Two reviewers (Han-Teng Yang and Hao Zhang) independently performed a literature search in the PubMed, EMBASE, China National Knowledge Infrastructure (CNKI) and Cochrane Central Register of Controlled Trials databases for all RCTs that assessed the impact of ERAS programs on patients following hepatectomy in comparison to traditional care, which had been published in 1966 through November 8, 2014. We did not apply language restrictions. The following search terms were used: fast track, fast-track, enhanced recovery, liver, hepatic, hepatocellular, hepatectomy, resection, surgery, surgical, randomized controlled trial, randomized, randomly and clinical trial. Synonyms of each of the terms were also used in the search. Abstracts from each of the studies were reviewed. In addition, the references from each of the retrieved articles were manually screened to identify other potential eligible RCTs.

Studies selected for the meta-analysis met all of the following inclusion criteria: (1) the study evaluated ERAS in comparison to traditional care in adult patients undergoing elective open or laparoscopic liver surgery; (2) the outcome measures included complications, recovery of bowel function, and hospital LoS; and (3) the study must be an RCT. The following exclusion criteria were applied: (1) the study was not an RCT; (2) the study did not compare ERAS with traditional care; (3) the study reported on emergency, non-elective or transplantation surgery; and (4) the study consisted of unpublished data with only the abstract presented at a national or international meeting. For different publications with overlapping data, the most complete publication was selected. Two authors (Tian-Gen Ni and Bo L) independently assessed the articles for compliance with the inclusion/exclusion criteria, resolved disagreements, and reached a unified decision.

Two researchers (Han-Teng Yang and Hao Zhang) independently reviewed and extracted the following information from each of the studies: first author’s name, publication year, country, total number of cases and controls, age, sex, type of surgery, outcome measures, and number of ERAS program items according to the guidelines established by the ERAS group[15]. The core elements of ERAS include preadmission information and counseling, preoperative bowel preparation, preoperative fasting and carbohydrate loading, preanesthetic medication, prophylaxis against thromboembolism, antimicrobial prophylaxis, standard anesthetic protocol, preventing and treating postoperative nausea and vomiting, laparoscopy-assisted surgery, surgical incisions, nasogastric intubation, preventing intraoperative hypothermia, perioperative fluid management, drainage of the peritoneal cavity, urinary drainage, preventing postoperative ileus, postoperative analgesia, postoperative nutritional care and early mobilization. If there were any disagreements between the two reviewers, a third reviewer (Hai-Peng Meng) was recruited and the issue was discussed until a consensus was achieved. When multiple publications reported on the same or overlapping data, we selected the study with the most complete dataset.

The methodology for each RCT was assessed independently by two reviewers (Tian-Gen Ni and Han-Teng Yang) according to the Cochrane Collaboration’s risk of bias tool[16,17], which analyzes the following criteria: (1) random sequence generation; (2) allocation concealment; (3) blinding of participants and personnel; (4) blinding of outcome assessment; (5) incomplete outcome data; (6) selective reporting; and (7) other bias. For each entry based on the risk of bias assessment guidelines, we made a judgment (low risk of bias, high risk of bias, or uncertain). All disagreements were resolved by discussion until a consensus was achieved.

Complications (defined according to the Dindo-Clavien classification[18]) were the primary outcome measure. Secondary outcome measures included: (1) hospital LoS (defined as the number of days in the hospital after surgery until discharge); and (2) time to first flatus.

Meta-analyses were performed by using risk ratios (RRs) for dichotomous outcomes, and weighted mean differences (WMDs) were used for continuous outcomes. Pooled estimates were presented with 95%CIs. If the included studies provided medians and interquartile ranges, we calculated the mean ± SD according to the methods outlined by Hozo et al[19]. When heterogeneity was found to be statistically significant (P < 0.05 or I² > 50%[20,21]), a random effects model was applied. Otherwise, a fixed effects model was adopted to calculate the pooled RRs or WMDs. Funnel plots were generated to determine the presence of publication bias. When a study presented with significant heterogeneity, sensitivity analyses were performed to assess how inferring standard deviations from medians and interquartile ranges from poor quality studies affected the overall results. We further identified sources of heterogeneity and assessed the robustness and consistency of statistical techniques used. For all other comparisons, statistical significance was defined by P < 0.05 and all tests were two-tailed. All statistical analyses were performed using the Review Manager (RevMan) software, version 5.3 from the Cochrane Collaboration (http://tech.cochrane.org/revman). Some outcomes were not analyzed but instead were presented as descriptive information.

After the initial literature search, a total of 101 potentially relevant studies were identified. The final meta-analysis, after application of inclusion and exclusion criteria, included five RCTs with a total of 723 patients[22-26]. In 10/723 cases, patients presented with benign diseases, while the remaining 713 cases had liver cancer. Of the five studies, three were published in English and two were published in Chinese. The PRISMA flow diagram is shown in Figure 1. Three hundred and fifty-four patients were in the ERAS group, while 369 patients were in the traditional care group. The sample size for the included studies ranged from 60 to 297 patients. The included RCTs were published between 2012 and 2014 and were conducted solely in adult patients. There were no multicenter trials. Characteristics of each included RCT are presented in Table 1. Each reviewer performed an assessment of risk of bias of each methodological component. The risk of bias summary for the included RCTs is presented in Figure 2.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Preferred reporting items for systematic reviews and meta-analyses flow diagram.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Risk of bias summary in included studies. The symbol of (-) red indicates that there is a high risk of bias, of (+) green indicates a low risk of bias and of (?) yellow indicates uncertainty.

Overall complications were reported in all included studies. Because we did not identify significant heterogeneity among the trials (I² = 0%, P = 0.93), a fixed-effects model was applied to this meta-analysis. In the ERAS group, there was a significant reduction in overall complications (RR = 0.66, 95%CI: 0.49-0.88; P = 0.005) (Figure 3). Using the Dindo-Clavien classification[18], information regarding grade I complications and grade II-V complications was available from the five studies. In the fixed-effects model, the ERAS group had significantly fewer grade I complications (RR = 0.51, 95%CI: 0.33-0.79; P = 0.003), without heterogeneity among the trials (I² = 0%, P = 0.44) (Figure 3). However, no differences in grade II-V complications were found between the ERAS and traditional care groups (RR = 0.94, 95%CI: 0.58-1.52; P = 0.80), without heterogeneity among the trials (I² = 0%, P = 0.58) (Figure 3).

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Forest plot of enhanced recovery after surgery programs vs traditional care for patient complications. ERAS: Enhanced recovery after surgery; TC: Traditional care.

In all included RCTs, hospital LoS was reported and was significantly shorter for the ERAS group than the traditional care group [WMD = -2.97 d, 95%CI: -3.18-(-2.76); P < 0.00001]. There was, however, significant heterogeneity among the trials (I² = 91%, P < 0.00001). Due to this heterogeneity, a random effects model was applied to the studies. This analysis demonstrated that the ERAS group had a shorter hospital LoS [WMD = -2.77 d, 95%CI: -3.87-(-1.66); P < 0.00001] (Figure 4A). After one study was excluded[24] due to standard deviations from medians and interquartile ranges, a sensitivity analysis did not substantially change the results of the original analysis.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Forest plots of enhanced recovery after surgery programs vs traditional care for hospital length of stay (A) and time to first flatus (B). ERAS: Enhanced recovery after surgery; TC: Traditional care.

Time to first flatus was reported in 4/5 studies. The heterogeneity among the trials was significant (I² = 99%, P < 0.00001), thus a random effects model was applied. In the ERAS group, time to first flatus was significantly reduced in comparison to those undergoing traditional care [WMD = -19.69 h, 95%CI: -34.63-(-4.74); P < 0.0001] (Figure 4B).

Because less than 10 studies were included in this meta-analysis, we did not evaluate the publication bias for ERAS programs in patients undergoing liver surgery. In accordance with the guidelines established by the Cochrane Handbook for Systematic Reviews, the test for publication bias is unreliable when less than 10 studies are included in a meta-analysis[27].

Enhanced recovery after surgery (ERAS) programs have been successfully implemented in different surgical fields to improve postoperative outcomes. Despite a number of studies evaluating ERAS programs in liver surgery, their safety and effectiveness have not been systematically evaluated.

In recent years, ERAS programs have been successfully implemented in colorectal, gastric and gynecologic surgical procedures. ERAS programs have also been used for hepatic surgery. This meta-analysis was performed to evaluate the safety and effectiveness of ERAS programs on liver surgery outcomes. The outcome measures in this study included complications, hospital length of stay, and the time to first flatus.

The present meta-analysis indicates that ERAS programs are safe and effective in liver surgery. ERAS reduces overall complication rates, accelerates postoperative recovery, and shortens hospital length of stay, without increasing surgical complication rates. Inconsistent and subjective outcome measures and ERAS programs that were not specific to the livery surgery field still represent limitations. Future studies should develop ERAS protocols and outcome measures optimized for the liver surgery field.

The results of this meta-analysis demonstrated the enhanced clinical effectiveness of ERAS programs in comparison to traditional care in liver surgery. ERAS programs result in a significant reduction in complications and recovery of intestinal function. Because the hospital length of stay is shortened after ERAS implementation, the authors infer that there is also a simultaneous reduction in associated hospital costs. Thus, the implementation of ERAS programs contributes to rapid postoperative recovery in liver surgery patients.

In this meta-analysis of randomized controlled trials, the authors found a positive impact of ERAS programs on liver surgery outcomes when compared with traditional care. This study was well conducted and meets all of the standard requirements for a meta-analysis. The study results are clear, reliable and clinically relevant. The findings from this study are novel and applicable to a wide readership audience.