{"paper_id":"24c433cb-635e-40b9-b563-71fa459e080c","body_text":"ORIGINAL ARTICLE\nComparison of minimally invasive surgical approaches\nfor hysterectomy at a community hospital: robotic-assisted\nlaparoscopic hysterectomy, laparoscopic-assisted vaginal\nhysterectomy and laparoscopic supracervical hysterectomy\nBang N. Giep • Hoang N. Giep • Helen B. Hubert\nReceived: 27 April 2010 / Accepted: 29 June 2010 / Published online: 10 August 2010\n/C211The Author(s) 2010. This article is published with open access at Springerlink.com\nAbstract The study reported here compares outcomes of\nthree approaches to minimally invasive hysterectomy for\nbenign indications, namely, robotic-assisted laparoscopic\n(RALH), laparoscopic-assisted vaginal (LAVH) and lapa-\nroscopic supracervical (LSH) hysterectomy. The total\npatient cohort comprised the first 237 patients undergoing\nrobotic surgeries at our hospital between August 2007 and\nJune 2009; the last 100 patients undergoing LAVH by the\nsame surgeons between July 2006 and February 2008 and\n165 patients undergoing LAVHs performed by nine sur-\ngeons between January 2008 and June 2009; 87 patients\nundergoing LSH by the same nine surgeons between\nJanuary 2008 and June 2009. Among the RALH patients\nwere cases of greater complexity: (1) higher prevalence of\nprior abdominopelvic surgery than that found among\nLAVH patients; (2) an increased number of procedures for\nendometriosis and pelvic reconstruction. Uterine weights\nalso were greater in RALH patients [207.4 vs. 149.6\n(LAVH; P\\ 0.001) and 141.1 g (LSH; P = 0.005)].\nDespite case complexity, operative time was significantly\nlower in RALH than in LAVH (89.9 vs. 124.8 min,\nP\\ 0.001) and similar to that in LSH (89.6 min). Esti-\nmated blood loss was greater in LAVH (167.9 ml) than in\nRALH (59.0 ml, P\\ 0.001) or LSH (65.7 ml, P\\ 0.001).\nLength of hospital stay was shorter for RALH than for\nLAVH or LSH. Conversion and complication rates were\nlow and similar across procedures. Multivariable regres-\nsion indicated that LAVH, obesity, uterine weight C250 g\nand older age predicted significantly longer operative time.\nThe learning curve for RALH demonstrated improved\noperative time over the case series. Our findings show the\nbenefits of RALH over LAVH. Outcomes in RALH can be\nas good as or better than those in LSH, suggesting the latter\nshould be the choice primarily for women desiring cervix-\nsparing surgery.\nKeywords Robotics /C1Laparoscopy /C1Hysterectomy /C1\nVaginal hysterectomy /C1Supracervical hysterectomy\nIntroduction\nAlthough abdominal hysterectomy still accounts for\napproximately two-thirds of all benign hysterectomies\nperformed in the USA [ 1, 2], minimally invasive tech-\nniques for this procedure have been slowly gaining\nacceptance. Recently published national data for 2005\nindicate that vaginal hysterectomies were performed in\nabout 22% of cases and that 14% of hysterectomies were\ndone laparoscopically [ 2]. Studies have shown that mini-\nmally invasive techniques for benign hysterectomy are safe\nfor the patient and result in decreased morbidity, shorter\nhospital stays, and a faster return to normal activity\ncompared to open procedures [ 3, 4]. However, with con-\nventional laparoscopic surgery, the surgeon generally\nexperiences some decrease in visual acuity, lack of tactile\nsensory input, and, thus, greater difficulty in mastering the\napproach. Difficulty with hand–eye coordination with\nB. N. Giep /C1H. N. Giep\nDepartment of Obstetrics and Gynecology, Spartanburg\nRegional Medical Center, Spartanburg, SC, USA\nH. B. Hubert\nDepartment of Medicine, Stanford University School\nof Medicine, Stanford, CA, USA\nB. N. Giep ( &)\nSpartanburg and Pelham P.A., 250 North Grove Medical Park\nDrive, Spartanburg, SC 29303, USA\ne-mail: bgobgyn@yahoo.com\n123\nJ Robotic Surg (2010) 4:167–175\nDOI 10.1007/s11701-010-0206-y\n\nlaparoscopy and the non-articulated instruments also\nmakes this approach more difficult to master. These factors\nmay explain a large part of the reluctance to adopt mini-\nmally invasive laparoscopic approaches over abdominal\nhysterectomy.\nImprovements in minimally invasive techniques were\nintroduced with 2005 Federal Drug Administration (FDA)\napproval of the da Vinci Surgical System (Intuitive Sur-\ngical, Sunnyvale, CA) for use in gynecologic procedures.\nSince that time, the use of the robotic system for hyster-\nectomy has been shown to shorten the learning curve for\nlaparoscopically naı ¨ve surgeons with improved three-\ndimensional (3D) visual acuity, articulated wrist-like\nmovement of instruments without tremor and ergonomic\nseating [5–7].\nFew studies have compared clinical outcomes of the\nrobot system with other minimally invasive techniques for\nhysterectomy. Payne and Dauterive demonstrated that\npatients who had undergone robotic-assisted compared to\ntotal laparoscopic hysterectomy had similar complication\nrates, but significantly less blood loss, fewer conversions to\nlaparotomy and shorter hospital stays [ 7]. A recent inves-\ntigation comparing robotic-assisted to traditional laparo-\nscopic hysterectomy confirmed these findings [ 8]. A\nsimilar comparative study showed shorter hospital stays but\nlonger operative times for patients who underwent a\nrobotic procedure [9]. A number of studies have compared\nlaparoscopic supracervical hysterectomy (LSH) to laparo-\nscopic-assisted vaginal hysterectomy (LAVH) or to total\nlaparoscopic hysterectomy. Results have repeatedly shown\nbetter outcomes with LSH than LAVH, i.e., shorter oper-\native time, less blood loss and lower complication rates\n[10–12]. To date, however, there have been no comparative\nstudies of robotic-assisted laparoscopic hysterectomy\n(RALH) to LAVH alone or to LSH. The goal of the study\nreported here was to provide such comparative data on\nclinical outcomes and learning curves for minimally\ninvasive procedures in a community-practice setting.\nMaterials and methods\nPatients and cases\nThe Spartanburg Regional Healthcare System is an inte-\ngrated healthcare system delivering services, including\nambulatory and inpatient surgery, to several counties in\nNorth and South Carolina. This study compares hysterec-\ntomies performed by multiple community surgeons at the\nSpartanburg Regional Medical Center, a facility within that\nsystem. Only two surgeons performed RALH during the\ntime period of this study (BG, HG), so the robotic cohort\n(n = 237) includes their entire consecutive RALH\nexperience beginning with the first patient in August 2007\nthrough to June 2009. Between July 2006 and February\n2008, these same surgeons performed their last 100\nLAVHs, transitioning almost exclusively to RALH in the\nlatter half of 2007. Thus, the LAVH cohort for this study\nincludes their last consecutive patients ( n = 100) along\nwith those of nine other gynecologists who performed\nLAVH at the same institution between January 2008 and\nJune 2009 ( n = 165), for a total of 265 LAVH cases. In\naddition, consecutive LSH patients of the nine other\ngynecologists ( n = 87) were identified during a similar\ntime period, January 2008 to June 2009. Patients were\nselected for LSH, rather than LAVH, if they preferred a\nprocedure that spared the cervix or if the surgeon believed\nit to be a better choice. There were too few total laparo-\nscopic hysterectomies performed at the hospital during this\ntime to form another comparison group. Every patient\nincluded in this study presented with a benign gynecologic\ncondition.\nSurgical procedures\nAll procedures were performed under general endotracheal\nanesthesia with preoperative antibiotics given. A steep\nTrendelenburg position was used with a gel pad underneath\nthe patient for stabilization. A Foley catheter and uterine\nmanipulator were placed for all surgeries.\nRobotic-assisted laparoscopic hysterectomy\nA 8.5-mm trocar was placed at the umbilicus and 8-mm\ntrocars were placed in the right and left lower quadrants.\nA 11-mm bladeless trocar was placed in the right upper\nquadrant. Docking of the robotic arms to the trocars fol-\nlowed by insertion of the 3-D camera was completed.\nA Gyrus PK Dissector (Gyrus ACMI, Maple Grove, MN)\nand monopolar scissors were inserted, at which time the\nconsole portion of the procedure began. Dissection of the\nbilateral round ligaments, adnexa and broad ligaments\nwas performed in the standard manner. The bladder flap\nwas created with a combination of sharp and blunt dis-\nsection. A colpotomy was performed with the monopolar\nscissors and carried circumferentially until the specimen\nwas amputated from the vagina. The specimen was either\nremoved vaginally or endoscopically, and the vaginal\ncuff was closed laparoscopically with interrupted figure-\nof-eight stitches of 0 Vicryl.\nLaparoscopic-assisted vaginal hysterectomy\nTo begin, the laparoscope was placed through a 5-mm\numbilical trocar. Three additional 5-mm ports were placed\nin the right and left lower quadrants and suprapubically.\n168 J Robotic Surg (2010) 4:167–175\n123\n\nA Maryland bipolar dissector was used. Dissection of the\nbilateral round ligaments, adnexa, and broad ligaments was\nperformed in the standard manner. The bladder flap was\ncreated with a combination of sharp and blunt dissection. A\nsmall anterior colpotomy was performed with monopolar\nscissors. The remainder of the procedure was completed\nvaginally. A posterior colpotomy was performed sharply,\nand then cardino-uterosacral ligaments and uterine vessels\nwere divided and sutured. After the specimen was removed\nvaginally, the vaginal cuff was closed with 0 Vicryl in a\nrunning locked fashion.\nLaparoscopic supracervical hysterectomy\nA 5-mm trocar was placed at the umbilicus and two\nadditional 5-mm trocars were placed in the right and left\nlower quadrants. Dissection of the bilateral round liga-\nments, adnexa and broad ligaments was performed in the\nstandard manner using either Harmonic ACE curved shears\n(Ethicon, Cincinnati, OH) or a LigaSure (Covidien,\nMansfield, MA) device. The bladder was dissected free\nfrom the lower uterine segment, and the uterine vessels\nwere coagulated and transected. The uterus was amputated\nfrom the cervix using either monopolar scissors or the\nHarmonic ACE curved shears and the endocervical canal\nfulgurated with the monopolar scissors. The specimen was\nthen extracted with a Gynecare Morcellex Tissue Morcel-\nlator (Ethicon).\nPatient characteristics and clinical procedures\nThe Institutional Review Board of Spartanburg Regional\nHospital approved this study. A standardized retrospective\nchart review was completed by the research nursing staff,\nand quality control procedures included verification of data\ninconsistencies and outliers using the medical records. The\nfollowing characteristics were obtained for all patients:\nage, body mass index (BMI), the presence or absence of\nprevious abdominopelvic surgery, gravidity, parity, and the\nprimary indication for surgery. Perioperative characteris-\ntics included concomitant procedures performed with the\nhysterectomy reflecting the complexity of the surgery,\nskin-to-skin operative time (defined as Foley catheter\ninsertion to skin closure), uterine weight, conversion to\nlaparotomy, estimated blood loss (EBL), length of hospital\nstay and intraoperative and postoperative complications up\nto 30 days post-discharge. Estimated blood loss was\ndetermined initially by canister collection as the differen-\ntial between aspirated and irrigated fluids and necessitated\nagreement between the surgeon and anesthesiologist.\nBlood collection was later done with the Stryker Neptune\nsystem that also enabled quantification of very small\namounts of blood loss that went undetectable using the\noriginal system. Thus, if no blood was detected by the\ncanister collection method, EBL was recorded as 25 ml.\nMinor complications following discharge (such as urinary\ntract infection) were captured on the patient follow-up visit\nfor robotic surgery. However, these were not available on\nall LAVH and LSH patients. Therefore, with regard to\npostoperative complications, only those patients requiring a\nvisit to the emergency room or a readmission to the hos-\npital were reported in this study.\nStatistical analyses\nData analyses included all pair-wise comparisons of min-\nimally invasive surgical methods, namely, RALH to each\nof the two other approaches (LAVH and LSH) as well as\nLAVH to LSH, using SAS software ver. 9.2.1 (SAS\nInstitute, Cary, NC). Continuous variables were compared\nusing two-sample t tests. Discrete variables were analyzed\nusing chi-squared tests or Fisher’s exact test with conti-\nnuity correction. In addition, multivariable regression\nanalyses were used to identify patient and procedure\ncharacteristics that had a significant impact on operative\ntime. To identify changes in operative time, blood loss and\nuterine weight that could be associated with surgical\nexperience, these parameters were compared over the\nconsecutive patient series for the RALH and LAVH pro-\ncedures performed by the same surgeons (BG, HG). In all\ninstances a P value \\ 0.05 was considered to be statisti-\ncally significant.\nResults\nComparisons of preoperative patient characteristics are\nshown in Table 1. The most frequent primary indication for\nbenign hysterectomy among these patients was abnormal\nuterine bleeding in RALH (50%) and LAVH (36%) and\nfibroids in LSH (39%). Endometriosis was the third most\nfrequent indication, ranging from 8% in RALH patients to\nalmost 21% in LSH cases. The age range of all patients was\n23–78 years. Those who underwent LSH were younger\non average than either RALH ( P = 0.067) or LAVH\n(P = 0.006) patients by 1.5 and 2.5 years, respectively, not\nsurprising given the desire for a cervix-preserving proce-\ndure. Differences in gravidity and parity followed similar\ntrends, with RALH patients having had significantly fewer\npregnancies and live births than either LAVH or LSH\npatients, with no differences between those in the LAVH\nand LSH groups. Means for BMI indicate that many\npatients were obese (BMI C 30) or bordering on obesity.\nDifferences in BMI by approach were small, and compar-\nisons were not statistically significant. A large proportion\nof patients had undergone prior abdominopelvic surgery,\nJ Robotic Surg (2010) 4:167–175 169\n123\n\nwith RALH patients exhibiting the highest percentage\n(83.1%), which was significantly greater than that in\npatients who underwent LAVH (73.2%; P = 0.01).\nPatients who underwent LSH did not differ from the other\ngroups with respect to prior surgery. Comparisons of pre-\noperative characteristics between LAVH patients of Drs.\nGiep versus the other nine surgeons show very similar\nbaseline profiles, with almost identical mean ages (42.6 vs.\n42.4 years, respectively) and very similar mean BMI (29.9\nvs. 29.5 kg/m\n2, respectively) and proportion with prior\nabdominopelvic surgeries (75.2 vs. 70.0%). These findings\nprovide reassurance that patients from the two groups of\nsurgeons combined as one LAVH cohort were very similar.\nExamination of intraoperative characteristics by\napproach indicates that concomitant procedures were more\noften performed with RALH (in 50.6% of surgeries) than\nwith either LAVH (in 26%) or LSH (in 20.6%; Table 2).\nSpecifically, the proportion of patients who had surgical\nprocedures for endometriosis or lysis of adhesions was\nsignificantly greater in the RALH group than in the LAVH\n(P\\ 0.001) or LSH ( P\\ 0.001) groups. Similar findings\nwere evident for pelvic reconstruction (RALH vs. LAVH\nP = 0.012, RALH vs. LSH P\\ 0.001). Most pelvic\nreconstruction involved robotically assisted uterosacral\nfixation, anterior and posterior repair and pubovaginal sling\nwith cystoscopy. These procedures would be expected to\nTable 1 Preoperative characteristics by minimally invasive approach to hysterectomy\nPreoperative characteristics RALH ( n = 237) LAVH ( n = 265) LSH ( n = 87) P value\nPrimary indication for surgery, n (%)\nBenign pelvic mass 1 (0.4) – –\nAbnormal uterine bleeding 119 (50.2) 96 (36.2) 28 (32.2)\nFibroids 68 (28.7) 89 (33.6) 34 (39.1)\nEndometriosis 19 (8.0) 27 (10.2) 18 (20.7)\nPelvic pain 11 (4.6) 14 (5.3) –\nAbnormal PAP – 1 (0.4) –\nPelvic relaxation 16 (6.8) 7 (2.6) –\nOvarian cyst – 2 (0.8) 3 (3.4)\nProlapse (uterine/vaginal) 3 (1.3) 17 (6.4) –\nOther – 12 (4.5) 4 (4.6)\nAge (years)\nMean (SD) 41.5 (8.4) 42.5 (10.6) 39.9 (6.3) 0.240\na\n95% CI 40.4–42.5 41.2–43.8 38.6–41.3 0.067 b\n0.006c\nGravidity, n (%)\n0 24 (10.1) 8 (3.0) 3 (3.4) 0.003 a\n1 38 (16.0) 40 (15.1) 8 (9.2) 0.009 b\nC2 175 (73.9) 217 (81.9) 76 (87.4) 0.387 c\nParity, n (%)\n0 28 (11.8) 13 (4.9) 4 (4.6) 0.007 a\n1 46 (19.4) 48 (18.1) 9 (10.3) 0.004 b\nC2 163 (68.8) 204 (77.0) 74 (85.1) 0.207 c\nBMI\nMean (SD) 30.3 (7.5) 29.9 (6.7) 31.2 (7.7) 0.530 a\n95% CI 29.4–31.2 29.0–30.7 29.6–32.8 0.343 b\n0.132c\nPrior abdominal or pelvic surgery, n (%) 197 (83.1) 194 (73.2) 69 (79.3) 0.010 a\n0.529b\n0.320c\nRALH Robotic-assisted laparoscopic hysterectomy, LAVH laparoscopic-assisted vaginal hysterectomy, LSH laparoscopic supracervical hyster-\nectomy, SD standard deviation, CI confidence interval\na Comparison of RALH to LAVH\nb Comparison of RALH to LSH\nc Comparison of LAVH to LSH\n170 J Robotic Surg (2010) 4:167–175\n123\n\ncontribute to an increase in operative time. Although the\nrates of pelvic reconstruction were low in both LAVH and\nLSH, they were statistically higher in LAVH (4.9%) than\nLSH (3.4%; P = 0.04). Uterine weights ranged from 24 to\n1,233 g. The mean uterine weight was similar in LAVH\nand LSH patients but differed significantly from that of\nRALH cases who had the largest uteri (207.4 ± 194.5\n(RALH) vs. 149.6 ± 118.7 (LAVH) g, P\\ .001; vs.\n141.1 ± 172.5 (LSH) g, P = 0.005]. The percentage of\npatients with uteri of at least 250 g was 23.2% in RALH\ncompared to 9.1 and 8.0% in LAVH ( P\\ 0.001) and LSH\n(P = 0.004), respectively. However, mean skin-to-skin\noperative time was 35 min longer for LAVH patients\ncompared to both RALH and LSH patients for whom\nprocedures took approximately 1.5 h on average (both\ncomparisons, P\\ 0.001). Similarly, estimated blood loss\nwas highest in LAVH, being over 100 ml greater than in\nRALH and LSH (both comparisons, P\\ 0.001). Length of\nTable 2 Intraoperative and perioperative characteristics by minimally invasive approach to hysterectomy\nIntraoperative and perioperative characteristics RALH ( n = 237) LAVH ( n = 265) LSH ( n = 87) P value\nConcomitant procedures, n (%)\nEndometriosis/lysis of adhesions 93 (39.2) 56 (21.1) 15 (17.2) \\0.001a\n\\0.001b\n0.520c\nPelvic reconstruction 27 (11.4) 13 (4.9) 3 (3.4) 0.012 a\n\\0.001b\n0.044c\nUterine weight (g)\nMean (SD) 207.4 (194.5) 149.6 (118.7) 141.1 (172.5) \\0.001a\n95% CI 182.5–232.6 134.9–164.3 105.2–177.1 0.005 b\n0.670c\nStratified uterine weight\n\\250 g 182 (76.8) 241 (90.9) 80 (92.0) \\0.001a\nC250 g 55 (23.2) 24 (9.1) 7 (8.0) 0.004 b\n0.867c\nSkin to skin operative time (min)\nMean (SD) 89.9 (37.5) 124.8 (48.7) 89.6 (38.0) \\0.001a\n95% CI 84.9–94.5 118.9–130.7 80.9–98.5 0.949 b\n\\0.001c\nEstimated blood loss (ml)\nMean (SD) 59.0 (75.7) 167.9 (146.0) 65.7 (60.7) \\0.001a\n95% CI 49.2–68.6 150.2–185.6 52.7–78.6 0.412 b\n\\0.001c\nLength of hospital stay (days)\nMean (SD) 1.0 (0.1) 1.2 (0.7) 1.2 (0.8) \\0.001a\n95% CI 1.0–1.03 1.1–1.2 1.0–1.4 0.022 b\n1.00c\nConversion, n (%) 4 (1.7) 1 (0.4) None 0.194 a\n0.577b\n1.00c\nIntraoperative complications, n (%) 1 (0.4) 1 (0.4) None NA\nPostoperative complications, n (%) 8 (3.4) 4 (1.5) 2 (2.3) 0.243 a\n1.00b\n0.640c\nRALH Robotic-assisted laparoscopic hysterectomy, LAVH laparoscopic-assisted vaginal hysterectomy, LSH laparoscopic supracervical hyster-\nectomy, SD standard deviation, CI confidence interval, NA Not available\na Comparison of RALH to LAVH\nb Comparison of RALH to LSH\nc Comparison of LAVH to LSH\nJ Robotic Surg (2010) 4:167–175 171\n123\n\nhospital stay was short for all approaches, but significantly\nshorter for RALH patients than either LAVH ( P\\ 0.001)\nor LSH ( P = 0.022) patients.\nThe rates of conversion to an open procedure were low,\nranging from zero to 1.7%, and did not differ by surgical\napproach (Table 2). The conversions in the RALH group\noccurred early in the learning curve. Two conversions were\nwithin the first 25 cases and another two within the first 150\ncases. These occurred in patients with large uteri where\nthere was difficulty in accessing the blood vessels. Simi-\nlarly, one conversion in the LAVH group occurred in a\npatient with a 16-week uterus where it was difficult to\nmaintain hemostasis while proceeding laparoscopically.\nThis patient was subsequently converted to an abdominal\nhysterectomy. Rates of intraoperative and postoperative\ncomplications within 30 days of surgery were low across\nthe three surgical approaches: the overall rates totaled 3.8%\nfor RALH, 1.9% for LAVH and 2.3% for LSH, with the\nmajority being minor in severity. No statistically significant\ndifferences were noted by surgical method. In both the\nRALH and LAVH cohorts, there was one intraoperative\ncystotomy repair with no further sequelae. There were also\ntwo patients with incisional infections and one with a\nbacterial infection in the RALH cohort; these were treated\nwith antibiotics during their hospital stay. One patient in\nthe robotic cohort and one in the LAVH cohort experienced\na pelvic abscess that was subsequently drained and treated.\nThere were no instances of cuff dehiscence requiring repair\nin any of the cohorts. There was one patient in the robotic\ngroup who was treated for cuff cellulitis, and two patients\nreported bleeding from the vaginal cuff in the RALH and\nLAVH cohorts. Both of these latter complications were\nresolved without the need for reintervention. During the\nperioperative period, one patient in the robotic group suf-\nfered a pulmonary embolism that was treated by antico-\nagulation with heparin and then enoxaparin (Lovenox).\nAtelectasis was noted in three patients, one from the LAVH\ncohort and two in the LSH cohort.\nComparisons of perioperative characteristics for the\nLAVH procedures performed by Drs. Giep versus the other\nsurgeons show similar patient uterine weights (153.4 ±\n124.5 vs. 147.0 ± 114.4 g, respectively), EBL (157.3 ±\n111.4 vs. 173.6 ± 163.3 ml, respectively) and length of\nhospital stay (1.1 vs. 1.2 days, respectively). However,\noperative time was significantly shorter for procedures\nperformed by Drs. Giep given their extensive experience\nwith laparoscopic procedures (101.5 ± 39.7 vs. 138.9 ±\n48.4; P\\ .001).\nMultivariable linear regression was used to identify those\ncharacteristics of the patients and the approach that were\nsignificant predictors of skin-to-skin operative time. The\ncharacteristics considered included patient age (dichoto-\nmized above or below the median age = 41 years), obesity\n(BMI \\30, C30), previous abdominal or pelvic surgery,\nuterine weight (\\250 g, C250 g), any concomitant proce-\ndure beyond hysterectomy with salpingo-oophorectomy and\napproach to surgery (RALH, LAVH, LSH). Younger age\n(P\\ 0.001), BMI \\30 (P = 0.02), uterine weight \\250 g\n(P\\ 0.001) and surgery other than LAVH (P\\ 0.001) were\nall independently associated with shorter operative times\n(data not shown). On average, the operative times were\nlonger for older patients versus younger ones (21 min), for\nobese patients versus non-obese patients (8 min), for greater\nuterine weight patients versus those with a smaller uterus\n(24 min) and for those undergoing LAVH surgeries vs. those\nundergoing LSH and RALH (35 min).\nData points, representing the means of 25 sequential\npatients, were generated for skin-to-skin times, uterine\nweights and EBL over the entire case series for the RALH\npatients and for the comparable 100 LAVH patients whose\nprocedures were performed by the same surgeons (HG,\nBG). Comparisons of the first 25 cases to the last showed\nno statistically significant differences for any of the LAVH\nparameters, although uterine weights generally increased\nwith longer surgical experience (data not shown). Similar\nfindings were evident in RALH for uterine weight and EBL\n(comparison of first and last data points, P = 0.189 and\nP = 0.875, respectively). However, operative skin-to-skin\ntime significantly decreased with surgical experience\n(comparison of first 25 to last 37 patients, P = 0.003,\nFig. 1). The first 25 RALH patients had a mean operative\ntime of 106.4 min, which is comparable to that of the\nLAVH procedures (101.5 min) and significantly different\nfrom the last patients’ operative time of 76.5 min.\nFig. 1 Learning curve for robotic-assisted laparoscopic hysterectomy\n(RALH). Solid black line connects the mean skin-to-skin operative\ntimes at 25-case intervals for RALH ( P = 0.003 for mean of first 25\nvs. last 37 patients). Dashed black line Mean skin-to-skin time for 100\ncomparable laparoscopic-assisted vaginal hysterectomy ( LAVH)\npatients (HG, BG only). There were no significant differences in\noperative time for the first 25 LAVH cases vs. the last 25 cases,\nP = 0.59. Gray bars Mean uterine weights corresponding to the each\nof the RALH 25-case intervals ( P = 0.189 comparing first 25 vs. last\n37 cases)\n172 J Robotic Surg (2010) 4:167–175\n123\n\nFurthermore, as suggested by the regression analyses\nabove, operative time varied with changes in uterine\nweight over time (Fig. 1).\nDiscussion\nThe results of this study show that while each of the three\nminimally invasive procedures had similar conversion and\ncomplication rates, there were distinct differences with\nregard to other perioperative outcomes. Although patients\nwho underwent RALH had a higher rate of prior abdomi-\nnopelvic surgeries than LAVH patients and more con-\ncomitant procedures performed and greater uterine weights\nthan either LAVH or LSH patients, they experienced sig-\nnificantly lower operative times and EBL than LAVH\npatients and a significantly shorter length of hospital stay\nthan either LAVH or LSH patients. Operative times were\n35 min longer and EBL 100 ml greater on average for\nLAVH patients than for those who underwent either the\nLSH or RALH procedures for whom these outcomes were\ncomparable.\nMultivariable analysis of the independent predictors of\noperative time confirmed the findings with regard to the\nimpact of LAVH compared to LSH and RALH and further\nshowed that older age, obesity, and large uteri ( C250 g) in\nparticular were also associated with longer surgical time.\nDespite the complexity of cases undertaken with RALH,\nincluding those with larger uteri, prior surgery and neces-\nsary concomitant procedures, the learning curve for oper-\native time continued to improve over the case series, with\nmean times dropping from 106 to 76 min over the\n23 months of surgery. The average operative time for the\nlast 100 LAVH cases performed by the same surgeons was\n102 min, with no noticeable decrease over the 20 months\nobserved.\nAs previously mentioned, there are few published\ncomparative studies on the outcomes of benign hysterec-\ntomy that include robotic procedures. Payne and Dauterive,\nwho compared total laparoscopic hysterectomy in their last\n100 patients to RALH in their first 100 patients, found\nlonger average operative times for RALH but significantly\nshorter times in their last 25 robotic cases [7]. As in the our\nstudy, patients undergoing the total laparoscopic procedure\nhad a greater EBL and a longer hospital stay than those\nundergoing the robotic hysterectomy. These researchers\nalso found low complication rates that were similar in the\ntwo patient groups. A recently completed study also\ndemonstrated significantly greater blood loss and longer\nhospital stay with standard laparoscopic hysterectomy\ncompared to RALH [ 8]. While complication rates were\nsimilar in the two groups, conversion rates were signifi-\ncantly higher in the conventional laparoscopic cohort.\nA similar comparative investigation showed that robotic\nhysterectomy patients had significantly shorter hospital\nstays but longer operative times than patients who under-\nwent total laparoscopic hysterectomy [ 9]. To date, no\ncomparative studies of RALH and LSH have been identi-\nfied in the Medline literature database. Our study findings\nalso indicate that outcomes in RALH, despite the increased\ncomplexity of the cases, were comparable to or better than\nthose in LSH, suggesting that LSH should be considered\nfor women primarily as a cervix-sparing surgery.\nStudies comparing other minimally invasive procedures\nfor hysterectomy are numerous and, as in our investigation,\ngenerally indicate better outcomes for patients undergoing\nLSH than LAVH. These better outcomes may be attributed\nto the fact that the LSH procedure avoids what can be a\ndifficult bladder dissection that is required in LAVH. Milad\nand colleagues found that LSH patients had significantly\nshorter operative times and hospital stays, less blood loss\nand a lower complication rate than LAVH patients [ 11].\nSimilar to the results in our study, a large multi-center\nstudy investigation with matched groups of patients dem-\nonstrated significantly less blood loss and shorter operative\ntimes (by 30 min) in the LSH patients compared to the\nLAVH patients and no difference in hospital stays [ 12].\nAnother large study of women undergoing a classic intra-\nfascial supracervical hysterectomy reported lower compli-\ncation rates and smaller amounts of EBL when compared\nto patients undergoing LAVH [ 13].\nSeveral studies of hysterectomy for benign indications\nhave also suggested that greater uterine weights or obesity\nare associated with longer operative times. Boggess et al.\nfound that among preoperative patient characteristics, only\nuterine weight [250 g was associated with increased\noperative time in RALH [ 14]. A comparative investigation\nof robotic and conventional laparoscopic hysterectomy\nreported significant independent effects of BMI and uterine\nweight on operative times [ 9]. Studies of robotic gyneco-\nlogic procedures in patients unselected for pathology have\nalso reported a relationship between greater uterine weight\nand longer duration of surgery or console time [6, 15]. This\nrelationship is usually attributed to the greater time needed\nfor morcellation and/or removal of larger specimens.\nInvestigations of other patients undergoing conventional\nlaparoscopic hysterectomy generally confirm our findings\nregarding the impact of large uterine weight on operative\ntime [ 16–\n18]. Several studies also have demonstrated a\nrelationship between obesity and longer operative times\nin robotic or conventional laparoscopic hysterectomy\n[15, 19].\nOther investigators have described the learning curve for\nrobotic hysterectomy. They have documented shorter\noperative time with greater experience of the primary\nsurgeon [5–7]. However, as surgeons become comfortable\nJ Robotic Surg (2010) 4:167–175 173\n123\n\nand competent in this technique, they also tend to take on\nmore complex cases, a factor that could negatively impact\noperative time over a case series. Our study shows a visual\nexample of the how operative time trends with the size of\nuteri in patients undergoing RALH over time.\nThe limitations to our study include the retrospective,\nobservational nature of the design, introducing the potential\nfor bias. However, with the exception of the small age\ndifference in LSH patients, baseline characteristics of the\nthree patient groups were fairly similar. RALH patients had\na greater proportion of prior abdominopelvic surgeries and\nconcomitant procedures and larger uterine weights than\nLAVH patients, but such differences would have only\nserved to introduce a bias toward better LAVH outcomes.\nOur findings with regard to RALH versus LAVH do not\nsupport such a bias, although the differences found\nbetween RALH and LAVH may be conservative as a\nresult. Misclassification errors, common in retrospective\nstudies, also may have served to weaken tests of differ-\nences between surgical approaches. Furthermore, compli-\ncation rates were low due to the exclusion of minor\nconditions, such as urinary tract infections, which may not\nhave been captured during clinic visits after hospital dis-\ncharge. These conditions were not routinely noted across\nall procedures and, therefore, could not be included.\nConclusion\nAlthough the RALH patients in this study presented as the\nmost complex cases in terms of prior pelvic surgery, con-\ncomitant procedures and uterine weight, they had better\nclinical outcomes with respect to operative time, EBL and\nlength of hospital stay than the patients undergoing LAVH.\nOn average, the operative time was 35 min longer and the\nEBL was 100 ml greater for LAVH patients than for either\nRALH or LSH patients. Length of hospital stay was sig-\nnificantly greater for both LAVH and LSH patients (mean\n1.2 days) than for RALH patients (mean 1.0 days). Other\noutcomes for LSH were similar to those for RALH. Con-\nversions to laparotomy and intraoperative and postopera-\ntive complications were similar in the three patient groups.\nOur analysis of the learning curves suggest that operative\ntime for RALH may continue to improve beyond the initial\npatient series despite the increased complexity of cases\nundergoing this procedure. The results of this study lend\nsupport to the small but growing body of literature\ndescribing the benefits of minimally invasive robotic-\nassisted laparoscopic over conventional laparoscopic-\nassisted vaginal hysterectomy. Our findings also demon-\nstrate that outcomes with robotic surgery are as good as or\neven better than those for laparoscopic supracervical hys-\nterectomy, suggesting that decisions regarding choice of\nthe latter approach be based primarily on the desire for a\ncervix-sparing procedure.\nAcknowledgments The authors wish to thank Usha Kreaden, MSc\nof Intuitive Surgical for statistical support during the completion of\nthis manuscript.\nConflicts of interest Drs. Bang and Hoang Giep are members of\nIntuitive Surgical’s Speakers’ Bureau, Proctor Cases, and have\nactively been running ‘ ‘Epicenter’’ programs at their institution.\nDr. Hubert is an epidemiologist who consults with Intuitive Surgical.\nOpen Access This article is distributed under the terms of the\nCreative Commons Attribution Noncommercial License which per-\nmits any noncommercial use, distribution, and reproduction in any\nmedium, provided the original author(s) and source are credited.\nReferences\n1. Whiteman MK, Hillis SD, Jamieson DJ et al (2008) Inpatient\nhysterectomy surveillance in the United States, 2000–2004. Am J\nObstet Gynecol 198:34.e1–34.e7\n2. Jacoby VL, Autry A, Jacobson G, Domush R, Nakagawa S,\nJacoby A (2009) Nationwide use of laparoscopic hysterectomy\ncompared with abdominal and vaginal approaches. Obstet\nGynecol 114:1041–1048\n3. Johnson N, Barlow D, Lethaby A, Tavender E, Curr L, Garry R\n(2005) Methods of hysterectomy: systematic review and meta-\nanalysis of randomized controlled trials. Br Med J 330:1478–1485\n4. Falcone T, Paraiso MFR, Mascha E (1999) Prospective ran-\ndomized clinical trial of laparoscopically assisted vaginal hys-\nterectomy versus total abdominal hysterectomy. Am J Obstet\nGynecol 180:955–962\n5. Pitter MC, Anderson P, Blissett A, Pemberton N (2008) Robotic-\nassisted gynaecological surgery–establishing training criteria;\nminimizing operative time and blood loss. Int J Med Robotics\nComput Assist Surg 4:114–120\n6. Lenihan FR Jr, Kovanda C, Seshadri-Kreaden U (2008) What is\nthe learning curve for robotic assisted gynecologic surgery?\nJ Minim Invasive Gynecol 15:589–594\n7. Payne TN, Dauterive FR (2008) A comparison of total laparo-\nscopic hysterectomy to robotically assisted hysterectomy: surgi-\ncal outcomes in a community practice. J Minim Invasive Gynecol\n15:286–291\n8. Landeen LB, Bell MC, Hubert HB, Bennis LY, Knutsen-Larson\nSS, Seshadri-Kreaden U (2010) Clinical and cost comparisons:\nrobot-assisted versus abdominal, standard laparoscopic and vag-\ninal hysterectomy. Gray J (submitted for publication)\n9. Shashoua AR, Gill D, Locher SR (2009) Robotic-assisted total\nlaparoscopic hysterectomy versus conventional total laparoscopic\nhysterectomy. J Soc Laparoendosc Surg 13:364–369\n10. Lalonde CJ, Daniell JF (1996) Early outcomes of laparoscopic-\nassisted vaginal hysterectomy versus laparoscopic supracervical\nhysterectomy. J Am Assoc Gynecol Laparosc 3:251–256\n11. Milad MP, Morrison K, Sokol A, Miller D, Kirkpatrick L (2001)\nA comparison of laparoscopic supracervical hysterectomy vs.\nlaparoscopically assisted vaginal hysterectomy. Surg Endosc\n15:286–288\n12. El-Mowafi D, Facharzt WM, Lall C, Wenger JM (2004)\nLaparoscopic supracervical hysterectomy versus laparoscopic-\nassisted vaginal hysterectomy. J Am Assoc Gynecol Laparosc\n11:175–180\n174 J Robotic Surg (2010) 4:167–175\n123\n\n13. Kim DH, Bae DH, Hur M, Kim SH (1998) Comparison of classic\nintrafascial supracervical hysterectomy with total laparoscopic\nand laparoscopic-assisted vaginal hysterectomy. J Am Assoc\nGynecol Laparosc 5:253–260\n14. Boggess JF, Gehrig PA, Cantrell L et al (2009) Perioperative\noutcomes of robotically assisted hysterectomy for benign cases\nwith complex pathology. Obstet Gynecol 114:585–593\n15. Payne TN, Dauterive FR (2010) Robotically assisted hysterec-\ntomy: 100 cases after the learning curve. J Robotic Surg 4:11–17\n16. Wattiez A, Soriano D, Fiaccavento A et al (2002) Total laparo-\nscopic hysterectomy for very enlarged uteri. J Am Assoc Gynecol\nLaparosc 9:125–130\n17. Seracchioli R, Venturoli S, Colombo FM et al (2003) GnRH\nagonist treatment before total laparoscopic hysterectomy for large\nuteri. J Am Assoc Gynecol Laparosc 10:316–319\n18. Bonilla DJ, Mains L, Whitaker R, Crawford B, Finan M, Magnus\nM (2007) Uterine weight as a predictor of morbidity after a\nbenign abdominal and total laparoscopic hysterectomy. J Reprod\nMed 52:490–498\n19. Chopin N, Malaret JM, Lafay-Pillet MC, Fotso A, Foulot H,\nChapron C (2009) Total laparoscopic hysterectomy for benign\nuterine pathologies: obesity does not increase the risk of com-\nplications. Hum Reprod 24:3057–3062\nJ Robotic Surg (2010) 4:167–175 175\n123","source_license":"CC0","license_restricted":false}