Abstract
Background
Thoracoscopic lung tumor surgery is associated with significant postoperative pain. While liposomal bupivacaine (LB) provides prolonged analgesia, its higher cost and extended duration profile have not been directly compared with ropivacaine, a commonly used long‑acting local anesthetic in paravertebral blockade. Whether these differences translate into meaningful clinical benefits remains uncertain. This study assessed the utility of LB as an alternative method for pain control through a comparison with ropivacaine.
Methods
In this prospective, double-blind, randomized controlled trial, 200 patients scheduled for elective thoracoscopic lung tumor surgery were randomly assigned to control [paravertebral block (PVB) with 10 mL of 0.75% ropivacaine; n=100] or experimental (PVB with 10 mL of 1.3% LB; n=100) groups. Postoperative patient-controlled intravenous analgesia (PCIA) with hydromorphone (10 mg/180 mL; background: 0 mL/h; bolus: 5 mL; lockout: 15 min) was initiated in all patients. Primary outcomes included pain scores of the Numerical Rating Scale (NRS) at rest and during movement at 6, 12, 24, 36, 48, and 72 hours post-operation; and the cumulative morphine milligram equivalent (MME) at 0–<24, 24–<48, 48–72, and 0–72 hours (including both PCIA and rescue analgesics). Secondary outcomes included PCIA doses, rescue analgesia, and adverse events within 72 hours. Exploratory 3‑month telephone follow‑up assessed pain presence during usual activities and breathing.
Results
Of 200 randomized patients, 184 completed in‑hospital follow‑up (92 per group) and were included in the primary analysis.Sixteen patients were excluded after randomization (3 due to cancelled surgery, 4 due to reoperation, 5 due to postoperative complications, and 4 due to withdrawal of consent). Compared with the control group, the LB group showed significantly lower resting NRS scores at 12, 24, 36, and 48 hours (P=0.04, 0.04, 0.02, and 0.01, respectively) and dynamic NRS scores at all time points (6, 12, 24, 36, 48, and 72 hours) (P=0.02, 0.001, 0.002, 0.04, 0.03, and 0.046, respectively). MME consumption was significantly reduced in the LB group during the 48- to 72-hour interval (P=0.03) and cumulatively over 0–72 hours (P=0.03). The number of delivered PCIA doses was significantly lower in the LB group during the 48- to 72-hour interval (P=0.02), with no significant differences in the 0 to 24- or 24- to 48-hour intervals. Rescue analgesia rates (3.3% vs. 5.4%, P=0.72) and the incidence of any adverse events (42.4% vs. 51.1%, P=0.30) did not differ significantly between groups. At 3 months, 175 patients completed follow-up (control: 90; LB: 85); fewer LB patients reported pain (17.6% vs. 21.1%), though no statistical comparison was performed.
Conclusions
In thoracoscopic lung tumor surgery, PVB with LB provides improved and sustained postoperative analgesia and reduces opioid consumption as compared with ropivacaine, without increasing the incidence of adverse events. It may represent a safe and effective single-injection alternative for multimodal analgesia.
Trial Registration
Chinese Clinical Trial Registry ChiCTR2400093464.
Keywords: Liposomal bupivacaine (LB), video-assisted thoracic surgery (VATS), paravertebral block (PVB), analgesia
Highlight box.
Key findings
• In this prospective, double-blind, randomized controlled trial (N=184), paravertebral block (PVB) with liposomal bupivacaine (LB) significantly reduced dynamic Numerical Rating Scale (NRS) scores at all time points (6–72 hours) and resting NRS scores at 12–48 hours as compared with ropivacaine. Morphine milligram equivalent consumption was significantly lower in the LB group than in the ropivacaine group during the 48- to 72-hour interval and cumulatively (0–72 hours). Doses of patient-controlled intravenous analgesia were significantly lower in the LB group only during the 48- to 72-hour interval. Rescue analgesia rates and the incidence of adverse events did not differ between the groups.
What is known and what is new?
• Single-injection ropivacaine PVB provides only 6–8 hours of analgesia, insufficient to cover the 72-hour acute pain peak, often requiring catheter infusion or supplemental opioids.
• This trial demonstrates that a single LB injection provides sustained analgesia up to 72 hours, particularly for dynamic pain, with significant late-phase opioid reduction and no increase in adverse events, even with demand-only patient-controlled analgesia.
What is the implication, and what should change now?
• LB PVB offers a simplified single-injection alternative to continuous catheter techniques for thoracoscopic surgery, reducing late-phase opioid needs while preserving patient satisfaction. It aligns with enhanced recovery principles and may benefit patients with opioid intolerance or early recovery goals. Further studies with validated chronic pain assessments are warranted to confirm long-term benefits.
Introduction
As the number of pulmonary procedures being performed continues to increase, it is likely that the burden of inadequate postsurgical pain management associated with these procedures will also rise. Postsurgical pain and its clinical consequences are recognized as critical issues in thoracic surgery. Inadequate pain control may lead to significant postoperative sequelae, including pneumonia, atelectasis, and deep venous thrombosis (1-3).Although thoracoscopic surgical techniques are gradually becoming less invasive—with single-port or dual-port approaches frequently employed—and the surgical incisions and trauma of traditional thoracotomy are being significantly reduced, patients may still experience substantial acute postoperative pain, with approximately 10% to 20% developing chronic pain (4-8). Adequate pain management after lung resection is essential for faster recovery and for the prevention of postoperative complications.
Various management strategies for pain after minimally invasive thoracic surgery are available, including epidural analgesia, thoracic paravertebral nerve block (TPVB), and intercostal nerve block. More recently, the erector spinae plane block (ESPB) has emerged as a simpler alternative, though available evidence suggests that it may provide less consistent analgesic coverage than TPVB for thoracic procedures. Systemic multimodal analgesia, including non-opioid adjuvants and patient-controlled intravenous analgesia (PCIA), is commonly used as an adjunct or as the primary analgesic strategy when regional techniques are contraindicated. By virtue of its excellent analgesic effect and favorable complication profile, TPVB has been widely adopted for postoperative pain management in thoracic surgery in recent years (9). Multiple studies have reported that the analgesic effect of TPVB is similar to that of the gold standard for postoperative pain management in thoracic surgeries—epidural analgesia (10). Furthermore, TPVB significantly reduces perioperative opioid consumption (11) and is associated with a lower incidence of hypotension and more stable hemodynamics as compared to epidural analgesia (12,13).
Ropivacaine and bupivacaine are commonly used for paravertebral block (PVB) due to their simplicity and low cost. However, the duration of blockade provided by these conventional local anesthetics is typically limited to 6 to 8 hours (14), which is insufficient to cover the peak acute pain period within the first 72 hours after surgery. This limitation often necessitates continuous infusion via a catheter or reliance on intravenous analgesia, increasing both management burden and opioid consumption. In our institution, as in many others worldwide, ropivacaine has become the standard choice for single-shot PVB because of its better cardiac safety profile, wider therapeutic index, and reduced motor blockade compared with bupivacaine—a practice supported by clinical guidelines and extensive prior experience in thoracic surgery.
Liposomal bupivacaine (LB) is a prolonged-release formulation of bupivacaine and is indicated for single-dose administration into the surgical site for effectuating postsurgical analgesia (15). With this formulation, a single injection is sufficient to provide analgesia for up to 72 hours, potentially reducing the need for catheters or supplemental opioids. Recently, LB has been used for pain management in various surgical procedures, including thoracic, orthopedic, colorectal, and plastic surgery (16-19). Studies have demonstrated that posterior intercostal nerve blocks with LB are safe and effective for patients undergoing thoracic surgery (20). Additionally, studies have compared LB to paravertebral catheters and to bupivacaine hydrochloride injection alone in patients undergoing thoracic surgery (21,22). However, the clinical question remains: does LB's higher cost and prolonged pharmacokinetic profile translate into meaningful analgesic or opioid-sparing benefit over ropivacaine in the specific setting of video-assisted thoracic surgery (VATS)? Evidence directly comparing LB with ropivacaine in this setting remains limited. Ropivacaine was chosen as the active comparator because it represents the current standard of care for single‑shot paravertebral blockade in our practice setting, as well as in many other institutions worldwide. This pragmatic design ensures that the comparison between LB and ropivacaine directly informs the clinical decision faced by practitioners in routine care. Therefore, the objective of this prospective double-blind randomized controlled trial was to compare the analgesic efficacy, opioid-sparing effects, and safety between LB and ropivacaine for TPVB, with an exploratory analysis of pain at 3 months post-operation. We present this article in accordance with the CONSORT reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-2185/rc).
Methods
Design
This single-center, prospective randomized, double-blind, active comparator-controlled clinical trial was approved by the Clinical Trial Ethics Committee of The First Affiliated Hospital, Zhejiang University School of Medicine ([2024] IIT Ethics Approval No. 067). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from all participants prior to enrollment. Preoperatively, patients completed the Brief Pain Inventory questionnaire to assess baseline chronic pain. All patients received preoperative instruction on the use of pain rating scales and patient-controlled analgesia (PCA) pumps.
Participants
Consecutive patients undergoing elective VATS lung resection at a tertiary referral hospital were enrolled, including adults aged 18 to 70 years with American Society of Anesthesiologists (ASA) physical status classification I to III who underwent video-assisted thoracoscopic pulmonary resection at The First Affiliated Hospital, Zhejiang University School of Medicine between October 2024 and February 2026. Procedures included VATS lobectomy, segmentectomy, or wedge resection. The exclusion criteria were as follows: emergency surgery; history of ipsilateral thoracic surgery; chronic opioid use for ≥6 months (defined as daily opioid intake for more than 3 months within the past 6 months); extensive opioid use within 1 month prior to screening; pre-existing neuropathic pain; known allergy to study drugs; severe hepatic dysfunction (Child-Pugh class B or C) or severe renal dysfunction (estimated glomerular filtration rate <30 mL/min/1.73 m2); andpregnancy or lactation. Patients with conditions that could interfere with pain status communication, such as expected intubation for >24 hours or altered mental status, were also excluded, as these conferred a high probability of postoperative morbidity. Patients who required conversion from VATS to thoracotomy during surgery, or who received additional regional blocks (e.g., intercostal nerve block, erector spinae plane block) beyond the study intervention, were also excluded from the final analysis.
Patients and the public were not involved in the design, conduct, reporting, or dissemination plans of this trial. No patient representatives were consulted for outcome selection or protocol development.
Randomization and blinding
A parallel-group design with a 1:1 allocation ratio was employed. According to a random number table, 100 patients were assigned to the ropivacaine TPVB group and 100 to the LB TPVB block group. The randomization sequence was generated by a biostatistician not involved in patient recruitment using a random number table. The sequence was 1:1, and there was no stratification or blocking. Allocation was concealed through the use of sequentially numbered, opaque, sealed envelopes. The site investigator enrolled participants and assigned them to groups by opening the next envelope. Both the test drug and control drug were identified by codes. Throughout the trial, patients, data collectors, outcome assessors, and the statistician remained blinded to group assignment. Drug codes were kept confidential by a designated person and were not unblinded until after trial completion unless required for patient safety.
As LB (a milky white suspension) is visually distinct from ropivacaine (colorless clear solution), to maintain patient blinding, all study drugs were prepared by an anesthesia assistant not involved in subsequent assessments using identical syringes wrapped with opaque tape. The injection volume was identical (10 mL) in both groups to ensure comparability of the interventions. Although the surgeon performing the block might have been able to distinguish the study drug by appearance, the surgeon was not involved in any postoperative assessments, and thus the blinding for study outcomes was not compromised.
Interventions
All patients underwent scheduled thoracoscopic lung tumor resection according to each surgeon’s standard practice. General anesthesia was induced, and patients were intubated with a double-lumen endotracheal tube for one-lung ventilation with the affected lung collapsed. PVBs were performed by four experienced thoracic surgeons, each with more than 5 years of thoracoscopic surgery practice and at least 50 prior TPVB procedures. All surgeons received standardized training on the LB injection technique before the trial commenced. Under video-assisted thoracoscopic guidance, the surgeon located the corresponding paravertebral space (typically at the T4–T5 or T5–T6 level, adjusted according to the surgical incision and port sites) and injected the local anesthetic.Patients randomized to the intervention arm received 133 mg of LB in 10-mL injections, while those in the control arm received 75 mg of ropivacaine (without epinephrine) in 10-mL injections. Both groups received the same injected volume (10 mL) via 22-gauge needles. Block success was confirmed by direct visualization of the anesthetic solution diffusing along the paravertebral space (Figure 1).
Figure 1.
Thoracoscopic view of local anesthetic injection and diffusion in the paravertebral space. (A) Needle position at the paravertebral space before injection (arrow). (B) Local anesthetic diffusion along the paravertebral space after injection (arrow).
Perioperative analgesia regimens were standardized and did not differ by group assignment. Postoperative PCIA was standardized for all patients and included 10 mg of hydromorphone diluted with normal saline to a total volume of 180 mL (concentration 55.6 μg/mL). The other parameters were as follows: background infusion, 0 mL/h; bolus dose, 5 mL (277.8 μg); and lockout interval, 15 minutes continued for 72 hours post-operation. PCIA was continued for 72 hours post-operation unless the patient requested removal (e.g., due to adverse events, perceived adequate analgesia, or early discharge) or if the medication was exhausted. Patients who experienced significant breakthrough pain also received intravenous tramadol (100 mg) as rescue analgesia.
Outcome measures
The baseline demographics recorded included age, gender, body mass index (BMI), ASA physical status, surgical procedure, surgical side (left or right), and comorbidities such as hypertension, diabetes mellitus, and chronic lung disease.
Pain intensity was assessed with the Numerical Rating Scale (NRS), an 11-point ordinal scale ranging from 0 (no pain) to 10 (worst pain imaginable) (23). NRS scores were evaluated at rest and during movement (coughing or deep breathing) at 6, 12, 24, 36, 48, and 72 hours post-operation.
Opioid consumption was recorded throughout the first 72 hours after the operation. All intravenous opioid administrations were documented, including opioid consumption via PCIA (during PCIA use) and rescue analgesics administered intravenously by physicians for breakthrough pain. Cumulative opioid consumption was calculated for the intervals of 0–<24, 24–<48, and 48–72 hours and for the total 0–72 hours. All intravenous opioid consumption was converted to the intravenous morphine milligram equivalent (MME) according to standard conversion factors (24), with MME serving as a primary outcome measure.
Secondary outcomes included PCA usage, rescue analgesia, and the incidence of adverse events. PCA usage was recorded as the number of delivered PCA doses for each interval (0–<24, 24–<48, and 48–72 hours). Rescue analgesia was defined as the proportion of patients requiring physician-administered intravenous rescue analgesia within the first 72 postoperative hours, and the total number of rescue events was also recorded. Adverse events were assessed systematically at each postoperative follow‑up time point (6, 12, 24, 36, 48, and 72 hours) using a structured data collection form that included common adverse events (e.g., nausea, vomiting, dizziness, pruritus, hypotension, respiratory depression, and arrhythmia) as prompts. However, the assessment was not limited to these events; all adverse events reported by patients or observed by study personnel were recorded. The following safety outcomes were specifically tracked: nausea, vomiting, dizziness, pruritus, hypotension, respiratory depression, arrhythmia, local anesthetic systemic toxicity, pneumothorax, block failure, infection, and neurological symptoms. The last four did not occur in either group. Causality assessment was not performed. Due to their strong clinical association and difficulty in being precisely differentiated, nausea and vomiting were combined into a single composite outcome of “nausea/vomiting”. A patient was considered to have experienced nausea/vomiting if they reported either nausea or vomiting (or both). Other adverse events—including dizziness, pruritus, hypotension, and bradycardia—were recorded separately.
As an exploratory analysis, patients were followed up by telephone at 3 months post-operation and were simply asked about the presence of pain in the wound or chest area during usual activities and breathing. The presence of pain (yes or no) was recorded. Of note, no standardized pain scales were used in this follow-up, and the results should be considered preliminary and exploratory.
Sample size calculation
The primary outcomes of this study included resting NRS, dynamic NRS, and MME. Sample size calculation was based on the 24-hour dynamic NRS score (during coughing), as dynamic pain has a greater impact on postoperative functional recovery and as previous studies have shown higher variability for this measure (25). Using this indicator provides a more conservative sample size estimate that is sufficient to power the analysis of other primary outcomes.
Based on the pilot study (N=20), the predicted postoperative 24-hour dynamic NRS score for the control group and the experimental group was 5.0±2.0 and 4.0±2.0, respectively. A between-group difference of 1.0 point was considered the minimum clinically meaningful effect size for the superiority comparison. With a two-sided α of 0.05 and a power (1-β) of 0.90, along with an assumption of equal sample sizes, the independent-samples t-test was used to calculate the required sample size, which indicated a minimum of 86 patients per group. With a 15% dropout rate being assumed, the final sample size was determined to be 100 patients per group, for a total of 200 patients. Although the primary analysis of NRS scores is conducted via a linear mixed-effects model, sample size calculation based on the t-test for the 24-hour time point can be considered conservative and valid, as the mixed-model comparison at this time point has similar statistical properties to the two-sample t-test and typically yields higher power.
No interim analysis was planned or performed during the trial. Therefore, no stopping guidelines were established. The trial was continued until the planned sample size was reached, and no early stopping rule was applied.
Statistical analysis
Statistical analyses were performed with SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables were tested for normality via the Shapiro-Wilk test. Normally distributed data were presented as the mean ±standard deviation (SD) and were compared via the independent samples t-test. Nonnormally distributed data were expressed as the median and interquartile range (IQR) and were analyzed with the Mann-Whitney U test. Categorical variables were presented as frequencies (%) and compared with the Chi-squared test or the Fisher exact test, as appropriate. All primary and secondary analyses were performed according to the Full Analysis Set (FAS) principle, which included all randomized patients who received the allocated intervention and had at least one post-baseline efficacy assessment. Of the 200 randomized patients, 184 (92 per group) met these criteria and were included in the primary FAS analysis for NRS scores and MME. The reasons for exclusion from the FAS were as follows: 3 patients due to cancelled surgery, 4 due to withdrawal of consent, 4 due to reoperation, and 5 due to postoperative complications (e.g., prolonged ventilation, severe bleeding) that precluded reliable pain assessment. For the exploratory 3‑month pain outcome, the 175 patients who completed the telephone follow‑up (90 controls and 85 LB-treated patients) were included. No per‑protocol, as‑treated analyses, or subgroup or sensitivity analyses were performed. All reported analyses were prespecified in the protocol.
NRS scores, as repeated measures, were analyzed with a linear mixed-effects model to account for within-patient correlations, and the effect size was recorded as the difference in estimated marginal means with the 95% confidence interval. Time was treated as a categorical variable (6, 12, 24, 36, 48, and 72 hours post-operation). The model included group, time point, and the group-by-time interaction as fixed effects, with the patient being a random intercept (i.e., only a random intercept, with no random slope). Restricted maximum likelihood estimation was used. Post hoc pairwise comparisons between groups at each time point were performed according to the estimated marginal means, with least significant difference adjustment for multiple comparisons being applied (no further adjustment across time points was applied).
MME, representing a cumulative consumption over predefined intervals, was compared between groups for each interval (0–<24, 24–<48, and 48–72 hours) and for the total amount of 0–72 cumulative hours via the t-test (for normally distributed data) or the Mann-Whitney U test (for nonnormally distributed data).
Secondary outcomes, including delivered PCA doses, rescue analgesia rate, and adverse events, were compared between groups with the same methods as those described above for continuous or categorical variables, as appropriate.
No data were missing for the primary outcomes (NRS scores and MME). Complete case analysis was therefore performed, and no imputation was used for missing data as none occurred. Statistical significance was defined as a P value <0.05. All tests were two-sided.
Data from the exploratory analysis were summarized descriptively without between-group statistical comparisons.
The trial protocol and statistical analysis plan were finalized before data lock and are available from the corresponding author upon reasonable request. A summary of the protocol has been registered with the Chinese Clinical Trial Registry (ChiCTR2400093464). No changes were made to the trial protocol after commencement, and all outcomes and analyses were prespecified.
Results
Participants
A total of 200 consecutive patients undergoing VATS were enrolled in the study, with 100 patients assigned to the LB group and 100 to the ropivacaine group. Sixteen patients were excluded from the final analysis due to surgery cancellation, withdrawal of consent, postoperative complications, or need for reoperation. Consequently, 184 patients (92 in each group) completed the in-hospital follow-up and were included in the primary and secondary outcome analyses (Figure 2). At the 3-month telephone follow-up, 2 patients in the control group and 7 in the LB group were lost to follow-up, resulting in 175 patients (90 in the control group and 85 in the LB group) being included in the exploratory analysis.
Figure 2.
CONSORT flow diagram of patient enrollment, randomization, and follow-up. LB, liposomal bupivacaine.
All 184 patients who underwent randomization received their allocated PVB as assigned, and no crossover occurred. For all patients, the actual injected volume was 10 mL, with block success being confirmed by direct visualization of anesthetic diffusion. The per-protocol adherence was 100%. Postoperative analgesia was standardized as described in the Methods section. All patients received demand-only PCIA with hydromorphone. Breakthrough pain was managed with intravenous tramadol (100 mg, as needed). No other analgesics (e.g., nonsteroidal anti-inflammatory drugs, acetaminophen, and gabapentinoids) were permitted during the 72-hour study period. The number of tramadol rescue doses was low (overall 4.3% of patients).
There were no significant differences between the two groups in terms of age, sex, BMI, ASA class, type of surgery, side of surgery, hypertension, diabetes, or chronic lung disease (all P values >0.05), indicating the comparability of the groups (Table 1).
Table 1. Baseline characteristics of the study population.
| Characteristic | LB (n=92) | Control (n=92) | P value |
|---|---|---|---|
| Age (years) | 57.5 (48.0, 62.0) | 58.0 (52.0, 62.8) | 0.39 |
| Sex | 0.76 | ||
| Male | 34 (37.0) | 37 (40.2) | |
| Female | 58 (63.0) | 55 (59.8) | |
| BMI (kg/m2) | 22.7±2.8 | 23.1±3.1 | 0.27 |
| ASA classification | 0.14 | ||
| ASA 1 | 22 (23.9) | 12 (13.0) | |
| ASA 2 | 62 (67.4) | 73 (79.3) | |
| ASA 3 | 8 (8.7) | 7 (7.6) | |
| Surgery procedure | 0.51 | ||
| Wedge resection | 64 (69.6) | 56 (60.9) | |
| Segmentectomy | 11 (12.0) | 15 (16.3) | |
| Pulmonary lobectomy | 17 (18.5) | 21 (22.8) | |
| Surgery side | 0.65 | ||
| Left | 33 (35.9) | 37 (40.2) | |
| Right | 59 (64.1) | 55 (59.8) | |
| Hypertension | 21 (22.8) | 22 (23.9) | >0.99 |
| Diabetes mellitus | 6 (6.5) | 4 (4.3) | 0.75 |
| Chronic lung disease | 6 (6.5) | 5 (5.4) | >0.99 |
Data are presented as mean ± standard deviation, median (IQR), or n (%). ASA, American Society of Anesthesiologists; BMI, body mass index; LB, liposomal bupivacaine.
Postoperative resting and dynamic NRS scores at each time point are presented in Table 2 and Figure 3. The linear mixed-effects model revealed a significant main effect of time (P<0.001) but no significant group-by-time interaction (resting NRS: P=0.25; dynamic NRS: P=0.62). Prespecified pairwise comparisons at each time point conducted via estimated marginal means revealed the following: For resting NRS, the LB group showed significantly lower scores only at 12, 24, 36, and 48 hours post-operation, with no significant differences at 6 or 72 hours. For dynamic NRS (during coughing/deep breathing), the LB group had significantly lower scores atall time points (6, 12, 24, 36, 48, and 72 hours).
Table 2. Postoperative resting and dynamic NRS scores.
| Variable | Time | LB (n=92) | Control (n=92) | Median difference (95% CI) | P value |
|---|---|---|---|---|---|
| Resting NRS | 6 h | 0 (0, 1.0) | 0 (0, 1.0) | −0.17 (−0.44, 0.10) | 0.21 |
| 12 h | 0 (0, 1.0) | 0 (0, 1.0) | −0.28 (−0.55, −0.01) | 0.04 | |
| 24 h | 0 (0, 1.0) | 0 (0, 1.0) | −0.28 (−0.55, −0.01) | 0.04 | |
| 36 h | 0 (0, 1.0) | 0 (0, 1.0) | −0.33 (−0.60, −0.06) | 0.02 | |
| 48 h | 0 (0, 0) | 0 (0, 1.0) | −0.35 (−0.62, −0.08) | 0.01 | |
| 72 h | 0 (0, 0) | 0 (0, 0) | −0.08 (−0.35, 0.19) | 0.58 | |
| Dynamic NRS | 6 h | 1.0 (1.0, 2.0) | 2.0 (1.0, 3.0) | −0.37 (−0.67, −0.07) | 0.02 |
| 12 h | 1.0 (1.0, 2.0) | 2.0 (1.0, 3.0 | −0.49 (−0.79, −0.19) | 0.001 | |
| 24 h | 1.0 (1.0, 2.0) | 2.0 (1.0, 3.0) | −0.48 (−0.78, −0.18) | 0.002 | |
| 36 h | 1.0 (1.0, 1.8) | 1.0 (1.0, 2.0) | −0.32 (−0.62, −0.02) | 0.04 | |
| 48 h | 1.0 (0, 1.0) | 1.0 (1.0, 2.0) | −0.34 (−0.64, −0.04) | 0.03 | |
| 72 h | 0 (0, 0) | 1.0 (0, 1.0) | −0.30 (−0.60, −0.01) | 0.046 |
Data are presented as median (IQR). P values from linear mixed-effects model with post hoc pairwise comparisons using estimated marginal means (least significant difference adjustment). CI, confidence interval; IQR, interquartile range; LB, liposomal bupivacaine; NRS, Numerical Rating Scale.
Figure 3.
Postoperative resting and dynamic NRS pain scores over time. (A) Resting NRS. (B) Dynamic NRS. Data are presented as the mean ± SD. *P<0.05, **P<0.01 vs. control group (linear mixed-effects model with post hoc pairwise comparisons). LB, liposomal bupivacaine; NRS, Numerical Rating Scale; SD, standard deviation.
The LB group had significantly lower MME consumption during the 48- to 72-hour and cumulative 0 to 72-hour intervals compared with the control group (P<0.05); no significant difference was found in the 0 to 24-hour and 24- to 48-hour interval (P>0.05) (Table 3).
Table 3. Mean morphine milligram equivalent administration.
| Interval | LB (n=92) | Control (n=92) | Median difference (95% CI) | P value |
|---|---|---|---|---|
| 0–<24 h | 16.7 (11.1, 22.2) | 16.7 (16.7, 27.8) | 0 (0, 5.6) | 0.053 |
| 24–<48 h | 11.1 (0, 16.7) | 16.7 (0, 22.2) | 0 (0, 5.6) | 0.09 |
| 48–72 h | 0 (0, 11.1) | 0 (0, 15.3) | 0 (0, 0) | 0.03 |
| Total | 33.3 (16.7, 50.0) | 44.4 (22.2, 61.1) | 10.0 (0, 16.7) | 0.03 |
Data are presented as median (IQR). Intervals are defined as left‑closed, right‑open; the 24‑hour measurement belongs to the 24–<48 h interval, and the 48‑hour measurement belongs to the 48–72 h interval. CI, confidence interval; LB, liposomal bupivacaine.
In a post-hoc exploratory analysis, the number of patients with zero MME (i.e., no intravenous opioid use) during the first 72 postoperative hours was 4 (4.3%) in the control group and 7 (7.6%) in the LB group (Fisher exact test P=0.36).
The number of delivered PCA doses was significantly lower in the LB group during the 48- to 72-hour interval (P=0.02) but not during the 0 to 24-hour or 24- to 48-hour intervals (P>0.05) (Table 4). Within 72 hours post-operation, rescue analgesia was required in 5 (5.4%) patients in the control group and 3 (3.3%) patients in the LB group, with no significant difference between the groups in this regard (P=0.72) (Table 5).
Table 4. PCA usage.
| Delivered PCA dose | LB (n=92) | Control (n=92) | Median difference (95% CI) | P value |
|---|---|---|---|---|
| 0–<24 h | 3.0 (2.0, 4.0) | 3.0 (3.0, 5.0) | 0 (0, 1.0) | 0.12 |
| 24–<48 h | 2.0 (0, 3.0) | 3.0 (0, 4.0) | 0 (0, 1.0) | 0.07 |
| 48–72 h | 0 (0, 2.0) | 0 (0, 2.8) | 0 (0, 0) | 0.02 |
Data are presented as median (IQR). Intervals are defined as left‑closed, right‑open; the 24‑hour measurement belongs to the 24–<48 h interval, and the 48‑hour measurement belongs to the 48–72 h interval. CI, confidence interval; IQR, interquartile range; LB, liposomal bupivacaine; PCA, patient-controlled analgesia.
Table 5. Rescue analgesia.
| Outcome | LB (n=92) | Control (n=92) | Relative risk (95% CI) | P value |
|---|---|---|---|---|
| Patients requiring rescue, n (%) | 3 (3.3) | 5 (5.4) | 0.60 (0.15, 2.44) | 0.72 |
CI, confidence interval; LB, liposomal bupivacaine.
Within 72 hours post-operation, adverse event occurred among the 47 (51.1%) patients in the control group and among the 39 (42.4%) patients in the LB group, with no difference between groups in this regard (P=0.30). Nausea/vomiting occurred in 25 (27.2%) and 23 (25.0%) (P=0.87) patients in the control and LB groups, respectively, while dizziness occurred in 35 (38.0%) and 26 (28.3%) patients, respectively (P=0.21). Other adverse events (pruritus, hypotension, and bradycardia) occurred at low frequencies, with the exact numbers presented in Table 6.
Table 6. Adverse events.
| Outcome | LB (n=92) | Control (n=92) | Relative risk (95% CI) | P value |
|---|---|---|---|---|
| Any adverse event | 39 (42.4) | 47 (51.1) | 0.83 (0.61, 1.13) | 0.30 |
| Nausea/vomiting | 23 (25.0) | 25 (27.2) | 0.92 (0.57, 1.50) | 0.87 |
| Dizziness | 26 (28.3) | 35 (38.0) | 0.74 (0.49, 1.13) | 0.21 |
| Other adverse events | ||||
| Pruritus | 0 (0.0) | 2 (2.2) | † | 0.50 |
| Hypotension | 8 (8.7) | 10 (10.9) | 0.80 (0.33, 1.93) | 0.81 |
| Bradycardia | 0 (0.0) | 1 (1.1) | † | >0.99 |
Data are presented as n (%) or n unless otherwise specified. †, due to the extremely low event rates, no relative risk was calculated for these events. CI, confidence interval; LB, liposomal bupivacaine.
Exploratory outcomes
At 3 months post-operation, 175 patients completed the telephone follow-up. In the control group, 19 (21.1%) patients reported persistent pain in the wound or chest area during usual activities or breathing; in the LB group, 15 (17.6%) patients reported pain. As this was an exploratory analysis, no statistical comparisons were performed between the groups.
Discussion
In this prospective, double-blind, randomized controlled trial, in which patients underwent thoracoscopic lung tumor surgery, those with PVB with LB had significantly lower dynamic NRS scores at all time points (6–72 hours) and lower resting NRS scores at 12, 24, 36, and 48 hours compared to those who received ropivacaine (Table 2). Opioid consumption (MME) was significantly lower in the LB group during the 48- to 72-hour interval and cumulatively (0–72 hours), supporting a sustained opioid-sparing effect. Adverse event rates and rescue analgesia did not differ between the groups.
The present study was designed as a pragmatic, clinically oriented trial rather than a formulation development study. While a head-to-head comparison with plain bupivacaine—with or without adjuncts—would be appropriate to isolate the effect of the liposomal delivery system, our goal was to answer a different question: in current clinical practice, in which ropivacaine has largely replaced bupivacaine for TPVB, does replacing the standard single shot local anesthetic with LB provide meaningful benefits? We therefore selected ropivacaine as the comparator to maximize the clinical generalizability of our findings. Although adjuncts such as dexamethasone or dexmedetomidine have been shown to prolong the duration of peripheral nerve blocks in some studies, their use in the paravertebral space remains off-label, non-standardized, and not sufficiently established to serve as a definitive reference comparator in a pivotal clinical trial.
Single-injection ropivacaine typically provides only 6–8 hours of analgesia (14), insufficient for the 72-hour pain peak. In our study, although the control group’s pain scores did not rise sharply after 24 hours, their opioid consumption during the 48 to 72-hour period was significantly higher than that of the LB group. This aligns with the known limited analgesic duration of ropivacaine, which requires additional opioids to cover the later analgesic gap. In contrast, LB maintained lower pain scores throughout 24–72 hours and provided significant late-phase opioid sparing. Its multivesicular liposome system slowly releases the drug, prolonging neural blockade. The lack of difference in opioid use during the first 24 hours reflects similar early efficacy, while the late-phase advantage (48–72 hours) is consistent with single-injection coverage of the entire acute pain period.
Clinically, LB eliminates the need for continuous infusion via catheter, simplifies analgesia, and aligns with the principles of enhanced recovery after surgery. In the exploratory post hoc analysis, 7 patients (7.6%) in the LB group did not require any opioids during the first 72 hours, while 4 patients (4.3%) in the control group did require opioids, but this did not represent a significant difference (P=0.36). Nevertheless, this numerical discrepancy is consistent with the overall opioid-sparing effect of LB and suggests that some patients may achieve opioid-free analgesia. Reduced opioid consumption lowers the risk of nausea, vomiting, and dizziness. The safety profile is acceptable and does not lead to a significantly higher incidence of adverse events. Notably, the benefit was most evident in the late phase, and thus LB addresses a gap that conventional local anesthetics cannot fill, making it valuable for patients with prolonged pain, opioid intolerance, or rapid recovery goals. It should be noted that LB is substantially more expensive than ropivacaine, and its availability varies across institutions and healthcare settings. Future cost-effectiveness analyses are warranted to determine whether the observed clinical benefits justify the additional cost and to inform guideline development for its use in routine practice.
All patients received PCIA without background infusion (demand only). Therefore, total opioid consumption (MME) directly reflects patient-initiated PCA demands plus rescue analgesia. The overall rescue analgesia rates (0–72 hours) were comparable between groups, but interval-specific comparisons were not performed due to low event rates; thus, we cannot exclude a potential contribution of interval-specific rescue analgesia to the difference in MME. However, the low overall rescue rate (<5%) and the significantly lower PCA demand in the LB group over 48–72 hours indicate that the MME reduction is largely attributable to decreased PCA boluses rather than rescue medication.
Regarding NRS scores, a distinct pattern emerged. Resting NRS was significantly lower in the LB group at 12, 24, 36, and 48 hours, with no differences at 6 or 72 hours; dynamic NRS was significantly lower at all time points (6–72 hours). This reflects dynamic pain’s higher sensitivity to blockade and the availability of on-demand PCA. The early (6 hours) dynamic benefit without resting difference is probably due to the higher sensitivity of dynamic pain, even during the onset phase. The middle phase (12–48 hours) showed differences in both pain types, representing peak sustained release. The loss of resting difference at 72 hours but the persistence of dynamic difference reflects waning block intensity, with resting pain “breaking through” earlier, while dynamic pain remains controlled for a longer period. Importantly, because every PCA dose was a direct response to perceived pain, the lower PCA demand in the LB group in the 48- to 72-hour interval indicates improved residual blockade at a point when ropivacaine’s effect dissipated.
The control group did not show the expected sharp rise in NRS after 24 hours, likely because patients could self-administer PCA boluses to compensate for the declining ropivacaine effect. However, the LB group required significantly fewer boluses during the 48- to 72-hour period, indicating that its prolonged action reduced the need for such compensatory opioid use. Therefore, the observed between-group differences are conservative estimates, and the true advantage of LB may be even greater.
In this study, all PVBs were performed by four experienced surgeons under general anesthesia using direct thoracoscopic visualization. Compared with preoperative ultrasound‑guided blocks by anesthesiologists, this technique has several advantages. Ultrasound‑guided TPVB can be technically challenging in patients with obesity or spinal deformity due to the deep anatomy and steep needle trajectory. In contrast, surgeons use anatomical landmarks under direct vision to reliably locate the target space, achieving better alignment between the incision and block level. Direct visualization also allows for the immediate confirmation of local anesthetic spread and the early detection of intravascular injection or pleural puncture, thus reducing complications such as pneumothorax and bleeding (26). The thoracoscopic transthoracic approach avoids repeated needle passes and provides a nearly zero‑risk anterior route with reliable block quality (27).
Performing the block under general anesthesia eliminates patient discomfort and preoperative anxiety, as it is administered after induction and lung collapse when the patient is pain‑free and relaxed, avoiding awake puncture and additional positioning. As noted by Tong et al., this approach is ideal for anxious, uncooperative, or ultrasound‑failed patients (26). It also improves workflow efficiency by requiring no extra ultrasound equipment, no extended preparation time, and no additional anesthesia team resources, thereby shortening block time and operating room turnover. Hong et al. reported that surgeon‑performed TPVB reduced block time while providing comparable efficacy and safety to the standard ultrasound‑guided technique performed by anesthesiologists (28).
Because LB and ropivacaine differ visually—with the former being milky white and the latter being colorless—blinding the operator was impossible. However, with surgeons who were not involved in postoperative assessments preserved blinding of patients, assessors, data collectors, and the statistician, ensuring outcome objectivity. Transferring the block to the surgical team standardized execution and maintained blinding integrity.
Moreover, the analgesic results achieved with surgeon‑performed TPVB in this study are consistent with those reported previously. The LB group showed significantly lower dynamic NRS at 6–72 hours, lower resting NRS at 12–48 hours, and reduced MME at 48–72 hours and cumulatively. Xu et al. found that thoracoscopic‑guided TPVB had higher first‑attempt success, a greater number of blocked segments, and longer analgesic duration than ultrasound‑guided TPVB (27). Studies on surgeon‑performed blocks also confirmed non-inferior 48‑hour opioid consumption compared to anesthesiologist‑guided blocks (29). In summary, surgeon‑performed intraoperative TPVB under direct thoracoscopic visualization is operationally convenient, efficient, amenable to blinding, and offers good efficacy and safety, offering a reliable paradigm for regional analgesia in minimally invasive thoracic surgery.
Limitations
Several limitations of this study should be considered. First, the single-center design may limit the generalizability of the findings, and multicenter studies are needed. Although a number of different surgeons performed the blocks, all received standardized training, and randomization likely distributed any operator variability evenly between the groups. The surgeons could not be blinded due to the distinct appearance of the study drugs, but they were not involved in any postoperative assessments; patients, assessors, data collectors, and the statistician remained blinded, preserving outcome objectivity. A formal assessment of blinding was not performed, which may be considered a limitation of the study design. Second, the primary analyses were based on the FAS, which excluded 16 patients for reasons unrelated to treatment response and determined before unblinding. A sensitivity analysis including all 200 randomized patients was not performed; however, given the low exclusion rate (8%) and the absence of missing data for the primary outcomes, the FAS analysis is unlikely to have introduced material bias. The PCA devices recorded only delivered doses (not total attempts), precluding analysis of ineffective attempts and analgesic demand intensity. Moreover, plasma concentration monitoring was not performed, but the double-blind design supports the validity of the observed clinical efficacy differences. Third, all patients received PCIA without background infusion (demand-only), reflecting common clinical practice and ensuring that opioid consumption directly reflected patient-initiated demands. However, this design might have increased the risk of breakthrough pain in the control group during the late phase, which was indeed captured by higher PCA usage and potentially by interval-specific rescue analgesia. The low rescue analgesia rate limited interval-specific power, and thus we used the overall 0 to 72-hour rescue rate as the primary analysis. Finally, we did not include a comparator arm with bupivacaine supplemented with dexamethasone or dexmedetomidine. While these adjuncts may prolong the duration of neural blockade, their use in the paravertebral space remains non-standardized and off-label, with considerable variation in dosing regimens and limited long-term safety data. A comparison with such a non-standardized regimen would have reduced the generalizability of our findings and is therefore better addressed in future dedicated studies.
Despite these limitations, this rigorously designed prospective, double-blind, randomized controlled trial provides valuable evidence for the relative benefit of LB to ropivacaine in TPVB. Both groups received identical standardized adjunctive analgesia, ensuring that between-group differences can be validly attributed to the study drugs. The core conclusion—that LB offers advantages in postoperative analgesic efficacy and opioid-sparing effect—remains robust and reliable.
Conclusions
For thoracoscopic lung tumor surgery, PVB with LB offers a superior postoperative analgesic effect as compared with ropivacaine, particularly for dynamic pain (coughing/deep breathing) throughout the 72-hour postoperative period, and it significantly reduces late-phase (48–72 hours) and cumulative opioid consumption without increasing the incidence of adverse events. These findings support the use of LB as an effective and safe component of multimodal analgesia, aligning with enhanced recovery protocols.
Supplementary
The article’s supplementary files as
Acknowledgments
None.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Clinical Trial Ethics Committee of The First Affiliated Hospital, Zhejiang University School of Medicine ([2024] IIT Ethics Approval No. 067). Written informed consent was obtained from all participants prior to enrollment.
Footnotes
Reporting Checklist: The authors have completed the CONSORT reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-2185/rc
Trial Protocol: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-2185/tp
Funding: This study was supported by the Analgesia Action Clinical Research Project, Medical Empowerment Special Fund of the Chinese Red Cross Foundation, and the Young Scientists Fund of the National Natural Science Foundation of China (NSFC Young Scientists Fund No. 82402555).
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-2185/coif). All authors report that this study was supported by the Analgesia Action Clinical Research Project, Medical Empowerment Special Fund of the Chinese Red Cross Foundation, and the Young Scientists Fund of the National Natural Science Foundation of China (NSFC Young Scientists Fund No. 82402555). The authors have no other conflicts of interest to declare.
Data Sharing Statement
Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-2185/dss
References
- 1. Elmore B, Nguyen V, Blank R, et al. Pain Management Following Thoracic Surgery. Thorac Surg Clin 2015;25:393-409.. 10.1016/j.thorsurg.2015.07.005 [DOI] [PubMed] [Google Scholar]
- 2. Ren P, Du Y, He G, et al. Efficacy and safety of general anesthesia combined with paravertebral blockade on postoperative recovery in patients undergoing pulmonary surgery: a systematic review and meta-analysis. J Thorac Dis 2022;14:431-42.. 10.21037/jtd-22-103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Singariya G, Kamal M, Paliwal B. Pain after thoracotomy: Conquered or to be conquered?. Indian J Anaesth 2023;67:S12-4.. 10.4103/ija.ija_90_23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Stammberger U, Steinacher C, Hillinger S, et al. Early and long-term complaints following video-assisted thoracoscopic surgery: evaluation in 173 patients. Eur J Cardiothorac Surg 2000;18:7-11.. 10.1016/s1010-7940(00)00426-7 [DOI] [PubMed] [Google Scholar]
- 5. Bendixen M, Jørgensen OD, Kronborg C, et al. Postoperative pain and quality of life after lobectomy via video-assisted thoracoscopic surgery or anterolateral thoracotomy for early stage lung cancer: a randomised controlled trial. Lancet Oncol 2016;17:836-44.. 10.1016/S1470-2045(16)00173-X [DOI] [PubMed] [Google Scholar]
- 6. Tuminello S, Schwartz RM, Liu B, et al. Opioid Use After Open Resection or Video-Assisted Thoracoscopic Surgery for Early-Stage Lung Cancer. JAMA Oncol 2018;4:1611-3.. 10.1001/jamaoncol.2018.4387 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Brescia AA, Harrington CA, Mazurek AA, et al. Factors Associated With New Persistent Opioid Usage After Lung Resection. Ann Thorac Surg 2019;107:363-8.. 10.1016/j.athoracsur.2018.08.057 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Tong Y, Wei P, Wang S, et al. Characteristics of Postoperative Pain After VATS and Pain-Related Factors: The Experience in National Cancer Center of China. J Pain Res 2020;13:1861-7.. 10.2147/JPR.S249134 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Park SJ, Choi EK. Regional analgesia for postoperative pain control after thoracic surgery: a narrative review. J Yeungnam Med Sci 2025;42:80.. 10.12701/jyms.2025.42.80 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Aiolfi A, Bona D, Shabat G, et al. Is paravertebral block the new standard of care for postoperative analgesia after thoracoscopic surgery?. J Thorac Dis 2024;16:2677-80.. 10.21037/jtd-23-1896 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kanani F, Nugzar R, Shimonov M, et al. Comparative Analysis of Perioperative Analgesia Methods in Thoracic Surgery: A Literature Systemic Review. J Clin Med 2025;14:2484.. 10.3390/jcm14072484 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Yeung JH, Gates S, Naidu BV, et al. Paravertebral block versus thoracic epidural for patients undergoing thoracotomy. Cochrane Database Syst Rev 2016;2:CD009121.. 10.1002/14651858.CD009121.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Qi X, Li Z, Zhou L, et al. Comparison of the efficacy and safety of thoracic epidural and paravertebral block in postoperative analgesia after thoracic surgery: a meta-analysis of randomized trials. Front Med (Lausanne) 2026;13:1747430.. 10.3389/fmed.2026.1747430 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Goravanchi F, Kee SS, Kowalski AM, et al. A case series of thoracic paravertebral blocks using a combination of ropivacaine, clonidine, epinephrine, and dexamethasone. J Clin Anesth 2012;24:664-7.. 10.1016/j.jclinane.2012.05.005 [DOI] [PubMed] [Google Scholar]
- 15. Grant GJ, Bansinath M. Liposomal delivery systems for local anesthetics. Reg Anesth Pain Med 2001;26:61-3.. 10.1053/rapm.2001.19166 [DOI] [PubMed] [Google Scholar]
- 16. DeClaire JH, Aiello PM, Warritay OK, et al. Effectiveness of Bupivacaine Liposome Injectable Suspension for Postoperative Pain Control in Total Knee Arthroplasty: A Prospective, Randomized, Double Blind, Controlled Study. J Arthroplasty 2017;32:S268-71.. 10.1016/j.arth.2017.03.062 [DOI] [PubMed] [Google Scholar]
- 17. Keller DS, Pedraza R, Tahilramani RN, et al. Impact of long-acting local anesthesia on clinical and financial outcomes in laparoscopic colorectal surgery. Am J Surg 2017;214:53-8.. 10.1016/j.amjsurg.2015.10.035 [DOI] [PubMed] [Google Scholar]
- 18. Weksler B, Sullivan JL, Schumacher LY. Randomized trial of bupivacaine with epinephrine versus bupivacaine liposome suspension in patients undergoing minimally invasive lung resection. J Thorac Cardiovasc Surg 2021;161:1652-61.. 10.1016/j.jtcvs.2020.01.112 [DOI] [PubMed] [Google Scholar]
- 19. Erotocritou M, Krishna V, Khajuria A. 751 Efficacy of Liposomal Bupivacaine (Exparel) for Post-Operative Analgesia in Plastic Surgery: A Systematic Review. Br J Surg 2024;111:znae163. [Google Scholar]
- 20. Chi Y, Su X, Liu S, et al. Liposomal bupivacaine intercostal nerve block for pain control in thoracoscopic surgery: a randomized controlled trial. Front Med (Lausanne) 2025;12:1647324.. 10.3389/fmed.2025.1647324 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Li H, Wang B, Wang F, et al. Effects of epidural analgesia with different concentrations of bupivacaine plus fentanyl on pain in patients undergoing thoracic surgery. Int J Clin Exp Med 2015;8:14123-6. [PMC free article] [PubMed] [Google Scholar]
- 22. Rice DC, Cata JP, Mena GE, et al. Posterior Intercostal Nerve Block With Liposomal Bupivacaine: An Alternative to Thoracic Epidural Analgesia. Ann Thorac Surg 2015;99:1953-60.. 10.1016/j.athoracsur.2015.02.074 [DOI] [PubMed] [Google Scholar]
- 23. Hartrick CT, Kovan JP, Shapiro S. The numeric rating scale for clinical pain measurement: a ratio measure?. Pain Pract 2003;3:310-6.. 10.1111/j.1530-7085.2003.03034.x [DOI] [PubMed] [Google Scholar]
- 24.ClinCalc LLC. Equivalent Opioid Calculator. Available online: https://clincalc.com/opioids/.
- 25. Myles PS, Myles DB, Galagher W, et al. Measuring acute postoperative pain using the visual analog scale: the minimal clinically important difference and patient acceptable symptom state. Br J Anaesth 2017;118:424-9.. 10.1093/bja/aew466 [DOI] [PubMed] [Google Scholar]
- 26. Tong Y, Wu J, Wu X, et al. Analgesic Efficacy of Thoracoscopic Direct-View Versus Ultrasound-Guided Thoracic Paravertebral Block in Multi-Port Video-Assisted Thoracoscopic Lung Surgery: A Randomized Controlled Non-Inferiority Study. Drug Des Devel Ther 2025;19:1825-38.. 10.2147/DDDT.S492040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Xu X, Xie YX, Zhang M, et al. Comparison of Thoracoscopy-Guided Thoracic Paravertebral Block and Ultrasound-Guided Thoracic Paravertebral Block in Postoperative Analgesia of Thoracoscopic Lung Cancer Radical Surgery: A Randomized Controlled Trial. Pain Ther 2024;13:577-88.. 10.1007/s40122-024-00593-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Hong SB, Hyun K, Choi H. Thoracoscopy-Guided vs. Ultrasound-Guided Paravertebral Block in Thoracoscopic Surgery: A Non-Inferiority Randomized Trial. J Clin Med 2025;14:8493.. 10.3390/jcm14238493 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Chenesseau J, Fourdrain A, Pastene B, et al. Effectiveness of Surgeon-Performed Paravertebral Block Analgesia for Minimally Invasive Thoracic Surgery: A Randomized Clinical Trial. JAMA Surg 2023;158:1255-63.. 10.1001/jamasurg.2023.5228 [DOI] [PMC free article] [PubMed] [Google Scholar]