This double-blind, randomised controlled trial was approved by the Institutional Scientific Research and Clinical Trials Committee of International Peace Maternity and Child Health Hospital (GKLW-A-024-032-01) and was registered in advance with the Chinese Clinical Trials Registry (Registration No.: ChiCTR2400087477).

Pregnant women scheduled for elective caesarean delivery under spinal-epidural anaesthesia were enrolled between 1 August 2024 and 15 April 2025. Eligible patients were women aged 20–40 years, classified as American Society of Anaesthesiologists (ASA) physical status II, carrying singleton pregnancies at 37–42 weeks of gestation, who requested postoperative analgesia. The exclusion criteria included contraindications to neuraxial anaesthesia, allergy to study medications (LB, ropivacaine, hydromorphone, oxycodone, or acetaminophen), body mass index (BMI) > 36 kg/m², gastrointestinal disorders, a history of substance abuse, or refusal to participate. Participants were withdrawn if protocol deviations occurred (e.g., epidural catheter failure, or noncompliance with oral analgesia).

Randomisation was performed using SPSS-generated random numbers (1:1 ratio), with assignments sealed in opaque envelopes. Envelopes were distributed to participants during preoperative assessments and opened by anaesthetists postoperatively to prepare study medications. Owing to visible differences between LB and saline, the anaesthetists administering the TAP block were unblinded, while the participants, research assistant in charge of follow-up, and data analysts remained blinded to the group allocation.

Participants were recruited by a research assistant on the day before the surgery, informed of study protocols, and provided written informed consent. The Chinese version of the 15-item Quality of Recovery (QoR-15) scale was used to assess the participant’s preoperative condition. Each participant was then provided with a sealed randomisation envelope. Participants were instructed to fast from midnight prior to surgery.

On arrival in the operating room, intravenous access was established, and standard monitoring including electrocardiography, non-invasive blood pressure measurement, and pulse oximetry was initiated. Combined spinal-epidural anaesthesia was administered to the L3–L4 interspace and 15 mg of 0.75% ropivacaine was injected intrathecally, followed by epidural catheter placement. Sensory block level was assessed using pin-prick to ensure that the block reached to or above T6 before surgery; otherwise a 5-mL dose of 1% lidocaine was administered through the epidural catheter.

During the surgery, obstetricians administer oxytocin, carbetocin, ergometrine, or prostaglandin-based uterotonic agents based on the patient’s uterine contractions, with oxytocin being the standard intraoperative choice. The number of uterotonic agents used is documented before the patient leaves the operating room.

Before surgical closure, 0.6 mg of epidural hydromorphone (diluted in 6 mL saline) was administered. Participants were provided with a patient-controlled intravenous analgesia (PCIA) device (ZZB-III, APON Corp, Nantong, Jiangsu, China) containing oxycodone 30 mg in 60 mL saline, provided as 2-mL PCIA boluses with a 10-min lockout interval and no background infusion, after the surgery.

Immediately after completing the surgery, the randomisation envelope was opened by the anaesthetist, and bilateral ultrasound-guided lateral TAP blocks were performed. The LB group received 133 mg LB in 20 mL fluid per side, whereas the placebo group received 20 mL saline per side. Postoperative analgesia included scheduled oral acetaminophen (500 mg every 6 h for 48 h). Participants were permitted to self-administer PCIA boluses when the numeric rating scale (NRS) score of their abdominal pain, measured on a scale of 0–10, was >3.

A research assistant documented the 24- and 48-h oxycodone consumption, and interval to the first PCIA, and assessed resting pain, bed mobility pain, uterine palpation pain, and ambulation pain using the NRS at 6, 12, 24, and 48 h postoperatively; collected QoR-15 scores at 24 and 48 h; and recorded analgesia-related adverse effects. The analgesic pump was discontinued 48-h postoperatively, and the analgesic satisfaction score was assessed using a 5-point Likert scale.

The primary outcomes were oxycodone consumption within the first 24- and 48 h postoperatively. The secondary outcomes included interval to first PCIA of oxycodone; resting and bed mobility NRS scores at 6, 12, 24, and 48 h postoperatively; worst NRS scores in postoperative 12 h during uterine palpation; 24- and 48-h ambulation NRS pain scores; side effects of analgesia; 24- and 48-h QoR-15 scores; and analgesia satisfaction 48 h after surgery.

The required sample size was calculated using PASS 2022. Due to the lack of prior data on oxycodone consumption amongst patients receiving epidural hydromorphone in conjunction with intravenous oxycodone following caesarean sections, sample size estimation was based on our clinical experience using the incidence of moderate-to-severe pain (NRS > 3) during ambulation at 24 h, which was the trigger for oxycodone consumption in our trial. In our clinic, the incidence of moderate-to-severe pain (NRS > 3) was about 60% amongst parturients during ambulation at 24 h. Assuming a 30% absolute reduction (from 60% to 30%), the required sample size was calculated as 53 patients per group (α = 0.05, power = 90%). Allowing for a 20% attrition rate, the target sample size was set at 128 participants, with 64 participants per group.

The “survminer” R software package was used to generate Kaplan–Meier curves and the log-rank test was used to compare the probability of participants having PCIA between the LB and placebo groups. Violin Plots of NRS for resting and bed mobility at 6, 12, 24, and 48 h postoperatively were generated for each group using the “ggplot2” R software package. The differences and effect sizes of outcomes related to postoperative analgesia effect were calculated using the “effsize” R software package. These statistical analyses were performed using R 3.4.4 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was defined as a two-tailed p-value < 0.05.

Continuous data were presented as the mean ± SD or median and interquartile range (IQR), and groups were compared using t-tests or Mann–Whitney U tests, as appropriate. Categorical variables were presented as frequencies and percentages and groups were compared using chi-square or Fisher’s exact tests. Cliff’s delta was used as a nonparametric measure of effect size of LB compared with placebo, with values close to 0 signifying no effect and values close to −1 or +1 signifying a greater effect. These statistical analyses were performed using SPSS for Windows, version 26.0 (IBM Corp, Armonk, NY, USA). Statistical significance was defined as a two-tailed p-value < 0.05.

The CONSORT flow diagram (Figure 1) outlines participant enrolment and allocation. A total of 154 potential were identified, of whom 26 were excluded because of preterm delivery, BMI > 36 kg/m², allergy to acetaminophen, or refusal to participate. After these exclusions, a total of 128 participants were enrolled in the study. Three participants in the placebo group did not receive TAP block after surgery (two cases with postpartum haemorrhage exceeding 1,000 mL did not undergo TAP block due to concerns of coagulopathy potentially causing local hematoma. One case was withdrawn due to researcher forgetfulness) and one participant in the placebo group was unable to score her pain degree. These four participants were excluded from the analysis, leaving 64 participants in the LB group and 60 participants in the placebo group.

The baseline demographic and clinical characteristics were comparable between the LB and placebo groups, with no significant differences in age, anthropometric measures, obstetric history, neonatal birth weight, or intraoperative parameters (all p > 0.05). The mean gestational age was slightly higher in the LB group than in the placebo group (39.0 ± 1.3 vs. 38.4 ± 1.7 weeks; p = 0.044) but this difference was not clinically significant. The types of uterotonic agents administered did not differ significantly between groups (Table 1).

The median (IQR) cumulative oxycodone consumption during the first 24 h after surgery was significantly lower in the LB group than in the placebo group [2 (0–5) mg vs. 4 (1–8) mg, p = 0.009]. The median (IQR) cumulative oxycodone consumption during the first 48 h after surgery was also significantly lower in the LB group than in the placebo group [8 (0–13) mg vs. 10 (4–18) mg, p = 0.022]. However, the median (IQR) oxycodone consumption during the period from 24 h to 48 h after surgery did not differ significantly between groups (Table 2).

The Kaplan–Meier curves of the probability and numbers of participants without PCIA at each postoperative time-point are presented in Figure 2. The median (IQR) time to the first PCIA was significantly delayed in the LB group compared with the placebo group [22 (12–48) h vs. 8 (4–18) h, p < 0.001].

The resting and bed mobility NRS scores at 6, 12, 24, and 48 h postoperatively are presented in Figures 3, 4. The resting and bed mobility NRS scores in the LB group showed significantly better pain control than that in the placebo group at all time-points evaluated. The median (IQR) worst pain during uterine palpation at 12 h was significantly lower in the LB group than in the placebo group [NRS: 2 (2–3) vs. 4 (3–6); p < 0.001]. The ambulation pain scores at 24 h [4 (2.25–5) vs. 5 (4–6), p < 0.001] and 48 h [2 (2–3) vs. 3.5 (3–5), p < 0.001] postoperatively were also significantly lower in the LB group than in the placebo group (Table 2).

The QoR-15 scores revealed a 5-point greater median (IQR) recovery quality in the LB group compared with the placebo group at both 24 h [138 (133–143) vs. 133 (122–138), p = 0.002] and 48 h [143 (138–146) vs. 137 (130–142), p < 0.001] postoperatively. The median (IQR) analgesia satisfaction score at 48 h postoperatively was significantly higher in the LB group than in the placebo group [5 (5–5) vs. 4 (4–5), p < 0.001] (Table 2).

The adverse event profiles were comparable between the LB and placebo groups, with no statistically significant differences in the incidence of dizziness, nausea, vomiting, degree of pruritus, or urinary retention between the LB and placebo groups (Table 3).

A strength of this study is the comprehensive evaluation of the analgesic effect of LB-enhanced TAP block using multiple endpoints: opioid rescue doses; interval to first PCIA demand; resting, bed mobility, ambulation and uterine palpation NRS pain scores; QoR-15; satisfaction score, and the use of Cliff’s delta to measure the effect size. These measures demonstrated the analgesic effect of LB-enhanced TAP block.

This study also has some limitations. First, since potential differences in practice settings (drug availability, analgesia protocols) that could influence outcomes, the single-centre design and modest sample size of the current study may limit generalizability, and control group with plain bupivacaine would have been valuable to evaluate the efficacy of the liposomal formulation. Further researches could be carried out to compare the LB and plain bupivacaine combined with epidural hydromorphone with multicentre or large cohort studies. Second, all participants fast from midnight according to the protocol of the study, however ERAS/ERAC recommendations clear fluids up to 2 h preoperatively. Third, the high cost of LB is a barrier to its widespread adoption, and further cost-effectiveness analyses are required. Fortunately, the price of liposomal bupivacaine in China has significantly dropped, and it now costs less than 54 USD. Finally, the duration of action of LB exceeds 72 h, a substantial proportion of women undergoing caesarean delivery are discharged prior to 72 h postoperatively. Consequently, we were unable to assess the analgesic efficacy of LB-enhance TAP block beyond 48 h. In future studies, postpartum pain situations can be followed up with telephone or online questionnaires for longer periods.