PREDICTING UTERINE WEIGHT BEFORE HYSTERECTOMY: ULTRASOUND MEASUREMNENTS VERSUS CLINICAL ASSESSMENT

T.S. Harb, R.A. Adam

Atlanta, GA

OBJECTIVE: To determine the most accurate method in estimating the preoperative uterine weight of enlarged nongravid uteri.

MATERIALS AND METHODS: We performed a retrospective review of 1238 patients who were premenopausal and underwent hysterectomy for benign indications between January 1993 and July 1999. 864 patients were selected to include only those that had both a reported bimanual assessment of preoperative uterine size and an ultrasonography report with all 3 estimated uterine dimensions. The number was further reduced to 718 to only include patients with a clinical uterine size between 9 and 20 weeks of estimated gestational age to allow comparison with published reports. Reported uterine sizes on bimanual examination were converted to clinical weight (CWT).  Two different calculations were used to estimate uterine weight from ultrasound measurements, utilizing a previously published formula and its commonly used simplification (UWT 1 & 2). Actual uterine weights (AWT) in pathology reports were then compared to the findings of bimanual assessment and the calculated weights to determine which method is the better predictor of AWT. Simple linear regression analysis and analysis of variance were used to measure and compare how closely the estimated weights, by clinical exam and ultrasound, predicted the actual weight. Predictive residual sum of squares (PRESS) statistic was then used to determine the best predictor of actual weight.

RESULTS: After exploring the data using linear modeling, all 3 estimated weights were significantly correlated to the actual weight when compared. UWT1 had a correlation coefficient r=.72 and a P-value < .0001. The second estimated weight UWT2 also had a significant correlation when compared to AWT (r=.71, P<.0001). Clinical weight, CWT, yielded a comparable significant correlation coefficient (r=.76, P<.0001). PRESS statistic, which looks at the effectiveness of a predictive model for predicting unobserved results, showed that the clinical weight estimate was superior by far compared to the other two. 

CONCLUSION: In this study, bimanual assessment was shown to be the most accurate method of preoperative uterine weight estimation. Since uterine size estimation often influences the decision for the route of hysterectomy, ultrasound examination for this purpose may not be routinely needed when deciding the route of hysterectomy.

Key Words: uterine weight, ultrasound measurements

Disclosure – Nothing to disclose.