Can differences in labor induction rates be explained by case mix?

To determine whether differences in case mix account for variations in labor induction rates between hospitals and whether case mix adjustment is helpful in evaluating hospital induction rates.

Using a New York State birth certificate database, factors associated with induction of labor were entered into a logistic regression to calculate labor induction probabilities. The probabilities were used to calculate expected induction rates at each of 16 regional hospitals. Each hospital's observed and expected induction rates were compared to determine the degree of residual variation.

Excluding women who delivered by cesarean section without labor, the regional labor induction rate was 20.8%, with a 3.9-fold difference between the highest and lowest hospital rates (P < .0001). Adjusting for risk factors explained only 12.6% of the variation in induction rates. Risk adjustment to compare labor induction rates had a minimal effect on rank ordering of rates but made clearer some hospitals' rates.

Differences in case mix do not explain labor induction rate variation, but adjustment for case mix may aid in interpreting rates at individual hospitals.