The commonest complication associated with vaginal hysterectomy is vault hematoma (1), which has been detected by ultrasound in 34–59% cases postoperatively (1,2). This has been shown to increase febrile morbidity, need for blood transfusions, longer hospital stay, and higher readmission rate (2). We therefore designed a randomized, controlled trial to determine whether nonclosure of the vaginal vault reduced the risk of febrile morbidity. Local ethics committee approval was obtained, and written informed consent was taken from all patients. Our primary outcome measure was to examine a 75% reduction in febrile morbidity (pyrexia) rate, i.e. febrile morbidity rate of 7% in the vault open compared to 31% in the closed group (2). At 80% power at a significance level of 0·05 and allowing for a 20% loss to follow-up, 120 women were required. Randomization was done at the time of the operation by opening a sealed opaque envelope sequentially numbered 1–120. No mention of the vault closure was made in the operative notes. In the open group, the vaginal vault edges and the peritoneal edges were overseen with a continuous interlocking vicryl suture, incorporating the uterosacral ligaments into the vault. This achieved hemostasis at the vault edges but allowed a drain site for any potential vault collection. In the closed group, the vault edges were opposed with interrupted vicryl sutures to completely close the vault. A single standard prophylactic antibiotic dose and antithrombotic agents (stockings and heparin) was given to all patients. Postoperative management was standard in order to reduce bias. If the patient was pyrexial after the first postoperative day, a midstream specimen of urine and a transvaginal pelvic ultrasound scan was performed. Patients were reviewed at outpatient's clinic after 6 weeks to determine whether the vault was closed or not. Operative time, blood loss, and length of hospital stay were compared with the Mann–Whitney U-test. Twenty-three women were recruited in the open vault group and 27 women in the closed vault group. There were no differences between the patient baseline characteristics. Table I shows the surgical outcomes. There were no significant differences in any of the other surgical outcomes, but this was probably due to the small sample size. The study was prematurely stopped due to two serious complications in the open group. One woman had loops of small bowel prolapse through the vault and required a laparotomy at day 4 in order to check the integrity and viability of the bowel. No gangrene of the bowel was noted. The vaginal vault was closed, and she made an uneventful postoperative recovery. Another patient in the open group had symptomatic tubal prolapse and needed laparoscopic salpingectomy 4 months later. Our earlier nonrandomized pilot study established the feasibility and potential efficacy of leaving the vaginal vault open at hysterectomy in reducing febrile morbidity (3). This pilot study did not show any increased morbidity with this surgical technique (3). In contrast, two cases of organ prolapse occurred in the experimental arm (open vaginal vault group) of this current randomized trial, and these complications were considered serious enough to warrant premature closure of the trial. Consequently, the trial lacks the necessary power to answer our primary research question about febrile morbidity or allow any meaningful conclusions to be drawn regarding other surgical outcomes. However, we believe that this paper is important, because it demonstrates the need for randomized, controlled trials to determine the effectiveness of therapies, to rigorously identify and assess adverse events and in those circumstances our obligation as clinical trialists to stop trials, and to report both good and bad outcomes in order to avoid publication bias. We thank Mr Eddie O'Donnell, Mr Khalid Khan who helped recruit some patients in the trial, Mr John Pogmore for advice regarding trial continuation and Sister Tracy Bingham for follow-up of patients. Conflicts of interest: none declared.
Search from 250 M+ research papersSearch from 250 M+ research papers
Apr 25, 2023
Kai Su + 5
Open Access