Study design:
A parallel randomized controlled pilot study conducted at Ain Shams University Maternity Hospital in the period between January 2017 and April 2018, in which fifty-two cases were recruited from those attending the outpatient gynecologic clinic, who were candidates for myomectomy.
Sample size calculation:
Sample size was calculated using the Power & Sample Size Calculator®, setting the power (1-b) at 0.8 and the type-1 error (a) at 0.05. The primary outcome of the current study is the mean intraoperative estimated blood loss (EBL). Reviewing the literature revealed no direct comparison between vertical and transverse uterine incision for abdominal myomectomy. A trial comparing the two types of incision in laparoscopic myomectomy was, however, found. This latter trial showed that the mean values for intraoperative EBL in transverse versus vertical uterine incision in laparoscopic myomectomy were 110.5 ± 81.7 ml and 136.4 ± 108.5 ml, respectively [9]. Therefore, transverse uterine incision would be assumed to reduce the mean intraoperative EBL by almost 18.9%. A previous study showed that the average blood loss during conventional abdominal myomectomy (which utilizes vertical uterine incision) was 621±121 ml [10]. The mean intraoperative EBL with transverse uterine incision would, therefore, be assumed to be 503 ml. Calculation according to these values, produces a minimal sample size of 17 women in each group. Considering an attrition rate of 35% (to compensate for drop-outs due to the limited surgical experience with the novel technique in difficult myomectomies), a total of 52 women were recruited.
Recruitment:
Criteria for inclusion in the study were as follows; single myoma measuring 5-10 cm in diameter, causing uterine pressure-related symptoms, infertility or recurrent pregnancy loss in women between 20 and 35 years of age, with BMI between 18.5-29.9 kg/m2. Candidates were excluded from the study if they had any of the following criteria; pregnancy, cases with bleeding tendency, prior laparotomies, cases having coexisting pelvic pathologies, as ovarian cysts. All candidates had a complete preoperative evaluation via clinical history taking, clinical examination and sonographic examination for confirmation of the exact site, size and number of uterine fibroids and to exclude any coexisting pelvic pathology, a venous blood sample for blood picture, liver and kidney functions as well as coagulation profile, as a part of anesthetic workup.
Convenience sampling was utilized. The process of recruitment and handling the study population during the course of the study is shown in the CONSORT flow diagram. Candidates were randomly assigned (with 1:1 allocation ratio) to one of two groups according to a computer generated sequence, distributed in sequentially numbered sealed opaque envelopes, allocated by the seventh author. Only the research candidates were blinded to the surgical technique.
Intervention:
All candidates received general anesthesia. Research group A included 26 women who were planned for transverse uterine incision in which a transverse elliptical uterine incision is accustomed within the false capsule in a manner that its transverse diameter is almost centralized over the myoma with a slight anterior shift. Its length is 2 cm shorter than the maximum diameter of the myoma and ends well away from the tubal ostia. Breadth of the ellipse is fashioned so that its edges wouldn't be under tension or redundant after closure, usually the breadth equals one third of its length [Figure 1]. Tissue planes are bluntly and/or sharply dissected and the myoma is resected with the overlying ellipse of myometrium and false capsule leaving a relatively smaller zone of dead space for closure [Figure 2]. Myometrium is re-approximated in two layers using absorbable suture [Figure 3]. The uterine serosa is sutured in a running fashion, exposing minimal suture material and raw area to minimize adhesion formation [Figure 4]. Research group B included 26 women who were planned for standard longitudinal uterine incision.[4] No vasopressin or any other hemostatics were used to avoid their confounding effect on blood loss. No peritoneal toilet was done and intra-peritoneal suction drains were left in all cases. Blood loss during the operation was estimated as follows: obtaining the volume difference of blood in suction bottle containers (in ml), obtaining weight difference of linen towels (in gm) [weight of soaked linen towels – weight of dry linen towels].[11] These differences were estimated from the beginning of the uterine incision till its closure. Estimated blood loss was also calculated by obtaining the difference between preoperative and 24 hours postoperative hematocrit using the following formula:
EBV x (Hi-Hf)
Hi
Where the Hi is the preoperative blood hematocrit, Hf is the postoperative one and EBV is the estimated blood volume. The estimated blood volume could be calculated by multiplying weight times average blood volume which is around 65ml/kg in females.[12]
After exclusion of three patients (as shown in the CONSORT flow diagram), a total of 25 women in the Transverse Uterine Incision group and 24 women in the Longitudinal Uterine Incision group were eligible for statistical analysis.
Postoperatively, all women were subjected to postoperative follow-up of vital data and suction drains for 24 hours, and received NSAIDs for analgesia.
Research Study Outcomes:
Intraoperative blood loss as primary outcome, operative time and postoperative fever as secondary outcomes.
Statistical analysis:
Conducted using SPSS for Windows version 20. Per protocol analysis was adopted to avoid dilution of treatment effect by excluded patients with surgically-difficult myomectomies.
Data was represented as mean+/- SD (95% CI) for continuous parametric variables and median (IQR) for non-parametric ones. Difference between two unrelated groups was analyzed using the independent Student’s t test as well as the mean difference [MD] and its 95% confidence interval [95% CI] (for numeric parametric variables); Mann-Whitney’s U test (for numeric non-parametric variables); paired t test (for paired numerical parametric variables); chi-squared test and Fischer exact test (for categorical variables). Significance level was set at 0.05.