocardiography, by Baker and Orsinelli, in the November 2004 issue /$30.00 doi: /j.echo (J Am Soc Echocardiogr 2004;17:1204-6), reference #8, Senni M, Redfield et al, was incorrectly cited. The reference is incorrectly listed as coming from the Journal of the American Society of Echocardiography. The correct journal citation is J Am Coll Cardiol 1999;33: