Analgesic Efficacy of Continuous Bupivacaine Infusion versus Liposomal Bupivacaine in Parasacral Ischial Plane Block After Total Knee Arthroplasty: An Open-Label, Randomized, Controlled, Non-Inferiority Study Authors: Xuan Pan, Peng Ye, Yawen Xie, Yingjie Chen, Hanliang Fan, Yongxin Huang, Chunying Zheng, Xiaochun Zheng, Ting Zheng Published: 2026 ABSTRACT AbstractBackgroundEffective management of posterior knee pain following total knee arthroplasty (TKA) remains a clinical challenge. Comparative analgesic efficacy of continuous parasacral ischial plane block (CPIPB) and liposomal bupivacaine (LB) is unclear. This study aimed to evaluate whether the analgesic efficacy of a single-shot PIPB using LB is non-inferior to CPIPB in patients undergoing TKA.MethodsThis open-label, randomized, controlled non-inferiority trial enrolled patients scheduled for unilateral TKA who were randomly allocated to either the CPIPB or LB group. Patient-controlled intravenous analgesia (PCA) with sufentanil was provided as rescue analgesia. The primary outcome was postoperative analgesic efficacy, which was indicated by the cumulative area under the curve (AUC) of the Numerical Rating Scale (NRS) scores obtained at rest and activity until 72 h postoperatively after leaving the post-anaesthesia care unit (PACU). A pre-specified non-inferiority margin of 20.7 was established. Other outcomes included NRS scores, rescue opioid consumption, rehabilitation parameters, PIPB outcomes and postoperative adverse events.ResultsA total of 100 patients were enrolled and randomly assigned to the CPIPB or LB group. Both per-protocol (n=90) and intention-to-treat (n=100) analyses confirmed non-inferiority, the upper limits of the 95% CI for the differences in cumulative NRS-AUCPACU-72h at rest (PP 18.03, ITT 16.82) and activity (PP 19.82, ITT 18.42) were strictly below the 20.7 margin. NRS differences were found at rest 60h and 72h, or at activity 72h (P<0.05). CPIPB was associated with lower cumulative rescue opioid consumption, a longer time to first rescue analgesia, and more favorable early rehabilitation parameters (including range of motion, 2-minute walking distance, and Western Ontario and McMaster Universities Osteoarthritis Index scores). The catheter-related complications of CPIPB had catheterization failure 4.1%, insertion site leakage 2.2%, catheter dislodgement 4.3% and catheter clogging 6.7%. The CPIPB group had lower rates of postoperative nausea, insomnia.ConclusionsA single-shot PIPB with LB met the prespecified non-inferiority criterion for pain relief compared with CPIPB after TKA, while CPIPB performs better in opioid consumption and early rehabilitation parameters.Keywords: total knee arthroplasty, liposomal bupivacaine, continuous parasacral ischial plane block, postoperative pain, non-inferiority trial Background Total knee arthroplasty (TKA) is the definitive and gold-standard treatment for end-stage degenerative knee osteoarthritis, effectively alleviating symptoms and restoring joint function.1 However, postoperative pain management remains a critical challenge. In particular, severe pain in the posterior aspect of the knee joint—which is predominantly innervated by the sciatic nerve and its branches—is frequently overlooked in conventional analgesic regimens.2–4 To address this specific challenge, continuous peripheral nerve blocks have been widely adopted. The parasacral ischial plane block (PIPB) is an advanced method for sciatic nerve block in the parasacral area. It involves injecting a local anesthetic between the fascia over piriformis and presacral fascia, acting as a fascial plane block to effectively manage posterior knee pain without directly engaging the nerve.5–7 Meanwhile, our previous studies have demonstrated that continuous PIPB (CPIPB) can reliably block the sciatic nerve, providing robust, prolonged analgesia with a favorable safety profile for patients undergoing TKA.8,9 While this continuous infusion modality has been proven to reduce pain scores and opioid consumption, thereby facilitating early postoperative recovery,10 it inherently relies on indwelling catheters. This reliance on catheter introduces specific risks and clinical burdens. Catheter dislodgement (3.7%), site leakage (7.4%), clogging (3.7%) occurred in previous study.9 Liposomal bupivacaine (LB) is an advanced, liposome-encapsulated formulation of local anesthetic designed to slowly release bupivacaine, extending its duration of action for up to 72 hours and thereby reducing reliance on adjunctive opioids.11 Emerging evidence confirms that a single-shot PIPB using LB can provide prolonged postoperative pain relief and circumvent the inherent drawbacks of catheter-related complications associated with CPIPB.7 Nevertheless, owing to the specific pharmacokinetic release profile of liposomes, the time-dependent fluctuation in local anesthetic concentration raises concerns that single-shot LB might yield a less consistent analgesic effect during the later postoperative period compared to the steady-state delivery of continuous infusions.12,13 Notably, a single administration of LB through intra-articular injection after TKA, has demonstrated similar efficacy in providing extended postoperative pain relief comparable to that achieved with continuous femoral nerve blocks and continuous intra-articular infusion.14,15 Consequently, uncertainty remained regarding whether a single-shot LB-based PIPB could preserve overall 72-h analgesic efficacy compared with CPIPB while avoiding the burden of an indwelling catheter. Therefore, we conducted this clinical trial. We hypothesized that the overall 72-hour postoperative analgesic efficacy of a single-shot PIPB with LB is non-inferior to that of a continuous bupivacaine infusion via CPIPB. Materials and Methods Study Design This open-label, randomized, controlled, non-inferiority trial was approved by the institutional review board of Fuzhou University Affiliated Provincial Hospital (NO. 2025–005-01; January 6, 2025) and conducted in compliance with the Declaration of Helsinki. This study followed the Consolidated Standards of Reporting Trials guidelines. Written informed consent was obtained from all patients. The trial was prospectively registered in the Chinese Clinical Trial Registry (https://www.chictr.org.cn/, No. ChiCTR2500103620) on June 3, 2025. Written informed consent was obtained from all participants before enrolment. Patient recruitment commenced on 5 June 2025. Participants We enrolled patients aged 18–80 years with ASA I–II who were scheduled for unilateral TKA for knee osteoarthritis. Further inclusion criteria were the ability to walk independently and normal sensory function in the distribution of the sciatic nerve. Exclusion criteria were patient refusal, known allergy to any study medication, long-term opioid use (defined as daily use for >3 months), pre-existing sciatic neuropathy, infection at the needle insertion site, or coagulopathy. Randomization and Masking Patients were randomly assigned to the CPIPB group or the LB group in a 1:1 ratio. Randomization was performed by third-party professional medical staff by generating a random number for each patient, which was then entered into a computer interactive response system to obtain the patient’s randomized group allocation. Similar-looking drug packages were prepared by third-party staff for both groups. Additionally, using a sham catheter might have subjected patients to unnecessary risks, so an open-label design was chosen.16,17 To minimize detection bias, all postoperative clinical assessments and data collection were performed by independent third-party staff who were blinded to the group allocation. The drug packages were prepared and distributed by third-party staff as follows. (1) The CPIPB drug package (Figure 1A and B): 20 mL of 0.5% bupivacaine was prepared for the first preoperative injection of CPIPB. Six bottles (72 mL each) of 0.125% bupivacaine were used in postoperative injections of CPIPB (6 mL/h). (2) The LB drug package (D): 20 mL of LB was prepared for the first preoperative injection of PIPB. Trial intervention diagram. (A) CPIPB group’s preoperative intervention. (B) CPIPB group’s postoperative intervention. (C) LB group’s preoperative intervention. (D) LB group’s postoperative intervention. Abbreviations: CPIPB, Continuous Parasacral Ischial Plane Block; LB, liposomal bupivacaine. Procedures All patients fasted for 8 hours preoperatively. The PIPB procedure was performed in the anesthesia preparation room under strict aseptic conditions. Patients were placed in a lateral decubitus position on the non-operative side. Real-time ultrasound guidance (2–5 MHz probe) was utilized by a senior anesthesiologist to identify the deep side of the greater sciatic foramen. The CPIPB procedure was consistent with our previously established methods.9 A 1.2×120 mm catheter (Baiyue Medical Technologies, AN-NI, China) was advanced between the piriformis muscle and the presacral fascia. An initial loading dose of 20 mL of 0.5% bupivacaine (Harvest Pharmaceutical Co., Ltd, Shanghai, China) was administered.18 Postoperatively, a continuous infusion of 0.125% bupivacaine was delivered via the catheter at a constant rate of 6 mL/h.19,20 For the LB group, a single-shot block was performed using a 100 mm needle (USG TYPE CCR, 22G, Hakko Co., Ltd, Osaka, Japan).7 Patients received 266 mg of liposomal bupivacaine (Jiangsu Hengrui Medicine Co., Ltd, Jiangsu, China) a total volume of 20 mL.7 The Numerical Rating Scale (NRS, scores: 0–10; a higher score indicates more pain) was used by the anesthesiologist to evaluate pain for puncture point 30 min before and after PIPB. Any adverse events or complaints were recorded by the anesthesiologist who performed the PIPB procedure. General anaesthesia was performed by another experienced anesthesiologist and administered by total intravenous technique using propofol, sufentanil (0.5 ug/kg), remifentanil, and cis-atricurium. To prevent postoperative nausea and vomiting, 5 mg tropisetron was intravenously administered 30 min before the surgery ended. According to the multimodal analgesia protocol, 1 g of acetaminophen and 400 mg of celecoxib were administered on the morning of surgery. All surgeries were performed by the same experienced orthopedic surgical team. Prior to the implantation of tibial and femoral prostheses, local infiltration anesthesia (300 mg ropivacaine, 0.3 mL epinephrine 1:1000, 10 mg morphine, and isotonic sodium chloride solution to 39.7 mL) was injected into the suprapatellar pouch, meniscus, and fat pad, excluding the posterior knee capsule. In the CPIPB group, postoperative injection of the given drugs was administered via fixed catheters until 72h (6mL/h, 0.125% bupivacaine) (Figure 1B).19,20 Immediately upon arrival in the post-anaesthesia care unit (PACU) after surgery, Patients received the rescue intravenous analgesia pump (2 μg/kg sufentanil and 20 mg of tropisetron diluted to 100 mL with 0.9% NS) without a background dose. The pump was administered when requested by the patient with the Numerical Rating Scale (NRS) scores ≥ 4, with a single dose of 2 mL bolus (sufentanil 0.04 μg/kg and tropisetron 0.4 mg) given through patient-controlled analgesia with a lockout time of 15 min (Figure 1B and D). Patients were monitored for consciousness, respiratory function, oxygen levels, and opioid side effects while using PCA, which was paused or stopped if significant sedation, respiratory problems, or serious side effects occurred.21 To ensure patient safety, a standardized safety monitoring protocol was pre-established and strictly implemented. All patients were monitored twice daily by ward nurses and the acute pain service for signs of local anesthetic systemic toxicity (LAST), including neurological symptoms (metallic taste, tinnitus, perioral numbness, confusion) and cardiovascular fluctuations.22 For catheter-related complications, standardized operational procedures were followed: puncture-site bleeding was managed by immediately replacing the sterile compression dressing; catheter clogging was addressed by flushing the catheter with 2–3 mL of sterile normal saline to restore patency; in cases of catheterization failure or catheter dislodgement, the continuous infusion was discontinued immediately, the catheter was removed, and pain management was transitioned to our standard oral and intravenous multimodal analgesic regimen (NSAIDs and acetaminophen). Postoperatively, patients received routine analgesia; 1 g acetaminophen every 6 h and 200 mg celecoxib every 12 h. The CPIPB catheters and all intravenous pumps were removed 72 h postoperatively. All patients received encouragement for early mobilization according to the ERAS guideline. Outcomes The primary outcome was to compare the overall postoperative 72-h analgesic efficacy as indicated by the cumulative area under the curve (AUC) of the pain rating according to the NRS scores and postoperative time; the co-primary outcome comprised cumulative NRS-AUCPACU–72h at rest and activity. The time points were set at PACU (0 h), 12 h, 24 h, 36h, 48h, 60h and 72h after leaving the PACU. The formula for AUC calculation was consistent with our previously established methods.7 NRS0h was the NRS at PACU, and NRS was scored on a scale of 0–10, with higher scores indicating more severe pain. The postoperative NRS score [PACU (0 h), 12 h, 24 h, 36h, 48h, 60h and 72h] at rest and at activity (active maximum knee flexion) was calculated and collected by third-party staff. Noninferiority margin was satisfied for both cumulative NRS-AUCPACU-72h at rest and at activity. PIPB onset was observed using the pin-prick test or 5 °C ice pack test (0: normal sensation, 1: delayed sensation, 2: sensory loss) on the metatarsal, dorsal, and lateral sides of the calf. Measurements were taken every 5 min after PIPB, and PIPB onset was considered present if the score was ≥ 1 at any site. The postoperative rescue opioid consumption every 24 h and the time of first rescue analgesic administration were also recorded. The rescue opioid consumption was converted to oral morphine equivalents (OME). Patient satisfaction with analgesia was assessed on the postoperative day (POD) 4 using a 5-point Likert scale (ranging from “very dissatisfied” to “very satisfied”). The Hospital Anxiety and Depression Scale (HADS) is a self-rating scale created in 1983 by A S Zigmond and R P Snaith to assess the emotional states of hospitalized patients. It consists of subscales for anxiety (HADS-A) and depression (HADS-D) (Scores: 0–21; a higher score indicates worse).23 Rehabilitation quality evaluation included the range of motion (ROM), two-minute and the Western Ontario and McMaster Universities OA Index (WOMAC). The patient was instructed to fully flex the knee, and an independent anesthesiologist measured the active ROM with a universal goniometer on POD4. The two-minute walking test was conducted at discharge. The WOMAC (scores: 0–96; a higher score indicates more worse) assessed on pre-operation and Week 2 after TKA, which is a valid and dependable disease-specific measure of knee pain, rigidity, and physical function.23 The length of hospital stay (LOS) was defined as the interval from admission to discharge. Postoperative adverse events, including nausea, vomiting, constipation, local anesthetics toxicity etc., were recorded. Discharge criteria comprised independent walking `capability, NRS<4, and absence of orthopedic complications. Statistical Analysis PASS 11.0 (NCSS, LLC, Utah, United States) was used to estimate the sample size based on our previous attempts in OMH and clinical considerations. The cumulative NRS-AUCPACU-72h at rest of our previous CPIPB procedures (mean 103.5, SD 13.3) and LB procedures (mean 117, SD 11.49), and the cumulative NRS-AUCPACU-72h at activity of the CPIPB procedures (mean 452.3, SD 12.09) and LB procedures (mean 461.3, SD 11.84) were calculated. Based on that, we set the threshold for noninferiority of the LB group for the primary outcome no higher than 20.7 from the CPIPB group in the mean cumulative NRS-AUCPACU-72h, which is within the range of 0–20% both at rest and at activity. This margin corresponds to an average difference of less than 0.5 points on the 0–10 NRS scale per assessment point over the 72-hour period,24 which is substantially smaller than the minimal clinically important difference of approximately 1–2 points reported for TKA.25–28 The choice of 0–20% threshold is also consistent with previous non‑inferiority trials in postoperative pain management and represents a clinically acceptable difference in analgesic efficacy.24,29,30 To control the overall type-I error rate for the co-primary outcomes (cumulative NRS-AUC at rest and during activity) without α inflation, the statistical analysis followed a conjunctive “AND” rule within an intersection-union testing framework. Under this framework, the overall null hypothesis was rejected, and non-inferiority was declared, only if both co-primary outcomes met their respective non-inferiority criteria. A total of 76 patients were calculated (1:1 ratio) to be needed (38 each group) for the primary analysis, with a significance level one side α=0.05 and a power of (1-β)=0.80. According to the larger sample size required for the two primary outcomes calculated respectively, and to allowing for patient drop out or inadequate data collection we added a 15% contingency factor and chose to enroll 45 patients in each group or 90 patients in total at least. A per-protocol (PP) analysis was conducted among participants who successfully completed 3 days of treatment without major protocol violation. All randomized participants were incorporated into an intention-to-treat (ITT) analysis. Microsoft Excel 2019 was used to establish the database, and SPSS software (version 25.0) was used for data analysis. The Shapiro–Wilk test for normality and Levene’s test for homogeneity of variance were conducted for continuous variables. Data meeting these criteria are expressed as mean (SD) and compared using two-sample t-tests. Data not meeting the criteria are expressed as median (interquartile range [IQR]) and compared using the Mann–Whitney U-test. Categorical variables were presented as frequencies (percentages) and compared using Pearson’s chi-square test or Fisher’s exact test. Pain scores at various time points were analyzed using a Generalized Estimating Equations. A Bonferroni correction was applied to adjust for multiple comparisons. Kaplan–Meier survival analysis described PIPB onset time in both groups, with comparisons conducted using the Log rank test. Statistical significance was set at P<0.05. Results Between June 5, 2025, and December 31, 2025, 100 patients were enrolled and randomly assigned to the CPIPB group (n=50) and the LB group (n=50). A total of 121 patients were screened for eligibility, 21 were excluded (15 did not meet the inclusion criteria and 6 declined to participate). The remaining 100 patients were randomly assigned in a 1:1 ratio to either the CPIPB group (n=50) or the LB group (n=50). All 100 patients were included in the Intention-To-Treat (ITT) analysis. During the study, 10 patients (5 in each group) were excluded due to discontinued interventions or protocol deviations, leaving 90 patients (45 per group) for the Per-Protocol (PP) analysis. The detailed Consolidated Standards of Reporting Trials (CONSORT) flow diagram is presented in Figure 2. Consolidated Standards of Reporting Trials (CONSORT) diagram of patient recruitment. Baseline demographics and clinical characteristics—including age, sex, body mass index (BMI), ASA physical status, surgical side, duration of surgery, tourniquet time, and preoperative HADS and WOMAC scores—were well-balanced between the two groups, with no significant differences observed (Table 1). Table 1CharacteristicVariablesGroup CPIPB (n=50)Group LB (n=50)PAge(year)70 (67, 74)72 (66, 76)0.346Sex, n (%)0.476Male13 (26)10 (20)Female37 (74)40 (80)Weight (kg)63.8 (11.3)64.3 (10.0)0.837Height (cm)160 (155, 166)158 (155, 163)0.225BMI (kg m-2)24.7 (4.1)25.4 (3.4)0.313ASA classification, n (%)>0.999I4(8)4(8)II46 (92)46 (92)Operative sideLeft/Right24/2626/240.689Duration of surgery (min)120 (25)124 (25)0.415Tourniquet Time66.5 (4.1)64.8 (5.4)0.096Cardiovascular comorbidities, No. (%)7(14)9(18)0.581Pulmonary comorbidities, No. (%)5(10)3(6)0.711Previous TKA,No. (%)2(4)1(2)>0.999HADSHADS-A10 (9,12)10 (9,11)0.640HADS-D7 (6,8)7 (6,8)0.590WOMAC65.3 (6.4)63.1 (6.1)0.078Notes: Values are number (proportion), mean (SD) or median (IQR).Abbreviations: ASA, American Society of Anesthesiologists. HADS, Hospital Anxiety and Depression Scale. HADS-A, HADS for anxiety. HADS-D, HADS for depression. WOMAC, the Western Ontario and McMaster Universities OA Index. Characteristic Notes: Values are number (proportion), mean (SD) or median (IQR). Abbreviations: ASA, American Society of Anesthesiologists. HADS, Hospital Anxiety and Depression Scale. HADS-A, HADS for anxiety. HADS-D, HADS for depression. WOMAC, the Western Ontario and McMaster Universities OA Index. Block Characteristics and Procedural Events The median procedural time was shorter in the LB group compared to the CPIPB group (8 min vs 26 min, P<0.001). Conversely, the onset time of the PIPB was longer in the LB group (30 min vs 15 min, P<0.001). Preoperative NRS scores in puncture point evaluated 30 minutes before and after the PIPB procedure showed no significant differences between the two groups (P>0.05) (Table 2). Table 2PIPB OutcomesVariablesGroup CPIPB (n=45)Group LB (n=45)PPIPB procedural time, min26 (24, 28)8 (7, 9)<0.001*PIPB onset time, min15 (12.5, 20)30 (30, 35)<0.001*NRS in Puncture pointNRS before block0(0)0(0)>0.999NRS after block1(1,2)1(0,2)0.134PIPB related complications, n (%)Suspected allergy0(0)2(4.4)0.475Puncture site haemorrhage45 (100)37 (82)0.010*Puncture site swelling5(11.1)1(2.2)0.205Rash0(0)3(6.7)0.240Cardiopalmus0(0)0(0)>0.999Dyspnea0(0)0(0)>0.999Shock0(0)0(0)>0.999Sciatic nerve injury0(0)0(0)>0.999Vascular puncture0(0)0(0)>0.999Death0(0)0(0)>0.999Catheter status during 0–72hFailure rate of catheterization, n (%)2(4.1) #––Insertion site leakage, n (%)1(2.2)––Catheter dislodgement, n (%)2(4.3) &––Catheter clogging, n (%)3(6.7)––Depth of catheterization, cm13.4 (1.0)––Notes: #, 49 patients underwent catheter insertion. &, The catheter was successfully inserted in 47 patients. Values are number (proportion), mean (SD) or median (IQR). *, Denotes statistical significance (P<0.05).Abbreviations: NRS, Numerical Rating Scale. PIPB, Parasacral Ischial Plane Block. PIPB Outcomes Notes: #, 49 patients underwent catheter insertion. &, The catheter was successfully inserted in 47 patients. Values are number (proportion), mean (SD) or median (IQR). *, Denotes statistical significance (P<0.05). Abbreviations: NRS, Numerical Rating Scale. PIPB, Parasacral Ischial Plane Block. The incidence of local site hemorrhage was higher in the CPIPB group (100% vs 82%, P=0.010). Catheter-related events were recorded across the 72-hour postoperative period in the CPIPB group, including catheterization failure rate of 4.1% (n=2), catheter dislodgement in 4.3% (n=2), catheter clogging in 6.7% (n=3), and insertion site leakage in 2.2% (n=1). No treatment-related deaths or severe systemic complications were observed in either group (Table 2). Primary Outcome In the PP analysis, the cumulative NRS-AUCPACU-72h at rest were 88.93 (14.60) in the CPIPB group and 100.93 (14.20) in the LB group, with a mean difference of 12.00 (95% CI, 5.97 to 18.03). During activity, the corresponding values were 466.00 (15.82) and 478.13 (20.58), respectively, with a mean difference of 12.13 (95% CI, 4.44 to 19.82). In the ITT analysis, the cumulative NRS-AUCPACU-72h were 89.40 (16.45) in the CPIPB group and 100.20 (13.78) in the LB group, with a mean difference of 10.80 (95% CI, 4.78 to 16.82). During activity, the corresponding values were 467.76 (16.08) and 478.92 (20.27), with a mean difference of 11.16 (95% CI, 3.90 to 18.42). Both the bound of the 95% CI upper limits for difference at rest and at activity both PP and ITT analyses below the prespecified noninferiority margin (20.7), indicating noninferiority of the postoperative analgesic efficacy in the LB group (Figure 3). The mean difference of the overall postoperative analgesic efficacy as indicated by the cumulative AUC of the NRS scores during 0–72 h between the CPIPB group and LB group. Abbreviations: CPIPB, Continuous Parasacral Ischial Plane Block; LB, liposomal bupivacaine; AUC, Area Under the Curve; NRS, Numerical Rating Scale; PP, per-protocol analysis; ITT, intention-to-treat analysis. Details of NRS at Each Time Point No significant differences of specific NRS scores at each time points in rest including the 0 h, 12 h, 24 h, 36 h, 48 h and at activity including the 0 h, 12 h, 24 h, 36 h, 48 h, 60h between two groups were observed respectively (P>0.05). Differences were found in NRS at rest 60h [0 vs 1, 95% CI: 0.33 (0.00, 0.66), P=0.043] and 72h [0 vs 0, 95% CI: 0.35 (0.05, 0.66), P=0.005], and at activity 72h [5 vs 5, 95% CI: 0.76 (0.19, 1.3), P=0.001] (Table 3, Figures S1 and S2). Table 3Pain Scores of PatientsNRS For Each TimeGroup CPIPB (n=50)Group LB (n=50)Difference (95% CI)PNRS scores at RestPACU0(0, 0)0(0, 0)0.0 (0.0, 0.0)>0.99912h0(0, 1)0(0, 1)0.02 (−0.38, 0.43)>0.99924h3(2, 4)3(3, 3)0.22 (−0.33, 0.77)>0.99936h2(2, 3)2(2, 3)0.11 (−0.32, 0.54)>0.99948h1(1, 2)2(1, 2)0.13 (−0.38, 0.64)>0.99960h0(0, 0)1(0, 1)0.33 (0.00, 0.66)0.043*72h0(0, 0)0(0, 1)0.35 (0.05, 0.66)0.005*NRS scores at ActivityPACU0(0, 0)0(0, 0)0.02 (−0.11, 0.15)>0.99912h5(5, 6)6(5, 6)0.38 (−0.41, 1.17)>0.99924h9(8, 9)9(8, 9)0.02 (−0.56, 0.61)>0.99936h9(8, 9)9(8, 9)0.02 (−0.37, 0.41)>0.99948h8(7.5, 8)8(8, 9)0.20 (−0.30, 0.70)>0.99960h6(6, 7)6(6, 7)0.00 (−0.51, 0.51)>0.99972h5(4, 5)5(5, 6)0.76 (0.19, 1.3)0.001*Notes: Values are median (IQR). *Denotes statistical significance (P<0.05).Abbreviation: NRS, Numerical Rating Scale. Pain Scores of Patients Notes: Values are median (IQR). *Denotes statistical significance (P<0.05). Abbreviation: NRS, Numerical Rating Scale. Secondary Outcomes The rescue opioid consumption (in mg of oral morphine equivalents) was lower in the CPIPB group vs LB group at 0–24 h [22 (15, 29) vs 26 (22, 31), P=0.021], at 24–48 h [40(34, 54) vs 51(43, 59), P=0.005], at 48–72h [4 (0, 6) vs 10 (5, 15), P<0.001] and 0–72h [70 (57, 84) vs 89 (77, 101), P<0.001]; moreover, the CPIPB group had six patients did not use rescue opioids. Within the first postoperative 72 h, the first time of rescue analgesia was longer in the CPIPB group (17.5 hours) than in the LB group (15.2 hours). Analgesia satisfaction scores were 5 (IQR, 5–5) in both groups (P=0.091) (Table 4). Table 4Secondary OutcomesVariablesGroup CPIPB (n=45)Group LB (n=45)POME, mg0-24h22 (15, 29)26 (22, 31)0.021*24-48h40 (34, 54)51 (43, 59)0.005*48-72h4 (0, 6)10 (5, 15)<0.001*0-72h70 (57, 84)89 (77, 101)<0.001*First time of rescue analgesia, hours17.5 (2.9) #15.2 (2.7)<0.001*Analgesia satisfaction score5 (5, 5)5 (5, 5)0.091LOS, days9 (7, 10)8.5 (7, 12)0.865Rehabilitation parametersROM, angle100.0 (10.8)93.4 (8.8)0.002*2 min walking test, meter39.3 (3.5)34.8 (3.4)<0.001*WOMAC50.4 (5.3)54.1 (6.4)0.003*Notes: # Six patients did not receive rescue analgesia, n=39. Values are mean (SD) or median (IQR). *, Denotes statistical significance (P<0.05).Abbreviations: OME, oral morphine equivalents; LOS, length of hospital stay; ROM, range of motion; WOMAC, the Western Ontario and McMaster Universities OA Index. Secondary Outcomes Notes: # Six patients did not receive rescue analgesia, n=39. Values are mean (SD) or median (IQR). *, Denotes statistical significance (P<0.05). Abbreviations: OME, oral morphine equivalents; LOS, length of hospital stay; ROM, range of motion; WOMAC, the Western Ontario and McMaster Universities OA Index. In rehabilitation parameters, the CPIPB group had better ROM (100.0 vs 93.4 angle, P=0.002), longer 2 min walking distance (39.3 vs 34.8 meters, P<0.001), lower WOMAC scores (50.4 vs 54.1, P=0.003), but there was no significant difference in LOS (P=0.865) (Table 4). Postoperative Adverse Events The CPIPB group had lower rates of nausea (6.7% vs 24.4%, P=0.042), insomnia (2.2% vs 20.0%, P=0.019) and higher rates of bleeding at the puncture site (15.5% vs 0%, P=0.018); however, there was no difference in vomiting, dizziness, local anesthetics toxicity, falling, pruritus, infection at the puncture site, infection in surgical incision, vein thrombosis, constipation, and poor appetite between the CPIPB group and LB group (P>0.05) (Table 5). Table 5Postoperative Adverse Events. Values are Number (Proportion)VariablesGroup CPIPB (n=45)Group LB (n=45)PNausea3 (6.7)11 (24.4)0.042*Vomiting2 (4.4)5 (11.1)0.431Dizziness3 (6.7)6 (13.3)0.482Local anesthetics toxicity0 (0)0 (0)>0.999Falling0 (0)0 (0)>0.999Pruritus2 (4.4)6 (13.3)0.266Bleeding at the puncture site7 (15.5)0 (0)0.018*Infection at the puncture site0 (0)0 (0)>0.999Infection in surgical incision0 (0)0 (0)>0.999Sciatic nerve injury0 (0)0 (0)>0.999Vein thrombosis0 (0)0 (0)>0.999Constipation3 (6.7)6 (13.3)0.482Insomnia1 (2.2)9 (20)0.019*Poor appetite3 (6.7)8 (17.7)0.108Notes: Values are number (proportion). *, Denotes statistical significance (P<0.05). Postoperative Adverse Events. Values are Number (Proportion) Notes: Values are number (proportion). *, Denotes statistical significance (P<0.05). Discussions Continuous peripheral nerve blocks (cPNBs) and long-acting local anesthetics are recognized as reliable, long-term analgesic strategies following TKA.10,19,31,32 While the management of anterior knee pain is well-established, severe posterior knee pain post-TKA—which is closely related to the sciatic nerve and its branches—remains a significant clinical challenge.33 Our previous studies have confirmed that both continuous parasacral ischial plane block (CPIPB) and liposomal bupivacaine (LB) in PIPB provide substantial analgesic benefits. Although statistical non-inferiority was demonstrated for our primary endpoints, the observed clinical trends—including slightly higher point-in-time pain scores, increased cumulative rescue opioid requirements, and less optimal early rehabilitation metrics in the LB group—suggest a subtle clinical inferiority of single-shot LB compared to the continuous, steady-state delivery of bupivacaine via CPIPB. This marginal disadvantage, however, is counterbalanced by the clinical convenience of a catheter-free approach. Single-shot LB may therefore find its strongest clinical indications in specific care pathways. For instance, in outpatient or day-surgery TKA protocols, home management of indwelling catheters is logistically challenging and carries safety risks. Additionally, institutions lacking a dedicated, 24-hour acute pain service to monitor and troubleshoot catheters, or patients with poor compliance or cognitive barriers who are at risk of accidental catheter dislodgement, would benefit significantly from the simplified, catheter-free workflow that LB provides. However, both techniques possess inherent limitations.7,9 Previous proof-of-concept study assessed the technical feasibility, catheter stability, preliminary analgesic efficacy, and safety of CPIPB in patients after TKA, thereby establishing a foundation for utilizing CPIPB as the control in the current investigation.9 To our knowledge, this is the first study to report that LB is noninferior to CPIPB in the postoperative pain management for TKA. The study demonstrated that in both the PP and ITT cohort analyses of NRS-AUCPACU-72h, whether at rest or during activity, the 95% confidence interval for the difference between the LB group and the CPIPB group did not surpass the non-inferiority threshold. Consequently, it can be inferred that within this 20% margin, the analgesic efficacy of LB is not inferior to that of CPIPB. Furthermore, this conclusion is corroborated by the multi-point NRS scores. No statistically significant differences were observed in NRS scores from 0 to 48 hours during rest and from 0 to 60 hours during activity. Nevertheless, the point NRS (60, 72h) indicated higher pain in the LB group. CPIPB was associated with lower rescue opioid consumption, longer time to first rescue analgesia, and better early rehabilitation outcomes, whereas LB required a shorter procedural time and provided a catheter-free analgesic strategy. In summary, PIPB with LB satisfied the predetermined non-inferiority criterion for pain relief in comparison to CPIPB; however, this result should not be construed as evidence of equivalence or superiority. Analgesic strategies should be individualized according to patient characteristics, the duration of analgesia required, catheter-related considerations, and available institutional resources. Several previous studies have reached conclusions similar to ours regarding the efficacy of LB non-inferior to continuous nerve blocks. The meta-analysis performed by Liu et al suggests that both single-injection LB and cPNBs exhibit similar efficacy in terms of pain relief and safety for patients undergoing TKA when administered via femoral nerve block (FNB) or intra-articular injections.14 A retrospective cohort study by Lonza GC19 found no statistical difference in pain scores on postoperative days (POD) 1–3 when comparing LB and continuous adductor canal block (cACB). Furthermore, Chen CM31 compared the analgesic effects of cACB and LB, concluding that the analgesic efficacy within the first 30 hours of a single LB injection is consistent with that of a continuous block. These findings are similar to our results, where no significant statistical differences were observed in single-timepoint resting and active pain scores within the first 48 hours. The MCID is commonly considered when selecting a non-inferiority margin. YaDeau et al used an MCID of 1 point in a non-inferiority study of ACB combined with iPACK,27 whereas Hinman et al used an MCID of 2 points in an ACB study.28 However, all of these were non-inferiority studies based on NRS pain scores measured at specific time points. In contrast, our study assessed the overall analgesic efficacy during the first 72 postoperative hours using the cumulative NRS. The non-inferiority margin of 20.7 represented 20.0% (0–20%) of the NRS-AUCPACU-72h at rest and 4.57% (0–10%) of the NRS-AUCPACU-72h at activity based on the findings from pilot study.29 When converted to a 72-hour time-weighted mean pain-score difference, this margin corresponds to 0.29 (20.7/72) NRS points, which is smaller than the commonly reported single-time-point MCID of NRS.25–28 Consequently, this margin establishes a rigorous criterion for assessing non-inferiority but the clinical acceptability requires further validation. The inherent advantage of continuous blocks lies in this constant drug concentration and the flexibility to adjust the infusion rate and catheter placement time, which caters to the prolonged analgesic demands post-TKA. However, these advantages depend on the safety and stability of catheter placement. Prior research has reported catheterization periods of up to two weeks, thereby offering substantial support for the three-day catheterization protocol employed in the current study.8 This advantage translated into superior analgesic efficacy and rehabilitation parameters in the CPIPB group. It is worth noting that statistical differences were noted in NRS scores at 60- and 72-hours during rest and at 72 hours during activity, the differences (0.33 and 0.35 for rest, and 0.77 for activity) did not exceed the MCID for TKA pain scores, which are 1.5 in the resting state and 1.8 in the active state.25,26 The MCID in morphine milligram equivalent (MME) is defined as an absolute reduction of 10 milligrams of intravenous morphine administered over a 24-hour period. Given the conversion ratio between OME and intravenous morphine is 3:1, the observed differences in OME consumption at 0–24 hours, 24–48 hours, and 48–72 hours post-surgery, correspond to reductions of 1.3 mg, 3.7 mg, and 2 mg in intravenous morphine consumption, respectively.25,34 Notably, the analgesic difference observed primarily emerged between 60 and 72 hours postoperatively. This late-phase difference is likely attributable to the pharmacokinetic release profile of LB, which may result in an insufficient local concentration of bupivacaine as the 72-hour mark approaches, which translate into a lesser duration of consistently sustained sensory block.35–38 In contrast, the CPIPB group received a continuous infusion of 0.125% bupivacaine at a constant rate of 6 mL/h, maintaining steady local anesthetic concentrations and a consistent dosage over time. Therefore, A single-injection interfascial LB block provides modest and clinically limited prolongation of analgesia, which is insufficient to enhance rehabilitation parameters.39,40 CPIPB is hindered by catheter-related challenges. In our study, the catheter failure rate was 4.1%, accompanied by insertion site leakage (2.2%), catheter dislodgement (4.3%), and catheter clogging (6.7%). These complication rates were comparable to those observed in our previous single-arm study.9 Even more, there have been studies reporting the possibility of catheter rupture.15 Bleeding at the puncture site was high (15.5%) in the CPIPB group, necessitating vigilant catheter care to prevent subsequent infections.41 These catheter-related challenges introduce specific risks and clinical burdens that may compromise the analgesic effectiveness and influence the clinical selection of the cPNBs.42 In contrast, the use of LB simplifies the block procedure, requiring less procedural time (8 vs 26 min, P<0.001) and avoiding the complexities of catheter placement. From an economics perspective, previous studies support the cost-saving advantages of LB over cPNBs.19 This study did not incorporate a cost analysis because the catheter, LB, bupivacaine, and other related materials were funded through research grants, and patients did not incur expenses voluntarily. Nonetheless, the economic implications of these two analgesic methods warrant further investigation. Moreover, the capability to perform procedures related to peripheral nerve catheterization is not available across all institutions, which contributes to the widespread popularity of LB. Regarding adverse events and safety, no cases of local anesthetics toxicity, sciatic nerve injury, falling, infection at the puncture site, or other serious adverse events were observed (0%). The incidences of vomiting, dizziness, pruritus, constipation, suspected allergy, rash and poor appetite did not differ significantly between two groups (P>0.05). The CPIPB group experienced less postoperative nausea (6.7% vs 24.4%) and insomnia (2.2% vs 20%), which may be attributed to the decreased opioid exposure and enhanced analgesic effectiveness (P<0.05). Although serum drug concentrations were not continuously monitored, extensive evidence supports the safety of the dosages used in this study.19,20 For CPIPB, the continuous infusion of 0.125% bupivacaine at 6 mL/h (0.082–0.178 mg/kg/h) is well below the recommended predictive toxicity threshold of 0.45 mg/kg/h.43 In our prior investigations, the safety of 266 mg of liposomal bupivacaine (LB) has been established.7 The mean peak concentration of LB during local infiltration for TKA at this dosage is 340 ng/mL, below the anticipated thresholds for cardiovascular and neurotoxicity, which range from 2000 to 4000 ng/mL.13,44 Concurrently, the maximum clinically approved dose of LB is 532 mg, with its peak plasma concentration remaining below established toxicity thresholds.45 Overall, both regimens demonstrated acceptable safety. This study has several limitations. First, the study was designed as a comparison of two clinically applicable PIPB-based analgesic strategies rather than as an equal-dose pharmacodynamic comparison. The total local anesthetic dosage was not perfectly matched. The CPIPB group received an initial 100 mg dose and a 540 mg continuous infusion of bupivacaine over 72 hours, whereas the LB group received a single 266 mg dose of a sustained-release formulation lasting 72 hours, possibly causing bias due to differing pharmacokinetics. Second, the visible nature of the catheter in the CPIPB group precluded blinding (open-label design), which might influence subjective pain reporting. Third, the routine use of patient-controlled rescue analgesia (PCA) might have blunted differences in postoperative pain scores, particularly at rest, potentially obscuring the true comparative efficacy of the two block techniques. Fourth, while no falls occurred, lower extremity nerve blocks carry a theoretical risk of motor weakness, necessitating careful postoperative nursing. Finally, this was a single-center study with a relatively small sample size. Future, larger multicenter trial designed with higher power are warranted to further validate these findings. Conclusion A single-shot PIPB with LB met the prespecified non-inferiority criterion for pain relief compared with CPIPB after TKA in both PP and ITT analyses. CPIPB performs better in opioid consumption and early rehabilitation parameters, while LB simplifies the process and eliminates catheter-related risks. Acknowledgments The authors appreciated all patients who participated in the study and all the research staff. Funding Statement This work was supported by the Natural Science Foundation of Fujian Province (grant number 2023J011194); Fujian provincial health technology project (grant number 2023CXA007). Natural Science Foundation of Fujian Province (grant number 2024J011647); Joint Funds for the innovation of science and Technology, Fujian province (grant number 2024Y9053). Research Registration Unique Identifying Number (UIN) The study protocol was registered in the Chinese Clinical Trial Registry (No. ChiCTR2500103620; dated June 3, 2025; https://www.chictr.org.cn/) before the first patient enrollment. Patient recruitment commenced on 5 June 2025. Data Sharing Statement The datasets generated during and/or analyzed during the current study are available from the corresponding authors, Ting Zheng (Email: zhengting1223@163.com) and Xiaochun Zheng (Email: fjslmzk7841@163.com), upon reasonable request. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors have no conflict of interest to declare. None of the authors has any financial, professional, or other relevant relationship with any company. References Associated Data Data Availability Statement The datasets generated during and/or analyzed during the current study are available from the corresponding authors, Ting Zheng (Email: zhengting1223@163.com) and Xiaochun Zheng (Email: fjslmzk7841@163.com), upon reasonable request.