PEDIATRICS Vol. 105 No. 6 June 2000, pp. 1184-1187

Clinical Auscultation Skills in Pediatric Residents

Peter R. A. Gaskin, MB, BS^*, Susan E. Owens, MD, Norman S. Talner, MD^*, Stephen P. Sanders, MD^*, and Jennifer S. Li, MD^*

From the Divisions of ^* Pediatric Cardiology and  General Pediatrics, Department of Pediatrics, Duke University Medical Center, Durham, North Carolina.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Objective.  The aim of this study is to determine the level of clinical auscultation skills in pediatric residents at Duke University Medical Center.

Methods.  Forty-seven residents from pediatrics and joint medicine/pediatrics training programs at Duke University Medical Center were enrolled in this study. They were asked to examine the cardiovascular patient simulator, Harvey, and report their findings. Five common conditions seen in the pediatric population were presented: ventricular septal defect, atrial septal defect, pulmonary valve stenosis, combined aortic valve stenosis and insufficiency, and innocent systolic ejection murmur. The responses were scored by the number of features and diagnoses accurately reported. Five pediatric cardiologists and cardiologists in training were also asked to participate in a manner similar to the trainees.

Results.  The mean score of features identified for the resident group was 11.4 ± 2.6 of a possible 19. The diagnostic accuracy was 33%. There was no significant difference between residents by year of training or by type of residency program, although there was a trend toward improved performance with more training. The difference in performance between the pediatric cardiology group and the residents group was striking. The condition that was most frequently misdiagnosed was the innocent systolic ejection murmur.

Conclusions.  The clinical auscultation skills of pediatric residents in this study were suboptimal. There was a trend toward improvement as training progressed, although not statistically significant. These skills are likely to improve further with increased exposure to patients with cardiovascular disease especially in the ambulatory care setting.  Key words:  auscultation, Harvey, murmur.

Cardiac auscultation has long been considered 1 of the cornerstones of the cardiovascular evaluation. Competence in auscultation is important for pediatricians because cardiac murmurs are very common in children. Up to 50% of children are reported to have innocent murmurs,¹ whereas structural heart defects are present in nearly 1% of births. Pediatricians must be competent at cardiac auscultation to screen these patients appropriately and cost-effectively, while not overlooking those with serious structural heart defects. Recently, the performance of clinicians in training at cardiac auscultation has come under scrutiny, and the reported results for residency programs in internal medicine have been disappointing.^2,3 The auscultatory performance of residents in pediatric training programs, however, has not been similarly evaluated. The purpose of this study was to assess the auscultatory performance during training of pediatrics and medicine/pediatric residents.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

This study was performed at Duke University Medical Center in Durham, North Carolina in the Department of Pediatrics by the Division of Pediatric Cardiology, where there is an approved 3-year fellowship training program in pediatric cardiology. The Medical Center also sponsors approved training programs in medicine and cardiovascular diseases. Forty-seven of the 64 residents in the pediatrics and medicine/pediatrics programs (>90% of onsite residents) were enrolled in this study of clinical auscultation skills on a voluntary basis. There was only 1 refusal. Testing was performed on each resident with the aid of the cardiovascular patient simulator, Harvey.² Harvey is an adult-sized mannequin that produces realistic simulations of arterial pulses, blood pressure, jugular venous pulsations, precordial activity, heart sounds, and respiratory sounds, alone or in synchrony.

We limited our assessment to clinical findings of 4 common congenital heart lesions and an innocent murmur. The congenital heart lesions included were pulmonary valve stenosis (with ejection click), ventricular septal defect, atrial septal defect, and combined aortic valve stenosis and insufficiency.

The testing was performed early in the academic year (August 1998) over a 5-week period yielding a cross-sectional survey of the current group of residents. The resident physician was given the opportunity to examine the mannequin for 5 minutes for each situation programmed. Each participant was then asked to complete a data response sheet by characterizing the programmed auscultatory findings. The only history provided was that "this asymptomatic child presented in their clinic for evaluation."

The pediatric cardiology group consisted of 5 pediatric cardiologists (2 attending and 3 cardiologists in training). They were tested in the same manner to confirm that the physical findings presented by the mannequin were appropriate. Three residents were excluded from the study. Two 4th-year medicine/pediatric residents were excluded because there were no pediatric residents at a similar level of training with whom to compare them. The other resident excluded was visiting from a British residency program.

Scoring

One point was scored for each key feature correctly identified for a maximum score of 19 points (Table 1).

Statistics

Results are reported as a mean score ± 1 standard deviation. Analysis of variance was used to test the significance of the difference in score among groups. A P value of .05 or less was considered significant.

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

The average score for pediatric residents was 11.1 ± 2.6 (58%), not significantly different for the average score for the medicine/pediatric residents, 10.4 ± 3.0 (55%; P = .5).

The mean number of correct diagnoses for the 5 conditions was 1.9 ± 1.3 for the pediatric residents and 1.5 ± .7 for the medicine/pediatric residents (P  .3).

The mean score for the pediatric cardiology group was 16.8 ± 1.9 (88%) and the mean number of correct diagnoses was 4.2 ± 1.1 (84%).

Mean scores for the residents, both composite (P = < .0005) and by training year (postgraduate year 1 [PGY1]: P < .0001; PGY2: P < .0001; and PGY3: P < .017), were significantly lower than for the pediatric cardiology group. Similarly, the number of correct diagnoses was significantly higher for the pediatric cardiology group than for the residents, both as a group (P < .0001) and by training year.

Figure 2 shows the diagnostic accuracy by level of training for each condition presented. The results ranged from 23% to 41% for the group as a whole. The systolic ejection innocent murmur was most frequently misdiagnosed.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Schematic representation of murmurs.

View larger version (42K): [in this window] [in a new window]   Fig. 2.   Diagnostic accuracy by year of training.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

Previous studies of internal medicine and family medicine residents have demonstrated less than optimal clinical auscultation skills^2,3 with diagnostic accuracy rates ranging from 20% to 54%.^2,3 The present study conducted in pediatrics and medicine/pediatrics programs at Duke University Medical Center demonstrates that there is similar room for improvement in their auscultation skills. The accuracy rate in this study is comparable to that for internal medicine residents, being 23% to 41% and 33% overall.

There was a trend noted toward improvement in auscultation skills as training progressed, but this did not reach statistical significance. Nonetheless, there was a wide separation between the skills of even senior residents and cardiologists in training or faculty members. Auscultation skills did not differ significantly by type of program (pediatrics vs medicine/pediatrics).

Only 6 residents completed a cardiology elective. Their score (11.6) was higher than their other colleagues (10.8), although not statistically different. The small size of this group may have hampered efforts to detect a difference.

One major limitation of this study is the adult size of the mannequin and the associated slower heart rate. However, this should have biased results to a higher score because the auscultatory findings are easier to discern at a slow heart rate. Further, Harvey is immobile and is not programmed to demonstrate auscultatory findings in the back, neck, or axilla. The ability to assess radiation of murmurs and the changes in character of the murmur with change in position may have limited the ability of the residents to make an accurate diagnosis. In contrast, it is certainly easier to examine the mannequin than a small child who might be moving or crying. It would be of interest to have similar studies of auscultatory skill repeated at other training institutions.

Of note in this study, the innocent systolic ejection murmur was the 1 most frequently misdiagnosed. This murmur, however, was a pulmonary flow murmur, not the typical Still's murmur commonly seen in childhood. A nonspecific, soft (grade 2/6) pulmonary flow murmur (without vibratory or musical quality) without either a wide fixed split in S2 or a tricuspid flow rumble is a common finding in young children with atrial septal defects. This may well have affected the diagnostic accuracy of the examiners. The innocent pulmonary flow murmur with normal splitting and a physiologic S3 is easily reproduced using the Harvey model. The difficulty in reproducing the musical or vibratory (Still's murmur) with this model, is a limitation of this study.

In this era of managed care, there is a greater focus on the cost of delivery of health care, while at the same time attempting to preserve quality of patient care. Pediatricians are faced daily with decisions about which patients should be referred for evaluation of a heart murmur. Smythe and associates⁵ also found that in 69% of the patients referred by pediatricians to pediatric cardiologists, the referral diagnosis was inaccurate, and only 30% of patients referred were found to have definite or possible underlying heart disease. Although it has been found to be more cost-effective for pediatricians to refer patients to a pediatric cardiologist for murmur evaluation than to request an echocardiogram,⁶ their ability to accurately screen for innocent or functional murmurs and avoid the referral to the subspecialist would certainly be more cost-effective. This requires more emphasis during residency training on developing auscultatory skills that would permit the physician to differentiate in most situations a murmur associated with underlying heart disease from 1 not associated with cardiovascular pathology.

Improved training in physical diagnosis should reduce the number of unnecessary referrals and may even enhance the accuracy of specific cardiac diagnosis. Greater skill in cardiac auscultation may also prove to be life saving for the occasional patient with a significant congenital heart defect. The pediatrician should be aware that the pediatric cardiologist always remains available as a consultation resource.

In the past, skills in the cardiovascular examination, including auscultation, were acquired during rotations on the pediatric inpatient service. The recent emphasis on decreasing the length of in-hospital stay severely limits the number of patients available for resident training. In addition, patients undergoing heart catheterization or cardiovascular surgery are frequently admitted on the day of the procedure and, therefore, are not readily available for examination by the house officer. To improve the situation, it would be appropriate and necessary to place more emphasis on training in the pediatric cardiology subspecialty clinic and private office settings.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

The clinical cardiac auscultation skills of pediatric residents were found in these studies to be suboptimal and comparable with what has been reported in internal medicine resident training programs. Of special concern was the finding that the pediatric house officer had considerable difficulty differentiating a common innocent murmur from 1 associated with cardiovascular pathology. Although not statistically significant, there was a slight trend toward improved scores and accuracy of diagnosis for the PGY3 group, compared with residents at the other levels of training. Despite this, there is still quite a difference between the performances of all residents and the pediatric cardiology group. The type of pediatric residency program (pediatrics and medicine/pediatrics) did not influence performance, suggesting no difference in exposure during training. Cardiologists affiliated with pediatric residency programs need to be more proactive in the teaching of residents to improve their auscultatory skills. Residency training programs should also recognize the need for house officers to spend more time in specialty clinics as an essential part of their training.

	FOOTNOTES

Received for publication May 28, 1999; accepted Oct 6, 1999.

Reprint requests to (P.R.A.G.) Division of Pediatric Cardiology, Box 3090, Duke University Medical Center, Durham, NC 27710. E-mail: gaski004{at}mc.duke.edu

	ABBREVIATIONS

PGY, postgraduate year.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Newburger JM, Rosenthal A, Williams RG, Fellows K, Miettinen OS Noninvasive tests in the initial evaluation of heart murmurs in children. N Engl J Med 1983; 308:61-64 [Abstract]
2. St Clair EW, Oddone EZ, Waugh RA, Corey GR, Feussner JR Assessing housestaff diagnostic skills using a cardiology patient simulator. Ann Intern Med 1992; 177:751-756
3. Mangione S, Nieman LZ Cardiac auscultation skills of internal medicine and family practice trainees: a comparison of diagnostic proficiency. JAMA 1997; 278:717-722 [Abstract]
4. Gordon MS, Ewy GA, DeLeon AC Jr, Harvey, the cardiology patient simulator: pilot studies on teaching effectiveness. Am J Cardiol 1980; 45:791-796 [CrossRef][Medline]
5. Smythe JF, Teixeira OHP, Vlad P, Demers PP, Feldman W Initial evaluation of heart murmurs: are laboratory tests necessary? Pediatrics 1990; 86:497-500 [Abstract/Free Full Text]
6. Danford D, Nasir A, Gumbiner C Cost assessment for the evaluation of heart murmurs in children. Pediatrics 1993; 91:365-368 [Abstract/Free Full Text]