Randomization

Suitable patients were randomized to have their surgical specimen removed either TV or through the umbilical port site (TU) using a computer-generated randomization list. This list was held centrally by a trial administrator. Randomization was carried out by telephone. The patient was assigned by the principal investigator to treatment group TU or TV according to the randomization. Once allocated, the treatment was revealed to the surgeon immediately prior to starting the procedure, while group allocation was concealed from patients and bedside clinicians. An opaque sterile dressing was applied postoperatively over the skin wounds to prevent women from looking at their incisions size.

Operative technique

Operative laparoscopy was performed under general anesthesia in all women. Standard anesthesia and perioperative care protocols were used. After the pneumoperitoneum was created (using a Veress needle), a 0-degree 5-mm laparoscope (Karl Storz, Tuttlingen, Germany) was introduced through the umbilicus. Two or three 3-mm ancillary trocars were inserted under direct vision in the lower abdomen. One 3-mm trocar was always inserted in the midline approximately 3-6 cm above the symphysis. The other trocars were inserted under laparoscopic vision laterally to the lateral umbilical ligament. Instrumentation included graspers, scissors, monopolar electrocautery, a bipolar PK System MoLly Forceps (Gyrus Medical Inc, Maple Grove, MN), and suction–irrigation. Pelvic washing for cytological examination was performed whenever needed. Tissue evacuation was always performed using specimen retrieval bags (EndoCatch II, Tyco Healthcare Group LP, Norwalk, CT). Preemptive infiltration of trocar sites or colpotomy site with a local anesthetic was never performed.

TU specimen removal

When the surgical specimen was freed, the laparoscope was then withdrawn from the umbilical trocar. The 5-mm umbilical port was replaced with a 10-mm port to allow insertion of the specimen bag. A specimen retrieval system was then disbanded to split the specimen pouch from the dispensing tube but not from the 40-cm long thread. To facilitate introduction of the retrieval bag into the peritoneal cavity, the bag was rolled up and with the help of a grasper was blindly introduced into the abdominal cavity through the umbilical port, leaving the thread of the specimen pouch protruding from the trocar, with the free end held outside the abdomen. The laparoscope was then reintroduced alongside the thread. Once in the peritoneal cavity, the bag was opened with laparoscopic atraumatic graspers and the surgical specimen was inserted under direct vision. Then, the free end of the thread was withdrawn pulling the bag into the trocar, and this was then removed together with the umbilical trocar. If the surgical specimen was too large to pass through the umbilical incision, or solid components were encountered, the mouth of the sac was brought out with the help of atraumatic graspers, and the specimen was carefully morcellated using Kocher clamps and scissors inside the bag to avoid intraabdominal spillage or loss of surgical debris. Even in the presence of large adnexal masses, gentle traction on the bag during the process allowed serial exteriorization of the bag, keeping the cyst components at the incision. In case of large cystic tumors, puncture of the mass and aspiration were performed within the retrieval bag. In the worst-case scenario the port site was expanded for specimen removal. During this procedure, an ancillary trocar served as the gas delivery port to avoid loss of the pneumoperitoneum. Umbilical incisions were closed with a short-term synthetic absorbable suture, while 3-mm wounds were approximated with adhesive skin closure strips.

TV specimen removal

The specimen retrieval device is introduced in the vagina and is gently pushed against the vaginal wall to define the posterior fornix between the uterosacral ligaments. A 1-cm length transverse TV posterior colpotomy was performed under laparoscopic control using a 3-mm monopolar hook set at 60W.

Grasping the bag orifice with ring forceps through the colpotomy, the specimen was then pulled into the vagina. The bag orifice was opened inside the vaginal canal and the specimen was delivered through the vagina. The vaginal opening was irrigated with a povidone-iodine solution. The colpotomy was closed TV with a running 2-0 synthetic medium-term reabsorbable braided and coated suture.

Operative times were recorded from first incision to last suture. Time for specimen retrieval was calculated from TU bag insertion to TU bag removal (TU group) and from insertion of the specimen retrieval device into the vagina to the end of colpotomy closure (TV group).

Postoperative evaluation

Postoperative pain was managed with 1 g of intravenous paracetamol every 8 hours to a maximum of 3 doses (the first dose was given approximately 30 minutes before skin closure). Rescue analgesia (paracetamol 1 g intravenously or ketorolac 30 mg intravenously) was administered on patient request. The timing of analgesic requirement and the total amount of medication were recorded. Each patient was asked to record the severity of her incisional pain (defined as a “pain located at the wound site”) on a 10-cm visual analog scale (VAS), with 0 being no pain and 10 the worst pain imaginable, at 1, 3, and 24 hours after incision closure. At each time point, women were asked to indicate the location of the most painful area. The patients were interviewed for pain evaluation by research assistants blinded with respect to the treatment group. At our institution, patients undergoing laparoscopy for benign gynecologic conditions are usually observed in hospital overnight and discharged on postoperative day 1. Postoperative instruction included a specific recommendation to avoid sexual intercourse for 2 weeks following the procedure.

Follow-up included an appointment 8 weeks postoperatively, at which time women were interviewed to identify any complication that may have occurred after hospital discharge and were asked to rate their overall satisfaction with the surgical procedure, cosmetic appearance of their scars, and dyspareunia on resumption of sexual intercourse on separate VAS. If the patient did not attend the follow-up appointment, a follow-up interview was held by telephone call.

Sample size

The primary outcome measure in this trial was postoperative pain. Secondary outcome measures included need for rescue analgesia, postoperative dyspareunia, overall satisfaction score, and cosmetic outcome. Previously published data by our group indicated a mean ± SD 1-hour postoperative pain score among women undergoing retrieval of a laparoscopically resected adnexal mass through the umbilical port-site incision of 2.3 ± 2.1. Since the average pain level in this group is low and it has been consistently reported that a difference of 13 mm on a VAS represents, on average, the minimum change in acute pain that is clinically significant, we hypothesized that TV specimen retrieval should at least reduce 1-hour pain score by two thirds (from 2.3-0.8) to be considered a valuable alternative. On the basis of this analysis, 32 patients were required in each arm to detect this difference with a 5% level of significance and 80% power. To allow for a 10% rate due to exclusion for intraoperative findings, a total of 70 women were sought for the study.

Statistical analysis

Statistical analysis was performed with software (GraphPad, version 5; GraphPad Software, San Diego CA). Proportions in the 2 groups were compared using χ ² and odds ratio with 95% confidence interval. The t test and the Mann Whitney U test were used to compare continuous variables sampled from a gaussian or non-gaussian distribution, respectively. Statistical significance was assumed when P < .05.