Enhanced recovery protocols after surgery: A systematic review and meta-analysis of randomized trials in cardiac surgery.

Background: Previous meta-analyses combining randomized and observational evidence in cardiac surgery have shown positive impact of enhanced recovery protocols after surgery (ERAS) on postoperative outcomes. However, definitive data based on randomized studies are missing, and the entirety of the ERAS measures and pathway, as recently systematized in guidelines and consensus statements, have not been captured in the published studies. The available literature actually focuses on "ERAS-like" protocols or only limited number of ERAS measures. This study aims at analyzing all randomized studies applying ERAS-like protocols in cardiac surgery for perioperative outcomes.

Methods: A meta-analysis of randomized controlled trials (RCTs) comparing ERAS-like with standard protocols of perioperative care was performed (PROSPERO registration CRD42021283765). PRISMA guidelines were used for abstracting and assessing data.

Results: Thirteen single center RCTs (N = 1704, 850 in ERAS-like protocol and 854 in the standard care group) were selected. The most common procedures were surgical revascularization (66.3%) and valvular surgery (24.9%). No difference was found in the incidence of inhospital mortality between the ERAS and standard treatment group (risk ratio [RR] 0.61 [0.31; 1.20], p = 0.15). ERAS was associated with reduced intensive care unit (standardized mean difference [SMD] -0.57, p < 0.01) and hospital stay (SMD -0.23, p < 0.01) and reduced rates of overall complications when compared to the standard protocol (RR 0.60, p < 0.01) driven by the reduction in stroke (RR 0.29 [0.13; 0.62], p < 0.01). A significant heterogeneity in terms of the elements of the ERAS protocol included in the studies was observed.

Conclusions: ERAS-like protocols have no impact on short-term survival after cardiac surgery but allows for a faster hospital discharge while potentially reducing surgical complications. However, this study highlights a significant nonadherence and heterogeneity to the entirety of ERAS protocols warranting further RCTs in this field including a greater number of elements of the framework.