Preview only show first 10 pages with watermark. For full document please download

The Learning Curve Of Total Laparoscopic Hysterectomy: Comparative Analysis Of 1647 Cases

The Learning Curve of Total Laparoscopic Hysterectomy: Comparative Analysis of 1647 Cases

   EMBED


Share

Transcript

  Laparoscopic hysterectomy is accepted as a safeand efficient way to manage benign uterine pathologyin selected patients, and is an acceptable alternativeto the standard abdominal hysterectomy. 1–8 Advan-tages to patients are well known. 1–4,6,9 To evaluate therole of laparoscopic hysterectomy and its rate of com-plications, the technique should be established, thelearning curve should reach a plateau, and methods of avoiding complications should be in place. Numerousreports attest to the feasibility of the procedure, and a339 August 2002, Vol. 9, No. 3The Journal of the American Association of Gynecologic Laparoscopists From the Department of Obstetrics and Gynecology and Reproductive Medicine, Polyclinique de L’Hôtel-Dieu CHU, Clermont-Ferrand, France (allauthors).Address reprint requests to A. Wattiez, M.D., Department of Obstetrics and Gynecology, Polyclinique de L’Hotel-Dieu CHU, 13 Boulevard Charles deGaulle, Clermont-Ferrand, France; fax 33 4 73931706.Accepted for publication February 28, 2002. Abstract Study Objective. To compare the frequency of complications of total laparoscopic hysterectomy performed in the first and more recent years of our experience, and based on that, offer ways to prevent them. Design. Retrospective, comparative study (Canadian Task Force classification II-2). Setting. University tertiary referral center for endoscopic surgery. Patients. During 1989–1995 and 1996–1999, 695 and 952 women, respectively, with benign pathology. Intervention. Total laparoscopic hysterectomy. Measurements and Main Results. No differences in patient characteristics were found between 1989–1995 and 1996–1999. Substantial decreases in major complication rates were noted, 5.6% and 1.3%, respectively. No major vessel injury occurred. Excessive hemorrhage (1.9%) and need for blood transfusion (2.2%) during the first period were statistically higher than in the second period (both 0.1%, p <0.005). Urinary complications (2.2%) includ- ing 10 bladder lacerations, 4 ureter injuries, and 1 vesicovaginal fistula occurred more frequently in the first period than in the second period (0.9%), when 6 bladder and 2 ureter lacerations and 1 vesicovaginal fistula occurred (p <0.005). One bowel injury and one bowel obstruction occurred in the first period, but no bowel complications in the second. Between periods, 33 (4.7%) and 8 (1.4%) conversions to laparotomy were necessary. During the first period there were nine reoperations; of six laparotomies, four were due to urinary injuries, one due to heavy vaginal bleeding, and one due to a vesicovaginal fistula; three diagnostic laparoscopies were required due to post- operative abdominal pain. Three reoperations during the second period were two laparoscopies due to heavy vagi- nal bleeding and one laparotomy due to a vesicovaginal fistula (p <0.005). Statistically significant differences in median (range) uterine weight 179.5 g (22–904 g) and 292.0 g (40–980 g) and operating times 115 minutes (40–270 min) and 90 minutes (40–180 min), respectively, were recorded (p <0.005). Conclusion. Laparoscopic hysterectomy was safe, effective, and reproducible after training, and with current tech- nique, had a low rate of complications. (JAm Assoc Gynecol Laparosc 9(3):339–345, 2002)  The Learning Curve of Total Laparoscopic Hysterectomy: Comparative Analysis of 1647 Cases A. Wattiez, M.D., D. Soriano, M.D., S. B. Cohen, M.D., P. Nervo, M.D., M. Canis, M.D., R. Botchorishvili, M.D., G. Mage, M.D., J. L. Pouly, M.D., P. Mille, M.D., and M. A. Bruhat, M.D.  few randomized controlled trials compared it withabdominal hysterectomy. 4–7 Most of them, however,were done early after its introduction into clinicalpractice and had small samples. 1–7,10–11 Others excludedwomen with very large uteri and those with severepelvic adhesions. 11 Materials and Methods From November 1989 to December 1999, 1647women underwent total laparoscopic hysterectomy(TLH) due to benign pathology. At the beginning of 1996 we introduced a new uterine manipulator to pre-vent complications. Therefore, we compared compli-cations rates before and after its introduction. In early(1989–1995) and recent years (1996–1999), 695 and952 women underwent TLH, respectively. Inclusioncriteria were benign uterine disorders and patient con-sent. Exclusion criteria were anesthetic contraindi-cations for laparoscopic surgery and total uterineprolapse. Women with endometriosis, postinflamma-tory disorders, previous abdominal or pelvic surgery,obesity, or very large uterus were not excluded. Noupper limit of uterine size was set. All women under-went preoperative pelvic ultrasonography and, whenindicated, endometrial sampling.Patients gave consent after receiving a detailedexplanation, including risks associated with anes-thesia, abdominal hysterectomy, and laparoscopy.Approval was not obtained from the institute’s humaninvestigation review board since the study was not anexperimental protocol, but a summary of outcomes of routine management. Data regarding patient charac-teristics and immediate and long-term postoperativeoutcome were available from retrospective reviews of hospital and outpatient records. Operative Procedure All laparoscopic procedures were done under endotracheal general anesthesia in modified dorso-lithotomy position. All patients received prophylacticcefazoline 2 g intravenously at the beginning of sur-gery and prophylactic anticoagulant therapy with low-molecular-weight heparin the evening before operationand continued for 14 days. They were continuouslymonitored for blood pressure, electrocardiogram, trans-cutaneous oxygen saturation, and end-tidal carbondioxide (CO2) pressure.AVeress needle was inserted through the umbili-cus and the abdomen was insufflated with CO2. After induction of pneumoperitoneum and insertion of thevideolaparoscope, three suprapubic cannulas wereintroduced for ancillary instruments (all equipment byKarl Stortz, Tuttlingen, Germany).After introducing the optic fiber, the pelvis andaccess to uterine pedicles were evaluated. In cases of no access, the laparoscopic procedure was stopped.When the laparoscopic phase was converted to lapa-rotomy, it was defined as laparoconversion of TLH.For coagulation and division of the left round lig-ament, the first assistant grasped the ligament whereit leaves the horn and pulled it toward the right andup. This revealed a triangle bordered by the round lig-ament at the top, iliac vessels laterally, and adnexalvein medially. By placing the round ligament under tension, the center of this triangle was revealed, madeup of the juxtaposition of two anterior and posterior peritoneal layers of broad ligament, the point at whichfenestration was performed later. The round ligamentwas divided in its middle by intermittent coagulationand section.Peritoneal capillaries were meticulously and pro-gressively coagulated, and peritoneum divided withcold scissors or by coagulating and cutting withmonopolar current. This dissection was stopped about1 cm short of midline.The right round ligament was coagulated. Theprevesical space was opened with bladder dissection.Peritoneal traction upward created a peritoneal falx,which was coagulated and sectioned. Section wasmade perpendicular to the uterus to avoid bladder injury. This action combined with pushing on theuterus allowed the vesicovaginal dissection plane tobe entered. Once the plane was broached, dissectionof the bladder continued downward by breaking tis-sues down.Internal pillars of the bladder were coagulatedand sectioned up to uterine pedicles. The two insidepillars, held under tension by pulling upward on thebladder, were coagulated and divided.Broad ligament capillaries were carefully coagu-lated and revealed the posterior layer of the broad lig-ament. This should appear gray, meaning that thereis no bowel behind it. The layer of peritoneum wasopened by breaking it down or by cutting. For hys-terectomy without adnexectomy, the opening wascontinued toward the uterosacral ligaments. For hysterectomy with adnexectomy it continued in thedirection of the infundibulopelvic (IP) ligament. Theleft IPligament (in cases of adnexectomy) or ovarian340 Learning Curve of Total Laparoscopic HysterectomyWattiez et al  ligament was opened by coagulation and dissection or by ligation. Coagulation and section of internal blad-der pillars and fenestration of the left broad ligamentwere performed on the left side.Posterior peritoneum and uterosacral and cardinalligaments were coagulated and dissected bilaterally,followed by coagulation and dissection of uterinevessels bilaterally. Coagulation was applied to theascending branch of the uterine artery after generalcoagulation at the superficial level. Then veins of theperiarterial uterine plexus were coagulated and theartery itself was readily visible and ready for coagu-lation and dissection. Once the artery was dissected,the surgeon carried on dissecting forward and back-ward to allow the pedicle to drop down below theedges of the vaginal fornix. Intrafascial dissection of the vagina was performed, and the vagina was openedthrough 360 degrees with pure monopolar cutting cur-rent by hook or spatula.In cases of very large uterus, hemisuction wasperformed with a laparoscopic cold knife. The uteruswas extracted through the vagina.The vagina was closed laparoscopically. Asurgi-cal glove filled with sponges was placed in the vaginato ensure pneumostasis for closure. The suture was 0or 1 polyglactin swedged to a curved 30-mm needle.The vagina was totally transfixed to achieve completehemostasis. After meticulous hemostasis, the integrityand mobility of the ureters were checked. In case of doubt, cystoscopy was performed. Afinal wash com-pleted the procedure. Bleeding from cut edges of thevagina was coagulated as necessary.Statistical analysis was performed with Student’s t test or χ 2test. Probability below 0.05 was consid-ered statistically significant. Results No differences were found between time periodsfor patient age, parity, body mass index (BMI), pre-menopausal status, and previous abdominal or pelvicsurgery (Table 1). There was no difference in indica-tions for surgery between periods. The most commonpathologic diagnoses in both periods were myoma,341 August 2002, Vol. 9, No. 3The Journal of the American Association of Gynecologic Laparoscopists TABLE 1. Patient Characteristics 1989–19951996–1999Characteristic(n = 695)(n = 952)Median (range) age (yrs)47 (31–75)50 (38–74)Median (range) parity2 (0–8)2.0 (0–7)Vaginal delivery (%)85.686.5Cesarean delivery (%)7.58.3Premenopause (%)84.780.9Previous abdominal surgery (%)Appendectomy45.336.0Laparotomy19.132.3Pelvic surgery22.626.9Median (range) BMI23.4 (17.6–33.7)23.8 (17.7–38.9)Indication for surgery a (%)Menorrhagia, metrorrhagia74.570.0Uterine leiomyoma29.244.3Dysmenorrhea, pelvic pain29.435.0Endometrial hyperplasia5.36.5Adnexal mass7.35.2Endometriosis7.94.2CIN2.21.2Total hysterectomy+ adnexectomy (%)54.348.2+ adhesiolysis (%)15.719.3BMI = body mass index; CIN = cervical intraepithelial neoplasia. a Most patients had more than one indication.  adenomyosis, and endometrial atrophy, with no sig-nificant differences between groups.Significant decreases in major complications(blood transfusion; urinary, bowel and neurologicinjuries; thromboembolism) occurred between earlyand late periods, 5.6% and 1.3%, respectively (p<0.005; Table 2). The most common sites of intra-operative bleeding were left uterine vessels, particu-larly uterine vein, back-flow from uterine horns andvessels, and cervicovaginal branches of the uterineartery.Overall there was a significant decrease in urinaryinjuries between the first and second periods, 2.3% and0.9%, respectively (p <0.005). During the first period11 (1.6%) bladder lacerations occurred. All but onewere diagnosed intraoperatively, and all but one weretreated laparoscopically with excellent outcomes.In the second period, six (0.6) bladder lacerationsoccurred, all treated laparoscopically (p <0.005). Sixwomen with bladder injury had a history of cesareandelivery and all of these injuries happened duringopening of the uterovesical fold in order to push thebladder away from the vagina. Obesity, previous pelvicsurgery, abdominal adhesions, and uterine size werenot risk factors for bladder injury.There were four (0.6%) ureteral injuries during theearly period. Only one was diagnosed intraoperatively,with no need for conversion to laparotomy. However,three of these injuries were diagnosed late and requiredlaparotomy. Two ureters (0.2%) that were laceratedduring the second period were managed intraopera-tively with no need for laparotomy. All but one of these injuries occurred on the right side at the termi-nal part of the ureter, and all but two were caused bythermal effect. Excessive bleeding (3 women), endo-metriosis (2), and very large uterus (1) were risk fac-tors for ureteral injury.Vesicovaginal fistula and bowel injury, one each,were diagnosed a week after operation during the firstperiod and were treated by laparotomy. One vesico-vaginal fistula was diagnosed 2 weeks postoperativelyin the second period, and ureteral reimplantation bylaparotomy was performed.The frequency of conversion to laparotomy was4.7% in the first period versus 1.4% in the secondperiod (p <0.005). No case of major vessel injuryoccurred during either period. There was a statisticallysignificant difference in the rate of conversions due tohemorrhage between periods: 13 cases (1.9%) versus2 (0.2%), respectively (p <0.005; Table 3). Technicalproblems that required conversion mainly concernedlimited exposure (severe adhesions, severe endome-triosis, very large myoma).During the first period nine (1.3%) reoperationswere necessary. Six laparotomies were performed,four due to urinary tract injuries, one due to heavy342 Learning Curve of Total Laparoscopic HysterectomyWattiez et al TABLE 2. Complications of Laparoscopic Hysterectomy 1989–19951996–1999Complication(n = 695)(n = 952)Excessive hemorrhage13 (1.9)1 (0.1) a Blood transfusion15 (2.2)1 (0.1) a Major vessel injury00Urinary tract injury16 (2.3)9 (0.9) a Bladder laceration:11 (1.6)6 (0.6)Ureter injury4 (0.6)2 (0.2)Vesicovaginal fistula1 (0.1)1 (0.1)Bowel injury1 (0.1)0Bowel obstruction1(0.1)0Neurologic injury4 (0.6)0Thromboembolism2(0.3)2 (0.2)Reoperation9 (1.3)3 (0.3) a Laparotomy60Laparoscopy33Abdominal wall hematoma105Vaginal cuff hematoma34Pyrexia144Vaginal cuff infection40Abdominal wall infection20Values are number (%). a p <0.005. TABLE 3. Conversions to Laparotomy 1989–19951996–1999Reason for Conversion(n = 695)(n = 952)Excessive hemorrhage13 (1.9)2 (0.2) a Anesthetic problems2 (0.3)0Emphysema4 (0.6)2 (0.2)Bladder laceration:1 (0.1)0Bowel injury1 (0.1)0Technical difficulties12 (1.7)9 (0.9)All laparoconversions33 (4.7)13 (1.4) a Conversion to LAVH29 (4.2)4 (0.4) a LAVH = Laparoscopic-assisted vaginal hysterectomy.Values are number (%). a p <0.005.  vaginal bleeding, and one due to a vesicovaginal fis-tula. In addition, three diagnostic laparoscopies wererequired for postoperative abdominal pain. Three(0.3%) reoperations were performed during the sec-ond period (p <0.005), one laparotomy due to vesico-vaginal fistula and two laparoscopies due to suspectedpostoperative intraabdominal hemorrhageFour neurologic complications were noted in thefirst period, compared with none in the second period.Clinical deep vein thrombosis was diagnosed byDoppler ultrasonography in two women in each perioddespite prophylactic anticoagulant treatment. No caseof pulmonary embolism occurred.Overall we noted a decrease in total complicationrates from 10.4% to 2.6% in the first and second peri-ods (p <0.005). Laparoscopic hysterectomy increasedfrom 68.0% of all nonvaginal hysterectomies per-formed in 1989–1995 to 94.4% performed from 1996– 1999 (p <0.005; Table 4). The median hospital staywas 4.3 days 3–11 and 3.2 days, respectively. 3–7 Discussion The present study, to our knowledge the largestreported to date from a single center, supports the fea-sibility and overall favorable outcome associated withTLH. Previous reports indicated that even experi-enced laparoscopists have complication rates of 5.8%to 11.5% and major complication rates of 2.2% to2.7% after laparoscopic hysterectomy for the man-agement of benign uterine pathology. 12,13 One meta-analysis reported a total complication rate of 15.6%, 14 which was comparable with the Finnish national reg-istry of laparoscopic hysterectomies. 15 We managed todecrease our total and major complication rates from10.4% and 5.6% to 2.6% and 1.3% in the first and sec-ond periods, respectively.Major vascular injuries are life-threatening com-plications, including perforations of the aorta, venacava, and iliac vessels. 15, 16 No major vascular injuryoccurred in our study. One of the most common major complications is bleeding and it is a main reason for laparoconversion. Bleeding rates of 1.9% of our earlyperiod and 0.1% in the recent period, compared withthe rate in another study in which 3.8% of patients hadexcessive bleeding and required blood transfusion. 15 The lower figures in our study were achieved by iden-tifying the cause and sites of bleeding. The round lig-ament should be coagulated in its middle part far fromthe uterine horn. The IPligament should be coagulatedonly after fenestration of the broad ligament, whichreveals 3 to 4 cm of free ligament for meticulous coag-ulation. The most common site of bleeding was theuterine pedicle. To prevent bleeding complications,this pedicle should be prepared properly.Amain difference between abdominal and TLHis in dissection of uterine vessels. During laparotomy,all tissues around the uterine vessels are grasped witha Kocher grasper and cut. During laparoscopic surgeryit is very important to dissect all tissues until one canclearly visualize the vessels. Only then can bipolar coagulation be effectively applied.The main cause of urinary tract injuries is gyne-cologic surgery, particularly hysterectomy. In the lit-erature the frequencies of urinary tract injury after laparoscopic hysterectomy were 2.7%, 15 2.8% in 318laparoscopic hysterectomies, 17 and 1.4% in the meta-analysis. 14 During the first period of our study our ratewas similar, 2.3%. However, in the second period thefrequency was reduced significantly to 0.9%, which343 August 2002, Vol. 9, No. 3The Journal of the American Association of Gynecologic Laparoscopists TABLE 4. Major Complications, Percentage of TLH of all Hysterectomies, Uterine Weights, Operating Times, and Numberof Surgeons 1989–19951996–1999Characteristic(n = 695)(n = 952)All complications (%)10.42.6Major complications a (%)5.61.3 b Total vaginal vs abdominal hysterectomies6894 b Mean (range) uterine weight (g)179.5 (22–904)292.0 (40–980) b Mean (range) surgery (min)115 (40–270)90.0 (40–180) b Number of surgeons918 a Blood transfusion, urinary, bowel, neurologic injuries, or thromboembolism. b p <0.005.