Estimation of Uterine Size: How Accurate Are We?

To evaluate the accuracy of gynecologic surgeons at estimating uterine dimensions and weight.

Six model uteri of various sizes were created to simulate the size and consistency of a uterus and displayed at 3 stations. The visual station (VS) comprised 2 specimens placed on an unmarked table. The laparoscopic station (LS) consisted of 2 model uteri, each placed in a separate simulated abdomen with a 0 degree laparoscope and 2 operative trocars with standard instruments. The blind weight station (BWS) consisted of blind palpation of 2 separately weighted models (heavy model [HM] and light model [LM]). Participants visually estimated the dimensions of each VS and LS models and blindly palpated the BWS models to estimate weight.

Participants included 15 residents, 27 attendings, and 6 medical students. There was no difference in estimation accuracy regarding gender and age. For the VS and LS groups, participants underestimated all dimensions (VS variance = -15.0%; P < 0.001 and LS variance = -31.9%; P < 0.0001). Laparoscopic estimation was less accurate than direct vision (P < 0.0001). Attendings and residents equally underestimated the 3 dimensions visually (P = 0.46), but attendings were more accurate at estimating laparoscopic dimensions (-25.8% vs -41.1%; P = 0.0001). All groups overestimated model weights (HM variance, 92.5%; P < 0.001 and LM variance, 132.0%; P < 0.0001), with attendings more accurate than residents (39.7% vs 167.6%; P = 0.015 for HM and 52.0% vs 238.5%; P = 0.035 for LM).

Gynecologic surgeons at all levels of training are inaccurate at estimating dimensions and weights. With surgical decisions often predicated on estimates, education is needed to improve estimation methods.