Use of Uterine Characteristics to Improve Fertility-Sparing Diagnosis of Adenomyosis

To describe patient demographics, determine accuracy of clinical diagnosis, and evaluate reliability of laparoscopic uterine characteristics in the diagnosis of adenomyosis. Enrollment included 117 patients undergoing laparoscopic hysterectomy for benign indications. Intraoperatively, the attending surgeon predicted uterine weight; evaluated the presence of fibroids; and commented on the uterus' shape, color, and consistency while probing it with a blunt instrument. A prediction was also made about whether final pathology would reveal adenomyosis. Standardized video recordings were obtained at the start of the case. Each video was viewed retrospectively twice by three expert surgeons in a blinded fashion. Uterine characteristics were reported again with a prediction of whether or not there would be a pathologic diagnosis of adenomyosis. These data were used to calculate inter-and intrarater reliability of diagnosis. Women with adenomyosis were more likely to complain of midline pain as opposed to lateral or diffuse pain (p = 0.048) with no difference in the timing of the pain (p = 0.404), compared to patients without adenomyosis. Uterine tenderness on examination was not an accurate predictor of adenomyosis (p = 0.566). Preoperative diagnosis of adenomyosis by clinicians was poor, with an accuracy rate of 51.7%. None of the intraoperative uterine characteristics were significant for predicting adenomyosis on final pathology, nor was any combination of the features (p = 0.546). Retrospective video reviews failed to reveal any uterine characteristics that generated consistent inter- or intrarater reliability (Krippendorff's α < 0.7) in making the diagnosis of adenomyosis. Clinical and video diagnosis of adenomyosis have low accuracy with no uterine characteristics consistently or reliably predicting adenomyosis on final pathology. (J GYNECOL SURG 34:183).