Methods: This is a retrospective study of children (8-20 years) who presented to a pediatric cardiology clinic between 01/1994 and 12/2011 with vasovagal syncope and no preexisting cardiac or chronic systemic disease. Data collected include: 1) demographics, 2) history and PE findings, 3) diagnostic tests, and 5) test results. Practice variation was evaluated based on providers experience (group 1 = 0-5 yrs, group 2 = 6-10 yrs, group 3 = >10 yrs) and subspecialty. Comparisons of demographic and clinical characteristics between groups were made with the Wilcoxon-Mann-Whitney and Kruskall Wallis tests for continuous data and the Fisher’s exact test for categorical data. Regression analysis was performed to assess factors leading to over-testing. The sensitivity and specificity of history, PE, and ECG to identify cardiac syncope were calculated.
Results: A total of 617 patients [males; 49%] with a median age of 14.5 [range 8.1-19.2] were studied. Electrophysiologists saw the majority of the patients [286, 46%] followed by generalists [187, 30%], imagers [104, 17%] and others [40, 6%]. A cardiac cause for syncope was found in 15 [2%]. The sensitivity and specificity for diagnosing a cardiac cause for syncope were 100% and 55% for the combination of concerning history, and abnormal PE and ECG. Of the 334 [54%] patients whose history was consistent with vasovagal syncope and had a normal PE and ECG, 140 [42%] echocardiograms, 53 [16%] exercise stress tests, 72 [22%] holter monitors, and 19 [6%] event monitors were obtained. In this cohort an average of $847 were spent on additional tests. On univariate analysis, providers with greater experience ordered slightly more tests [median 3 tests in group 3 vs. 2 tests in groups 1 & 2, p = 0.005]. Subspecialty did not influence type or number of testing. On multivariate analysis, factors that influenced testing included more syncopal episodes [RR 0.96, p = 0.0083] and follow-up visits [RR 1.24, p = 0.0001] and history of psychiatric medication use [RR 0.77, p = <0.0001].
Conclusion: Clinical history, PE, and ECG detected all cardiac causes of syncope. In pediatric vasovagal syncope, despite a negative screen, pediatric cardiologists over-test. Experience or subspecialty of the provider did not influence type or number of tests. Increased number of syncopal events and clinic visits resulted in more testing. Pediatric specific guidelines may result in targeted testing and judicious resource utilization in pediatric vasovagal syncope.