Enhanced recovery after surgery program in Gynaecologic Oncological surgery in a minimally invasive techniques expert center

Background Enhanced Recovery After Surgery Programs (ERP) includes multimodal approaches of perioperative patient’s clinical pathways designed to achieve early recovery after surgery and a decreased length of hospital stay (LOS). Methods This observational study evaluated the implementation of ERP in gynaecologic oncological surgery in a minimally invasive techniques (MIT) expert center with more than 85% of procedures done with MIT. We compared a prospective cohort of 100 patients involved in ERP between December 2015 and June 2016 to a 100 patients control group, without ERP, previously managed in the same center between April 2015 and November 2015. All the included patients were referred for hysterectomy and/or pelvic or para-aortic lymphadenectomy for gynaecological cancer. The primary objective was to achieve a significant decrease of median LOS in the ERP group. Secondary objectives were decreases in proportion of patients achieving target LOS (2 days), morbidity and readmissions. Results Except a disparity in oncological indications with a higher proportion of endometrial cancer in the group with ERP vs. the group without ERP (42% vs. 22%; p = 0.003), there were no differences in patient’s characteristics and surgical procedures. ERP were associated with decreases of median LOS (2.5 [0 to 11] days vs. 3 [1 to 14] days; p = 0.002) and proportion of discharged patient at target LOS (45% vs. 24%; p = 0.002). Morbidities occurred in 25% and 26% in the groups with and without ERP and readmission rates were respectively of 6% and 8%, without any significant difference. Conclusion ERP in gynaecologic oncological surgery is associated with a decrease of LOS without increases of morbidity or readmission rates, even in a center with a high proportion of MIT. Although it is already widely accepted that MIT improves early recovery, our study shows that the addition of ERP’s clinical pathways improve surgical outcomes and patient care management. Electronic supplementary material The online version of this article (10.1186/s12893-017-0332-9) contains supplementary material, which is available to authorized users.


Background
Enhanced Recovery Programs (ERP), or "fast-track" programs, are a multimodal approach of perioperative patient's clinical pathways designed to achieve early recovery after surgery and decrease length of hospital stay (LOS). This patient-centered management was defined in the 1990's by Kehlet et al. [1,2], and refers to a specific organization in which an informed patient plays an active role. The key elements of ERP include preoperative counsel, optimization of nutrition, standardized analgesic and anaesthetic regimens and early mobilization. Although ERP have become an important focus of perioperative management after many type of surgeries [3], initial studies in gynaecology surgery only included hysterectomies for benign indications [4][5][6] or precancerous lesions [7,8]. Only few studies have evaluated ERP role in gynaecological oncology [9][10][11], all with the same conclusion of achieving shorter LOS and enhancing recovery. Furthermore, most publications are limited to one disease or one type of procedure, and didn't report the implementation of ERP in a established high proportion of minimally invasive techniques (MIT) context. Our institution introduced ERP in 2015 for gynaecological cancer patients (cervical, endometrial and ovarian cancers) admitted for MIT as for open surgery. MIT, which constitutes by itself one of the most important component of ERP's clinical pathways [12], has been shown to be safe and feasible with the same oncological results than open procedures with better cosmetic results and obviously, less postoperative pain, and shorter hospital stays [13].
This study evaluated the implementation of ERP in gynaecologic oncological surgery in a MIT expert center with more than 85% of procedures done with MIT. The primary objective was to achieve a significant decrease of median LOS in the ERP group. Secondary objectives were decreases of prevalence of postoperative complications and readmissions, and increase of patients achieving a target LOS of 2 days.

Study design and data source
This prospective observational study was conducted at the Institut Paoli-Calmettes which is one of the 18 French comprehensive cancer centers, located in Marseille, FRANCE, between December 2015 and June 2016. Consecutive patients undergoing hysterectomy and/or pelvic or para-aortic lymphadenectomy for gynaecological cancer (cervical, endometrial, ovarian cancer without bowel resection or other including borderline ovarian tumor, endometrial hyperplasia and cervical intraepithelial neoplasia) were identified. The only exclusion criterion was post-operative admission to intensive care unit (ICU) for more than one night.
All staff involved, including surgeons, anaesthesiologists, nurses and dieticians, received specific training regarding three defined phases of Enhanced Recovery Pathway management (Table 1), including latest published recommendations [14][15][16][17][18]. Written procedures were established and spread to all the actors of the clinical pathway. Further description of Paoli-Calmettes Institute's ERP is available in Annex 1 (Additional file 1). Since association between improved protocol adherence and improved postoperative outcomes is described [19,20], we evaluated compliance criteria which are summarized in Table 2. Good compliance was defined as ≥70% score per criteria. After hospital discharge, follow-up was conducted at home to day 7 through a network of nurses, in order to detect early perioperative complications and to evaluate the compliance to the ERP and post-operative instructions. In addition, the postoperative nurse coordinator organized telephone interviews at days 1, 7 and 30, to record all occurrences of readmission in other hospitals and/or long-term post-operative complications. Per and post-operative complications were collected according to the Clavien-Dindo classification [21].
To confirm the interest of ERP in terms of hospital's length of stay (nights spent at hospital, including the night before surgery) and morbidity (including readmission), we compared our prospective cohort to a 100 patients control group, without ERP, matched on Age, American Society of Anesthesiologists (ASA) score and type of procedure (conventional laparoscopy or robotically assisted laparoscopy or open surgery) from a historical cohort, previously managed in the same center between April 2015 and November 2015.
All procedures performed in this study involving human participants were done in accordance with the French ethical standards and with the 2008 Helsinki declaration. All included patients provided written informed consent before surgery. This work was approved by our institutional review board (IPC Comité d'Orientation Stratégique).

Statistical analysis
Categorical variables were described using counts and frequencies, and quantitative variables were described using mean, medians and ranges. Patients' characteristics and distribution were compared with Mann-Whitney U and χ2 tests. The level of statistical significance was set at α = 0.05. Statistical analyses were carried out with the SAS® software version 9.3. We followed the reporting recommendations specified in the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) Statement [22].

ERP compliance
Compliances with perioperative ERP's criteria are reported in Fig. 1. Overall compliance rate was 90%. Each criteria compliance varies from 68 to 100%. During the post-operative period, only one patient required to maintain nasogastric intubation and 12% of patients had abdominal drainages that were removed on day 2 according to our recommendations. At day 1, urinary catheter was removed in 94% patients, 97% had returned to a regular diet, 87% patients were able to walk in the corridors and 92% had undergone early mobilization. Only 10% suffered from immediate post-operative nausea. Post-operative level of haemoglobin was >10 g/dL for 76% patients and thromboembolism prophylaxis has been administrated in 100% of cases.

Characteristics of the study population and surgical procedures
Patients' characteristics before and after ERP's implementation are summarized in Table 3. No difference was observed between matching criteria: Mean age, ASA score and surgical procedures were similar between the 2 groups. We observed a significant difference concerning oncological indications in relation with a higher proportion of endometrial cancer in the ERP group (42% vs. 22% before ERP; p = 0.003). The others indications were distributed as follow: cervical cancer 39% vs. 42%, ovarian cancer 14% vs. 22% and others 5% vs. 14%, without significant difference. Most of the procedures were performed by conventional or robotic assisted laparoscopy (87% after ERP vs. 88% before ERP) and only few of the procedures were performed by open surgery (13% after   Table 4).
Lengths of stay (in nights spent) before and after ERP are summarized in Fig. 2.

Postoperative complications and readmissions
Overall complication rate was 25% in the ERP group vs. 26% in the group without ERP up to 30 days after discharge, including per and post operative complications (Table 5). Considering only grade III-IV complications, we observed 3 grade III postoperative complications in both groups. No grade IV occurred. The readmission rate in relation with postoperative complications up to 30 days after surgery was 6% in the ERP group and 4% in the group without ERP. An extended description of complications that have occurred only in patients who have undergone minimally invasive techniques is available in Additional file 2: Table S1.

Discussion
The present study demonstrates that ERP implementation in gynaecologic oncological surgery is associated with LOS decrease without increases of morbidity or readmission rates, even in a high-volume cancer center where MIS is developed and routinely used since 15 years. Although ERP have become an important focus of perioperative management after many type of surgeries [3,23], the recovery benefits following gynaecologic oncological surgery remain uncertain when MIT are already implanted. Recently, De Groot et al. published the design of a stepped strategy that aims at the nationwide implementation of the ERP in major gynaecological surgery in the Netherlands [24]. This ambitious trial is ongoing with the objective to Analgesia Loco-regional analgesia performed (injection at incision site or bilateral TAP block).

Postoperative
Opioid-sparing multimodal analgesia Use of opioid-sparing strategies including, abdominal trunk block and oral analgesia: acetaminophen, NSAIDs (unless complication), Level II opioids.
Free diet on POD 0 Patient received one meal with regular food by POD 0.

Early mobilization out of bed on POD0
Patient mobilized out of bed after surgery by POD 0.

Early ablation of IV fluid infusion
Ablation of intravenous fluid infusion by POD 0.

Early removal of urinary drainage
Removal of urinary catheter by POD 1.
TED prophylaxis TED prophylaxis with low molecular weight heparin.

Avoidance of nasogastric tube
Nasogastric tube removed at the end of general anaesthesia.
Abbreviations: MBP mechanical bowel preparation, PONV prevention of nausea and vomiting evaluate effectiveness and costs of a stepped implementation approach that is characterized by tailoring the intensity of implementation activities to the needs of organizations and local barriers for change, in comparison with the generic breakthrough strategy that is usually applied in large-scale improvement projects. In accordance with De Groot's hypothesis, the high proportion of MIT already established in our department could have participated to the success of our study. One of the most important difficulties observed to implement such a program is the need of collaboration between different health professionals and different specialties. ERP required a multidisciplinary team involving surgeons, anaesthesiologists and nurses to establish standardized clinical pathway, train and encourage the whole team to participate. ERP's compliance must be evaluated during the implementation phase because the success of this program is directly linked with the percentage of adherence [19]. The selection of our ERP's preoperative, peroperative and post-operative criteria's has been made according to the current recommendations that are reproducible, acceptable for patients and adapted to our practices. In our series, overall compliance rate was at 90% (68 to 100% according the different criteria) and led to a significant reduction of LOS. There has been a recent widespread international 'paradigm shift' to new perioperative systems, based on a multidisciplinary team providing an integrated process of care. Lee et al. synthesize the evidence supporting these new systems moving to a more proactive, coordinated and teambased approach [25]. In addition, it seems that the adding of home visits to telephone interviews, as we provide in our ERP, are more effective than telephone interviews alone to prevent readmission [26]. In relation with these observations, it seems that a postoperative coordination is necessary to develop perioperative surgical at home in a safe way for our patients. There are multiple considerations in how to optimize ERP implementation. The most crucial at our institution was to invest in a dedicated nurse to oversee the process from hospitalization to home. In addition, to support this strategy we have organized an "at home" follow up with a network of care with nurses, from the day of discharge to the day 7 after discharge, to detect early perioperative complications, and to evaluate the compliance to the ERP and postoperative instructions. In a retrospective study, Shiavone et al. [27] evaluated 128,634 women who had undergone a laparoscopic hysterectomy and showed a 4.0% rate of re-evaluation within 60 days for women discharged on the same day of surgery, 3.6%, for those discharged after a 24-h stay, and 5.1% for those whose length of stay was 2 days or longer, respectively (p < 0.001). Moreover, patients discharged on post-operative day 1 were 11% less likely to require revaluation than women discharged on the day of surgery. Current constraints in France are due to the lack of availability of operating rooms and hospital beds; ERP appears to be a solution to improved healthcare system without impairing patients quality of life or satisfaction [8,28,29,20]. The LAFA Study suggest that the success of patient's centred clinical pathway is linked to the surgical approach [12]. MIT was the only independent predictive factor to reduce LOS in this study. We showed here that ERP has participated to a quiet small, but statistically significant, reduction of primary hospital stay, even with a high rate of MIS. Additionally, if the most important gain in term of LOS in subgroups analysis appears to be in open surgery, this reduction didn't reach the significance level, probably due to the low number of patients. Complication rate was consistent with those reported in literature, in Total Hysterectomy** with pelvic and aortic lymphadenectomy 9 9 Isolated Lymphadenectomy: 37 27 Pelvic lymphadenectomy 12 4 Para-aortic lymphadenectomy 21 18

Both 4 5
Abbreviations: BMI body mass index, SD standard deviation, CIN Cervical Intraepithelial Neoplasia *significative difference in oncological indications in relation with the rate of endometrial cancer, higher in the group after ERP **total hysterectomy was always associated with bilateral oophorectomy relation with surgical complications and particularly lymphatic dissections [30]. Despite careful methodology, our study has some limits, lying mainly in the small number of patients and in the retrospective analysis of the control group with potential selection bias and lack of standardization in the monitoring of post-operative complications. However, this last point would rather lead to an underestimation of the complications in the group without ERP and thus to increase the difference in favour of ERP implementation. Our study has also strengths as the matching of the two populations and the high rate of MIT procedures in both groups, which allowed us to conclude on ERP implementation in a center where the practices of MIT are already extended.  Ethics approval and consent to participate This work was approved by our institutional review board (IPC -Comité d'Orientation Stratégique). Table 5 Analysis of morbidity before and after implementation of ERP Before ERP(n = 100) After ERP(n = 100) p-value