The authors concluded that the available evidence was limited, but implementation of an Enhanced Recovery After Surgery programme in pancreatic resection seemed feasible and safe. The evidence was limited and varied, and of generally low quality, suggesting that the findings may not be reliable.

To assess the benefits and safety of implementing an Enhanced Recovery After Surgery (ERAS) protocol in pancreatic surgery.

PubMed, EMBASE, CINAHL, and The Cochrane Library were searched, for articles in English, Dutch, or German, from 1966 up to December 2012. Search terms were reported. Reference lists of relevant reviews and articles were manually searched.

Eligible for inclusion were studies that assessed an enhanced recovery programme, with at least four different perioperative elements of the ERAS protocol, as described in the review. Eligible patients were adults undergoing major elective resection of the pancreas, including distal pancreatectomy, total pancreatectomy, and pancreaticoduodenectomy. The primary outcome of interest was the postoperative length of hospital stay. Secondary outcomes included time to normal function, postoperative morbidity (as defined in the review) and mortality, and readmission rates. Total hospital costs were considered if available.

The mean age of patients in the included studies ranged from 53.6 to 65 years. Studies contained from nine to 14 ERAS protocol preoperative, intraoperative and postoperative elements, in different combinations, with 19 different elements overall. The most frequently reported elements were preoperative thromboembolic prophylaxis, perioperative antibiotics, early removal of nasogastric tubes and catheter, and early mobilisation after surgery. Case-control studies included historical controls receiving usual care.

Two reviewers independently screened studies for inclusion. Discrepancies were referred to a third reviewer.

The included studies were assessed for methodological quality using MINORS criteria. Comparative studies were assessed on 12 criteria (maximum score 24) and non-comparative studies were assessed on eight criteria (maximum score 16).

The authors did not explicitly state how many authors assessed quality.

Hospital length of stay in days was extracted from individual studies, but it was unclear whether means or medians were reported. Binary outcome data were extracted to calculate risk differences and 95% confidence intervals for controlled studies. Study authors were contacted for missing data.

Two reviewers independently extracted the outcome data.

Hospital length of stay was presented as a narrative synthesis and in tables. Where studies of pancreaticoduodenectomies were sufficiently homogeneous, risk differences and 95% confidence intervals were pooled using a random-effects model. The results were presented as 100 times the risk difference. Statistical heterogeneity was assessed using Χ² and Ι².

Sensitivity analyses were performed to assess the impact of methodological quality and the number of ERAS elements on the findings. Publication bias was explored using a funnel plot and Egger's test.

Eight studies were included in the review (1,558 patients). Five were case-control studies, and three were non-comparative studies (one prospective and two retrospective case series). The mean MINORS score for case-control studies was 13.4. The prospective case series scored 14, and the two retrospective case series scored 12 and 13. Follow-up was 30 days in all studies.

Postoperative length of stay: Four of the five case-control studies, of any type of pancreatic resection, showed a statistically significant reduced postoperative length of hospital stay with Enhanced Recovery After Surgery (ERAS) elements (6.7 to 13.5 days) compared with control patients (eight to 16.4 days).

Morbidity: There was a statistically significant reduction in overall morbidity for ERAS patients (RD 8.2%, 95% CI 2.0 to 14.5; four case-control studies) and no evidence of statistical heterogeneity (Ι²=0). Sensitivity analyses indicated that the results were influenced by studies of better methodological quality and studies with more than 13 ERAS elements.

Meta-analyses showed no statistically significant differences, between ERAS and control groups, in mortality (four case-control studies) and readmission rates (four case-control studies). There was no evidence of statistical heterogeneity for either outcome.

Data on functional recovery were not reported. Other outcomes, such as protocol adherence, were reported. The funnel plot indicated some potential for bias, but only four studies were included.

Four case-control studies reported costs before and after the ERAS elements. Three showed statistically significant reductions in costs after intervention. Before intervention the costs ranged from $26,393 to $240,242, and after intervention they ranged from $22,806 to $126,566.

The available evidence was limited, but implementation of an Enhanced Recovery After Surgery programme in pancreatic resection, particularly pancreaticoduodenectomies, seemed feasible and safe.

The review question and supporting criteria were clearly stated. A satisfactory literature search was undertaken, but there was some potential for language and publication bias. Study selection and data extraction were performed in duplicate. Study quality was assessed, but only aggregate scores were reported and the case-control studies did not appear to be of high quality. Few study and participant characteristics were reported, but it was clear that the surgical techniques and ERAS protocols varied across studies. Sample sizes were generally small and the study designs had inherent limitations. The authors acknowledged these limitations.

Meta-analysis was conducted almost entirely in patients undergoing pancreaticoduodenectomies, and it was unclear whether the results were applicable to other surgical techniques. There was no evidence of statistical heterogeneity, but there were some differences in the direction of the effect for secondary outcomes. It was unclear whether the length of stay was mean or median, but the studies suggested a reduced length of stay with the ERAS elements. The data on hospital costs were limited, but indicated a large difference in the costs of the ERAS protocol elements across studies. These differences were not explained.

The authors' conclusions were generally cautious, but the evidence was limited and varied, and of generally low quality, suggesting that the findings may not be reliable.

Practice: The authors did not state any implications for practice.

Research: The authors stated that future studies should report on predefined discharge criteria, and the time to functional recovery, to assess whether actual postoperative recovery was accelerated.

None.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.