Initial result of unselective implementation of enhanced recovery after surgery in a low-volume bariatric unit: a feasibility and safety study


 BackgroundThe feasibility and safety of unselectively applying an enhanced recovery after surgery (ERAS) protocol in a low-volume bariatric unit were determined. MethodsWe retrospectively reviewed all patients undergoing bariatric surgeries under a single surgeon between January 2015 and December 2018. Our ERAS protocol initiated in January 2017 with all patients enrolled unselectively. For those receiving non-primary procedures or with BMI<32.5 kg/m2 were excluded from this analysis. Demographic features and all 30-day outcome measures, including operation time, length of stay (LOS), ER visits, readmissions, reoperations were collected and compared between the ERAS (2017–2018) and control (2015-2016) groups. Results One hundred eighty-four consecutive patients underwent bariatric surgeries during the study period. Of those fulfilling the inclusion criteria, 62 (40.8%) were treated before and 90 (59.2%) were treated after ERAS implementation. No differences in baseline demographics were found between the groups except ERAS group had more Roux-en-Y gastric bypass procedures (58.9% vs. 12.9%). A markedly reduced operation time (101 min vs. 147 min; p<0.001) and shortened LOS (2.6 days vs. 3.3 days; p<0.001) were observed, with significantly more ERAS patients achieving POD1 discharge (45.6% vs. 1.6%; p<0.001). There were no significant differences in terms of ER visits (2.2% vs. 8%), readmissions (1.1% vs. 4.8%) or total complication rates between the groups (5.5% vs. 9.7%). ConclusionUnselective ERAS implementation in low-volume units is feasible and safe, with significantly reduced operation times and a shortened LOS without increased complications.


Abstract Background
The feasibility and safety of unselectively applying an enhanced recovery after surgery (ERAS) protocol in a low-volume bariatric unit were determined.

Methods
We retrospectively reviewed all patients undergoing bariatric surgeries under a single surgeon between January 2015 and December 2018. Our ERAS protocol initiated in January 2017 with all patients enrolled unselectively. For those receiving non-primary procedures or with BMI<32.5 kg/m 2 were excluded from this analysis. Demographic features and all 30-day outcome measures, including operation time, length of stay (LOS), ER visits, readmissions, reoperations were collected and compared between the ERAS (2017-2018) and control (2015-2016) groups.

Results
One hundred eighty-four consecutive patients underwent bariatric surgeries during the study period. Of those ful lling the inclusion criteria, 62 (40.8%) were treated before and 90 (59.2%) were treated after ERAS implementation. No differences in baseline demographics were found between the groups except ERAS group had more Roux-en-Y gastric bypass procedures (58.9% vs. 12.9%). A markedly reduced operation time (101 min vs. 147 min; p<0.001) and shortened LOS (2.6 days vs. 3.3 days; p<0.001) were observed, with signi cantly more ERAS patients achieving POD1 discharge (45.6% vs. 1.6%; p<0.001).

Conclusion
Unselective ERAS implementation in low-volume units is feasible and safe, with signi cantly reduced operation times and a shortened LOS without increased complications.

Background
While obesity is being recognized as a global epidemic [1], the commensurate rising prevalence of obesity and multiple comorbidities, such as hypertension (HTN), type II diabetes mellitus (DM), obstructive sleep apnea (OSA), dyslipidemia, etc., resulting in mounting pressure to offer proper treatment for this worldwide health hazard [2]. Interventions such as diet, exercise and intensive medical therapy generally fail to achieve durable effectiveness [3]. Instead, bariatric surgeries currently serve as a standard solution to settle this problem because of their well-recognized long-term effectiveness for weight reduction and comorbidity remission [4]. However, to widespread introduction of such a relatively complex treatment is not without risks [5]. Conventionally, cumulative surgical experience or high-volume practice generally yields superior outcomes [6]. However, it takes a long time to build up adequate experience and achieve this goal. Therefore, for low-volume practices, it is important to utilize alternative measures that are able to maintain safety pro les and facilitate quality improvement.
Clinical pathways such as enhanced recovery after surgery (ERAS) concept and alike, which consists of multimodal recommendations, has successfully been shown to be a valuable modality to attenuate perioperative stress and achieve faster convalescence for various surgical disciplines [7]. Systemic reviews and meta-analyses demonstrate that ERAS concepts could optimize perioperative care with improved results for bariatric surgeries [8]. Only technique pro ciency is generally believed to be a prerequisite to conduct such a practice [9]. In addition, there are no clear patient selection or exclusion criteria. For instance, individuals with an age beyond the regular limits (e.g., <18 or >60 years); super obese ; American Society of Anesthesiologists class >3 [10]; organ failure such as chronic kidney disease or congestive heart failure, etc. [11], are traditionally deemed unsuitable for ERAS enrollment. In contrast, other clinicians have considered that the presence of multiple comorbidities should not preclude patients from bene ting from such protocols [12]. In summary, because of the aforementioned factors, the literature remains scarce with respect to the feasibility and results to conduct such a plan in a low-volume practice. The primary aim of this study was to verify e cacy of ERAS in a low volume setting. All 30 days' adverse events are carefully audit and served as the secondary outcome measures. By means of an unselective approach, our goal was to increase its generalizability.

Methods
Data were retrieved retrospectively from a prospectively maintained database between January 2015 and December 2018 and all procedures performed in this study were in accordance with the ethical standards in the 1964 Declaration of Helsinki and its later amendments. The local institutional review board approved this study. The requirement to obtain formal consent was waived for this kind of study. All patients met the regional criteria [13] and were enrolled unselectively. Those who underwent non-primary surgery or for metabolic purpose that had a body mass index (BMI) of <32.5 kg/m 2 were excluded from this analysis. The periods analyzed were two years before and after the implementation of the ERAS protocol. Patients were strati ed into control (2015-2016) or ERAS groups (2017-2018) based on case sequence. Procedure selection was conducted through a shared decision-making process after full clearance of e cacy and risks based on literature at the time. Affected by growing concern regarding long-term sequelae, such as anemia and bile re ux after one-anastomosis gastric bypass (OAGB), as the study progressed [14][15][16]; we modi ed our approach to preferably suggesting Roux-en-Y gastric bypass (RYGB) for younger patients (<40 yrs.) and those with gastroesophageal re ux disease or DM in ERAS group despite the lack of established guidelines to support such practice. Preoperative routine evaluations included blood tests, chest radiographs, electrocardiograms, echocardiograms, abdominal ultrasound and upper gastrointestinal endoscopy.

ERAS protocol
A pragmatic ERAS protocol in accordance with the available guidelines at that time that utilized basic and essential changes was fully initiated in January 2017 [17]. Interventions consisted of proper preoperative education and preparations, intraoperative care, and early ambulation and oral intake postoperatively. Table 1 highlights the differences between this protocol and the previous standard of care. Before admission, patients underwent thorough instruction with emphasis on clear expectations and goals in addition to general information regarding the risks and bene ts of the surgery as per surgical consent. Premedication was also provided for ERAS patients. Both groups received thromboprophylaxis according to the individual risk pro le. Routine nasogastric tube, abdominal drainage, or urinary catheter placement were discontinued after ERAS implementation.

Anesthesia
A standard regimen for general anesthesia was given after full preoxygenation in the ramped position. In the interim, positive end-expiratory pressure was utilized to prevent the formation of atelectasis, with concerted efforts focused on maintaining normothermia, euglycemia, and euvolemia. Apart from standard care, bolus dexamethasone (10 mg) and droperidol (0.625 mg) were introduced for postoperative nausea and vomiting (PONV) prophylaxis in the ERAS group. As per the multimodal analgesic regimen, a particular emphasis was placed on the total elimination of opioid medication and its equivalent usage under the aid of laparoscopically guided transversus abdominis plane (TAP) block. Additionally, an optimal muscle tension monitor (train of four) was utilized to facilitate deep neuromuscular blockage plus sugammadex (2-4 mg/kg) administration for its reversal. Postoperatively, ERAS patients treated with routine and on-demand multimodal analgesics and antiemetic agents (i.e., Dynastat, acetaminophen, ondansetron). The clinical pathway included avoidance of uid overload (<2L/day) and rigorous early ambulation.
Liquid diet was commenced on the rst postoperative day in the ERAS group with no further need for a swallow study. Discharge criteria included no surgical complications, no fever, stationary hemodynamic parameters, a negative routine laboratory survey, tolerable oral uid intake and no need for further IV analgesics; this status was granted on POD1 or POD2 in the ERAS group. Each patient received a printed instructional handout when discharged.

Surgical technique
All surgical procedures were conducted laparoscopically. RYGB was performed by constructing a 30-mL vertical gastric pouch over a 32 Fr. calibrating tube, followed by an average 100-cm antecolic alimentary, 100-cm biliary limb; linear stapled gastrojejunostomy, jejuno-jejunostomy and suture closure of mesenteric defects via non-absorbable suture.
For OAGB, the technique involved rst stapling via Crow's foot and subsequent multiple rings alongside a Fr. 32 calibration tube with stapled gastrojejunostomy. The average length of the biliary limb was 200 cm. Laparoscopic sleeve gastrectomy (SG) involved multiple rings of the proper linear stapler 4-6 cm from the pylorus along a Fr. 32 calibration tube.
Data collection and statistical analysis Demographic features, including age, gender, BMI, type of operation and history of common comorbid conditions, e.g., HTN, DM, and dyslipidemia, together with all relevant outcome measures, including operating times, length of stay (LOS), emergency room (ER) visits, readmissions, reoperations, and any 30-day complications, such as unplanned procedures/interventions and mortality, were collected and analyzed between groups. Prolonged stay was de ned as an LOS over 3 days. Any adverse events including outside facility ER visits were obtained by inquiring directly from all our patients. Postoperative thirty-day complications were recorded and graded according to the Clavien-Dindo classi cation (CD) [18], and CD grade ≥ IIIa was considered a major complication. Statistical analyses were performed using Statistical Package for the Social Sciences software version 20.0 (SPSS, Inc., Chicago, Illinois, USA). Data are reported as the mean ± standard deviation (SD) or as counts and percentages when appropriate. Chi-square tests or Fisher's exact tests were used to compare two categorical variables. Tests for statistical signi cance were two-sided with a level of signi cance of 0.05.

Results
From Jan 2015 to Dec 2018, a total of one hundred and eighty-four consecutive patients underwent bariatric surgery at our hospital. In the ERAS group, 26 patients receiving metabolic surgery with BMI<32.5 kg/m 2 and another four patients who underwent revision surgeries were excluded. Among the control group, two patients for metabolic purpose with BMI<32.5 kg/m 2 were excluded, leaving a total of 152 patients enrolled in this study. Of these patients, 90 (59.2%) were in the ERAS group, and 62 (40.8%) were in the control group.
Five patients in the ERAS group experienced 30-day complications, with details described in Table 4. Of these, four minor complications were identi ed, including a patient with focal abdominal wall hematoma.
Two other patients visited the ER after discharge; one for self-limited abdominal pain and another patient with hematemesis who visited the ER at an outside hospital and was arranged to be readmitted. The fourth patient had gastrointestinal bleeding that required a total of 4 days of in-hospital observation. Last, one particular patient who had to abort the index operation due to an obscured surgical eld was classi ed as having major complications. This speci c patient underwent a second surgical attempt four months later uneventfully after vigorous diet control.
In the control group, six patients experienced complications. Among these were four minor complications, including one case of transient liver dysfunction. The other three patients visited the ER, with one presenting mild fever and another reporting self-limited abdominal pain. The third patient had PONV and was readmitted for hydration. Two major complications were identi ed, namely, one patient was readmitted through the ER for an intra-abdominal hematoma that required image-guided drainage; another patient of anastomosis stenosis that required readmission via the ER for balloon dilatation under general anesthesia, which comprised the second unplanned intervention.

Discussion
Our analysis demonstrates that unselective implementation of ERAS protocol can be safe in a low volume unit and realize its advantage by a marked decrease in operation time, shortened LOS and with more patients securely discharge on POD1 without an increase in complications.
Since the concept of a multimodal approach to control postoperative pathophysiology and improve recovery was rst introduced in 1997 [19], relevant protocols have evolved. The rationale is to hasten convalescence by reducing perioperative stress. From then onwards, guidelines have been established regarding recommendations for integral ERAS components in bariatric eld [17]. Recently, several systemic reviews and meta-analyses have demonstrated the superiority of enhanced recovery protocols in terms of carrying out more e cient surgical procedures, lessening the length of hospitalization and effectively reducing overall morbidities compared with standard care [8,20,21]. That being said, to generally implement the ERAS regimen remains questionable since this approach is not without risks [9,22]. For instance, Rebibo et al. reported a cohort with a 40% increase in the readmission rate (from 4% to 5.6%) [23] and others observed an increase in the ER visit rate [24]. Using the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program data set, Inaba et al. found a signi cantly higher morbidity (3.76% vs. 1.54%) and mortality rate (0.94% vs. 0.05%) when comparing same-day discharge to POD1 discharge [25]. Similarly, Morton et al. discovered that an LOS of ≤1 day was associated with a signi cantly increased risk of 30-day mortality (OR 2.02) for RYGB patients [26]. In fact, ERAS be implemented under the setting of specialized high-volume centers is frequently deemed a prerequisite to conduct these projects safely [9]. For example, McCarty et al. proposed that 84% of patients can be discharged within 23 hours postoperatively with a readmission rate as low as 1.7% for 2000 consecutive RYGB patients [27]. Similarly, Jacobsen et al. published an ERAS cohort based on a single high-volume center with a signi cantly shortened hospital stay (from 3 days to 2 days), an early complication rate as low as 2.8% and a readmission rate of only 1.9% [28]. In this regard, it seems reckless to conduct ERAS in a non-accredited low volume unit with fewer than 50 perennial cases.
Moreover, unlike our study background, a vast accumulating experience can usually be found when referring to preceding research that has been conducted with low case numbers. Hahl et al. analyzed data from 318 patients who underwent RYGB during a 4-year period with excellent results showing a mean LOS of only 1.3 days and 83% of patients discharged on POD1 [29]. Notably, their accumulative experience at that time was already far more than ve hundred cases. Similar to our study, Awad et al. introduce a series of 226 cases that comprised various bariatric procedures. With a low 30-day complication rate of only 4.4% and a readmission rate of 2.7% [12]; however, this particular study were conducted at a regional high-volume tertiary referral center. Therefore, whether these superior outcomes merely re ect clinician pro ciency remains unclear. Some other studies were notably undertaken via independent patient-selection or procedure-selection process. Such as Sasse et al. presented a 38 RYGB case series with a 100% POD1 discharge rate and a low 30-day complication rate of only 2.6% [30]. However, their study group represented fewer than 3% of their total RYGB cases as a result of stringent patient selection. Likewise, Fares et al. accomplished a high POD1 discharge rate of up to 94.8% from a consecutive 96 RYGB case series in a small, teaching community hospital with only 5.2% complications [31]. Nevertheless, the study group comprised 55% of patients selected from a total caseload of 173. Lam et al. reported their recent work which yielding remarkable results in terms of 83.1% of patients achieving POD1 discharge with a 1.5% readmission rate whereas they selected 130 out of 240 total cases in their research, and all subjects were receiving SG [32]. Even so, there was no coherent criterion for subject selection and risk strati cation among these studies. On the other hand, a noticeable tendency toward an increase in adverse outcomes was found in several studies that were speci cally endorsed with an unselective approach. For example, Geubbels et al. followed a cohort of 360 unselected RYGB patients and found an increasing early complication rate (from 17.3% to 18.3%) and an increasing readmission rate (from 4.8% to 8.1%) after ERAS conduction [33]. Similarly, Mannaerts et al. reported a signi cantly higher minor complication rate (20.7% vs. 16.1%) and a higher ER visit rate (16.8% vs. 12.5%) [34]. By unselective approach, our series comprised various susceptible groups; for example, 5 (5.6%) patients were aged >65 yrs. (range, 65-74 yrs.), 4 (4.4%) patients had a BMI of >50 kg/m2 (range, 52.5-73.7 kg/m2), and two other patients had a wheelchair-bound status. Unlike the aforementioned studies, a tendency of decreasing 30-day ER visits (2 (2.2%) vs. 5 (8%)), readmissions (1 (1.1%) vs. 3 (4.8%)), 30-day complication rates (5 (5.5%) vs. 6 (9.7%)) was found via our ERAS regimen despite none of these values reaching statistical signi cance. Compared to an LOS ranging from 1 day to 2.9 days, readmission rates between 1.7% and 8.1%, early complication rates between 2.8% and 18.3%, and mortality rates of up to 0.7% across former, high-volume studies [12,[27][28][29]35], a total LOS of 2.6 days was accomplished in our series and fall in line with these valuable large studies.
It is clear that a major difference in procedures was noted between groups, with signi cantly more patients who underwent RYGB in the ERAS group. Subgroups analysis revealed similar trend of advantages after ERAS implementation in terms of operation time (RYGB, 128 min vs. 104 min; OAGB, 150 min vs.99 min), LOS (RYGB, 3.4 days vs. 2.5 days; OAGB, 3.3 days vs. 2.8 days) and accomplishment of POD1 discharge (RYGB, 0% vs. 54.7%; OAGB, 1.9% vs. 33.3%) with no increment of complications. Though there were lacks of comparison study regarding impact of ERAS for individual procedure, the index protocol bene ts both procedures in current study. Another interesting nding in our study was OAGB (n=53) took signi cant longer op time than RYGB (n=8) in control group. In addition to the research sample is too small, our results can be partially explained by differences in the demographic characteristics between groups with RYGB comprising younger patients (31.3 yrs. vs. 39.8 yrs.) with a lower BMI (37.5 kg/m2 vs. 42.0 kg/m2) for these are well -identi ed factors that affect the operation time.
Our study shows that ERAS can be safely performed unselectively under low-volume setting and provided with bene cial effects usually reported from high-volume, specialized centers.

Limitations
Our analysis has several limitations that should be taken into consideration.
First, inherent to its retrospective nature, quality measures such as pain score, nausea episodes, etc., cannot be thoroughly collected. In addition, we report only 30-day morbidities, thus leaving long-term complications unreported. However, the strengths of this study reside in the efforts to capture of all complications since no patients dropped out within 30 days postoperatively.
Second, the marked improvement can be interpreted to be caused by the increase in expertise rather than the positive impact of the ERAS protocol. Nevertheless, like all preceding studies conducted across two time frames, it is hard to total eliminate time bias. Having said that, consider current study was based on such a small sample size within a short period and comprised only a single surgeon's work; we believe biases from major technique progress and aws of individual practice can be avoided.

Page 9/16
Our study veri es the safety and feasibility of conducting the ERAS protocol over a wide spectrum of patients in the context of mixed procedures under a low volume setting. Given the small study population, more relevant data and follow-up are required to elucidate the continual bene cial effect and long-term results.

Consent for publication
Written informed consent for publication was waived because no clinical details and/or clinical images regarding the cases were disclosed.

Availability of data and materials
The datasets generated during and/or analyzed during the current study are not publicly available due to restrictions from local Institutional Review Board but are available from the corresponding author on reasonable request and with permission from the local Institutional Review Board.

Competing interests
The authors declare that they have no competing interests.

Funding
The study was not sponsored and funded by any funding.
Authors' contributions HC designed the study, performed the surgical procedures, followed the patients, and participated in the data analysis and writing of the manuscript. WK contributed to the data analysis. YC, AC and YT participated in patient anesthesia. The authors read and approved the nal manuscript.