Indications for hysterectomy: variation within and across hospitals.

This study investigates the factors associated with the probability of finding specific clinical indicators (benign or malignant tumor, cancer in situ, fibroid, abscess/empyema, or positive culture of salpinx, fallopian tube, fetus, or uterus) that validate necessity for hysterectomy. Data for the 4,660 cases in the study come from 42 Pennsylvania hospitals. The probability that validating indicators were present varied significantly at the hospital level but not at the level of individual surgeons within hospital, suggesting that physicians in different hospitals adopted different practice styles. The results at the hospital level indicate that higher hysterectomy volume increased the probability of validating findings, whereas presence of an OB/GYN program was associated with lower probability of validating findings. The policy and management implications of these results are discussed.