APGO

Objective: Evidence-based medicine has become the standard of care in clinical practice. To fulfill the CanMEDS role of Medical Expert and Scholar defined by the Royal College of Physicians and Surgeons of Canada, treatment decisions should be based on sound interpretations of best available evidence. There is little data on epidemiology/biostatistics training in Canadian postgraduate OB/GYN programs or on understanding/confidence with fundamental biostatistical and epidemiological principles. The objectives were to: (1) Assess postgraduate OB/GYN trainees’ training and confidence with critical appraisal; and (2) Assess the knowledge of fundamental epidemiological/biostatistical principles among Canadian OB/GYN trainees.

Methods: All Canadian OB/GYN residents (years 1 to 5) were invited to complete an anonymous cross-sectional survey to determine confidence levels with critical appraisal. This was performed during a national standardized in-training examination administered to all Canadian OB/GYN residents in November 2011. Second, fifteen critical appraisal knowledge questions were integrated into the same standardized examination. Third, the residency program directors of the sixteen OB/GYN residency programs in Canada were asked to complete a 5 question survey, assessing the state of epidemiology training within their residency program. Primary outcome measures were level of confidence interpreting biostatistical results and applying research findings to clinical practice, desire for more biostatistical/epidemiological training in residency, and percentage of biostatistics/epidemiology questions correct. Descriptive statistics were created for all study variables. Cochran-Armitage Trend test was used to test if there was any trend with an increase in postgraduate year (PGY) training level. For dichotomized outcomes, logistic regressions were applied. For continuous outcomes, linear regression was used. A p-value of <.05 was considered statistically significant. All analyses were conducted using SAS for Windows, Version 9.2, SAS Institute Inc.

Results: Of 355 residents taking the examination, 301 residents consented to complete the survey (response rate = 84.8%). The majority (76.74%) had little/no confidence interpreting research statistics. Confidence was significantly higher with increased seniority (OR=1.93), previous epidemiology/statistics course (OR=2.65), and personal history of publication (OR=1.82). Many (68%) had little/no confidence applying research findings to clinical practice. Confidence was increased with higher training year (p<0.0001) and with formal epidemiology training during residency (OR=2.01). All 355 residents completed the epidemiology/biostatistical knowledge assessment questions on the standardized examination. Average mark was 69.8%. Increasing seniority was associated with improved overall test performance (p=0.023). On knowledge questions, the odds ratio was correctly interpreted by 92.7% and the p-value by 88.7%. Poorer performance topics were analytical study method (9.86%), analytical study design (36.9%), sample size (42.0%), and number needed to treat (53.8%). The survey provided to the residency program directors had a 68.75% response rate. Ninety percent responded that they a have formal clinical epidemiology curriculum incorporated into their residency program. The majority (73%) of program directors responded that they felt that the clinical epidemiology training within their program was sufficient to prepare their residents to practice evidence based medicine. Most (84.39%) residents wanted more epidemiology teaching.

Conclusion: Canadian OB/GYN residents may have the biostatistical/epidemiological knowledge to interpret many of the results published in the literature but lack confidence applying these skills in a clinical setting. Most residents want additional training in these areas and residency programs should include training, perhaps in a formalized curriculum, to improve their confidence and prepare them for a lifelong practice of evidence-based medicine.