The effect of physician characteristics on clinical behavior: Cesarean section in New York State

This study examines 65,784 obstetric deliveries occuring in 1986 in New York State, attended by 1740 different physicians. Cesarean section rates, and rates of reporting of dystocia and fetal distress, are calculated by physicians' year of graduation from medical school, U.S. or foreign medical graduate (FMG), physician sex, board certification, and professional appointment. Probit regressions are estimated, in which the dependent variable is whether an individual delivery is vaginal or cesarean section. Crude cesarean section rates, cross-tabulations, and probit regressions all show physician characteristics to influence cesarean section rates significantly. FMGs are significantly more likely to deliver by cesarean section, both overall and when all other relevant factors are controlled for. Cesarean section rates by FMGs vary somewhat according to U.S. vs non-U.S. birthplace, and country or area where educated. Unlike others, the authors fail to find more recently graduated physicians to have higher cesarean section rates. Indeed, their rates for dystocia are significantly lower, when other factors are controlled for. Women physicians have a slightly lower cesarean section rate overall when other factors are controlled for, but a higher rate for dystocia. Board certified obstetricians have significantly higher, and physicians with professorial appointments lower cesarean section rates, though the direction of causation is not always clear. The authors conclude that the effect of physician characteristics on clinical behavior is a fruitful line of inquiry, and that in particular the results pertaining to FMGs warrant further investigation.