A meta-analysis of fast track surgery for patients with gastric cancer undergoing gastrectomy

RCTs included in the meta-analysis met the following criteria: (1) comparison of the FTS pathway with standard care for patients with gastric cancer undergoing elective gastrectomy; (2) reporting of at least one relevant outcome (length of primary postoperative hospital stay, hospitalisation expenditure; and/or short-term morbidity, readmission; and/or mortality rates); and (3) clear documentation of a multimodal FTS pathway composed of at least seven elements. Given the lack of a consensus guideline for FTS programme implementation in patients undergoing gastric surgery, we made the arbitrary decision that such programmes should include at least 7/20 items addressed in the guidelines published by a working group related to FTS.¹⁰ This number of FTS elements was selected because fewer elements might represent ‘modern’ traditional care.

When two articles reported on the same clinical trial, the publication providing the most information was selected. Non-randomised studies, case-controlled trials, cohort studies, retrospective studies and other studies that did not fulfil the inclusion criteria were excluded.

The primary endpoints were the length of primary postoperative hospital stay and hospitalisation expenditure. The secondary endpoints were short-term morbidity, readmission and mortality rates.

This meta-analysis was performed in accordance with the recommendations of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement.²² Two reviewers evaluated all retrieved studies independently to determine selection, assess study quality and extract data. Disagreements were resolved through discussion and consensus of the study team.

Two reviewers independently assessed RCT quality and risk of bias using tools provided by the Cochrane Collaboration.²³ The reviewers examined randomisation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, baseline comparability of groups, completeness of outcome data and selective outcome reporting. Baseline comparability of groups was assessed using six matching criteria: age, sex, body mass index (BMI), ASA (American Society of Anesthesiologists) grade or Eastern Cooperative Oncology Group score, type of reconstruction and pathological TNM (tumour, lymph nodes, metastasis) stage. Baseline incomparability was defined as non-matching in at least three of these six criteria.

The risk of bias in each domain was categorised as low, high or unclear. Trials with high risk of bias in at least four of seven domains were not included in the meta-analysis.

Statistical analyses were performed using RevMan version 5.0 (Nordic Cochrane Centre, Copenhagen, Denmark). Weighted mean differences (WMDs) with 95% confidence intervals (CIs) were calculated for continuous variables (length of postoperative hospital stay and hospitalisation expenditure) and odds ratios (ORs) with 95% CIs were calculated for dichotomous variables (short-term morbidity, readmission and mortality). A random effects model was used to pool studies with significant heterogeneity, as determined by the chi-squared test (p≤0.10) and the inconsistency index (I²≥50%). Quality of life data could not be pooled because of incompatibility in measurement among studies but they are presented descriptively.

Sensitivity analysis was also performed by excluding one trial at a time to assess its effect on the overall estimate. Bias exploration using a funnel plot was attempted but was not possible because of the small number of trials included in this review. Subgroup analyses were performed of: 1) FTS versus conventional care with/without laparoscopy, to explore the role of laparoscopy in FTS for patients undergoing gastrectomy, and 2) FTS versus conventional care in patients undergoing total, distal, proximal or mixed gastrectomy, to explore the role of the extent of gastric resection in FTS.

All seven RCTs reported this outcome measure^11–17 but data from only five trials were available for meta-analysis.^{11–13,16,17} Postoperative hospital stays were significantly shorter for patients receiving FTS treatment than for those receiving conventional perioperative care (WMD: -2.62 days; 95% CI: -3.59–-1.65 days, p<0.00001), with significant heterogeneity among studies (I²=71%, p=0.009) (Fig 3).

Six articles reported hospitalisation expenditure.^11,13–17 Meta-analysis demonstrated that expenditure was significantly lower for the FTS group than for the conventional perioperative care group (WMD: -0.39 × 10⁴ Chinese yuan, 95% CI: -0.52–-0.26 × 10⁴ Chinese yuan, p<0.00001), with significant heterogeneity among studies (I²=57%, p=0.03) (Fig 4).

Postoperative morbidity rates were reported in all seven studies.^11–17 Meta-analysis demonstrated comparable postoperative morbidity between patient groups (OR: 0.82, 95% CI: 0.40–1.66, p=0.58), with significant heterogeneity among studies (I²=57%, p=0.02) (Fig 5). Readmission rates, reported in four studies,^12,14,16,17 were also comparable (OR: 2.01, 95% CI: 0.36–11.29, p=0.43), with no significant heterogeneity among studies (I²=0%, p=0.85) (Fig 6). No death occurred during the follow-up period in any RCT.

Wang et al ¹⁴ and Kim et al ¹⁶ found that the FTS pathway was significantly superior to the conventional care system, in terms of quality of life at discharge and two weeks thereafter respectively. Wang et al ¹⁴ and Feng et al ¹⁷ reported lower pain intensity in the FTS group than in the conventional care group on postoperative days 1–5, as measured with a visual analogue scale. Kim et al reported similar pain intensities in both groups but less need for additional pain control in the FTS group.¹⁶

Sensitivity analysis determined that no single RCT significantly affected the overall results of the meta-analysis. Subgroup analyses of FTS versus conventional care with/without laparoscopy yielded similar results (Figs 3–6). Subgroup analyses of FTS versus conventional care focusing on the extent of gastric resection also yielded similar results for postoperative hospitalisation, hospitalisation expenditure and short-term readmission rates although the short-term morbidity showed some differences. The short-term morbidity was significantly less for patients who received FTS treatment than for patients who received conventional perioperative care in the total gastrectomy subgroup. However, this parameter was comparable for the distal gastrectomy subgroup and the mixed gastrectomy (including distal, total and proximal gastrectomies) subgroup. Although there was a selection bias (only one trial was included in the total gastrectomy subgroup),¹⁷ the results suggest that patients with gastric cancer undergoing a total gastrectomy might also benefit from the FTS pathway.

This meta-analysis had some limitations. First, the sample of RCTs was small and the meta-analysis was limited by the presence of bias in all included studies. Second, the studies showed significant heterogeneity in the length of postoperative hospital stay, hospitalisation expenditure and postoperative morbidity. On the other hand, this heterogeneity probably represents an acceptable degree of diversity in medical and economic status in different global regions. For example, Feng et al noted that patients worried about their safety after discharge owing to the traditional belief that surgery could cause great damage.¹⁷ In addition, the cost of the same medical service differs among regions. Such factors are likely to be responsible for the heterogeneity observed. However, individual FTS groups showed improvements in length of postoperative hospital stay and hospitalisation expenditure, and no difference in postoperative complications from conventional care groups.