Enhanced Recovery After Surgery for Noncolorectal Surgery?

A Systematic Review and Meta-analysis of Major Abdominal Surgery

Anthony Visioni, MD; Rupen Shah, MD; Emmanuel Gabriel, MD, PhD; Kristopher Attwood, PhD; Moshim Kukar, MD; Steven Nurkin, MD


Annals of Surgery. 2018;267(1):57-65. 

In This Article


Selection of Studies

A systematic review of published literature was performed in accordance with Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) guidelines.[9] A list of studies was obtained by searching the following databases: PubMed, Scopus, and ClinicalTrials.gov. The following search was used for each database: "fast track" or "enhanced recovery after surgery" or "ERAS" but not "colorectal," "colon," and "rectal." Studies were limited to English language and the last search date was September 1, 2016. Data from identified studies were extracted independently by 2 authors (AV and RS) using a standardized format.

Inclusion and Exclusion Criteria

Studies were included in the systematic review if the following criteria were met: studies that involved abdominal/pelvic surgery, approached either open or with minimally invasive techniques, studies that outlined an enhanced recovery pathway that included at least 5 perioperative items that were distinct from traditional care, studies that included either a historical or randomized control group; and studies that reported outcomes such as total length of hospital stay, complication rates, readmission rates, time to passage of flatus, and/or costs.

Studies were excluded for any of the following: studies of colorectal, vascular, emergency surgeries, studies that focused on only a single aspect of the enhanced recovery pathway, studies that were reviews, meta-analyses, case reports, letters, or protocols, studies that did not report key outcomes; and those in which data was inadequate for interpretation via meta-analysis.

Assessment of Bias

Identified studies were broadly grouped into 1 of 2 types, either randomized trials or cohort studies. Cohort studies were assessed for bias using the Newcastle-Ottawa Scale[10] and in accordance with the accompanying coding manual. Randomized trials were assessed based on the Cochrane Handbook.[11] Two authors (AV and RS) were independently responsible for assessments of bias. Funnel plots were performed for all outcome measures to evaluate for possible publication bias.

Selection of Outcomes

Primary endpoints for this study included LOS and proportion of patients who experienced a complication. Studies must have reported at least one of these metrics in a suitable format to be included in the meta-analysis. For LOS, the format mean +/- standard deviation was utilized. For complications, we used percentage of patients with any complication as defined by the authors of the original study (readmissions were not included and, as such, evaluated separately). Secondary endpoints were time to first flatus, percent of patients readmitted, and total hospital costs. Hospital costs were determined according to the authors of the original studies and were converted into US dollars and adjusted for inflation as of September 2016. For any studies that were missing data or did not have data in the correct format, corresponding authors were emailed and the missing data was requested.

Statistical Analysis

The study outcomes (mean difference in LOS, time to first flatus and cost; and odds ratios (OR) for complications and readmissions) were reported by study using 95% confidence intervals (CI) obtained using standard techniques; and displayed graphically using forest plots. The results of the studies were combined in a meta-analysis using standard random effects models, from which estimates of the average mean difference (ERAS – Control) or odds ratio (ERAS vs Control) were obtained, with 95% CI. To assess the heterogeneity of the study outcomes, the Q and I2 statistics were obtained. The Q statistic evaluates the homogeneity assumption (ie, all studies have the same common mean difference or OR); whereas the I2 statistic represents the amount of variability in the meta-analysis attributed to study heterogeneity.

All analyses were conducted in SAS v9.4 (Cary, NC) at a significance level of 0.05.