Enhanced Recovery after Emergency Surgery: A Systematic Review

Introduction

Emergency Surgery (ES) is a key hospital service, with the highest proportion of cases in General Surgery. Surgical mortality is a major concern, with reports of rates as high 80% of all surgical mortality being as a result of emergency surgical interventions [1]. There are currently strong recommendations that the delivery model of ES needs to be changed in order to improve efficiency and quality of care [2]. Despite the issues being appreciated and discussed, there is uncertainty about how best to proceed. One of the proposed measures to improve outcomes has been the recommendation to implement enhanced recovery programmes (ERAS) [1,3].
ERAS programmes are evidenced-based protocols designed to standardize and optimize perioperative care in order to reduce surgical trauma, perioperative physiological stress and organ dysfunction [4]. Although published initially for colorectal surgery in 2005, they are now well established for many other surgical conditions (http://www.erassociety.org). There is already substantial evidence in the literature demonstrating the effectiveness of adopting ERAS based protocols in elective surgery [5-9], resulting in a change of clinical practice. Intuitively, ERAS could benefit ES patients due to its design to reduce surgical stress and return functional status more efficiently. The aim of our work was to evaluate the current scientific evidence for the applicability, safety and effectiveness of Enhanced Recovery pathways in ES.

Materials and Methods

Protocol and registration
The review has been registered in PROSPERO (International prospective register of systematic reviews, http://www.crd.york.ac.uk/PROSPERO/searchadvanced.php) with the Registration number: CRD42016049268 and was reported in accordance with PRISMA statements (http://prisma-statement.org).

Eligibility criteria and search
We undertook a search using PubMed and Cochrane databases for ERAS protocols in emergency cases. The search was restricted to the last 10 years in order to avoid pre ERAS guideline studies and to allow for greater homogeneity in the studies to be reviewed. No language restrictions were applied.
The following search string was used for PubMed and adapted for Cochrane: (enhanced[All Fields] AND recovery[All Fields] AND ("emergencies"[MeSH Terms] OR "emergencies"[All Fields] OR "emergency"[All Fields]) AND ("surgery"[Subheading] OR "surgery"[All Fields] OR "surgical procedures, operative"[MeSH Terms] OR ("surgical"[All Fields] AND "procedures"[All Fields] AND "operative"[All Fields]) OR "operative surgical procedures"[All Fields] OR "surgery"[All Fields] OR "general surgery"[MeSH Terms] OR ("general"[All Fields] AND "surgery"[All Fields]) OR "general surgery"[All Fields])) AND ("2006/01/01"[PDat]: "2016/10/16"[PDat] AND "humans"[MeSH Terms]).
An additional search using “fast-track” OR “multimodal” AND “emergency”, with no date restrictions, did not produce further relevant studies.

Study selection
Titles and abstracts were scrutinized; duplicates and citations were removed and full text articles of studies matching search criteria were included. Papers focused on ES that were other than abdominal were excluded. References of relevant studies were then reviewed for possible additional papers. ERAS guidelines recommend a total of 20 elements (divided into preoperative, intraoperative and postoperative) however not all of these are feasible for emergency patients and no restriction was placed on the number of elements applied as part of the protocol in each study. (Appendix 1 - Guidelines for perioperative care in elective colonic surgery: ERAS Society recommendations). After the search, study selection was independently performed by two authors (MP, LP) and disputes were resolved by discussion or the judgement of a third reviewer (PS) as to which papers should be included if required.

Quality and risk of bias assessment
Two reviewers (MZ, IMC) independently assessed the quality and risk of bias of the papers selected using SIGN levels of evidence and grades of recommendation (http://www.sign.ac.uk/methodology/checklists.html).

Data collection
Data extracted for analysis consisted of: patient age, type of surgery performed, ERAS elements implemented, surgical outcomes in terms of postoperative complications, mortality, length of stay (LOS) and readmission rate.

Results

The search on ERAS and ES generated 65 titles. After eliminating the papers not meeting initial search criteria (55 papers), we selected 10 abstracts for screening; of these, 5 were eligible and one additional article was retrieved from backward chain of references. The flow chart in Figure 1 gives a summary of the article selection process.
One of the 5 papers initially considered eligible was an Editorial, evaluated as relevant. Although this could not be used for the findings due to not providing outcome data, it was included for the discussion, offering experts’ opinion (Level 4 of evidence) [10].
Quality and risk of bias assessment is given in Table 1. Three out of the five papers were cohort studies rated as acceptable quality (level of evidence 2+), one cohort study as high (2++) and one RCT as poor quality (1-).

Fig. 1. Flow chart of the study.

Table 1. Quality and risk of bias assessment (SIGN)

Study characteristics
Baseline data and results from each study are shown in Table 2. The impact of ERAS on a total number of 311 emergency patients was assessed, in comparison to 605 patients consisting of 235 emergency patients receiving Conventional care (CC) and 370 elective patients receiving ERAS. Outcomes reported were based on 30 day follow up in the majority of studies. Outcomes reported by study are summarized in Table 3.

Table 2. Baseline characteristics of selected studies.

^aERAS: Enhanced Recovery after Surgery; ^bCC: Conventional care; ^cES: Emergency surgery
* In relation to Appendix 1 - Guidelines for perioperative care in elective colonic surgery: ERAS Society recommendations

Table 3. Outcomes reported by study.

Analysis of findings
Three studies compared ERAS to CC in emergency surgery (1 RCT and 2 cohort studies) [11-13]. All studies showed post-operative complication rate reduction in patients receiving ERAS, with a statistically significant reduction in major complications in one study [13].
LOS was similarly reduced by 2-3 days in 2 studies [11,12], with statistical significance and mortality rates did not increase or were improved upon (0 vs. 3.8%) [11]. In the cohort studies readmission rates were not increased by the implementation of ERAS, however in the RCT an increase from 7.6 to 19% was reported [11] although this was not considered statistically significant.
Two studies [14,15] compared emergency to elective post operative outcomes for colorectal surgery within an ERAS pathway. There was no statistically significant difference in the results between intervention and control groups for complication and readmission rates, although Verheijen et al., [15] only reported anastomotic leaks for complication rates. Mortality was not reported by either study.
Current ERAS Guidelines (Appendix 1) were used as the measure against which the enhanced recovery protocols reported to be applied in each study were evaluated. The total number of items per study ERAS protocol ranged from 8 to 18 (Table 2).

Grade of recommendation (SIGN)
Summarizing the current evidence in the studies analyzed and using SIGN Revised grading system for recommendations in evidence based guidelines (Appendix 2), grade C of recommendation was givenbased on: 1 Cohort study level 2++ and one Cohort study level 2+, comparing ERAS to CC in ES; and 2 Cohort studies level 2+ comparing ES to Elective surgery with ERAS. The RCT (level 1-) could not be used for establishing the grade of recommendation due to its a high risk of bias.

Discussion

To our knowledge this is the first review that evaluates the evidence with regard to the feasibility and effectiveness of ERAS in ES. The studies we found focusing on ERAS in ES were scarce, with two distinct comparators of either conventional care or elective surgery, different pathologies (abdominal, perforated ulcer and colorectal) and different mean ages.In addition, the heterogeneity of the scales used to report postoperative complications in the studies (Clavien-Dindo [16], major-minor complications, superficial-organ/space type SSI) and the fact that only 3 out of the five studies reported mortality rates, limited the comparative analysis between studies. Despite this, some valuable general observations could be made.
Complication rates were reduced in four out of the five studies and readmission rates were equal or not increased significantly; the exception was Gonenc et al., [11] with a higher readmission rate for the intervention group, but no explanation was offered for the difference. This could be in relation to the low LOS achieved in this group.
LOS was significantly reduced with ERAS in comparison to CC in 2 out of the 3 studies. Advance age is common in patients requiring emergency surgery and the 3 studies with higher mean age had the higher LOS; this observation, however, is not identified by the individual studies. Although LOS was measured by all five studies, it has been argued that it is not a reliable measure when evaluating the effectiveness of ERAS and that the return to functional status is a more valid one [16].
Mortality is a key issue in ESand has been specifically identified as one to be addressed. The three studies that reported mortality outcomes for both intervention and control groups found rates to be the same or better [11] with ERAS application. Co-morbidities are well known contributing factors to mortality. Two studies excluded higher risk patients, as assessed by ASA and POSSUM [12,14] which could have contributed to their low mortality rates.  
The quality of the studies analysed in this review is mainly acceptable, with 3 cohort studies assessed as level 2+ and one as 2++. The only RCT was assessed as having a high risk of bias (level of evidence 1-), being non blinded, randomization being made at the end of the surgical procedure [11], employing many exclusion criteria and has been proved to deviate from clinical trial protocol [18]. The results from this study therefore need to be used with caution.
The application rate of ERAS items demonstrated in the studies analysed was between 11 [15] and 18 [14], with the exception of Gonenc et al. applying only 8 elements. The study undertaken by Verheijen et al., [15] was focused on several patient groups (emergency-elective, younger-elderly, open-laparoscopic, benign-malignant) and the reporting of ERAS elements was generic. 
Current ERAS guidance recommends the implementation of 20 items in order to provide a comprehensive pathway leading to better outcomes. We found no obvious correlation between the number of items applied and improved outcomes in the studies we reviewed.
It has been stated that in ES the implementation of all ERAS preoperative components may not always be feasible [3]. Pre-operative optimisation by cessation of smoking and alcohol consumption four weeks before surgery is clearly not achievable in ES cases. We found that of the seven pre-operative elements (Appendix 1), the range of implementation was from 1 to 6; from the six intra-operative items the implementation ranged from 2 to 6; and from the seven post-operative items the implementation range was from 4 to 6. We also observed that there was variability in the way some elements were applied, most notably for early post operative oral feeding and mobilisation.
Although we did not analyse the possible correlation between specific elements applied in the studies reviewed and their outcomes, the impact of individual elements of an ERAS programme on post-operative results have been undertaken. The ERAS study group[19]identified 2 key elements which had an independent positive impact on post operative outcomes: perioperative intravenous fluid management and preoperative carbohydrate treatment and Brandstrup et al.,[20] has demonstrated the important impact of fluid management on post-operative outcomes; however neither of these studies was specific to ES.The trial undertaken by Gonenc et al.,[11] stated that 3 key elements produced better outcomes in their intervention group: non NGT usage, early oral feeding and use of NSAIDs. More research is still needed in relation to identifying which elements of ERAS might have greater impact and whether individual influence plays a more significant role than the number of elements applied. It is also necessary to consider how these factors might vary in emergency surgery and with different patient groups within it.
A separate issue to the application of ERAS items is whether patients are able to comply with individual element application. Only one study in our review looked specifically at patient compliance to elements of ERAS; Roulin et al., [14] reported an overall patient compliance of 57% in ES. This was compared to 77% in elective ERAS patients. Difficulties identified were: pre-operative carbohydrate loading, NGT early removal/non routine use, postoperative fluid management, nutrition and early mobilization. However, the difference was no longer evident from the first postoperative day and functional recovery was similar in both ES and elective patients following an ERAS pathway [14]. Wisely et al.,[13] identified the ERAS elements considered to be appropriate for ES patients. Based on their study findings, most elements were considered appropriate in varying degrees. Laparoscopic surgery, avoiding resection‐site drain and general anaesthetic ± epidural anaesthesia was identified as appropriate only for some ES patients [13]. These findings could provide a baseline for further investigation for ES patients.
The main limitation of this review is the heterogeneity and quality of the studies evaluated. This is due to the fact that there is little information on ERAS programmes in ES and therefore we did not eliminate any of the studies which met our inclusion criteria. Quiney et al., [10] attributed the small number of studies evaluating the impact of ERAS in ES partially to the difficulty to apply many of the ERAS principles. However this difficulty does not prohibit the use of evidence-based practice, on which ERAS is based. 
In conclusion, the studies reviewed agreed that ERAS in ES was feasible and safe with generally better outcomes, but needs to be adapted for this patient group as compliance with all ERAS elements can be difficult to achieve. A tailored ERAS pathway would better serve this population along with a multidisciplinary team approach. The limited number of trials and studies focusing on ERAS in ES clearly indicates that this is still a new area to explore. More evidence is required as to what can improve outcomes and how this can be formulated into an effective care pathway for the heterogeneous ES patient.

Acknowledgments

Melanie Radcliff, BA, for assisting with the English translation.

Conflicts of Interest: None declared.