Hand-assisted laparoscopic surgery versus open surgery for colorectal disease: a systematic review and meta-analysis

Transcription

1 The American Journal of Surgery (2014) 207, Review Hand-assisted laparoscopic surgery versus open surgery for colorectal disease: a systematic review and meta-analysis Jie Ding, M.D. a,b, Yu Xia, M.D. c, Guo-qing Liao, M.D. b, *, Zhong-min Zhang, M.D. a, *, Sheng Liu, M.D. b, Yi Zhang, M.D. b, Zhong-shu Yan, M.D. b a Department of Gastrointestinal Surgery, Guizhou Provincial People s Hospital, Guiyang, China; b Department of Gastrointestinal Surgery, Xiangya Hospital, Central South University, Changsha, China; c Department of Stomatology, Guizhou Provincial People s Hospital, Guiyang, China KEYWORDS: Hand-assisted laparoscopic surgery; Laparoscopy; Open surgery; Colorectal; Colon; Meta-analysis Abstract BACKGROUND: Laparoscopic colorectal surgery remains one of the most challenging techniques to learn. METHODS: The authors collected studies that have compared hand-assisted laparoscopic surgery (HALS) and open surgery for the treatment of colorectal disease over the past 17 years. Data of interest for HALS and open surgery were subjected to meta-analysis. RESULTS: Twelve studies that included 1,362 patients were studied. In total, 2.66% of HALS procedures were converted to laparotomy. Compared with the open surgery group, blood loss, rate of wound infection, and ileus in the HALS group decreased, and incision length, recovery of gastrointestinal function, and hospitalization period were shorter. There were no significant differences in operating time, hospitalization costs, mortality, and complications, including urinary tract infection, pneumonia, and anastomotic leak, between the groups. CONCLUSIONS: HALS has the advantages of minimal invasion, lower blood loss, shorter incision length, and faster recovery, and it can shorten the length of hospitalization without an increase in costs. The drawbacks are that a small number of patients who undergo HALS may need to be converted to laparotomy, and the oncologic safety and long-term prognosis are not clear. Ó 2014 Elsevier Inc. All rights reserved. Laparoscopic colorectal surgery was first described in Because of its advantages of minimal invasion, faster postoperative recovery, and shorter hospital stays, it The authors declare no conflicts of interest. * Corresponding author. Tel.: ; fax: addresses: guoqingliao@126.com, zhangzhongmin@medmai. com.cn Manuscript received November 29, 2012; revised manuscript April 10, 2013 has been widely applied to colorectal surgery over the past decade. 2 But with its wide implementation, most surgeons have realized that laparoscopic colorectal surgery is one of the most challenging techniques to learn; it has been estimated that 20 to 62 laparoscopic colectomy cases are needed to achieve proficiency with laparoscopic techniques. 3 6 The reasons for this steep learning curve include difficulty in exposing the colon and a lack of tactile feedback. 7,8 Hand-assisted laparoscopic surgery (HALS) is a hybrid laparoscopic approach by which the surgeon inserts a hand /$ - see front matter Ó 2014 Elsevier Inc. All rights reserved.

2 110 The American Journal of Surgery, Vol 207, No 1, January 2014 inside the abdomen to facilitate the laparoscopic dissection without disturbing the pneumoperitoneum. 9 The potential advantages of HALS include the restoration of tactile feedback and proprioception, the ability to perform blunt dissection, rapid control of unexpected bleeding, and a potential reduction in the number of trocars and instruments required to perform the resection. 10 HALS has been introduced as an alternative surgical technique, essentially bridging both open and laparoscopic approaches, 11,12 and may be a better option for surgeons early in their laparoscopic careers. 13 But whether intra-abdominal placement of a hand during HALS abrogates the benefits of minimally invasive techniques remains to be established. A study by Aalbers et al 14 showed a significant shorter time to flatus and length of hospital stay after HALS than after open surgery (), and the number of harvest lymph nodes, postoperative complications, mortality rates, and hospitalization costs were similar between the 2 groups. Therefore, they concluded that HALS has the advantages of laparoscopic surgery over, especially for indications in which an incision to extract the resection specimen is required. Unfortunately, only 7 studies with a small number of cases were included in their study, and no subcategory analysis of complications was performed, 14 making objective evaluation of the safety of HALS difficult. More important, the included studies contained benign and malignant disease, colon and rectal surgery, and randomized controlled trials and nonrandomized controlled trials. Therefore, sensitivity analysis should be performed to evaluate the stability of the results, and a meta-analysis of more studies with a larger sample size and objective appraisal of complications is necessary. To that end, we collected all studies published since 1995 that compared HALS with for the treatment of colorectal disease to perform a meta-analysis and investigated the value of HALS for treatment of colorectal disease. Methods Search strategy The publications were identified by searching the major medical databases, such as MEDLINE, EMBASE, and the Cochrane Library, for relevant reports published between January 1995 and May The search string was as follows: (hand OR hand-assisted OR hand-assisted laparoscopic OR manual OR manually) AND (colon OR colorectal OR sigmoid OR rectal OR rectum OR colectomy OR hemicolectomy OR proctectomy). Inclusion and exclusion criteria Inclusion criteria were (1) comparison of HALS with for treatment of colorectal disease between 1995 and 2012; (2) inclusion of patients with primary colorectal disease; and (3) presence of raw data including most of the following: conversion rate, operative time, blood loss, incision length, number of harvested lymph nodes, time to first flatus, length of hospital stay, complications, mortality, and hospitalization costs. Exclusion criteria were (1) no group as a control; (2) nonprimary colorectal diseases; and (3) duplicate publication or provision of insufficient data. 134 Publications identified by computerized search: last search done on 25 April Excluded by abstract review 45 Articles screened in complete form 26 Excluded for non comparative studies 19 Comparative studies 12 studies included in final meta-analyses: 5 Randomized controlled trials 7 Retrospective studies 7 Excluded for: 1 included emergency operation 1 duplicate publication 2 did not provide sufficient data 3 included laparoscopic assisted surgery Figure 1 Systematic search and selection strategy.

3 J. Ding et al. HALS versus : a meta-analysis 111 Table 1 Study Characteristics and conversion rates of the 12 selected clinical studies Location Study style Group Gender Type of disease Extent of resection Male Female BD MD RH LH SR RR TC PC Conversion cases Dunker et al 17 (2000) The Netherlands RNT HALS Kang et al 9 (2004) Taiwan RCT HALS Maartense et al 18 (2004) The Netherlands RCT HALS Zhang et al 19 (2006) China RNT HALS Chi et al 20 (2005) China RNT HALS Anderson et al 21 (2007) United States RNT HALS (6%) Chuang et al 22 (2007) Hong Kong RCT HALS (7%) Polle et al 23 (2007) RCT HALS Osarogiagbon United States RNT HALS d d (10%) et al 24 (2007) d d Liu et al 25 (2010) China RCT HALS Orenstein et al 8 (2011) United States RNT HALS d 52 d d d d d 228 d d d d Zhang et al 26 (2011) China RNT HALS Total (2.66%) BD 5 benign disease; HALS 5 hand-assisted laparoscopic surgery; LH 5 left hemicolectomy; MD 5 malignant disease; 5 open surgery; PC 5 proctocolectomy; RCT 5 randomized controlled trial; RH 5 right hemicolectomy; RNT 5 retrospective nonrandomized trial; RR 5 rectal resection; SR 5 sigmoid resection; TC 5 total colectomy. Quality of studies We used a star scoring system 15 on the basis of criteria related to study design, comparability of patient groups, and outcome assessment to assess literature quality. The total score was 9 stars, and the quality of each study was graded as level 1 (0 to 5 stars) or level 2 (6 to 9 stars). Data extraction Three researchers extracted data from each study by using a structured sheet and entered the data into a database. The extracted information included author, year of publication, source journal, sample size, research design, conversion rate, operative time, blood loss, number of Table 2 Quality assessment scoring of studies Selection Comparability of groups Outcomes Author Total Dunker et al 17 * * * ** ** * * 9* Kang et al 9 * * * * ** * 7* Maartense et al 18 * * * ** ** * 8* Zhang et al 19 * * * * * * 6* Chi et al 20 * * * * ** * * 8* Anderson et al 21 * * * * ** * 7* Chuang et al 22 * * * ** ** 7* Polle et al 23 * * * * ** * 7* Osarogiagbon et al 24 * * * ** ** 7* Liu et al 25 * * * ** ** * * 9* Orenstein et al 8 * * * * * * 6* Zhang et al 26 * * * * ** * 7* Selection for treatment: 1 5 inclusion criteria reported; 2 5 generalizability of patients undergoing hand-assisted laparoscopic surgery to population undergoing surgery for colorectal disease; 3 5 generalizability of patients undergoing open surgery to population undergoing surgery for colorectal disease. Comparability between groups (if yes to all, 2 stars; if 1 of these characteristics was not reported, 1 star; if the 2 groups differed, no stars): 4 5 age, sex, and body mass index; 5 5 lesion location, type of disease, and extent of resection. Outcome assessment: outcomes of interested clearly recorded (1 star); 7 5 adequacy of follow-up (1 star if follow-up.90%).

4 112 The American Journal of Surgery, Vol 207, No 1, January 2014 Study H ALS WMD (random) Weight WMD (random) or sub-category N Mean (SD) N M ean (SD) % Test for heterogeneity: Chi?= , df = 4 (P < ), I?= 96.4% Test for overall effect: Z = 0.34 (P = 0.73) Figure 2 Forest plot showing operation time between HALS and. harvested lymph nodes, postoperative pain, time to first flatus, length of hospital stay, number of complications, mortality, and hospital costs. Mortality was defined as 30- day mortality. Statistical analysis Statistically, it was not possible to combine the analysis of mean and median, and only data in the form of means and standard deviations were included in the meta-analysis. Therefore, continuous variables such as median or range were excluded. In this meta-analysis, weighted mean differences (WMDs) were used for the analysis of continuous variables, and odds ratios (ORs) were used for dichotomous variables. Random-effects models were used to identify heterogeneity among the studies. 16 Heterogeneity was assessed using the chi-square test. Ninety-five percent confidence intervals (CIs) were calculated. P values %.05 were considered to indicate statistical significance. Statistical analyses were performed using Review Manager version 4.2 (Cochrane Collaboration, Copenhagen, Denmark). Results According to the search strategy and inclusion criteria, a total of 12 clinical studies 8,9,17 26 that included 1,362 colorectal surgery cases (536 HALS and 826 ) were considered suitable for meta-analysis. The search and exclusion strategy is displayed in Fig. 1. The characteristics of the 12 clinical studies (5 randomized controlled trials 9,18,22,23,25 and 7 retrospective studies 8,17,19 21,24,26 ) are shown in Table 1. Quality assessment of the reported studies is shown in Table 2, and each study has a score of.6 points. Conversion rate All 12 clinical studies showed data on conversion rate, and a total of 13 cases were converted to ; in other words, the conversion rate was 2.66%. The conversion rate ranged from 0% to 10.26%, and 9 studies 8,9,17 20,23,25,26 showed no conversion cases (Table 1). Reasons for conversion to included dense adhesions (2 cases), 21,22 ureteral injury (1 case), 22 and phlegmon and bleeding (6 cases). 21 Four cases did not indicate reasons for conversion. 24 Operative time Five clinical studies 9,20,21,25,26 included data on operative time. The heterogeneity test showed heterogeneity in the operative times of the 5 clinical studies; therefore, the random-effects model was used to combine the data. This model showed no significant difference between HALS and in operative time (WMD, 3.51;, to 23.50; P 5.73; Fig. 2). Blood loss Four studies 9,20,25,26 included data on intraoperative blood loss, and all showed significantly less blood loss with HALS than with. The random-effects model was used to combine the data because of evident heterogeneity. Study H ALS WMD (random) Weight WMD (random) or sub-category N Mean (SD) N M ean (SD) % Test for heterogeneity: Chi?= 10.96, df = 3 (P = 0.01), I?= 72.6% Test for overall effect: Z = 6.36 (P < ) Figure 3 Forest plot showing blood loss between HALS and.

5 J. Ding et al. HALS versus : a meta-analysis 113 Study H ALS WMD (random) Weight WMD (random) or sub-category N Mean (SD) N M ean (SD) % Test for heterogeneity: Chi?= , df = 1 (P < ), I?= 99.1% Test for overall effect: Z = 3.96 (P < ) Figure 4 Forest plot showing incision length between HALS and. In the pooled data, intraoperative blood loss for HALS was less than that for (WMD, ;, to ; P,.05; Fig. 3). Incision length Two studies 9,25 included data on incision length, and both showed a significantly shorter length of incision with HALS than with. The random-effects model was used to combine the data because heterogeneity was evident. In the pooled data, the incision length with HALS was significantly shorter than with (WMD, 28.79;, to 24.44; P,.05; Fig. 4). Number of harvested lymph nodes Four studies included data on the number of harvested lymph nodes, 3 of which 22,25,26 showed no significant difference between HALS and and 1 of which 24 showed significantly more harvested lymph nodes with HALS than with. However, only 1 study 26 included data in the form of means and standard deviations, so a metaanalysis was not appropriate. Postoperative pain Because the presentation of data was heterogeneous, with a lack of a standardized protocol for analgesia, a metaanalysis was not appropriate. Of the included studies, 3 reported less postoperative pain after HALS than after, 9,22,25 and 1 showed that the difference was not significant. 18 Time to first flatus Four studies 9,20,25,26 demonstrated the time to first flatus, and all showed a significantly shorter time with HALS than with. The random-effects model was used to combine the data because heterogeneity was evident. In the pooled data, the time to first flatus in HALS was significantly earlier than with (WMD, 20.94;, to 20.65; P,.05; Fig. 5). Postoperative period of hospital stay Six studies 9,17,20,21,25,26 showed data on the postoperative period of hospital stay, and all showed a significantly shorter time with HALS than with. The randomeffects model was used to combine the data because of evident heterogeneity. In the pooled data, the length of hospital stay after HALS was significantly shorter than after (WMD, 23.22;, to 22.57; P,.05; Fig. 6). Number of complications Data on postoperative complications, including wound infection, 8,9,18,20 22,24 26 urinary tract infection, 9,18,21,24 pneumonia, 18,20 22,24,26 anastomotic leak, 8,18,21,22,25,26 and ileus, 8,9,18,20,21,25,26 were collected and analyzed. The fixed-effects model was used to combine the data because heterogeneity was not evident. The HALS group and group had similar incidences of urinary tract infection (OR, 0.58;, 0.15 to 2.20; P 5.43; Fig. 7), pneumonia (OR, 0.46;, 0.16 to 1.35; P 5.16; Fig. 8), and anastomotic leak (OR, 0.95;, 0.40 to 2.27; P 5.91; Study H ALS WMD (random) Weight WMD (random) or sub-category N Mean (SD) N M ean (SD) % Test for heterogeneity: Chi?= 9.06, df = 3 (P = 0.03), I?= 66.9% Test for overall effect: Z = 6.53 (P < ) Figure 5 Forest plot showing time to flatus between HALS and.

6 114 The American Journal of Surgery, Vol 207, No 1, January 2014 Study H ALS WMD (random) Weight WMD (random) or sub-category N Mean (SD) N M ean (SD) % Dunker 2000 Test for heterogeneity: Chi?= 12.06, df = 5 (P = 0.03), I?= 58.5% Test for overall effect: Z = 9.70 (P < ) Figure 6 Forest plot showing hospital stay between HALS and. Fig. 9), while the rates of wound infection (OR, 0.45; 95% CI, 0.23 to 0.87; P 5.02; Fig. 10) and ileus (OR, 0.35; 95% CI, 0.16 to 0.74; P 5.006; Fig. 11) were significantly less in the HALS group than in the group. Mortality rate Eight studies 8,9,17,20 25 had data on mortality rate and showed no significant difference between HALS and. A total of 11 patients died. The major causes of death were cardiopulmonary events (8 cases) and sepsis (3 cases). The fixed-effects model was used to combine the data because heterogeneity was not evident. In the pooled data, there were no significant differences in the mortality rate between HALS and (OR, 0.68;, 0.19 to 2.36; P 5.54; Fig. 12). Hospitalization costs Only 2 studies 8,18 had data on hospitalization costs. One study 18 showed no significant difference between the HALS and groups (V16,728 vs V13,405, P 5.095), and another 8 showed a significantly lower hospital charges in the HALS group despite lower operative charges for procedures ($23,132 vs $33,150, P,.01). Assessment of publication bias The funnel plot of standard of error by effect size for the measurements of wound infection and pneumonia showed an equal distribution of studies around the middle line, indicating that publication bias was not evident (Fig. 13). Sensitivity analysis Six outcomes were selected for sensitivity analysis, and the results are listed in Table 3. This analysis showed that the OR or WMD and the level of significance for 5 outcomes (operative time, time to first flatus, length of hospital stay, pneumonia, and ileus) were not significantly affected, whereas the WMD and the level of significance for wound infection were significantly different in randomized controlled trials and studies with only malignant disease. Additionally, the wound infection was not significantly different in the HALS group compared with the group. Comments HALS was first reported in and has been widely used in treatment of patients with colorectal disease. Several systematic reviews and meta-analyses have indicated that HALS can provide a more efficient segmental colectomy regarding operating time and conversion rate compared with laparoscopic surgery and suggested that it should be considered a valuable addition to the laparoscopic armamentarium to avoid conversion and speed up complicated colectomies. 27,28 However, compared with, it remains to be established whether HALS can maintain the advantages of minimally invasive techniques. Study HALS Weight or sub-category n/n n/n % Maartense 2004 Osarogiagbon 2007 Total events: 2 (HALC), 5 (OC) Test for heterogeneity: Chi?= 1.56, df = 3 (P = 0.67), I?= 0% Test for overall effect: Z = 0.80 (P = 0.43) Figure 7 Forest plot showing urinary tract infection between HALS and.

7 J. Ding et al. HALS versus : a meta-analysis 115 Study HALS Weight or sub-category n/n n/n % Maartense 2004 Chung 2007 Osarogiagbon 2007 Total events: 3 (HALC), 9 (OC) Test for heterogeneity: Chi?= 2.91, df = 5 (P = 0.71), I?= 0% Test for overall effect: Z = 1.41 (P = 0.16) Figure 8 Forest plot showing pneumonia between HALS and. In this meta-analysis, we collected 12 clinical studies that compared HALS with to investigate the value of HALS in the treatment of colorectal disease. The heterogeneity of different lesion sites, different lesion types, and different study types was overcome by sensitivity analysis to combine the data more effectively. Meanwhile, 5 common complications, rather than the overall complication rate, were selected for analysis in this meta-analysis for a better evaluation of the safety of HALS. A total of 12 studies involved 13 cases of conversion to ; the conversion rate was 2.66% and ranged from 0% to 10.26% in the other studies, which was significantly lower than that in laparoscopic-assisted surgery. 27,28 The most common reasons were dense adhesions (2 cases), ureteral injury (1 case), and phlegmon and bleeding (6 cases). Remarkably, 9 studies showed no conversion cases in the 495 patients who underwent HALS colorectal resection. Anderson et al 21 even reported that part of laparoscopicassisted surgery can be converted to HALS to complete the operation, which indicates that HALS does have a particular advantage over conventional laparoscopic surgery on conversion rate. This meta-analysis of 5 clinical studies revealed that the operative time was not significantly different between the HALS and groups, which is not in accordance with Aalbers et al s 28 conclusion. The authors considered that it is the maturity of hand-assisted laparoscopic skills lead to the different conclusion, with the surgeon s accumulation of cases, familiarity with the endoscopic instruments, and cooperation of the whole therapeutic team, the mean operative time for HALS would decrease significantly. Furthermore, the sensitivity analysis also proved the similar operative time between HALS and. Our meta-analysis revealed significantly less blood loss during HALS than during. This might be attributed to its advantages of a minimally invasive and intuitive approach and the ability to rapidly control unexpected bleeding. In addition to this, we believe that the use of special equipment, including ultrasonic scalpels and ligatures, also contributed to this outcome. Orenstein et al 8 even reported that the probability of blood transfusion was significantly lower in the HALS group than in the group (0% vs 3.8%). Blood loss of.250 ml and blood transfusion during surgery were risk factors for overall mortality of patients with colon cancer. 32 Areduction in the amount of bleeding in patients with colorectal cancer is desirable, because it reduces the likelihood or amount of transfusion and reduces the immunosuppression caused by transfusion and cancer recurrence. 33,34 The meta-analysis also revealed that the incision length for HALS was significantly shorter than that for. Typical open incisions approximately double those needed for HALS, while the incision for HALS generally averaged only 4 to 8 cm to accommodate the surgeon s hand and allow for specimen removal. 8,26 The small incisions can not Study HALS Weight or sub-category n/n n/n % Maartense 2004 Chung 2007 Orenstein 2011 Total events: 9 (HALC), 16 (OC) Test for heterogeneity: Chi?= 2.57, df = 4 (P = 0.63), I?= 0% Test for overall effect: Z = 0.12 (P = 0.91) Figure 9 Forest plot showing anastomotic leak between HALS and.

8 116 The American Journal of Surgery, Vol 207, No 1, January 2014 Study HALS Weight or sub-category n/n n/n % Maartense 2004 Chung 2007 Osarogiagbon 2007 Orenstein 2011 Total events: 13 (HALC), 38 (OC) Test for heterogeneity: Chi?= 2.93, df = 8 (P = 0.94), I?= 0% Test for overall effect: Z = 2.40 (P = 0.02) Figure 10 Forest plot showing wound infection between HALS and. only cause less abdominal wall trauma, and achieve better cosmetic results, but they also carry a smaller risk for infection, hernia, and dehiscence. 20,35 Given the same therapeutic efficacy, patients are more willing to accept smaller incision. Because of the smaller incisions and lower trauma of HALS, patients felt less pain after surgery and took fewer anesthetic drugs. Our meta-analysis revealed that the time to first flatus and the postoperative period of hospital stay were both shorter for HALS than for. Time to oral intake as a measure of recovery of gastrointestinal function is prone to bias because surgeons with an interest in laparoscopic surgery are likely to feed their patients earlier than those who perform. 36 Although time to flatus is an objective indicator, it indicated that HALS was more conducive to recovery of gastrointestinal function. Many complications may occur after colectomy. Relatively common complications were selected in this study, including wound infection, urinary tract infection, pneumonia, anastomotic leak, and ileus. Our meta-analysis revealed that the HALS group and group had similar incidences of urinary tract infection, pneumonia, and anastomotic leak, while the rates of wound infection and ileus were significantly less in the HALS group than in the group. Some studies have suggested that the hand port device may serve as a barrier to wound contamination, although more investigation is needed to support this hypothesis. 7,37 Many factors can lead to postoperative ileus, 38 but no selected studies explained reasons for the low incidence of ileus in the HALS groups. Possibly the lower inflammation response and the use of fewer narcotic analgesics after HALS led to this outcome. Perhaps the faster recovery of gastrointestinal function, and the lower incidence of wound infection and ileus, contributed to the shorter period of hospitalization. Our meta-analysis revealed no significant difference in mortality; thus, we can conclude that HALS is safe and feasible and even has the advantage of reducing the incidence of complications. Hospitalization costs remain an important issue in current health care. Only 2 studies 8,18 included data on hospitalization costs. One study 18 showed no significant difference between the 2 groups, and another 8 showed significantly lower hospital charges in the HALS group. Although the costs of disposable surgical equipment (eg, trocars, hand port, energy sources) led to the higher operative charges for the hand-assisted procedures, the total hospital charges were similar or even less for the HALS group, which may be due to fewer complications and shorter hospital stays in Study HALS Weight or sub-category n/n n/n % Maartense 2004 Zhang LY 2006 Orenstein 2011 Total events: 6 (HALS), 41 () Test for heterogeneity: Chi?= 4.47, df = 6 (P = 0.61), I?= 0% Test for overall effect: Z = 2.74 (P = 0.006) Figure 11 Forest plot showing ileus between HALS and.

9 J. Ding et al. HALS versus : a meta-analysis 117 Study HALS Weight or sub-category n/n n/n % Dunker 2000 Chung 2007 Osarogiagbon 2007 Polle 2007 Orenstein 2011 Total events: 2 (HALS), 10 () Test for heterogeneity: Chi?= 2.67, df = 3 (P = 0.45), I?= 0% Test for overall effect: Z = 0.61 (P = 0.54) Figure 12 Forest plot showing mortality between HALS and. Review: Hand-assisted Laparoscopic Surgery Versus Open Surgery for Colorectal Disease: A Systematic Review and Meta-analysis Comparison: 05 Complications Outcome: 01 Wound infection 0.0 SE(log OR) Review: Hand-assisted Laparoscopic Surgery Versus Open Surgery for Colorectal Disease: A Systematic Review and Meta-analysis Comparison: 05 Complications Outcome: 03 Pneumonia 0.0 SE(log OR) Figure 13 Funnel plot for the results from all studies comparing wound infection and pneumonia in patients undergoing HALS versus for colorectal disease.

10 118 The American Journal of Surgery, Vol 207, No 1, January 2014 Table 3 Sensitivity analysis of included studies No. of patients Outcomes HALS No. of studies OR/WMD P Randomized controlled trials Operative time to Time to first flatus to 2.84,.05 Length of hospital stay to 22.29,.05 Wound infection to * Pneumonia to Ileus to Studies with only colon disease Operative time to Time to first flatus to 2.39,.05 Length of hospital stay to 22.16,.05 Wound infection to Pneumonia to Ileus to Studies with only malignant disease Operative time to Time to first flatus to 2.65,.05 Length of hospital stay to 22.18,.05 Wound infection to * Pneumonia to Ileus to CI 5 confidence interval; HALS 5 hand-assisted laparoscopic surgery; OR 5 odds ratio; 5 open surgery; WMD 5 weighted mean difference. *No statistically significant difference. HALS group. Furthermore, many energy sources and stapling devices previously used only for the laparoscopic techniques currently are used routinely for open procedures, thus narrowing the difference in operative instrumentation costs between standard and minimally invasive procedures. 8 There were some limitations to this meta-analysis. First, data on the number of harvested lymph nodes, positive margin rate, recurrence rate, and 5-year survival rate were seldom provided, so it is impossible to assess the oncologic safety and long-term prognosis of HALS. Second, different studies may have had different defining criteria for the outcomes we were interested in, which may not be reported in the study methodology. To account for this heterogeneity, we have performed sensitivity analysis. Third, because of the inclusion of nonrandomized studies, there is an inherent selection bias in the 2 groups. Finally, there is the limitation of nonpublication and selective reporting bias, which cannot be accounted for. 39 Conclusions This study suggests that HALS is a safe technical alternative to for patients with colorectal disease. It involves minimal invasion, lower blood loss, shorter incision length, faster postoperative recovery, and fewer complications such as ileus, and it can shorten the length of hospitalization without increasing costs. Furthermore, this technique is not associated with increased operative time compared with the open procedure. The drawbacks are that a small number of patients who undergo HALS may need to be converted to laparotomy, and the oncologic safety and long-term prognosis are not clear. In addition, whether HALS has a lower wound infection rate remains to be confirmed. Most of the studies included in our meta-analysis were retrospective studies, which could have led to some selection bias. Therefore, more well-designed, multicenter, prospective randomized controlled trials are expected to be published to allow a more convincing evaluation. Acknowledgments This study was supported by the National Natural Science Foundation of China (No and No ), and the Science and Technology Fund of Guizhou Province (No. [2013]2178). References 1. Jacobs M, Verdeja JC, Goldstein HS. Minimally invasive colon resection (laparoscopic colectomy). Surg Laparosc Endosc 1991;1: Schwenk W, Haase O, Neudecker J, et al. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005;3:CD Schlachta CM, Mamazza J, Seshadri PA, et al. Defining a learning curve for laparoscopic colorectal resections. Dis Colon Rectum 2001;43: Tekkis PP, Senagore AJ, Delaney CP, et al. Evaluation of the learning curve in laparoscopic colorectal surgery: comparison of right-sided and left-sided resections. Ann Surg 2005;242: Liang JT, Lai HS, Lee PH. Laparoscopic pelvic autonomic nervepreserving surgery for patients with lower rectal cancer after chemoradiation therapy. Ann Surg Oncol 2007;14:

11 J. Ding et al. HALS versus : a meta-analysis Li JC, Hon SS, Ng SS, et al. The learning curve for laparoscopic colectomy: experience of a surgical fellow in an university colorectal unit. Surg Endosc 2009;23: Meijer DW, Bannenberg JJ, Jakimowicz JJ. Hand-assisted laparoscopic surgery: an overview. Surg Endosc 2000;14: Orenstein SB, Elliott HL, Reines LA, et al. Advantages of the handassisted versus the open approach to elective colectomies. Surg Endosc 2011;25: Kang JC, Chung MH, Chao PC, et al. Hand- assisted laparoscopic colectomy vs open colectomy: a prospective randomized study. Surg Endosc 2004;18: Milsom JW, Böhm B, Hammerhofer KA, et al. A prospective, randomized trial comparing laparoscopic versus conventional techniques in colorectal cancer surgery: a preliminary report. J Am Coll Surg 1998;187: Kusminsky RE, Boland JP, Tiley EH, et al. Hand-assisted laparoscopic splenectomy. Surg Laparosc Endosc 1995;5: Ballaux KE, Himpens JM, Leman G, et al. Hand-assisted laparoscopic splenectomy for hydatid cyst. Surg Endosc 1997;11: Ozturk E, Kiran RP, Remzi F, et al. Hand-assisted laparoscopic surgery may be a useful tool for surgeons early in the learning curve performing total abdominal colectomy. Colorectal Dis 2010;12: Aalbers AG, Doeksen A, Van Berge Henegouwen MI, et al. Hand-assisted laparoscopic versus open approach in colorectal surgery: a systematic review. Colorectal Dis 2010;12: Athanasiou T, Al-Ruzzeh S, Kumar P, et al. Off-pump myocardial revascularization is associated with less incidence of stroke in elderly patients. Ann Thorac Surg 2004;77: DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7: Dunker MS, Bemelman WA, Slors JF, et al. Laparoscopic-assisted vs open colectomy for severe acute colitis in patients with inflammatory bowel disease (IBD): a retrospective study in 42 patients. Surg Endosc 2000;14: Maartense S, Dunker MS, Slors JF, et al. Hand-assisted laparoscopic versus open restorative proctocolectomy with ileal pouch anal anastomosis: a randomized trial. Ann Surg 2004;240: Zhang LY. Hand-assisted laparoscopic vs. open total colectomy in treating slow transit constipation. Tech Coloproctol 2006;10: Chi P, Lin HM, Chen YC, et al. Feasibility of lymphadenectomy with skeletonization in extended right hemicolectomy by hand-assisted laparoscopic surgery. Zhonghua Wei Chang Wai Ke Za Zhi 2005;8: Anderson J, Luchtefeld M, Dujovny N, et al. A comparison of laparoscopic, hand-assist and open sigmoid resection in the treatment of diverticular disease. Am J Surg 2007;193: Chung CC, Ng DC, Tsang WW, et al. Hand-assisted laparoscopic versus open right colectomy: a randomized controlled trial. Ann Surg 2007;246: Polle SW, Dunker MS, Slors JF, et al. Body image, cosmesis, quality of life, and functional outcome of hand-assisted laparoscopic versus open restorative proctocolectomy: long-term results of a randomized trial. Surg Endosc 2007;21: Osarogiagbon RU, Ogbeide O, Ogbeide E, et al. Hand-assisted laparoscopic colectomy compared with open colectomy in a nontertiary care setting. Clin Colorectal Cancer 2007;6: Liu FL, Lin JJ, Ye F, et al. Hand-assisted laparoscopic surgery versus the open approach in curative resection of rectal cancer. J Int Med Res 2010;38: Zhang H, Li M, Zhan TC, et al. Comparison of short-term postoperative outcomes between hand-assisted laparoscopic and conventional sigmoidectomy. Zhonghua Wei Chang Wai Ke Za Zhi 2011;14: Moloo H, Haggar F, Coyle D, et al. Hand assisted laparoscopic surgery versus conventional laparoscopy for colorectal surgery. Cochrane Database Syst Rev 2010;6:CD Aalbers AG, Biere SS, van Berge Henegouwen MI, et al. Hand-assisted or laparoscopic-assisted approach in colorectal surgery: a systematic review and meta-analysis. Surg Endosc 2008;22: Cima RR, Pattana-arun J, Larson DW, et al. Experience with 969 minimal access colectomies: the role of hand-assisted laparoscopy in expanding minimally invasive surgery for complex colectomies. J Am Coll Surg 2008;206: Meshikhes AW. Controversy of hand-assisted laparoscopic colorectal surgery. World J Gastroenterol 2010;16: Cima RR, PendlimariR, Holubar SD, etal. Utility and short-termoutcomes of hand-assisted laparoscopic colorectal surgery: a single-institution experience in 1103 patients. Dis Colon Rectum 2011;54: Mörner ME, Gunnarsson U, Jestin P, et al. The importance of blood loss during colon cancer surgery for long-term survival: an epidemiological study based on a population based register. Ann Surg 2012;255: Amato A, Pescatori M. Perioperative blood transfusions for the recurrence of colorectal cancer. Cochrane Database Syst Rev 2006;25:CD Mortensen FV, Jensen LS, Sørensen HT, et al. Cause-specific mortality associated with leukoreduced, buffy coat-depleted, or no blood transfusion after elective surgery for colorectal cancer: a posttrial 15-year follow-up study. Transfusion 2011;51: Senagore AJ, Stulberg JJ, Byrnes J, et al. A national comparison of laparoscopic vs open colectomy using the National Surgical Quality Improvement Project data. Dis Colon Rectum 2009;52: Aziz O, Constantinides V, Tekkis PP, et al. Laparoscopic versus open surgery for rectal cancer: a meta-analysis. Ann Surg Oncol 2006;13: Kurian MS, Patterson E, Andrei VE, et al. Hand-assisted laparoscopic surgery: an emerging technique. Surg Endosc 2001;15: Johnson MD, Walsh RM. Current therapies to shorten postoperative ileus. Cleve Clin J Med 2009;76: Venkat R, Edil BH, Schulick RD, et al. Laparoscopic distal pancreatectomy is associated with significantly less overall morbidity compared to the open technique: a systematic review and meta-analysis. Ann Surg 2012;255: