This study was approved by the Ethical Committee of the Korea University Guro Hospital Institutional Review Board, Seoul, Republic of Korea (ref.: 2019GR0294) on 21 August, 2019, and registered at the ClinicalTrials.gov (ref.: NCT04082039). Between September 2019 and June 2020, female patients aged 19–75 years with American Society of Anesthesiologists (ASA) physical status I–II, who were scheduled to undergo elective TLH under general anesthesia, were enrolled in this study after providing written informed consent. Patients were excluded if they had a body mass index >30.0 kg/m2, known hypersensitivity to the drugs used in this study, significant liver or renal dysfunction, or a history of drug abuse or dependence, recent major procedure, or surgery.

Participants were randomly assigned to either a 1-channel or a 2-channel group using a web-based computer-generated list and were unaware of their assignment. The randomized numbers were kept by a non-blinded researcher (SKO) in sequentially numbered envelopes, and the envelope was opened in the operating room only by the non-blinded anaesthesiologist (SKO) who was responsible for setting the PCA pump as per the study protocol. Other investigators who provided anesthesia and assessed the study endpoints after the operation in the post-anesthesia care unit (PACU) and ward were blinded to the group assignment. All PCA devices were applied to patients with labels only denoting the patient’s study number, so that neither the patient nor medical care providers and investigators could recognize the allocation.

Following the patient’s arrival in the operating room, routine physiological monitoring, such as pulse oximetry, electrocardiography, and noninvasive arterial blood pressure measurements, were performed. The bispectral index (BIS VISTA™; Aspect Medical Systems Inc., Norwood, MA, USA) was used to monitor the depth of hypnosis and maintained between 40 and 60 during surgery.

For preemptive pain management and maintenance of hemodynamic stability during the initial phase of the operation, 50 μg of fentanyl was administered intravenously in both groups immediately before the induction of anesthesia. Anesthesia was induced using 2 mg/kg of intravenous propofol, followed by 0.6 mg/kg rocuronium, which was administered after the loss of consciousness (BIS < 60) for endotracheal intubation and maintained with desflurane and 50% oxygen in the air. The concentration (vol%) of desflurane was titrated using the bispectral index value. The esophageal temperature was maintained at approximately 36°C using a warm air blower. Five minutes after the uterus incision, 50 μg of fentanyl was injected to control the pain from surgical stimuli perioperatively. An addition, 50 μg of fentanyl was administered 30 min after the last dose of fentanyl until the laparoscopic procedure ended. At the end of the surgical procedure, neuromuscular blockade was reversed using 0.4 mg glycopyrrolate and 10 mg pyridostigmine. After confirming the response to verbal commands and spontaneous respiration, tracheal extubation was performed. If the mean arterial pressure decreased or increased by more than 30% from baseline, 4 mg of ephedrine or 0.5 mg of nicardipine, respectively, were administered.

The Bellomic Duo® silicone balloon infuser has a dual-channel infusion elastomeric pump. The continuous channel is a line with a fixed flow rate of 2 mL/h for basal infusion, and the selector-bolus channel can be set to 0, 1, and 2 mL/h with an adjustable flow rate and bolus function. The bolus volume is 1 mL, and the lock-out period is 10 min. The volume of each chamber was 100 mL, and the total volume was 200 mL. In this study, the Bellomic Duo® was used for both groups, but the same regimen of drugs was administered by dividing it into different channels. The methods were as follows for each group.

In the 1-channel group, fentanyl 16 μg/kg, ketorolac 2 mg/kg, and ondansetron 12 mg with normal saline to reach a total volume of 100 mL were administered in the selector-bolus channel at a rate of 2 mL/h, and 100 mL of normal saline only was present in the continuous channel to blind the group allocation. Therefore, only the selector-bolus channel was utilized, but the continuous channel was not utilized in the 1-channel group.

Meanwhile, both channels were utilized in the 2-channel group. In the 2-channel group, fentanyl 16 μg/kg with normal saline to a total volume of 100 mL was administered via the selector-bolus channel at a rate of 2 mL/h, and ketorolac 2 mg/kg, ondansetron 12 mg with normal saline to a total volume of 100 mL was administered via the continuous channel.

A schematic drawing of the PCA device and its application is shown in Fig. 2.

Schematic of the PCA device and its application. A In the 1-channel group, only the selector-bolus channel was actually utilized. B In the 2-channel group, both the selector-bolus channel and continuous channel were utilized by dividing the drugs between the two. The adjustable flow dial has a range of 0, 1, and 2 mL/h, but only 0 or 2 mL/h were used in both groups per our study protocol

In the PACU, the level of the patient’s pain was assessed using a verbal pain score (VPS; 0–3; 0 = no pain, 1 = mild pain, 2 = moderate pain, 3 = intense pain) (Du Manoir et al. 2003) at 10-min intervals, and the sedation score (0–3; 0 = clearly conscious; 1 = temporarily drowsy; 2 = drowsy but responsive to verbal communication; and 3 = drowsy without response to verbal communication) (Du Manoir et al. 2003) was also evaluated every 10 min. Twenty-five micrograms of fentanyl was administered if the patient had a VPS ≥ 2, respiratory rate ≥ 10 per minute, SpO2 ≥ 95%, and sedation score ≤ 1. When the VPS and sedation score were ≤ 1, the PCA device was connected.

After transferring the patients to the ward, the pain was assessed using a numeric rating scale (NRS; 0–100; no pain [0] to worst pain imaginable [100]) at 6, 12, 24, 36, and 48 h after surgery. The cumulative consumption of PCA over 48 h and the occurrence of adverse effects, such as nausea and vomiting, dry mouth, dizziness, urinary retention, headache, sedation, itchiness, shivering, respiratory depression, confusion, hypotension, and bradycardia, were monitored. Respiratory depression was defined as a respiratory rate < 10 per min or oxygen saturation < 90% for >1 min (Tsui et al. 1997). The patient’s overall satisfaction score (0–10; from unsatisfied [0] to fully satisfied [10]) was also assessed. If the postoperative pain control in the ward was insufficient (NRS score for pain > 40) or if an additional analgesic was requested by a patient, paracetamol 1 g or ketorolac 30 mg was planned as a rescue analgesic.

Nausea was assessed (0 = none, 1 = mild, 2 = moderate, and 3 = severe), and 4 mg of ondansetron was planned as a rescue antiemetic agent if a patient suffered severe nausea or retching/vomiting (Choi et al. 2017).

In both groups, the flow rate of the selector-bolus channel was initially set at 2 mL/h. When opioid-related side effects such as nausea, vomiting, or dizziness occurred, infusion of the selector-bolus channel was suspended by adjusting the flow rate to 0, while the continuous channel remained infused. After the symptoms were relieved and/or pain control was demanded by patients, the selector-bolus channel was re-infused at a rate of 2 mL/h and re-suspended in the recurrence of unbearable side effects. The decision of rate change was made by the evaluating doctors who round the patients at 6, 12, 24, 36, and 48 h after surgery and depending on the case the nurse in the ward gives a notice.

Postoperative quality of recovery was evaluated by the QoR-40, which has 40 questions in five domains (Choi et al. 2017). Each question that assesses patient recovery uses a five-point Likert scale (1 = none of the time, 2 = some of the time, 3 = usually, 4 = most of the time, and 5 = all of the time). The five domains are physical comfort, pain, physical independence, psychological support, and emotional state. The global QoR-40 scores range from 40 to 200. All patients answered the questionnaire on the day before surgery and 24 h after surgery.

The outcomes measured included (1) QoR-40 score on the day before surgery and at postoperative 24 h (primary endpoint), (2) VPS (0–3) measured every 10 min at the PACU, (3) the sedation score (0–3) measured every 10 min at the PACU, (4) rescue fentanyl administration in the PACU, (5) NRS score (0–100) at rest and while coughing in the ward, (6) cumulative PCA consumption over 48 h (mL), (7) nausea and vomiting (0–3) for 48 h after PCA connection, (8) occurrence of other side effects in the PACU and in the ward for 48 h after PCA connection, (9) incidence of the selector-bolus channel at a rate of 0 mL/h due to severe side effects, (10) patient’s overall satisfaction score (0–10; 0 = unsatisfied to 10 = fully satisfied), and (11) length of postoperative hospital stay.

The primary endpoint was the QoR-40 score at 24 h postoperatively. The sample size calculation was based on a mean QoR-40 score of 167 ± 23 at 24 h after surgery, which was derived from a previous study (Myles et al. 2000). We considered that a difference of more than 10% between groups with respect to the QoR-40 score was clinically significant following a previous study (Choi et al. 2017). Based on the assumption that the allocation ratio of the two groups was 1, 40 subjects were required in each group to achieve an alpha value of 0.05 and power of 80%. We aimed to assign 45 patients to each group after accounting for a 10% dropout rate.

Statistical analyses were performed using SPSS software (version 20.0; IBM, Corporation, Armonk, NY). The normal distribution of continuous data was first evaluated using the Shapiro–Wilk test (P > 0.05). Normally distributed data were analyzed using Student’s t test, and non-normally distributed data were analyzed using the Mann–Whitney U test. Student’s t test was used to compare the age, height, weight, anesthesia time, and operation time, while the Mann–Whitney U test was used for the QoR-40 score, rescue fentanyl dose, and PCA connection time at the PACU.

Ordinal parameters, including the VPS and sedation score at the PACU and patient’s overall satisfaction score, were compared using the Mann–Whitney U test, while categorical variables, including the ASA classification and incidence of adverse events, were compared using a chi-squared test or Fisher’s exact test. The change in the QoR-40 score from the preoperative period to the postoperative 24 h in each group was analyzed using the Wilcoxon signed-rank test, and the changes over time in the NRS score at rest and while coughing at 6, 12, 24, 36, and 48 h after surgery, and cumulative PCA consumption in the ward were compared using repeated-measures analysis of variance.

Data are expressed as the mean ± SD, median [IQR], or number of patients (%). P values were two-tailed, and a P value of less than 0.05 was considered statistically significant.