2.1 Protocol

This study adhered to the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.²² Ethical approval and informed consent were not required, as the study utilized data exclusively from previously published research.

2.2 Eligibility and literature search

Included studies were selected according to the Patients/Intervention/Comparator/Outcome/Study design criteria: (1) “patients” were those undergoing hand or wrist surgery, (2) “intervention” was LB, (3) “comparator” was a placebo or another analgesic, and (4) “outcomes” were postoperative pain scores and opioid consumption.²²

All RCTs without any limitations on language or publication year were included. Observational studies, prospective non-randomized studies, systematic reviews, meta-analyses, conference abstracts, economic analyses, studies unrelated to the research question, and any RTCs not reporting our primary measures were excluded.

A comprehensive search of PubMed, Embase, Cochrane, Web of Science, Scopus and Google Scholar (pages 1–20) until July 2025 was conducted to identify randomized controlled trials (RCTs) evaluating LB use in hand and wrist surgeries. Search terms included combinations of “liposomal bupivacaine” or “Exparel” with “hand surgery,” “wrist surgery,” “distal radius,” “carpal tunnel,” “trigger finger,” “upper extremity surgery,” “analgesia,” “pain,” and “opioid.” Detailed search queries and outcomes are outlined in Appendix 1 (available online on the Journal's website). After completing the initial electronic search, a manual review of relevant articles and their cited references was also performed.

2.3 Study selection

All article titles and abstracts were screened by two independent reviewers (J.A.G and J.J) for eligibility for a thorough full-text review. If the titles and abstracts provided insufficient data for evaluation, the complete article was reviewed. Any disagreements on inclusion were resolved by consensus. No automation tools were used.

2.4 Data extraction

For each RCT, the published year, country, sample size, percent female, mean age, procedure type, intervention details (dose and administration method of LB), control details (dose and administration method), postoperative day (POD) zero to five pain scores (Visual Analog Scale and Numeric Rating Scale scores), POD zero to five opioid consumption, POD one to four Overall Benefit of Analgesia Score (OBAS) if reported, and adverse events were extracted in the form reported by each independent reviewer into a single Microsoft Excel sheet. Missing or incomplete data were requested from primary authors; studies were excluded if no response was received. Opioid consumption was converted to morphine milligram equivalents (MME) and weighted means of pain scores and opioid consumption were calculated for POD 0–5 across all included patients.

2.5 Risk-of-bias assessment

Study quality and potential biases were evaluated using the Cochrane Risk of Bias (RoB 2) tool. Several key domains were analyzed, including random sequence generation, allocation concealment, blinding of participants and study personnel, blinding of outcome assessment, inadequate outcome data, selective reporting, and other potential sources of bias such as baseline imbalances, early stopping, or conflicts of interest.²³ (Fig. 2) Trials were classified as having a high risk of bias if they demonstrated significant concerns in at least one domain. Conversely, studies with a low risk of bias in all domains were categorized as having a low risk of bias. If neither condition was met, the trial was considered to have an unclear risk of bias due to potential limitations in methodology or reporting.²³

Close

Fig. 1 PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart for study selection.

Export to PPT

Close

Fig. 2 Cochrane risk of bias for included studies.

Export to PPT

2.6 Certainty of evidence assessment

A formal certainty of evidence assessment using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was not performed because only four RCTs were included, with substantial heterogeneity and variability in reported outcomes, which limited the feasibility of applying GRADE criteria meaningfully in this review.²⁴

2.7 Data synthesis and statistical analysis

Weighted means of pain scores and opioid consumption were calculated for POD 0–5 across all included patients. For the two studies that reported OBAS, weighted means were calculated for POD 1–4. Subgroup analyses compared outcomes between CMC arthroplasty and distal radius ORIF. A 95 % confidence interval (CI) was applied, and a random-effects model was used when significant heterogeneity was detected (p ≤ 0.05 on the Q test or I² > 50 %). Statistical analysis was conducted using Review Manager 5.4 (The Cochrane Collaboration) to compute mean differences (MD) and standardized mean differences (SMD) between LB and control groups.²⁵ SMD was used to standardize pain score metrics when different scales were reported. Results were visualized using forest plots.

3.1 Systematic review

The initial literature search identified 2308 articles. After removing 1307 duplicates, 1001 unique studies remained. Title and abstract screening excluded 988 articles, leaving 13 studies for full-text evaluation. Following full-text review, nine studies were excluded, resulting in four studies that met the inclusion criteria.^12–15 (Fig. 1).

3.2 Study and patient characteristics

All four studies included were RCTs, with three conducted in the United States,^12–14 and one in China.¹⁵ A total of 211 patients were included, with 105 in the LB group and 106 in the control group. The mean age across studies was 60 years (range: 57–65), and 63 % of participants were female. Ninety patients (LB = 45, Control = 45) underwent CMC arthroplasty,^12,13 and 121 (LB = 60, Control = 61) underwent distal radius ORIF.^14,15 Among the 105 patients in the LB group, 65 received local LB injections at the surgical site,^12–14 while 40 received a supraclavicular LB block.¹⁵ Pain was reported using the Numeric Rating Scale (NRS) in three studies,^12,13,15 and Visual Analog Scale (VAS) in one study.¹⁴ Opioid consumption was measured as the number of pills consumed per day in all studies.^12–15 OBAS was reported in two studies from POD 1–4.^12,15 Pain and opioid consumption were measured through POD 4 in one study,¹² POD 5 in two studies,^13,14 and POD 7 in one study.¹⁵ Only one study reported a complication, nausea/dizziness on POD 1 in the control group.¹⁵ Study characteristics are summarized in Table 1.

3.3 Meta-analysis

Postoperative Pain Outcomes: Pooled and subgroup analyses comparing CMC arthroplasty and distal radius ORIF for POD 0–5 pain scores were conducted (Table 2, Appendix 2Fig. 1A–F). Across all time points, there were no statistically significant differences in pain scores between LB and control groups. The pooled SMD on POD 0 was −1.06 (95 % CI: −2.82, 0.70; p = 0.24, I² = 97 %). The pooled SMD on POD 1 was −1.73 (95 % CI: −3.74, 0.28; p = 0.09, I² = 97 %). The pooled SMD on POD 2 was −0.44 (95 % CI: −1.54, 0.66; p = 0.43, I² = 93 %). The effect size continued to diminish on subsequent days (POD 3: −0.24 [−0.78, 0.31]; p = 0.40, I² = 74 %, POD 4: −0.34 [−1.22, 0.54]; p = 0.44, I² = 89 %, POD 5: −0.38 [−1.50, 0.75]; p = 0.51, I² = 91 %). Subgroup analyses showed no significant differences between CMC arthroplasty and distal radius ORIF.

Postoperative Opioid Consumption Outcomes: Pooled and subgroup analyses for opioid consumption from POD 0–5 were conducted (Table 3, Appendix 2Fig. 2A–F). Across all time points, no significant differences in opioid consumption were observed between LB and control groups. On POD 0, the pooled MD was −0.01 (95 % CI: −0.80, 0.79; p = 0.99, I² = 81 %), suggesting no reduction in opioid use in the LB group. On POD 1, the MD was −3.02 (95 % CI: −6.53, 0.49; p = 0.09, I² = 89 %), reflecting a trend toward reduced opioid consumption in the LB group that did not reach statistical significance. No meaningful differences were seen on POD 2 (MD: −1.10 [95 % CI: −4.17, 1.97]; p = 0.48, I² = 91 %), POD 3 (MD: −0.04 [95 % CI: −0.54, 0.46]; p = 0.87, I² = 0 %), POD 4 (MD: 0.07 [95 % CI: −0.54, 0.67]; p = 0.83, I² = 64 %), or POD 5 (MD: −0.10 [95 % CI: −0.27, 0.07]; p = 0.25, I² = 91 %). No significant differences were found between CMC arthroplasty and distal radius ORIF groups.

Postoperative Overall Benefit of Analgesia Outcomes: Pooled and subgroup analyses for OBAS scores from POD 1–4 were conducted for two studies (Table 4, Appendix 2 Fig. 3A–D). LB was associated with significantly lower OBAS scores on POD 2 and 3, indicating reduced patient satisfaction with pain management compared to standard anesthetic techniques. On POD 1, there was no significant difference between groups (MD: −0.08 [95 % CI: −0.67, 0.52]; p = 0.80, I² = 0 %). However, by POD 2, the pooled MD was −0.89 (95 % CI: −1.46, −0.32; p = 0.002, I² = 0 %), suggesting lower satisfaction in the LB group. This trend persisted on POD 3 (MD: −0.91 [95 % CI: −1.43, −0.40]; p = 0.0005, I² = 11 %) but was no longer significant by POD 4 (MD: −0.22 [95 % CI: −2.13, 1.69]; p = 0.82, I² = 69 %). The decline in OBAS was more pronounced in distal radius ORIF patients compared to CMC arthroplasty patients.

4.1 Strengths and limitations

This systematic review is the first to synthesize evidence specifically on the use of LB for postoperative pain management in hand and wrist surgery. A strength of this analysis is the inclusion of only RCTs and the use of standardized methods for risk-of-bias assessment and quantitative synthesis.

This study has several limitations. First, only four RCTs met the inclusion criteria, limiting both the statistical power and generalizability of our findings. Second, substantial heterogeneity was observed in pain scores and opioid consumption outcomes, especially during the early postoperative period (POD 0–2). This variability likely stems from differences in surgical techniques, anesthetic protocols, pain assessment methodologies, timing of pain assessments, inconsistently reported concurrent use of non-opioid analgesics, inconsistent reporting of total opioids prescribed and dosing schedules, and varying thresholds for analgesic interventions used across studies. Third, the mechanism of LB administration varied significantly among the included trials. Three studies employed local infiltration techniques, while one study by Chan et al. utilized a supraclavicular nerve block, reporting improved early postoperative pain control but lower patient satisfaction scores.¹⁵ This finding suggests that alternative administration routes may affect drug distribution, analgesic duration, and overall efficacy differently. Moreover, limited detail was provided regarding the specific anatomical sites targeted during local LB infiltration. Given the complex innervation of the CMC joint and distal radius, incomplete coverage of relevant nerve branches may have contributed to variability in analgesic effectiveness. These disparities in administration methods and the resulting inconsistencies limit our ability to clearly determine the isolated effects of LB independent of technique. Additional studies are required to specifically assess the efficacy of LB administered via supraclavicular nerve blocks compared to local infiltration methods.

Moreover, while LB offers extended analgesia lasting up to 72 h, the potential for prolonged paresthesias raises questions about its impact on patient satisfaction, an area that remains insufficiently studied. In a multicenter safety evaluation, Ilfeld et al. found that LB was generally well tolerated in peripheral nerve blocks, though some patients experienced prolonged sensory changes, including hypoesthesia and numbness.³⁵ These events were typically mild and resolved over time; however, the study did not assess how these symptoms affected patient-reported satisfaction. Similarly, Eberle and Newman evaluated patient perceptions of postoperative pain after LB use in plastic surgery and found favorable ratings for overall pain control, but did not specifically examine the impact of extended numbness or sensory disturbance.³⁶ Further research examining patient-reported outcomes specifically related to sensory disturbances from LB is warranted to better understand this relationship and its implications for clinical practice.

Another frequently cited limitation of LB is its significantly higher and sometimes controversial cost compared to standard bupivacaine, raising concerns about its overall economic value, particularly given its marginal or inconsistent clinical benefits. A single dose of LB may cost approximately $300, whereas an equivalent dose of plain bupivacaine ranges from $5 to $30​.^37,38 While this cost disparity is notable, concerns have been somewhat alleviated by Medicare's 2019 decision to reimburse its use in ambulatory surgery centers, though coverage remains variable across states and facilities.³⁹ Additionally, some studies have demonstrated that LB can reduce overall hospital expenditures in procedures such as spine, hip, and knee surgeries, primarily by shortening hospital stays and facilitating earlier rehabilitation.^33,40 However, in the context of hand and wrist surgery, where procedures are often outpatient and recovery is typically rapid, these downstream cost benefits are less applicable. As such, the high upfront cost of LB is difficult to justify in this setting, particularly in the absence of clear superiority over conventional analgesics.