Predictors of Clinical Outcomes in the Laparoscopic Management of Adnexal Masses

Abstract

This paper presents the Duke University Medical Center's experience with 396 patients who underwent laparoscopic surgical evaluation of a pelvic mass between 1994 and 2001. From a review of the medical records of these patients, 7 different patient or treatment characteristics were identified and analyzed for their ability to predict 6 different clinical outcomes. The predictive variables included mass size, hysterectomy status, prior abdominal surgery, concurrent serious medical condition, age at surgery, training of surgeon (gynecologic oncologist or other gynecologist), and date of surgery. These variables were evaluated for their association with estimated blood loss, length of hospital stay, perioperative complications, rupture of cyst other than intentional draining, conversion to laparotomy, and length of operating room time. Patients were 12 to 87 years of age with a median age of 43 years. Thirty-seven percent were postmenopausal. The incidence of prior abdominal surgery ranged from 0 to 5 (median, 1) of which 21.2% were hysterectomy. A further 18% had a concurrent hysterectomy. Median size of the pelvic mass was 5.2 cm (range, 0.5-17). At the time of operation, 21.5% of women had hypertension, 2.2% had diabetes, 2.9% had cardiac disease, and 4.4% had pulmonary disease. Sixty-one percent had no confounding medical conditions. The operating surgeon was a general obstetrician-gynecologist in 54% and a gynecologic oncologist in 46% of cases. The most common pathologic diagnoses were ovarian cystadenoma (17.2%), endometrioma (16.2%), functional cyst (13.2%), mature teratoma (12.2%), and simple cyst (11.9%). Five women (1.3%) had invasive ovarian cancer, 4 (1.0%) had borderline ovarian tumors, and 1 was diagnosed with invasive fallopian tube cancer. Three of the patients with borderline disease were converted to laparotomy for additional staging after a positive frozen section. All of the women with cancer were converted to laparotomy, 3 before removal of the adnexae and 3 after frozen section evaluation of the adnexal mass. Perioperative complications occurred in 8% of cases. Seven patients (1.8%) had a wound infection, 3 (0.8%) had urinary retention, 5 (1.3%) developed a partial small bowel obstruction or prolonged ileus, and 7 (1.8%) required transfusion for blood loss. One or 2 patients each had perioperative complications, including cystectomy, bowel injury, uterine perforation, nerve injury, or subcutaneous emphysema. In addition, 5 patients underwent reexploration for treatment of infected trocar-site hematoma, vaginal cuff abscess, unrecognized bowel injury, hernia repair, and hemorrhage from ovarian vessels. The estimated blood loss was greater than 500 mL in 18 women, occurring in 21% (14 of 63) of patients undergoing concurrent hysterectomy and 1 % (4 of 317) of the remaining women. Concurrent hysterectomy was the only predictive variable associated with blood loss greater than 500 mL (18 women) (P <.001). Concurrent hysterectomy was also associated with length of hospital stay, (P <.001), as was prior hysterectomy (P <.001), mass size (P =.01 for 1 cm increase), prior abdominal surgery, (P =.009), and coexisting serious medical conditions (P =.007). Concurrent hysterectomy and mass size were predictive of perioperative complications (P =.01 for both). In addition, concurrent hysterectomy emerged as a predictive variable for rupture of the adnexal mass (P =.003).