Clinical Outcomes and Predicted Model of Surgical Patients With Enhanced Recovery After Surgery (ERAS): A Multicenter Prospective Clinical Study

Enhanced recovery after surgery (ERAS) has been accepted widely in the whole world. However, clinical effects of ERAS in China have not been systematically reported, and it is still unclear whether there is key component with the present ERAS program to secure enhanced recovery. Patients who were undergoing operations with ERAS program were included. All the perioperative information were collected via a website and a nomogram to predict postoperative complication was conducted. 950 subjects from 59 hospitals were included in this study. Illness of cardiovascular (22.6%) and endocrine system (11.1%) were the top two coexisted diseases preoperatively. The recovery time of ability of drinking water after surgery was 6 (4-8) h, and almost half of patients could do active exercises in bed within 6 h postoperatively. The overall incidence of complications within 1 month postoperatively was 11.1%. Preoperative creatinine and bilirubin, intraoperative maximum systolic blood pressure and NRS scores at rest at postoperative 3 days were independent risk factors for complications within 1 month postoperatively. However, minimally invasive surgery was associated with a decrease probability of the complications. This study rstly indicates preoperative hepatorenal function, intraoperative systolic blood pressure, minimally invasive surgery and postoperative pain control can independently inuence the prognosis of surgical patients.


Introduction
Enhanced recovery after surgery (ERAS) is a multimodal perioperative care pathway to reduce the physiological and psychological stress of surgical patients and achieve early recovery after surgery 1 . The concept rested on several elements: a multidisciplinary team, evidence-based perioperative management to prevent and resolve complications 2 . The advantages of ERAS has been strongly proved by many studies [3][4][5][6][7] . It was reported that ERAS could reduce the need for hospital stay by about 30% and major complications by as much as 40% after major abdominal surgery 8 . Nowadays, ERAS is becoming standard care for many surgical patients in the whole world.
About 16-25 evidenced-based interventions of ERAS were summarized and guided by ERAS Society for different types of surgery (http://erassociety.org). More and more studies found and highlighted the importance of compliance of ERAS program: the better compliance, the better outcomes 9,10 . However, it was di cult to adhere strongly to all ERAS items in most medical centers, in fact, the elements were implemented in part 11 . And it is still debatable that all the item of ERAS program are important equally and whether there are obligatory elements for the recovery of patients 12 . In addition, with the ERAS program, the effects of other perioperative factors on the prognosis is still unclear.
ERAS was carried out for more than 10 years in China and increasing achievement were reported in the recent 5 years 13 . However, to the best of our knowledge, the clinical outcomes of ERAS in China have not been examined, and the comprehensive understanding the effect of all the perioperative elements on the recovery of patients has not been established. And understanding the corrections of different elements and outcomes of surgical patients may help us to optimize the ERAS program further. Therefore, we conducted this multicenter clinical trial to investigate the current status and clinical outcomes of ERAS and its' contributed element in Chinese population.

Results
Demographics of the study population suggested that the vital signs (heart rate, breathe, blood pressure, temperature) of patient were relative stable during the operation although a certain uctuation in heart rate and blood pressure was shown.
The volume of uid therapy was also reported and the input were 1100 (800-1600) ml, crystalloid was the main component. The volume of output were 350 (85-600) ml (Table 2).
Compliance and clinical outcomes with ERAS program ERAS program was consisted of 16 items in our study. The highest compliance was counseling and education, the lowest compliance was belonged to the item of carbohydrate intake. And there were 3 items which the compliance was over 90%. However, the compliance of 4 items, including carbohydrate intake, prevention of DVT, prevention of PONV and multimodal analgesia, were less than 70% ( Figure 1).
The time of anal exhaust and defecation were 12 h and 30 (18-50)h postoperatively. Moreover, about 50% patients could drink water within 6h postoperatively, and a quarter of patients even could drink water within 4h postoperatively. A half of patients could returned to liquid diet within 11h postoperatively and gradually converted to normal diet within 30h postoperatively. Otherwise, 25% of patients could not fully recovered to normal diet at postoperative 48h. The time of active exercises in bed for patients postoperatively was 6(2-10)h. About 75% patients had the ability to stand without assist within 48h postoperatively. The time of walking with assist was 24 (13-55)h postoperatively while the time of walking without assist was 33 (20-72)h postoperatively (Table 3).
106 cases of postoperative complication were reported within 1 month postoperatively in this study, and the incidence of complication was 11.1%. others (3 cases, 0.3%). The median of hospital stay postoperatively was 6 days, about 75% patients could discharge within postoperative 10 days. And the median of hospitalization expenses were 13000 CNY, and the satisfaction scores was 9 (Table 3).

Clinical risk factors of complications within 1 months postoperatively
The risk factors were summarized with univariate and multivariate logistical regression analysis in table 4. Univariate analysis revealed that 20 potential perioperative factors (age, preoperative hemoglobin, hematocrit, red blood cell, creatinine, total protein, albumin, bilirubin, and brinogen, mode of anesthesia, duration of anesthesia, intraoperative maximum of heart rate, pulse and systolic blood pressure, NRS at rest and in motion at 3 days postoperatively, time of activity in bed, minimally invasive surgery, no drainage placed as usual and early exercise) were associated with postoperative complications. And the nal multivariate logistical regression showed 5 signi cant independent predictors: preoperative creatinine (OR 1.012, 95%CI 1.002-1.021) and bilirubin (OR 1.017, 95%CI 1.001-1.033), intraoperative maximum systolic blood pressure (OR 1.015, 95%CI 1.001-1.029), NRS scores at rest at postoperative 3 days (OR 1.43, 95%CI 1.111-1.84), minimally invasive surgery (OR 0.539, 95%CI 0.292-0.995). Results of the Hosmer-Lemeshow test suggested a good t of the model (χ2=11.869, df=8, P=0.157).

Prediction model for complications within 1 months postoperatively
To identify the high-risk patients of complications after surgery, a predictive nomogram model based on the results of multivariate logistical regression analysis was conducted to present the probability of complications within 1 months postoperatively. "Points" were calculated with the corresponding independent predictors. "Total points" were summed with each single points, and we draw a vertical line and projected it from the "Total points" scale to the "Probability" scale to predict the probability of complications within 1 months postoperatively. Results of ROC curve suggested a good discrimination with AUC 0.702 (95%CI 0.624-0.779). At the optimal cutoff value of 0.241, the sensitivity and speci city were 57.7% and 73.2%, respectively. The model showed a relatively good calibration because the actual line was not signi cantly deviated from the ideal line ( Figure 3).

Discussion
In this multicenter, prospective study, we found that the most common preoperative comorbidity was cardiovascular disease among the patients. With ERAS program, almost half of patients could drink water and active exercises in bed within 6h postoperatively with a good control of postoperative pain. For the compliance of ERAS program, there were 3 items (counseling and education, preoperative fast and hypothermia prevention) which the compliance was over 90%, however, 4 of 16 items (carbohydrate intake, prevention of DVT, prevention of PONV and multimodal analgesia) were less than 70%. The overall incidence of complications within 1 month postoperatively among the patients with ERAS program was 11.1%, and preoperative creatinine and bilirubin, intraoperative maximum systolic blood pressure and NRS scores at rest at postoperative 3 days were independent risk factors for the complications, conversely, minimally invasive surgery was associated with a decrease probability of complications within 1 month postoperatively.
It is well known that preoperative general condition and comorbidities of the patient were related to the safety of surgery and postoperative prognosis 15,16 . In this study, we found that 22.6% of patients were coexistent with cardiovascular diseases (CVDs). CVDs remained a major cause of health loss in the whole world 17 . And according to the latest report from national center for cardiovascular disease, there were an estimated of 290 million CADs patients in China 18 . Perioperative management of cardiac patients undergoing noncardiac surgery was still a challenge to the anesthesiologists and surgeons although a series of guidelines have been published [19][20][21] . In our study, a certain uctuation in heart rate and blood pressure was found in a small group of patients. Furthermore, intraoperative maximum systolic blood pressure was found to be an independent risk factor for complications, which was in line with the results of previous study 22 . However, recent studies reported that intraoperative hypotension, but not hypertension, was associated with shortened survival 23,24 . Several factors might explain this difference: (1) the de nition of intraoperative hemodynamic aberration or intraoperative hypertension/hypertension was, for awhile, loosely de ned. (2) speci c blood pressure target management from group data cannot be extrapolated to be optimal for all patients 25 . (3) In addition to the cutoff value of intraoperative hypertension/hypotension, other factors [26][27][28] (including duration of hemodynamic aberration, improper treatment during hemodynamic aberration) could also impacted the outcomes. It was widely accepted that a smooth intraoperative course with minimal blood pressure aberrations was related to better prognosis although a certain variation of blood pressure seemed to be inevitable virtually 29 . To the best of our knowledge, the effects of variation of intraoperative blood pressure on the outcomes of the surgical patients were still controversial. And further researches are needed to identify whether and to what extent hemodynamic aberration impacts the prognosis of surgical patients with ERAS program.
In this study, the results suggested a moderate compliance to ERAS program, but 4 items (carbohydrate intake, prevention of DVT, prevention of PONV and multimodal analgesia) showed poor compliance. The item of lowest compliance was carbohydrate intake. Preoperative carbohydrate loading has been theorized to reduce insulin resistance and enhance patient comfort 30 . However, in addition to overcoming the inertia to change the outdated fasting guideline, several concerns were still unsolved and might affected widely used in clinical. Effects of preoperative carbohydrate loading on the clinical outcomes and prognosis were unidenti ed 31,32 , Smith and the colleagues reviewed all the clinical studies and no positive result was found about the relationship between preoperative carbohydrate and postoperative complication 33 . In addition, the safety of carbohydrate use in patients with diabetes remained uncertain [34][35][36] . Furthermore, our previous study found that the composition proportion of carbohydrate could in uence tumor growth and malignance in mice 37 . Thus, the impact of carbohydrate on the speci c populations, such cancer groups, should be taken into consideration. Nowadays, the ERAS program was recommended in general populations for different types of surgery, and present ERAS procedure should be further optimized in special populations based on high-quality researches.
Growing evidence has been proved that compliance of ERAS program was associated with clinical outcomes of patients 38,39 . Actually, full implementataion of ERAS program seems impossible for most surgical patients because of the complex procedure and labor intensive 40 . Although many authors stressed the important of compliance, little was known about the impact of number of used components.
For those reasons, a simpli ed ERAS program named RAPID (remove, ambulate, postoperative analgesia, introducing diet) was suggested by Lloyd et al, and patients with RAPID protocol also showed shorter hospital stay 41 . In our study, we found that several factors, including minimally invasive surgery, one of the elements in our ERAS program, appeared to be more in uential than others. This results are agreed with the study by Demartines who observed that minimally invasive surgery was the single most ERAS program should be more valuable 46 . In addition to minimally invasive surgery, preoperative creatinine and bilirubin, intraoperative maximum systolic blood pressure and NRS scores at rest at postoperative 3 days were also included in the predictive model for complications. The component of this model con rmed the core concept of ERAS which revealed that the better perioperative management, the better clinical outcomes for patients. More narrowly, it is necessary to optimize the function of liver and kidney preoperatively, keep a stable hemodynamics with minimally invasive approach intraoperatively, control the postoperative pain well to reduce the postoperative complications.
However, several limitations exist in this study. First, all participating hospitals are located in the centralwestern region of China, it may not represent other areas in China. Second, the sample size was relatively small. Finally, the follow-up is completed at 1 months postoperatively, and the long-term outcomes was not explored in this study.
In conclusion, this study revealed that preoperative cardiovascular diseases was the most common comorbidity for the patients with ERAS in western region of China. Most patients got rapid and effective recovery after surgery with our ERAS program. And the overall incidence of complications within 1 month postoperatively was 11.1%. Preoperative creatinine and bilirubin, intraoperative maximum systolic blood pressure, NRS scores at rest at postoperative 3 days and minimally invasive surgery were independent predictors for the complications.

Methods
This is a multicenter, observational study, which was conducted between February 2019 and June 2020.

Subjects
All the participants were informed about the study and consent was obtained. Surgical patients who were undergoing operations with ERAS program were recruited. Participants were excluded if they had serious mental illness, speech impediment, participate rejection. Participating hospitals were amenable to recruit one patients per month at least.

ERAS program
According to the regional medical practice and guidelines of ERAS, 16 items were summarized and applied in this study: 1) Perioperative counseling and patient education; 2) Oral intake of 400ml carbohydrate drink started 3 h before surgery and consumed 2 h prior to surgery; 3) Intake of clear uid up to 2h and solids up to 8h before anesthesia. 4) Prevention of deep venous thrombosis with active or passive physical exercise, and usage of low molecular weigh heparin in necessity; 5) antibiotic prophylaxis 30min before incision; 6) optimized anesthesia management, especially for administration of lung protective ventilation strategy, depth of anesthesia monitor, neuromuscular monitor, hemodynamics monitor and combined different methods of anesthesia. 7) minimally invasive surgery or laparoscopic surgery; 8) conduct of intraoperative goal-directed uid therapy; 9) prevention of hypothermia (body temperature less than 36℃); 10) multimodal prevention of postoperative nausea and vomiting (combined with usage of dexamethasone, 5-HT 3 receptor antagonist, metoclopramide and haloperidol); 11) no surgical drains placement as far as possible; 12) multimodal management of postoperative pain (combined with usage of nonsteroidal anti-in ammatory drugs, peripheral never block analgesia, patients controlled analgesia, incision in ltration analgesia, opioids in necessity); 13) Removal of surgical drains within 3 days postoperatively; 14) Early oral feeding (drinking water within 6 hours or/and liquid diet feeding within 24h postoperatively); 15) Early and scheduled mobilization (active exercise in bed within 6 hours and/or out-of-bed activity with assist within 24h postoperatively); 16) Perioperative nutrition support.

Data collection and export
All the required data were collected from the electronic medical record by an anesthesiologist in each participating hospital and then were uploaded via website (https://218.207.2.100:10601/), which was established with the help of Department of Information Technology of the First Affiliated Hospital of Chongqing Medical University. Each participating hospital reported data via the unique account belonged to their hospital, and each hospital except for the leading hospital of this study can only view and download its own dataset in this website. Electronic medical record, which was used for checking whether there was any mistake of the dataset, was also required to uploaded. When eligible patient was admitted to hospital, he/she was announced with the purpose and process of this study, the basic information of the patient should be uploaded to the online database after the consent was obtained from the patient. Data of surgery and anesthesia was required to uploaded within 24h postoperatively, and the follow-up data should be recorded within 24h after the visit. All the important perioperative information were collected, including characteristic of the patients (age, BMI, nationality, history of smoking, alcohol intake and illness, blood biochemical examination), the information of surgery and anesthesia (types of surgery and anesthesia, ASA and NYHA classi cation, duration of anesthesia and surgery, intraoperative vital signs and volume of input-output), compliance of ERAS program, complications within 30 days postoperatively, pain scores at rest and in motion within 30 days postoperatively, gastrointestinal function recovery and ambulation time, as well as hospital stay, fees and satisfaction. All the data in the database were exported in the form of Excel by a network engineer.

Statistical Analysis
All data were expressed as median (interquartile range) for continuous variables and total number (percent frequency) for categorical variables. Risk factors for major complications were identi ed with odds ratio (OR) and 95% con dence interval (CI) using a univariate logistical regression analysis, the variables with P values ≤0.1 were included in multivariable model to determine the independent factors by using the method of backward stepwise (conditional). The nomogram performance was assessed by discrimination and calibration 14 . The discrimination was evaluated by receiver operation characteristic (ROC) curve. A value of area under curve (AUC) closer to 1 suggested a good performance of the model, an AUC 0.7 was generally considered to be relatively good discrimination. The calibration was assessed by a visual calibration plot comparing the predicted and actual probability of major   Drink water 6(4-8) Liquid diet 11 (6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24) Semi-liquid diet 20  Normal diet 30 (