Factors associated with hospital-specific cesarean birth rates

The variation in hospital-specific cesarean birth rates was examined using a multiple regression analysis in a national data base of deliveries in 1977. This data base included 222,285 singleton births in 282 hospitals representing all regions of the United States. The 32 independent variables included measures of medical risk, technologic sophistication, demographic characteristics and economic incentives. The incidences of medical risk factors, nonwhite race and ratio of obstetricians to fertile women were associated positively with hospital-specific cesarean delivery rates. Participation in health maintenance organizations and numbers of deliveries were correlated negatively with those rates. Hospital-specific cesarean delivery rates were not associated with technologic sophistication, malpractice premiums or reimbursement differences between routes of delivery. Pay source stratified models explained 45-63% of the variation in the hospital-specific cesarean rates. Sensitivity analysis revealed that even a 50% change in the incidence of any independent variable would change the cesarean delivery rate by less than 2%.