ABSTRACT
OBJECTIVE: We aimed to determine whether a threshold number of forceps deliveries in residency predicts use of forceps in independent practice. STUDY DESIGN: We surveyed obstetrics and gynecology residency graduates of 2 academic programs from 2008 through 2012 regarding the use of operative vaginal delivery in practice. At these programs, residents are trained in both forceps and vacuums. Individual case log data were obtained with the number of forceps deliveries performed by each respondent during residency. Respondents were grouped as currently using any forceps or vacuums alone. A logistic regression model estimated the probability of forceps use, predicted by the number of residency forceps deliveries. From the resulting receiver-operating characteristic curve, we assessed sensitivity, specificity, positive predictive value, and area under the curve. RESULTS: The response rate was 85% (n = 58) and 90% (n = 52) practice obstetrics. Seventy-nine percent (n = 41) use forceps in practice. The mean number of forceps performed during residency was 22.3 ± 1.3 (mean ± SE) in the any-forceps group and 18.5 ± 2.1 in the vacuums-only group (P = .14). Although the model performed only moderately (area under the curve, 0.61, 95% confidence interval [CI], 0.42-0.81), more than 13 residency forceps deliveries corresponded to a 95% sensitivity (95% CI, 84-99) and a positive predictive value of 83% (95% CI, 69-92) for using forceps in practice. The specificity of this threshold is 27% (95% CI, 6-61). CONCLUSION: Although exceeding 13 forceps deliveries made it highly likely that obstetricians would use them in practice, further study is necessary to set goals for a number of resident forceps deliveries that translate into use in practice.