Data were collected from May 2017 to February 2019. A total of 105 patients were enrolled for evaluation of eligibility, and 90 patients entered the ERAS protocol and were ultimately analyzed (Table 1). The mean age of our patients was 59.7 years; 54% were male and 46% were female. The mean body mass index (BMI) was 24.6 (SD = 3.8), and ASA classification II accounted for the majority of our patients (ASA classifications 1, 2, and 3 = 22.2%, 71.1%, and 26.7%). No significant difference was observed among the 3 study stages in terms of BMI or ASA classification. Robotic surgery was performed on 68 (75.6%) patients. Among the operations, low anterior resection (LAR) was the most common procedure, performed on 64 patients (71.1%); the next most popular was anterior resection (AR) with left hemicolectomy (LH), used on 15 patients (16.7%). The proportion of robotic surgery versus laparoscopic surgery and for the procedures implemented (right hemicolectomy [RH], AR and LH, or LAR) were similar for all stages. In total, 63 patients (70%) were classified as having rectal cancer, and the proportion of patients receiving concurrent chemoradiotherapy (CCRT) in our study was relatively high. Synchronous metastatic disease was detected in 4 (4.4%) patients preoperatively.
All patients in all 3 stages were nonsmokers or had not smoked within a minimum of 24 h before surgery (Table 2). No patient undergoing RH received MBP preoperatively. All patients but one who underwent AR/LH/LAR consumed clear liquid diets for the 24 h preceding surgery and took commercial Bowklean® powder suspension (magnesium oxide + sodium picosulfate + citric acid anhydrous; Genovate Biotechnology, Taiwan) in split doses for MBP. Oral metronidazole and neomycin were added to the MBP. Routine prophylaxis with intravenous antibiotics (first generation cephalosporin) were given to all patients except one, who had received therapeutic antibiotics for a preexisting intraabdominal infection. A standard anesthetic protocol was applied to every patient in our study, and normothermia was strictly maintained intraoperatively. PCA with intravenous fentanyl was used in 10, 13, and 20 (33.3%, 43.3%, and 66.7%) patients in stage 1, 2, and 3, respectively, for 3 days (range 2–4 days) postoperatively. Widespread use of PCA was discouraged mostly due to personal consideration of patients. Multimodal analgesia was applied to all patients to avoid excessive opioid use. However, 9 (30%), 2 (6.7%), and 4 (13.3%) patients in stages 1, 2, and 3, respectively, received additional opioid treatment postoperatively for pain relief. Significantly lower proportions of patients received additional opioid treatment in stages 2 and 3 compared with the proportion of patients receiving opioid treatment in stage 1 (P = 0.0452). Multimodal prophylaxis and treatment of PONV were used for 29, 25, and 29 (96.7%, 83.3%, and 96.7%) patients in stages 1, 2, and 3, respectively. Moreover, prevention of postoperative ileus was implemented in the majority of our patients (93.3%, 96.7%, and 86.7% in stages 1, 2, and 3, respectively). Removal of transurethral bladder drainage 1 day after colonic surgery and 3 days after rectal surgery was recommended. At least 90% of patients in each stage followed this recommendation, and only one episode of urinary retention was recorded, namely in a 62-year-old woman who had received robotic LAR for low-lying rectal cancer. The prohibition of routine NG tube use was adhered to. If decompression was indicated, the NG tubes were removed before reversal of anesthesia. This was undertaken in 23, 24, and 29 (76.7%, 80%, and 96.7%) patients in stages 1, 2, and 3, respectively, and compliance was significantly improved (P = 0.0405) after the implementation of ERAS. In terms of diet, 87 of 90 (96.7%) patients resumed early enteral feeding, starting with a clear liquid diet 24 h after operation. The goal of a full diet within 48 h postoperatively was achieved for 86.7% to 93.3% of patients. Early mobilization was observed in 20 (66.7%) patients in stage 1, and this proportion increased to 80% in stage 2. Of the patients in stage 3 wearing smartbands, 25 (83.3%) achieved early mobilization. Compliance with respect to early mobilization was not significantly increased in patients with smartbands compared with patients without them. After improvement to ERAS protocol, compliance was greater than or equal to 80% for all items, except for the PCA ratio after stage 2 (stage 2 and 3).
The mean total LOS in stage 1 was 11.4 days (SD = 2.8 days), which significantly decreased to 10.1 days (SD = 1.0 days; P = 0.0230) in stage 2 and to 10.0 days (SD = 1.3 days; P = 0.0159) in stage 3 (Table 3 and Figure 2a). The total LOS in stage 3 and in stage 2 was almost identical (P = 0.9896). After cases with complications were excluded, the LOS in uncomplicated patients decreased from 10.7 days (SD = 2.1 days) in stage 1 to 10.0 days (SD = 0.9 days) in stage 2, without statistical significance (P = 0.1800). However, the LOS of uncomplicated patients wearing smartbands in stage 3 exhibited a further decrease to 9.8 days (SD = 0.7 days) compared with stage 1 (P = 0.0489). Theoretical LOS also significantly shortened form 8.9 days (SD = 1.6 days) in stage 1 to 7.8 days (SD = 1.4 days) in stage 3 (P = 0.0091). Meanwhile, theoretical LOS in stage 2 was 8.5 days (SD = 1.0 days) without significant decreasing compared with that in stage 1 (P = 0.4396). The time to recovery of bowel function, including time to flatus passage and time to stool passage, was not shortened after implementation of the ERAS protocol or after introduction of smartbands during perioperative care. Patients had flatus passage after 1.5 and 1.6 days, on average, after MIS in stages 1 and 2 and in stage 3, respectively. In stage 1, stool passage was observed at a minimum of 2.1 days (SD = 1.0 day) after operation. Patients started to drink clear liquid 1 day after MIS in stages 1 and 2. A full diet was resumed 2.6 days (SD = 1.4 days) postoperatively in stage 1 and 2.1 days (SD = 0.6 days; P = 0.1265) in stage 2. Patients in stage 3 had similar outcomes in early enteral feeding compared with those in stage 2.
The number of complications in each stage in our study tended to decrease after implementation of ERAS (Figure 2b). The total complication rate was 13.3%, 10.0%, and 6.7% in stages 1, 2, and 3, respectively (P = 0.6860). In stage 1, 4 cases had complications. Symptomatic anastomotic leakage was observed 6 days after robotic LAR in a 46-year-old male patient, and he developed deep surgical site infection (SSI) and sepsis. An urgent defunctioning stoma was created to manage complications, and the patient was discharged with a total LOS of 19 days. Another 3 patients developed chylous ascites with ileus, urinary tract infection, and pneumonia, respectively. In stage 2, 3 medical complications were observed. A 60-year-old female patient who received CCRT and robotic LAR exhibited intraabdominal infection (IAI) and ileus after being discharged from her ward (postoperative LOS, 6 days). She was readmitted for antibiotic treatment. SSI and prolonged postoperative ileus developed in a 61-year-old male patient with diabetes. Acute urinary retention is another example of complications occurring in stage 2. A 59-year-old female patient with a BMI of 30 who received laparoscopic RH in stage 3 developed anastomotic leakage and severe IAI. Reoperation was performed, and she received reanastomosis and a temporary ileostomy with a total LOS of 16 days. A patient in stage 3 also experienced prolonged ileus.
To evaluate the quality of recovery after surgery and anesthesia, QoR-40 was used to measure each patient’s health status preoperatively, at POD1, at POD3, and on the day of discharge. Scores were compared only between stages 2 and 3 when the enhanced recovery program had been well established and ERAS skills were enhanced in our surgical team. In the stage 2 group, the preoperative total baseline score was 178.3 points, and this decreased to a minimum of 156.8 points at POD1 (−21.5 points, P < 0.0001; Table 4 and Figure 3). All dimensions decreased synchronously except for psychological support. The total score recovered to 170.4 points (−7.9 points) by POD3, but this was still significantly lower than the baseline score (P = 0.0463). The scores for emotional state (35.1 points, −2.9, P = 0.0370) and physical independence (19.6 points, −3.8, P < 0.0001) remained much lower than they had been preoperatively. However, patients were restored to baseline physical comfort scores (52.4 points, +0.2, P = 0.8716) and pain scores (32.1 points, −0.8, P = 0.1261) by POD3. Scores in all dimensions returned to or were better than baseline scores by the day of discharge. The total score was 181.9 points (plus 3.6, P = 0.3692) at the day of discharge in the stage 2 group.
QoR-40 score was also recorded in stage 3 when patients were assigned to wear smartbands postoperatively. The total preoperative score was 182.1 points; similarly, this significantly declined to a minimums of 161.9 (−20.2, P < 0.0001) by POD1. At POD3, the total score and scores for all 5 dimensions recovered to the baseline level. The scores for emotional state (37.5 points, −1.9, P = 0.1749) and physical independence (22.0 points, −1.2, P = 0.1822) in POD3 seemed to improve more rapidly postoperatively after the introduction of smartbands into ERAS perioperative care. Patients experienced slightly worse pain scores (31.9 points, −1.5, P = 0.0549) on POD3. Compared with the total QoR-40 score of patients in stage 2, that of patients in stage 3 recovered to an adequate level by POD3 (175.4 points, −6.8, P = 0.0988). The average score on the day of discharge was 183.5 points, 1.4 points higher than the baseline score (P = 0.7409), with all dimensions exhibiting recovery to baseline of higher scores.