An Institutional Experience of Introducing an Enhanced Recovery After Surgery (ERAS) Program for Pancreaticoduodenectomy

Pancreaticoduodenectomy (PD) remains the mainstay of surgical treatment for hepatobiliary and pancreatic malignancy and offers the only chance of long-term survival.^1,2 However, the morbidity after PD has been reported to range from 30% to 65%, and the complications are sometimes fatal.^3–8 Moreover, previous studies have demonstrated that the development of postoperative complications increases the risk of both disease recurrence and overall survival.^9,10 Therefore, it is important to determine the most appropriate perioperative care.

Enhanced recovery after surgery (ERAS), fast-track, or clinical pathway programs are multimodal strategies that aim to attenuate the loss of or to improve the restoration of functional capacity after surgery.^11,12 ERAS programs have many elements, including preoperative education, preoperative carbohydrate loading, omission of bowel preparation, epidural analgesia without opioids, early postoperative enteral feeding, and early mobilization of patients.¹³ ERAS is considered to reduce the rate of morbidity and shorten the length of hospital stay.^12–14 Several studies have reported that ERAS programs are feasible and useful in gastric cancer surgery and colorectal surgery.^12,15

Recently, an international working group within the ERAS society published a comprehensive and evidence-based framework to enhance perioperative care in patients undergoing PD.¹⁶ However, few studies have assessed the implementation of an ERAS program following PD.^17–19 Some studies have reported the effect of ERAS on reducing the postoperative length of hospital stay and hospital costs. Other studies have demonstrated lower morbidity, mortality, and readmission rates following the implementation of fast-track programs. However, the safety and feasibility of ERAS has not been well evaluated in patients undergoing PD.

The aim of this study was to evaluate the safety and feasibility of implementing an ERAS program for patients undergoing PD in a high-volume center.

Patients and Methods

Preoperative care

Preoperative counseling was held in the outpatient clinic before hospitalization and in the ward after admission. Patients were allowed to eat a normal diet until dinner of the day before surgery. Magnesium oxide was administered on the day before surgery (Table 1).

Table 1  Time table of the enhanced recovery after surgery program

View Fullscreen

Results

Patients

One hundred nine patients were eligible for this study and underwent PD following the introduction of an ERAS protocol. Patient background and operative data are shown in Table 2. There were 59 men and 50 women. The median age was 70 years (range, 49–86 years). The American Society of Anesthesiologists physical status classification was 2 in most patients. The median body mass index (BMI) was 21.5 (range, 14.9–31.2).

Table 2  Clinicopathological features

View Fullscreen

Surgical morbidity and mortality

Postoperative complications were found in 56 of the 109 patients (51.4%). Surgical mortality was observed in 2 patients (1.8%) as a result of pneumonia. The details of the complications are shown in Table 3. Pancreatic fistula was the most frequently diagnosed complication, followed by abdominal abscess, and DGE. Grade 2 complications occurred in 64.2% of the patients, grade 3 in 28.6%, grade 4 in 3.6%, and grade 5 in 3.6%.

Table 3  Details of complications

View Fullscreen

Discussion

This study evaluated the safety and feasibility of the ERAS program in patients who underwent PD. In our study, the overall incidence of morbidity was 51.4% and the incidence of mortality was 1.8%. These results obtained with our ERAS program are safe compared with morbidity and mortality rates without the ERAS program (30%–65% and 1%–5%, respectively). Our ERAS program included 12 elements. Among the 12 elements, 4 of the preoperative elements and 2 of the intraoperative elements seemed feasible. However, of the 5 postoperative elements, only 1 parameter and NGT removal seemed feasible, while the rest of the parameters seemed unfeasible. Moreover, the median postoperative hospital stay was 23 days. The median length of stay was significantly shorter than our pre-ERAS value, while the median length of stay was longer than previous ERAS reports.

In the present study, the postoperative complications were found in 56 of the 109 patients (51.4%) and surgical mortality was observed in 2 patients (1.8%). Further, pancreatic fistula was the most frequently diagnosed complication, followed by abdominal abscess and DGE. The overall incidence of postoperative complications and mortality were similar to other PD fast-track studies. For example, Robertson et al evaluated the safety and efficacy of implementing an ERAS program for 50 consecutive patients undergoing PD.¹⁸ They found that the postoperative complication rate was 46% (23 patients), and the mortality rate was 4% (2 patients). DGE was the most common complication (14%). In addition, Braga et al assessed compliance with an ERAS protocol and its impact on short-term outcome in 115 consecutive patients undergoing PD.¹⁷ They found the overall postoperative complication rate was 60.0% (69 patients), and the mortality rate was 3.5% (4 patients). Complications of grade 2–5 according to the Clavien-Dindo classification affected 60 patients (52.2%). Moreover, Balzano et al compared a large group of patients treated according to a fast-track program group (n = 252) with a conventionally treated group after PD (n = 252).¹⁹ They found that the rates of pancreatic fistula and other intra-abdominal complications were similar in the 2 groups. In addition, DGE was significantly reduced in the fast-track group (13.9% versus 24.6%, respectively; P = 0.004). These results suggest that the ERAS program is safe and feasible without an increase in postoperative complications and mortality, and the application of the ERAS program was not associated with any harmful effect.

ERAS is a structured, multimodal, perioperative strategy aimed at reducing surgical stress and improving patient functional capacity and quality of life after surgery.³⁰ It provides the patient with preoperative education regarding their expectations in recovery, and it is also intended to give a structured timeline for health care personnel in the management of patients. Preoperative elements and intraoperative elements were met by the majority of patients in the present study. However, postoperative elements, especially early removal of catheters, were more challenging. The number of patients having their urinary catheter removed on POD 3 was lower than expected. This was mainly because the epidural catheter was removed on POD 5 in a majority of patients. Urinary catheters were often removed only after epidural analgesia had been stopped. Why was the epidural catheter removed on POD 5? One possibility is that neither a nonsteroidal antiinflammatory drug (NSAID) nor acetaminophen was routinely used in the postoperative course in this study. Baseline NSAID analgesics could be sufficient for the prevention of postoperative pain regardless of the short-term action of the epidural analgesic.³¹ Less pain facilitated early mobilization. Moreover, Lassen et al previously recommended the use of NSAIDs or acetaminophen oral multimodal analgesia in their guidelines for perioperative care for PD.¹⁶ A revised ERAS protocol should include the use of NSAIDs or acetaminophen and earlier epidural catheter and urinary catheter removal. Another concern was the removal of intra-abdominal drains. When comparing the previous reports and our present study, the removal of intra-abdominal drains were trend to later in the present study. There is still debate on whether or not drains should be routinely placed following PD, and their use is based on low-level evidence. However, a randomized controlled trial found significantly decreased rates of pancreatic fistula, abdominal, and pulmonary complications with early drain removal (POD 3) compared with late drain removal (POD 5 and later) in patients at low risk for pancreatic fistula (amylase values of <5000 U/L on POD 3).^32,33 A revised ERAS protocol should perhaps include earlier analysis of drain fluid amylase, with the target of achieving earlier drain removal.

The overall median length of stay was 23 days. Although the median length of stay was significantly shorter than our pre-ERAS value, the median length of stay was longer than previous ERAS reports. This discrepancy can be explained by several factors. One possible reason is the late removal of intra-abdominal drains. A randomized controlled trial found that a prolonged period of drain insertion is associated with increased hospital stay.¹⁵ A second possible reason is that physicians are reluctant to discharge PD patients too soon after surgery, since major complications can even occur with some delay after surgery. In a recent study in patients undergoing colonic surgery, only half of the patients were discharged when all discharge criteria were met, and 43% of this group was kept in hospital because of surgeons' judgment. For this reason, several authors advise to at least add data on time-to-recovery compared with length-of-stay as outcome of ERAS programs.³⁴ A third possible reason is that the medical system after discharge is different in Japan than in other countries. Almost all the patients were discharged to home in this study, while discharge with home health care or discharge to a nursing facility might be used in other studies in other countries.³⁵

There were many limitations in this study. First, this was a retrospective single-center study with a relatively small sample size. A randomized controlled trial comparing ERAS with traditional care in PD would be the strongest way to assess the true effect of ERAS. However, this brings up ethical questions, since the majority of ERAS protocol elements have been separately proven to be effective in randomized controlled trials and meta-analysis, and several ERAS protocol elements have now become standard care.³⁶ Second, our ERAS program did not include fluid management, which is one of the key elements of an ERAS program. Third, the optimal period of starting nutrition support was unclear in the present study. Oral intake of solid food stimulates pancreatic gland secretion of pancreatic juice; pancreatic surgeons have largely been conservative in starting early enteral feeding for pancreatic surgery patients. One concern of pancreatic surgeons is how early to start patients on a solid diet without causing a severe adverse event. It seems appropriate in settling the clinical question. However, there is a lack of evidence to support the optimal period for starting solid food, and early diet has been shown to be safe following major upper GI and hepatobiliary-pancreatic surgery, including PD. Therefore, we chose this period from the previous ERAS program for PD. However, it is unclear whether this period is fast or not. Fourth, most patients had good performance status. Patients with poor performance status (e.g., Eastern cooperative oncology group performance status ≥3, severe dementia, swallowing difficulty) could not be treated in our hospital because we specialize in cancer treatment. This could be selection bias.

In conclusion, this study showed the feasibility and safety of an ERAS program following PD. However, some targets, especially postoperative elements, were challenging. Further modification of the protocol and additional education of patients and health professionals may help to improve recovery, shorten hospital stay, and enhance the return of normal function.