Background: Performance on the annual in-training examination (ITE) for emergency medicine (EM) residents has been shown to correlate with performance on the American Board of Emergency Medicine (ABEM) qualifying examination. As such, significant planning is often committed to ITE preparation, from an individual resident and a residency program perspective. Online specialty-specific question banks (QBanks) represent a popular medium for ITE preparation; however, the impact of QBanks on ITE performance is unclear. We sought to assess the impact of QBank participation on EM resident ITE performance.
Methods: ITE and QBank performance results were collated over 2 academic years, 2019 and 2020, from a 3-year EM residency program. The QBank was provided as a self-study option in 2019 and incorporated as a mandatory component of the curriculum in 2020. ITE raw scores and percentile rank for training level scores were compared with performance on the QBank, including QBank average performance score as well as number of QBank questions completed. The Pearson correlation coefficient was used to measure association between ITE performance and QBank correlates. Additional descriptive demographics, to include gender, PGY level, and USMLE step 1 and 2 scores were also considered.
Results: Sixty-two sets (30 residents in 2019, 32 residents in 2020) of ITE performance data and QBank correlates were included. Overall, raw ITE scores and number of QBank questions completed were found to have a significant, positive correlation, (r(60) = 0.34, p < 0.05); correlation was stronger for 2019 (r[28] = 0.39, p < 0.05) compared to 2020 (r[30] =0.25, p = 0.16). Overall, ITE percentile rank for training level scores were also found to have a significant, positive correlation with number of QBank questions completed (r(60) = 0.35, p < 0.05); correlation was again stronger for 2019 (r(28) = 0.42, p < 0.05) compared to 2020 (r(30) = 0.29, p = 0.12). Finally, ITE percentile rank for training level correlated positively with QBank average performance (as a percentage), albeit weakly, and was not found to be significant overall (r[60] = 0.20, p = 0.16); in this instance, 2019 did not show a correlation (r[28] =0.12, p = 0.54); however, 2020 did (r[30] =0.55, p < 0.05). Academic year 2020 raw ITE scores also demonstrated a significant correlation with QBank average performance (r[30] = 0.66, p < 0.0001) while 2019 did not (r[28] = 0.08, p = 0.68).
Conclusion: Participation and engagement in a QBank are associated with improved EM resident performance on the ABEM ITE. Utilization of a QBank may be an effective mode of ITE preparation for EM residents.
© 2021 Society for Academic Emergency Medicine.