Introduction

Effective perioperative analgesia is essential for optimizing enhanced recovery after arthroscopic shoulder surgery.1 ISB is a well-established technique for postoperative pain control following shoulder surgery,2–4 as it effectively reduces opioid consumption and promotes early functional rehabilitation. However, due to the anatomical adjacency of the ISB injection site to the phrenic nerve, inadvertent cephalad diffusion of local anesthetic frequently induces phrenic nerve blockade, leading to hemidiaphragmatic paralysis and postoperative dyspnea.3,5–7 Such respiratory complications are particularly unfavorable for patients with impaired respiratory reserve, such as those with chronic pulmonary disease or obesity, who are more susceptible to postoperative respiratory deterioration.8,9 In addition, studies have reported that the incidence of rebound pain after single-shot ISB in patients undergoing arthroscopic shoulder surgery is as high as 82.9%.10

The ITMB is a novel regional block technique proposed based on recent anatomical investigations.11 Cadaveric staining experiments demonstrated that the dye injected via all 10 ITMB procedures diffused bidirectionally along the fascial plane: cephalad infiltration extended to the infraspinatus fossa and spinoglenoid notch to fully surround the lateral trunk of the suprascapular nerve, whereas caudal spread penetrated the quadrilateral space to completely bathe the main trunk of the axillary nerve. Dye extravasation into the posterolateral glenohumeral joint capsule was observed in 8 specimens, and subacromial staining was noted in 4 specimens; no contrast solution diffused to extra-articular branches innervating the anterosuperior shoulder, including the subscapular nerve and lateral pectoral nerve.11 Anatomically, the spinoglenoid ligament and scapular spine together form a fibro-osseous tunnel that acts as a natural anatomical barrier.12 This structure restricts excessive cephalad spread of local anesthetic toward the supraspinatus fossa and facilitates drug accumulation within the pericapsular space. Meanwhile, injectate fills the quadrilateral space in a posterior-to-anterior direction to achieve complete axillary nerve blockade. A spacious loose pericapsular plane lies deep to the infraspinatus and teres minor muscles. Following ITMB injection, local anesthetic simultaneously blocks the articular branches of both the suprascapular and axillary nerves, and pericapsular drug diffusion further augments analgesia of the shoulder. Compared with the conventional two-point combined suprascapular and axillary (SSAX) nerve block, ITMB accomplishes blockade of both nerves with a single puncture, simplifying procedural performance. Additionally, its posterior injection site is anatomically distant from the phrenic nerve, which theoretically substantially reduces the risk of hemidiaphragmatic paralysis frequently encountered with ISB. Previous randomized studies and case reports have demonstrated the analgesic efficacy of ITMB for shoulder surgery.13,14 However, direct comparisons between ITMB and ISB, the current reference standard for shoulder analgesia, remain limited. Therefore, further randomized evidence is needed to directly compare these two techniques with respect to analgesic efficacy, opioid consumption, and phrenic nerve-sparing effects. Local anesthetic infiltration within this intermuscular fascial plane reliably blocks both the suprascapular and axillary nerves, yielding satisfactory postoperative analgesia while markedly lowering the incidence of respiratory adverse events secondary to diaphragmatic dysfunction.

Given the limited direct comparative evidence between ITMB and ISB, we conducted a prospective, participant-and outcome assessor-blinded, two-arm parallel-group randomized non-inferiority trial to directly compare the postoperative analgesic efficacy and safety profiles of ITMB versus ISB in patients undergoing shoulder arthroscopy, with particular emphasis on opioid consumption and phrenic nerve- sparing effects. The primary hypothesis was that the 24-hour postoperative oxycodone consumption in the ITMB group would not exceed that in the ISB group by a non-inferiority margin of 5 mg.

Methods

This prospective, participant- and outcome-assessor-blinded, two-arm parallel randomized non-inferiority trial aimed to evaluate whether ITMB yields non-inferior postoperative analgesia relative to ISB in patients undergoing arthroscopic shoulder surgery.

The trial received ethical approval from the Ethics Committee of Nanjing First Hospital (IRB reference KY20240123-09), and written informed consent was acquired from every enrolled participant prior to any study-related procedures. This trial was prospectively registered on the Chinese Clinical Trial Registry before participant recruitment commenced (registration ID: ChiCTR2400084716, principal investigator: Yong Zhang, registration date: 23 May 2024; registration link: https://www.chictr.org.cn/hvshowproject.html?id=285004&v=1.3). All study procedures were performed in strict accordance with the pre-specified trial protocol, and the manuscript was reported in line with the CONSORT 2010 guidelines (eTable 1 in Supplement 1). A dedicated Data and Safety Monitoring Board oversaw trial progress and periodically reviewed masked safety data throughout recruitment and follow-up.

Eligible candidates were adults aged 18–65 years scheduled for shoulder arthroscopy, with an American Society of Anesthesiologists (ASA) physical status of I–III and a body mass index (BMI) ranging from 18 to 30 kg/m2. Exclusion criteria comprised contraindications to peripheral nerve blocks, severe chronic pulmonary disease, chronic opioid consumption (oral morphine equivalent dose (OMED) ≥ 5 mg per day for ≥ 1 month or continuous opioid use exceeding 3 months), hepatic or renal insufficiency, pregnancy, pre-existing cervical disc herniation, ipsilateral upper limb neuropathy, cervical myelopathy, prior receipt of alternative perioperative analgesia regimens, or inability to adhere to trial protocols. All eligible patients attended a preoperative consultation at least 24 hours before surgery, where trial risks and procedures were fully explained. Baseline shoulder pain severity was assessed using the 11-point Numeric Rating Scale (NRS-11; 0=no pain, 10=worst pain imaginable).

For statistical analysis, two analytical populations were defined in accordance with the pre-specified protocol. The PP cohort consisted only of participants who fully adhered to all trial procedures and possessed complete primary endpoint data (24-hour total oxycodone consumption). Predefined criteria for exclusion from the PP set were: (1) unplanned intraoperative conversion; (2) unplanned intensive care unit (ICU) admission secondary to adverse events unrelated to the assigned nerve block, resulting in incomplete primary outcome documentation; (3) failure to receive the allocated regional block intervention; (4) complete missingness of 24-hour postoperative oxycodone consumption data. Consistent with CONSORT recommendations, all randomly assigned patients—regardless of PP eligibility—were retained in the ITT population for the primary outcome and all safety-related analyses.

Randomization and Blinding

Randomization sequences were produced using SPSS version 22.0 by an independent researcher excluded from patient recruitment, block procedures and data acquisition. Eligible patients were allocated in a 1:1 ratio using block randomization with a block size of 2, receiving either ISB or ITMB. Sequentially numbered, opaque sealed allocation envelopes were exclusively prepared by this independent statistician. Each envelope was opened only by the anesthesiologist performing nerve blocks immediately before block administration, after the patient was placed in the lateral decubitus position and sedated. All nerve blocks were performed by a single experienced anesthesiologist (ZC) who took no part in intraoperative anesthesia management or outcome evaluation to preserve blinding integrity. Multiple standardized measures were implemented to maintain blinding of participants: all patients were placed in the lateral decubitus position for block procedures; sedation regimens (midazolam combined with sufentanil), intraoperative monitoring, and skin disinfection protocols were identical across groups. Sterile drapes covered patients’ heads to prevent them from viewing the ultrasound screen and procedural operations. During block performance, the operator and assistant only communicated using neutral, standardized language to avoid revealing group allocation. Intraoperative general anesthesia was administered by an independent attending anesthesiologist who was unaware of treatment assignments and did not participate in nerve block execution or data collection. All patients, outcome assessors, ward nurses, surgeons, and anesthetic staff excluding the block operator remained blinded to group allocation throughout the entire trial period.

Preoperative Preparation

Patients were placed in a lateral position with their surgical side up during nerve block procedures, and they could either drape the arm across their chests or leave it hanging naturally. All patients received 1 mg midazolam and 5 µg sufentanil for sedation with standard ASA monitors and supplemental oxygen.

Ultrasound-Guided Infraspinatus-Teres Minor Interfascial Block

Infraspinatus-teres minor interfascial blocks were performed using the technique described by Kim SH.11 The patient was placed in the lateral decubitus position, and a pillow was wedged under the anterior shoulder and upper arm to avoid the over protraction of the scapula. Under aseptic conditions, a high-frequency linear ultrasound probe (6–13 MHz, Sonosite) was positioned over the humeral head in the sagittal plane to visualize the infraspinatus and teres minor muscles. The probe was then swept mediocaudally to track the interfascial plane between these two muscles alongside the glenohumeral joint and scapular neck. We chose the interfascial plane between infraspinatus and teres minor on the posterior surface area of the neck of the scapula for the target site of the ITMB because this area is located between the spinoglenoid notch and quadrilateral space. A 22 G 2-inch block needle (B. Braun Medical Inc.) was inserted after identifying infraspinatus and teres minor on the posterior surface of the scapula’s neck, and an in-plane technique under real-time ultrasound guidance was used to advance the needle towards the interfascial plane between the infraspinatus and teres minor, then 25 mL ropivacaine 0.375% was administered into infraspinatus teres minor interfascial space as divided dose (Figure 1).

Figure 1 Ultrasound image and corresponding anatomical illustration of infraspinatus-teres minor interfascial block. (A) Red arrowheads indicate the interfascial plane between the two muscles, and white arrowheads indicate the needle trajectory direction. (B) The dashed line represents the placement of the ultrasound probe, and the colored regions highlight the relevant anatomical structures.IS = infraspinatus; S = scapula; Tm = teres minor.

Ultrasound-Guided ISB

The patient was placed in the lateral decubitus position, and a high-frequency linear ultrasound probe (6–13MHz, Sonosite) was placed at the level of the cricoid cartilage at the level of the interscalene groove. After scanning and confirming the C5 and C6 nerve roots, 25 mL of 0.375% ropivacaine hydrochloride was injected between the C5-6 nerve roots using the needle from back to front with in-plane technique.

Sensory Block and Recovery Quality Assessment

Thirty minutes after block completion, sensation to ice was assessed on the top of the shoulder (axillary nerve) and the posterior surface of the scapula (suprascapular nerve) bilaterally at the deltoid area using a 4-point numeric scale (0, no cold sensation; 1, severely decreased cold sensation; 2, mildly decreased cold sensation; 3, normal cold sensation). The sensory score < 3 was considered a successful block. A block failure was considered if none of the deltoid area had decreased sensation at the surgical side. Recovery quality was assessed with the Chinese version of the 15-item quality of recovery (QoR-15) scale,15,16 a well-validated patient outcome measure after surgery. The QoR-15 scale was assessed the day before surgery and 24 hours after surgery.

Perioperative Management

All patients received standard general anesthesia care during the intraoperative period. Neuromuscular blocking agents, airway management, and ventilation were managed at the anesthesiologist’s discretion. For analgesia, flurbiprofen axetil 50 mg was injected. All patients received an intraoperative intravenous combination of 10 mg dexamethasone, 4 mg tropisetron and 1 mg droperidol to prevent postoperative nausea and vomiting. Following skin closure, patients were transferred to the post-anesthesia care unit (PACU). The extubation time and the length of stay in PACU were recorded. All patients received an intravenous 0.5 mg/kg oxycodone patient- controlled analgesia (PCA) pump (volume 150 mL, bolus=5 mL, lockout interval=8 min, no basal infusion, maximum dose 35 mL/h). The pump was initiated when NRS score≥4 or per the patient’s request. In cases where the NRS score persisted at ≥4, intravenous oxycodone 1 mg bolus was delivered as physician bolus. Discharge from the PACU required a modified Aldrete score of >9.

Diaphragm Function Assessment

Diaphragmatic excursion was assessed at two time points: prior to nerve block implementation and 30 minutes after block completion. All subjects received identical standardized sedation with midazolam combined with sufentanil throughout the ultrasound assessment. All diaphragmatic function evaluations were performed by DK, who was fully blinded to group allocation. Patients were positioned supine, and a 3.5–5 MHz convex Sonosite ultrasound probe was placed along the anterior axillary and midclavicular lines. The right hemidiaphragm was visualized via the liver acoustic window, while the left hemidiaphragm was visualized through the spleen. M-mode ultrasonography was utilized to quantify craniocaudal diaphragmatic excursion during standardized deep breaths. Each discrete measurement captured the full diaphragmatic excursion of one complete respiratory cycle. Three independent replicate measurements were acquired per patient, and the maximum excursion value across the three single-cycle recordings was extracted for statistical analysis. A relative reduction in diaphragmatic movement of > 25% compared to baseline was defined as hemidiaphragmatic paralysis. Partial hemidiaphragmatic paralysis was classified as a diaphragmatic excursion of 25–75% of the pre-block baseline value, whereas complete hemidiaphragmatic paralysis was defined as excursion less than 25% of baseline.17,18

Primary Outcome

The primary outcome was total oxycodone consumption (all rescue oxycodone boluses were included) within 24 hours postoperatively. The hypothesis was that the 24-hour oxycodone consumption in the ITMB group after surgery would not exceed that of the ISB group by more than 5 mg, demonstrating non-inferiority of ITMB compared with ISB in acute pain management for Arthroscopic Shoulder Surgery.

Secondary Outcomes

Data on the following secondary outcomes were recorded: NRS-11 scores at 1, 6, 24, and 48 hours postoperatively. Oxycodone consumption at 0–12, 12–24, and 24–48 hours postoperatively. The following data were also collected within the first 24 hours:the worst resting NRS pain score within 24 hours, the incidence of rebound pain (defined as the transition from well-controlled pain (11-point NRS ≤ 3) while the block is working to severe pain (NRS ≥ 7) within 24 hours of block performance),19 the incidence of diaphragmatic paralysis, the number of patients with NRS ≥ 4, the number requiring rescue analgesia, the number of patients experiencing dyspnea (defined as shortness of breath as described by patient20), QoR-15 score, the incidence of nausea and vomiting, the incidence of Horner’s syndrome and complications related to the block. Other endpoints included the extubation time and the length of stay in PACU. All data collectors received training before the study initiation.

Sample-Size Determination

This trial was designed to test non-inferiority with respect to the primary endpoint, 24-hour postoperative oxycodone consumption. On the basis of preliminary data, mean 24-hour postoperative oxycodone consumption was 15.0 (SD 4.4) mg for patients receiving ISB and 17.2 (SD 3.8) mg for patients allocated to ITMB. A recent scoping review established a provisional between-group minimal clinically important difference (MCID) of 5 mg intravenous morphine equivalents for cumulative rescue opioid consumption within the first 24 h after acute pain surgery.21 Per standard opioid conversion criteria, intravenous oxycodone exhibits a 1:1 analgesic equivalence ratio to intravenous morphine.22 Accordingly, we predefined a non-inferiority margin of 5 mg for oxycodone. We reasoned that a maximum 5 mg higher total 24-hour oxycodone consumption in the ITMB group would not translate to a clinically meaningful loss of analgesic efficacy, while the technique’s key clinical advantage of reduced phrenic nerve-related adverse events would still be retained. Sample-size calculations were performed with PASS 15 software, assuming a one-sided type I error of 0.025 and 80% power. Under these assumptions, 70 patients were required. To allow for an anticipated attrition rate of 15%, the target enrollment was 82 patients, assigned in a 1:1 ratio (41 per group).

Statistical Analysis

The primary outcome was assessed using a one-sided non-inferiority test by comparing the 95% CI of the between-group mean difference (ISB minus ITMB) in total oxycodone consumption within 24 h postoperatively with the predefined non- inferiority margin, such that negative values indicate greater oxycodone consumption in the ITMB group. A one-sided P<0.025 was considered statistically significant. To evaluate the robustness of the non-inferiority conclusion, sensitivity analyses were performed using more conservative non-inferiority margins of −3 mg and −2 mg, in addition to the prespecified margin of −5 mg. For continuous variables, normality was evaluated using the Shapiro–Wilk test and homogeneity of variance was assessed with Levene’s test. Continuous variables are reported as mean (SD) or median (IQR). Normally distributed data were presented as mean (SD) and analyzed using independent samples t-tests. Non-normally distributed data were expressed as median (IQR [range]) and compared using Mann–Whitney U-test. We calculated the median difference (95% CI) with the Hodges–Lehmann estimator. Categorical variables are described as numbers (proportion) and analyzed with χ2 tests or Fisher’s exact tests, as appropriate. Linear mixed-effects models (LMMs) were constructed to analyze longitudinally collected measurements of oxycodone consumption, surgical-side diaphragmatic excursion, and serial resting pain scores. Group, time, and the group-by-time interaction term were specified as fixed effects, while the patient was included as a random intercept to account for within-subject correlation. Model assumptions including normality of residuals and homogeneity of variance were verified prior to statistical inference. Outcomes were presented as adjusted marginal mean differences together with their corresponding 95% CI and P values for the group-by-time interaction effect. Analyses of secondary outcomes were considered exploratory, P<0.05 was considered statistically significant. Statistical analysis followed the PP principle, with ITT analysis additionally performed for primary outcomes and safety endpoints. One patient required ICU admission due to intraoperative antibiotic-related anaphylaxis and had partially missing data for 24- hour oxycodone consumption. This patient remained included in the ITT analysis, and the missing oxycodone consumption data were imputed using the mean value of the corresponding treatment group for the ITT sensitivity analysis. Although not initially planned in the protocol, we carried out a post-hoc exploratory analysis of the QoR-15 score 24 h after surgery. SPSS 22.0 version (IBM SPSS Inc, Chicago, IL, USA) was used for analysis.

Results

Between May 29, 2024, and December 26, 2024, a total of 100 patients were assessed for eligibility. Of these, 18 were excluded due to preexisting medical conditions (n=12), refusal to participate (n=3), or language barriers (n=3) (Figure 2). The remaining 82 patients were randomized, with 41 patients allocated to the ISB group and 41 to the ITMB group, and all underwent the assigned nerve block procedure. Two participants assigned to the ITMB group were excluded from the PP analysis. One patient underwent intraoperative conversion to open surgery, representing a major protocol deviation, while the other required intensive care unit (ICU) admission due to intraoperative antibiotic-induced anaphylaxis with transient hemodynamic instability. This adverse event was considered unrelated to the ITMB block. The patient’s complete safety data were retained in the ITT safety analysis. After these exclusions, the PP population consisted of 41 patients in the ISB group and 39 patients in the ITMB group. Baseline characteristics, including age, sex, BMI, ASA physical status, and preoperative QoR-15 scores, were comparable between groups (Table 1). The results of the ITT analysis, including all 82 randomized patients (41 in each group), were consistent with those of the PP analysis (eTable 2 in Supplement 1).

Table 1 Patient Characteristics and Baseline Assessments

Figure 2 Flow diagram of patient enrollment.

Primary Outcome

The oxycodone consumption within 24 hours postoperatively was (mean [SD]) 11.26 [4.38] mg (95% CI:9.89 to 12.65) in the ISB group and 10.92 [3.48] mg (95% CI: 9.79 to 12.05) in the ITMB group. The mean difference between groups was 0.34 mg (95% CI: −1.42 to 2.11), which did not exceed the predefined non-inferiority margin of −5 mg (ISB minus ITMB)(P<0.001 for non-inferiority) (Table 2 and Figure 3). Using narrower non-inferiority margins of −3 mg and −2 mg, the lower boundary of the 95% CI for the treatment difference (ISB minus ITMB) remained above both margins (−1.42 mg vs −3 mg and −2 mg), confirming the robustness of the non-inferiority conclusion. ITT analysis: The mean 24 h oxycodone consumption was (mean [SD]) 11.26 [4.38] mg (95% CI, 9.89 to 12.65) in the ISB group and 11.00 [3.43] mg (95% CI, 9.92 to 12.09) in the ITMB group (mean difference: 0.26 mg; 95% CI: −1.46 to 1.99; P < 0.001 for non-inferiority) (Table 2 and Figure 3). Using narrower non-inferiority margins of −3 mg and −2 mg, the lower boundary of the 95% CI for the treatment difference (ISB minus ITMB) remained above both margins (−1.46 mg vs −3 mg and −2 mg), confirming the robustness of the non-inferiority conclusion.

Table 2 Primary Outcome Analysis: 24-Hour Oxycodone Consumption (Mg) in the per-Protocol (PP) and Intention-to-Treat (ITT) Populations

Figure 3 Noninferiority comparison in the per-protocol (PP) and intention-to-treat (ITT) populations. Mean difference (95% confidence interval) for oxycodone consumption within 24 hours postoperatively. ITMB is noninferior to ISB for the oxycodone consumption within 24 hours postoperatively. All between-group differences are calculated as ISB group value minus ITMB group value. The dashed black line designates the noninferiority margin of −5 mg on oxycodone consumption. The red error bars designate the 95% CI of the difference between the groups. (a) Noninferiority comparison in the per-protocol (PP) populations. (b) Noninferiority comparison in the intention-to-treat (ITT) populations. ITMB=Infraspinatus-teres minor interfascial block; ISB=interscalene plexus block.

Secondary Outcomes Pain Scores

Linear mixed-effects models were used to analyze postoperative resting NRS pain scores and oxycodone consumption in both groups. The model revealed statistically significant main effects of time and significant group-by-time interaction (both P<0.001). At postoperative 1 h, the adjusted marginal mean difference (ISB minus ITMB) in resting NRS score was −1.55 (95% CI: −1.9 to −1.2, P<0.001), indicating superior early analgesia in the ISB group. At postoperative 6 h, the between-group mean difference was −1.2 (95% CI: −1.5 to −0.9, P<0.001), and the ISB group still exhibited lower pain intensity. The analgesic advantage reversed at 24 h postoperatively: the mean difference was 0.7 (95% CI: 0.4 to 1.0, P<0.001), demonstrating better pain control in the ITMB group. No significant intergroup difference was detected at 48 h, with a mean difference of 0.1 (95% CI: −0.2 to 0.4, P=0.426). For the 0–12 h, the adjusted marginal mean difference (ISB minus ITMB) in oxycodone consumption was −3.9 mg (95% CI: −4.6 to −3.3, P<0.001), which meant the ISB group consumed less rescue opioid in the early postoperative stage. During 12–24 h, the mean difference reversed to 3.80 mg (95% CI: 3.1 to 4.5, P<0.001), showing lower oxycodone demand in the ITMB group. No intergroup difference was found from 24 to 48 h, with a mean difference of 0.2 mg (95% CI: −0.6 to 0.9, P=0.632) (eTable 3 in Supplement 1). The worst resting NRS pain score within 24 h postoperatively was lower in the ITMB group (median[IQR]: 3.0[3.0−4.0]) compared to the ISB group (4.0[3.0−7.0], median difference:1.0; 95% CI: 0 to 1.0), P<0.001) (Table 3).

Table 3 Secondary Outcomes

Rebound Pain and Rescue Analgesia

The incidence of rebound pain was 2.6% in the ITMB group versus 26.8% in the ISB group (P=0.006). Rescue analgesia requirements were less frequently in the ITMB group (7.7% vs 29.3%, P=0.016)(Table 3).

Diaphragmatic Function and Adverse Events

The incidence of hemidiaphragmatic paralysis was lower in the ITMB group (2.6%) compared to the ISB group (90.2%, P<0.001) (Table 3). Linear mixed-effects models were adopted to compare surgical-side diaphragmatic excursion measured 30 minutes after nerve block. The mean diaphragmatic excursion was 3.6 cm (SD 0.8) in the ITMB group versus 1.9 cm (SD 1.1) in the ISB group, with a marginal mean difference of −1.6 cm (95% CI: −2.1 to −1.2, P<0.001). (eTable 3 in Supplement 1).

Recovery Parameters

Extubation time and PACU stay were shorter in the ITMB group than in the ISB group (extubation time: 14.9 [3.4] vs 19.8 [5.3] minutes; mean difference:4.9 minutes; 95% CI: 2.9 to 6.9, P<0.001; PACU stay: 46.0 [3.8] vs 49.7 [5.9] minutes; mean difference, 3.7 minutes; 95% CI: 1.4 to 5.9; P=0.002) (Table 4).

Table 4 Intraoperative Data

Quality of Recovery

Although total QoR-15 scores at 24 h did not differ between groups ((mean [SD]) 118.9 [10.3] (95% CI:115.6 to 122.1) in the ISB group and 122.7 [9.5] (95% CI:119.7 to 125.8) in the ITMB group, P=0.085) (Table 3), the ITMB group exhibited significantly higher scores in the pain subscale (median:18.0 vs 15.0, median difference: −2.0; 95% CI: −3.0 to −1.0, P<0.001). This exploratory post-hoc analysis only demonstrates a between-group difference limited to the pain subscale of QoR-15, with no predefined hypothesis testing for global recovery outcomes. Any observed association between higher pain-domain scores and total QoR-15 values remains merely speculative, and we cannot draw definitive causal conclusions linking improved pain relief to enhanced comprehensive recovery.

Discussion

In this randomized non-inferiority trial involving patients who received arthroscopic shoulder surgery, ultrasound-guided ITMB achieved non-inferior outcomes for 24 h postoperative opioid consumption versus conventional ISB. Cumulative oxycodone use on postoperative 24 h was similar across groups, with a mean difference of 0.34 mg (95% CI: −1.42 to 2.11; P<0.001 for non-inferiority). The MCID for NRS pain scores has been established as 1 point in prior literature.23 The median difference in the worst resting NRS pain score between the ITMB and ISB groups was 1.0 (95% CI, 0 to 1.0). While statistical significance was observed, the between-group difference did not exceed the predefined MCID, suggesting a lack of clinical importance. In addition to equivalent total 24-hour opioid utilization, ITMB conferred some clinical advantages over ISB: It lowered the rates of early hemidiaphragmatic paralysis (2.6% vs 90.2%, P<0.001) and postoperative rebound pain (2.6% vs 26.8%, P=0.006), while demonstrating statistically lower worst resting NRS pain scores within 24 hours after surgery; however, the magnitude of this difference did not exceed the predefined MCID.

Linear mixed-effects regression revealed prominent group-by-time interactions for serial resting NRS scores and cumulative interval oxycodone consumption, demonstrating markedly disparate time-dependent analgesic profiles between ITMB and ISB. During the acute postoperative phase (0–12 h), ISB provided greater early analgesic efficacy and reduced opioid rescue demand, reflected by significantly lower resting pain intensity and diminished cumulative oxycodone utilization within this early window. This may be explained by anatomical findings11 that the ITMB can simultaneously block the articular branches of the suprascapular and axillary nerves, yet may provide incomplete coverage of the glenohumeral joint, which is innervated by the subscapular nerve. The pronounced short-term analgesic potency of ISB gradually waned 12 h following block performance, accompanied by a reciprocal shift in comparative pain burden: ITMB conferred superior pain relief and less opioid consumption throughout the 12–24 h postoperative interval. These countervailing temporal analgesic effects across the first 24 postoperative hours reciprocally offset one another, culminating in equivalent total 24-hour oxycodone exposure and satisfying the predefined non-inferiority criterion for the primary analgesic endpoint.

A prominent temporal divergence in analgesic profiles emerged between the two block modalities, a pattern consistent with prior comparisons of single-shot ISB versus combined suprascapular and axillary nerve block reported by Dhir et al24. Barry et al19 hypothesized that the near-complete sensory blockade achieved by ISB may contribute to central hyperalgesia after the resolution of local anesthetic effects. This proposed mechanism may partly account for the markedly lower incidence of rebound pain observed in the ITMB group; however, it remains speculative and requires further investigation. Nevertheless, this mechanistic interpretation remains speculative; our trial lacked dedicated testing to validate this pathophysiological cascade, and targeted prospective studies are required to corroborate this causal pathway. One previous RCT evaluating10 rebound pain after arthroscopic shoulder surgery reported incidence rates of 37.1% and 86.9% following ISB with 12 mL 0.5% ropivacaine, with and without perineural dexamethasone, respectively, and most rebound episodes arose 12–18 h post-block. A meta-analysis further noted25 that ISB’s analgesic advantage over isolated suprascapular nerve block is confined to the PACU, whereas our data demonstrated sustained superior pain control with ISB across the initial 6 postoperative hours. The comparatively mild rebound pain observed in our ISB group is likely mediated by our standardized intraoperative 10 mg intravenous dexamethasone regimen, a well-documented intervention to mitigate rebound hyperalgesia.10,19,26 This steroid administration constitutes an important confounding variable; extrapolation of our findings to clinical settings without routine perioperative glucocorticoid prophylaxis requires cautious adjustment for this modifying factor.

In terms of safety, ITMB demonstrated potential advantages. Unintentional phrenic nerve involvement is a well-recognized adverse effect of ISB and may result in symptomatic hemidiaphragmatic paralysis. In contrast, ITMB was associated with a substantially lower incidence of hemidiaphragmatic paralysis at 30 minutes after block placement. However, the large between-group difference in diaphragmatic dysfunction should be interpreted within the context of the anesthetic volume used in this study. We administered a standardized volume of 25 mL of 0.375% ropivacaine for both techniques, which exceeds the volumes commonly used in contemporary low-volume ISB protocols. The relatively high injectate volume may have facilitated cephalad spread toward the phrenic nerve and consequently increased the observed incidence of hemidiaphragmatic paralysis in the ISB group. Consistent with this interpretation, meta-analytic evidence suggests that ISB using injectate volumes ≥15 mL is associated with a substantially increased risk of phrenic nerve blockade. Low-volume ISB techniques using 5–10 mL ropivacaine can reduce the incidence of hemidiaphragmatic paralysis by approximately 38%, although this benefit may be accompanied by shorter analgesic duration and increased postoperative opioid requirements.27 A prospective double-blind randomized controlled trial further demonstrated the volume-dependent respiratory effect of ISB: even with an extrafascial approach, administration of 20 mL of 0.75% ropivacaine resulted in hemidiaphragmatic paralysis in 80% of patients at 30 minutes after the block, whereas reducing the volume to 10 mL decreased this incidence to 19%, with a shorter duration of analgesia and increased 24-hour morphine consumption (20 mg).28

In this trial, the same local anesthetic volume was deliberately applied to both groups to minimize volume-related confounding when evaluating analgesic non-inferiority between ITMB and ISB. Nevertheless, this high-volume regimen likely contributed to greater cephalad local anesthetic spread in the ISB group, which may partly explain the high incidence of hemidiaphragmatic paralysis observed in our control cohort (90.2%). In addition to reduced diaphragmatic dysfunction, no Horner’s syndrome was observed in the ITMB group (0% vs 29% in the ISB group).

We also performed a post hoc exploratory analysis of overall postoperative recovery using the QoR-15 scale. This instrument evaluates five domains—pain, physical comfort, physical independence, psychological support, and emotional status—and provides a comprehensive assessment of perioperative recovery quality.15,16 No significant difference was observed in total QoR-15 scores at 24 hours postoperatively between groups (118.9±10.3 in the ISB group vs 122.7±9.5 in the ITMB group; P=0.085), suggesting comparable overall recovery after the two block techniques. Analysis of individual domains showed that the ITMB group achieved better pain- domain QoR-15 scores (median: 18.0 vs 15.0, median difference: −2.0, 95% CI −3.0 to −1.0, P<0.001), which was consistent with our secondary pain outcomes. Importantly, QoR-15 assessment was a post hoc exploratory endpoint and was not prespecified as a primary outcome. Furthermore, the sample size was calculated solely based on the primary endpoint of 24-hour cumulative oxycodone consumption. Therefore, findings from secondary outcomes should be interpreted as exploratory and require confirmation in adequately powered prospective trials specifically designed with these outcomes as primary endpoints. Accordingly, these results should be considered supportive rather than definitive evidence of improved recovery with ITMB, and further dedicated studies are warranted.

Several limitations of the present trial should be acknowledged. First, this was a single-centre study, and the generalisability of our findings may be limited by the institutional setting and the experience of the block operator. In addition, our eligibility criteria restricted enrollment to patients aged 18–65 years with a BMI of 18–30 kg/m2 and excluded those with severe pulmonary disease. Consequently, patients with limited respiratory reserve, chronic pulmonary disease, or obesity were not represented in this trial. Whether the potential phrenic-sparing advantages of ITMB translate into clinically meaningful benefits in these higher-risk populations remains unknown and should be investigated in future studies. Second, postoperative pain assessments were limited to resting NRS scores without formal evaluation of movement-evoked pain. This approach was consistent with standard orthopedic protocols that restrict aggressive early shoulder mobilization to protect repaired tendinous and soft-tissue structures. Therefore, our analgesic conclusions apply primarily to resting pain during the first postoperative 48 hours. Although ITMB resulted in non-inferior 24-hour oxycodone consumption and lower peak resting pain scores, we cannot determine whether it provides equivalent analgesia during physical therapy or active rehabilitation. Given that ITMB and ISB differ in their nerve coverage and local anesthetic spread patterns, their effects on dynamic pain during functional recovery may differ. Future studies should incorporate standardized early mobilization protocols and assess both resting and movement- evoked pain to comprehensively characterize their analgesic profiles. Third, the 25 mL volume of ropivacaine used for both block techniques differs from contemporary clinical practice, in which lower volumes (approximately 10–15 mL) are commonly used for interscalene and superior trunk blocks. Therefore, the effects of lower local anesthetic volumes on analgesic efficacy and phrenic nerve-sparing properties require further investigation. Fourth, diaphragmatic function was assessed only at a single time point 30 minutes after the block, without serial longitudinal ultrasound monitoring. This early assessment primarily reflects the immediate effects of the blocks on diaphragmatic excursion and may not detect delayed phrenic nerve blockade or characterize the temporal evolution of diaphragmatic weakness, including its onset, peak effect, and spontaneous resolution. Therefore, our conclusions regarding the phrenic-sparing effect of ITMB are limited to the immediate perioperative period. Serial ultrasound assessments over longer follow-up periods are needed to determine the duration and durability of this diaphragmatic-sparing effect. In addition, baseline diaphragmatic excursion was measured after sedation, which may have influenced respiratory patterns and diaphragmatic motion. Furthermore, using the maximum of three measurements rather than the mean may have increased measurement variability and affected the precision of the baseline assessment. Fifth, the sample-size calculation was based on data from an independent single-center pilot trial rather than similar published studies. Although the pilot data were collected independently of the present trial, their single-center origin may have introduced uncertainty and potential optimism in the estimation of the treatment effect, which may have affected the robustness of the sample-size calculation and limited the generalizability of the study findings. Finally, the use of a fixed block size of 2 in the randomisation sequence may have introduced a theoretical risk of allocation predictability. Nevertheless, allocation concealment was maintained using sequentially numbered, opaque, sealed envelopes, which helped minimize this potential risk.

In summary, ultrasound-guided ITMB provides non-inferior 24-hour postoperative opioid analgesia compared with ISB and reduces the incidence of early hemidiaphragmatic paralysis in patients undergoing arthroscopic shoulder surgery. The comparable total 24-hour oxycodone consumption reflects distinct time-dependent analgesic profiles: ISB provides greater analgesic efficacy during the immediate postoperative period, whereas ITMB offers more sustained analgesia thereafter. Given the exploratory nature of secondary outcomes, further adequately powered studies are warranted to confirm the broader clinical benefits of ITMB.

Declaration of Generative AI and AI-Assisted Technologies in the Writing Process

During manuscript drafting, generative AI (ChatGPT) was solely used for linguistic polishing of grammar and sentence structure. All study design, patient enrollment, data collection, statistical analysis, and interpretation were completed independently by the authors without AI involvement in core research content.

Data Sharing Statement

Anonymized raw study data can be obtained from the corresponding author Xiaoliang Wang upon reasonable written request via email: [email protected].

Acknowledgments

We wish to thank Wengbo Yang. Wengbo Yang, Department of Orthopedics, Nanjing First Hospital, Nanjing Medical University, we thank him for his help with anatomy and anatomical illustration. Hailing Yin and Chen Zhang contributed equally to this work and share first authorship.

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.

Hailing Yin:Conceptualization; Methodology; Investigation; Data curation; Formal analysis; Visualization; Writing-original draft.

Chen Zhang (co-first author):Conceptualization; Methodology; Investigation; Data curation; Formal analysis; Visualization; Writing -original draft.

Wenwen Zhang:Investigation; Data curation; Validation; Visualization.

Yong Zhang: Methodology, Supervision.

Ke Ding:Investigation; Data curation; Validation.

Hongyu Wang:Investigation; Resources; Project administration.

Zhenhong Wang:Investigation; Resources.

Yihong Gao:Data curation; Validation.

Jiabin Liu:Methodology;Data curation;Writing-review and editing.

Hao Cheng (corresponding author):Data curation, Formal analysis, Methodology, Software, Validation, Writing-review and editing.

Xiaoliang Wang (senior corresponding author):Conceptualization; Funding acquisition; Methodology; Project administration; Resources; Supervision;Writing-review and editing.

Disclosure

The authors report no conflicts of interest in this work.

References

1. Kumara AB, Gogia AR, Bajaj JK, Agarwal N. Clinical evaluation of postoperative analgesia comparing suprascapular nerve block and interscalene brachial plexus block in patients undergoing shoulder arthroscopic surgery. J Clin Orthop Trauma. 2016;7:34–15. doi:10.1016/j.jcot.2015.09.003

2. Hughes MS, Matava MJ, Wright RW, Brophy RH, Smith MV. Interscalene brachial plexus block for arthroscopic shoulder surgery: a systematic review. J Bone Joint Surg Am. 2013;95:1318–1324. doi:10.2106/JBJS.L.01116

3. Fredrickson MJ, Krishnan S, Chen CY. Postoperative analgesia for shoulder surgery: a critical appraisal and review of current techniques. Anaesthesia. 2010;65:608–624. doi:10.1111/j.1365-2044.2009.06231.x

4. Toma O, Persoons B, Pogatzki-Zahn E, Van de Velde M, Joshi GP, PROSPECT Working Group collaborators. PROSPECT guideline for rotator cuff repair surgery: systematic review and procedure-specific postoperative pain management recommendations. Anaesthesia. 2019;74:1320–1331. doi:10.1111/anae.14796

5. Helmy MA, Gamal B, Milad LM, et al. Single-shot versus slow infusion interscalene block and its impact on diaphragmatic function in patients undergoing shoulder surgery, a double blind randomized controlled trial. Anaesth Crit Care Pain Med. 2026;45(4):101668. doi:10.1016/j.accpm.2025.101668

6. Büyükgebiz Yeşil B, Emmez G, Kanatlı U, et al. Impact of interscalene versus superior trunk blocks as sole anesthetics on respiratory outcomes for shoulder arthroscopy: a randomized controlled trial. BMC Anesthesiol. 2026;26(1):106. doi:10.1186/s12871-025-03575-0

7. El-Boghdadly K, Chin KJ, Chan VWS. Phrenic nerve palsy and regional anesthesia for shoulder surgery: anatomical, physiologic, and clinical considerations. Anesthesiology. 2017;127:173–191. doi:10.1097/ALN.0000000000001668

8. Marty P, Ferré F, Basset B, et al. Diaphragmatic paralysis in obese patients in arthroscopic shoulder surgery: consequences and causes. J Anesth. 2018;32:333–340. doi:10.1007/s00540-018-2477-9

9. Ferré F, Cugnin N, Martin C, et al. Regional anesthesia with noninvasive ventilation for shoulder surgery in a patient with severe chronic obstructive pulmonary disease: a case report. AA Case Rep. 2017;8:261–264. doi:10.1213/XAA.0000000000000488

10. Woo JH, Lee HJ, Oh HW, Lee JW, Baik HJ, Kim YJ. Perineural dexamethasone reduces rebound pain after ropivacaine single injection interscalene block for arthroscopic shoulder surgery: a randomized controlled trial. Reg Anesth Pain Med. 2021;46:965–970. doi:10.1136/rapm-2021-102795

11. Kim SH, Yeo IS, Jang J, Jung HE, Chun YM, Yang HM. Infraspinatus-teres minor (ITM) interfascial block: a novel approach for combined suprascapular and axillary nerve block. Reg Anesth Pain Med. 2024;49:67–72. doi:10.1136/rapm-2023-104738

12. Chan C, Peng PWH. Suprascapular nerve block: a narrative review. Reg Anesth Pain Med. 2011;36:358–373. doi:10.1097/AAP.0b013e3182204ec0

13. He Y, Zhao W, Ze Z, et al. Comparison between infraspinatus-teres minor (ITM) interfascial block and superior trunk block in shoulder arthroscopy: a randomized non-inferiority trial. J Clin Anesth. 2026;109:112106. doi:10.1016/j.jclinane.2025.112106

14. Zhu M, Zhang L, Zhang L, et al. Analgesic effect of infraspinatus-teres minor interfascial block in scapular fracture surgery: a case report and literature review. J Perianesth Nurs. 2026. doi:10.1016/j.jopan.2025.12.006

15. Stark PA, Myles PS, Burke JA. Development and psychometric evaluation of a postoperative quality of recovery score: the QoR-15. Anesthesiology. 2013;118:1332–1340. doi:10.1097/ALN.0b013e318289b84b

16. Bu XS, Zhang J, Zuo YX. Validation of the Chinese version of the quality of recovery-15 score and its comparison with the post-operative quality recovery scale. Patient. 2016;9:251–259. doi:10.1007/s40271-015-0148-6

17. Ferré F, Pommier M, Laumonerie P, et al. Hemidiaphragmatic paralysis following ultrasound-guided anterior vs posterior suprascapular nerve block: a double-blind, randomized control trial. Anaesthesia. 2020;75:499–508. doi:10.1111/anae.14978

18. Kim HJ, Koh KH, Park JI, et al. Comparison of the analgesic efficacy between arthroscopically placed continuous suprascapular nerve block and ultrasound- guided continuous superior trunk block: a double-blinded randomized controlled trial. Anesthesiology. 2023;139(5):591–601. doi:10.1097/ALN.0000000000004691

19. Barry GS, Bailey JG, Sardinha J, Brousseau P, Uppal V. Factors associated with rebound pain after peripheral nerve block for ambulatory surgery. Br J Anaesth. 2021;126:862–871. doi:10.1016/j.bja.2020.10.035

20. Harbell MW, Kolodzie K, Behrends M, et al. Extraplexus versus intraplexus ultrasound-guided interscalene brachial plexus block for ambulatory arthroscopic shoulder surgery: a randomized controlled trial. PLoS One. 2021;16(2):e0246792. doi:10.1371/journal.pone.0246792

21. Saito A, Laigaard J, Bjersand Sunde P, et al. Minimal important difference for rescue opioid consumption in adults with acute pain: a scoping review. Eur J Pain. 2025;29(10):e70151. doi:10.1002/ejp.70151

22. Raff M, Belbachir A, El-Tallawy S, et al. Intravenous oxycodone versus other intravenous strong opioids for acute postoperative pain control: a systematic review of randomized controlled trials. Pain Ther. 2019;8(1):19–39. doi:10.1007/s40122-019-0122-4

23. Reed DE, Stump TE, Monahan PO, Kroenke K. Comparable minimally important differences and responsiveness of brief pain inventory and PEG pain scales across 6 trials. J Pain. 2024;25:142–152. doi:10.1016/j.jpain.2023.07.028

24. Dhir S, Sondekoppam RV, Sharma R, Ganapathy S, Athwal GS. A comparison of combined suprascapular and axillary nerve blocks to interscalene nerve block for analgesia in arthroscopic shoulder surgery: an equivalence study. Reg Anesth Pain Med. 2016;41:564–571. doi:10.1097/AAP.0000000000000436

25. Hussain N, Goldar G, Ragina N, Banfield L, Laffey JG, Abdallah FW. Suprascapular and interscalene nerve block for shoulder surgery: a systematic review and meta-analysis. Anesthesiology. 2017;127:998–1013. doi:10.1097/ALN.0000000000001894

26. Yang X, Su B, Chen Y, et al. Adjuvants and rebound pain following peripheral nerve block in adult surgical patients: a systematic review and network meta-analysis of randomized controlled trials. Anesthesiol Perioper Sci. 2025;3:30. doi:10.1007/s44254-025-00111-8

27. Oliver-Fornies P, Aragon-Benedi C, Gomez R, et al. Hemidiaphragmatic paralysis after ultrasound-guided brachial plexus blocks for shoulder surgery: a systematic review and meta-analysis of randomized clinical trials. J Clin Anesth. 2025;105:111874. doi:10.1016/j.jclinane.2025.111874

28. Renard Y, Grape S, Gonvers E, et al. Respiratory impact of local anaesthetic volume after extrafascial interscalene brachial plexus block: a randomised controlled double-blinded trial. Br J Anaesth. 2025;134(4):1153–1160. doi:10.1016/j.bja.2024.12.010