Clinical efficacy of ultrasound-guided serratus posterior superior intercostal plane block in unilateral cervicothoracic myofascial pain syndrome: a retrospective study Authors: Sefa Tan, Banu Gokcen Baydogan Tan, Ruhiye Reisli, Sema Tuncer Uzun Published: 2026 ABSTRACT BackgroundMyofascial pain syndrome (MPS) of the cervicothoracic region is a common cause of chronic neck and upper back pain that often shows limited response to conservative therapies. The recently described serratus posterior superior intercostal plane block (SPSIPB) targets the fascial plane between the serratus posterior superior and intercostal muscles and may provide effective regional analgesia. However, its long-term clinical efficacy in MPS has not been systematically evaluated. This study aimed to investigate the effectiveness and safety of ultrasound-guided SPSIPB in patients with unilateral cervicothoracic MPS.MethodsThis single-center retrospective cohort study evaluated the effectiveness of ultrasound-guided SPSIPB in patients with unilateral cervicothoracic MPS. A total of 75 patients treated between June 2024 and May 2025 were included. Pain intensity (Numerical Rating Scale [NRS]) was assessed at baseline, 1 h, 4 weeks, 12 weeks, and 6 months. Functional disability (Neck Disability Index [NDI]) and quality of life (EQ-5D-5 L) were evaluated in a predefined subset. A linear mixed-effects model including all available observations was used for primary analysis, with post-baseline co-interventions incorporated as a time-varying covariate. Sensitivity analyses included per-protocol, complete-case, and conservative missing-not-at-random (MNAR) approaches.ResultsNRS scores decreased significantly over time, with the greatest reduction observed at 1 h and continued improvement at 6 months (adjusted mean change − 2.73, 95% CI − 3.18 to − 2.28; p < 0.001). Clinically meaningful improvement, defined as ≥ 50% reduction in NRS, was observed in 90.7%, 34.7%, and 28.0% of patients at 1 h, 4 weeks, and 6 months, respectively. Functional disability improved significantly, with NDI reductions of − 7.83, − 7.37, and − 6.67 at 4 weeks, 12 weeks, and 6 months, respectively (all p < 0.001). EQ-5D-5 L scores also improved significantly across all follow-up time points (all p < 0.001). Sensitivity analyses demonstrated consistent findings across all models, including conservative MNAR assumptions. No major complications were observed.ConclusionsUltrasound-guided SPSIPB was associated with rapid pain reduction and continued improvements in disability and quality of life up to six months in this retrospective cohort. No major complications were documented. These findings support further prospective controlled evaluation of SPSIPB as a potential interfascial approach for cervicothoracic MPS.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12871-026-03956-z. Background Myofascial pain syndrome (MPS) is a common musculoskeletal condition characterized by localized pain, muscle stiffness, and tenderness in specific areas of the muscle and surrounding fascia [1]. Although widely recognized in clinical practice, the exact pathophysiological mechanisms underlying MPS remain incompletely understood. Several theories have been proposed, including neuromuscular junction dysfunction, the integrated trigger point hypothesis, neurogenic inflammation, and fascial densification—all of which attempt to explain the development and persistence of myofascial trigger points [2]. The cervicothoracic region, due to its biomechanical load and postural stress, is frequently affected by MPS, particularly in individuals with sedentary lifestyles or repetitive occupational strain. In this region, postural stabilizing muscles such as the trapezius and rhomboid are among the most commonly involved in myofascial pain [3]. In addition to these well-recognized muscles, the serratus posterior superior (SPS) muscle has also been identified as a potential source of pain. Early descriptions by Travell and Simons identified trigger points within the SPS as contributors to referred pain in the shoulder and upper extremity [4, 5] while subsequent studies demonstrated that tenderness near the superior scapular angle frequently originates from this muscle and may respond to targeted interventions [6]. Beyond individual muscle involvement, recent anatomical and histological studies have highlighted the critical role of fascial structures in the development and persistence of myofascial pain [7]. These insights have shifted the focus of MPS treatment toward targeting the fascial structures themselves. The superficial and deep fasciae, interconnected through retinacular ligaments and surrounding the muscles, serve not only as mechanical support structures but also as key sites of nociceptive input due to their rich innervation by Aδ and C fibers [8, 9]. In addition to this dense innervation, fascial dysfunction may arise from reduced fascial gliding due to mechanical densification, increased mechanical stress on embedded sensory nerve fibers, and amplification of nociceptive signaling through neuroinflammatory pathways [10, 11]. Accordingly, the fascia can be considered an integrated peripheral substrate linking muscular and neural components in MPS. Effective management of MPS typically requires a multimodal approach integrating both pharmacological and non-pharmacological therapies. Although conservative therapies such as physical therapy, pharmacologic agents, and dry needling are commonly employed in clinical practice, a considerable number of patients with MPS remain refractory to these approaches [12]. In recent years, ultrasound-guided interfascial plane interventions have emerged as a promising treatment approach [13–15]. Local anesthetics injected into the interfascial plane tend to spread along low-resistance pathways within the fascial compartments, while avoiding the potential toxic effects associated with intramuscular administration [16]. This allows for broad and safe analgesic coverage with a single injection. The serratus posterior superior intercostal plane block (SPSIPB) is a relatively new regional anesthesia technique first introduced by Tulgar et al. [17] in 2023 through a cadaveric investigation and a case series involving five patients with MPS. This technique targets the fascial plane between the SPS muscle and the underlying intercostal muscles. Anatomical and clinical findings suggest that SPSIPB can provide effective sensory coverage across the posterior hemithorax [18, 19], shoulder girdle, and posterior cervical region [17, 20], making it a promising option for managing pain originating from cervicothoracic myofascial structures. The analgesic effect of SPSIPB is presumed to result from the spread of local anesthetic within this interfascial plane, facilitating modulation of the dorsal rami of the cervical and upper thoracic spinal nerves [17], rather than targeting a single discrete nerve. Although the efficacy of blocks such as the erector spinae plane block (ESPB) [21, 22] and rhomboid intercostal plane block (RIPB) [23, 24] has been investigated in MPS, the clinical effectiveness of the SPSIPB remains unknown. To date, no longitudinal studies have examined its sustained analgesic effect, safety profile, or patient-level predictors of response in cervicothoracic MPS. We hypothesize that SPSIPB may improve outcomes in cervicothoracic MPS by combining fluid-mediated restoration of fascial mobility with broader analgesia through spread of local anesthetic along the interfascial plane, affecting the dorsal rami. Therefore, the aim of this study was to evaluate the longitudinal clinical outcomes of ultrasound-guided SPSIPB in patients with unilateral cervicothoracic MPS, focusing on its effects on pain intensity, functional disability, and quality of life over a six-month follow-up period. Methods Study design and participants This study was conducted as a single-center, retrospective observational cohort study based on deidentified electronic medical records and the procedural logbook of patients with unilateral cervicothoracic MPS who underwent ultrasound-guided SPSIPB between June 1, 2024, and May 1, 2025, at the Pain Medicine Clinic of Necmettin Erbakan University Faculty of Medicine. The study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [25]. The study population was identified by screening the procedural logbook for all SPSIPB procedures performed during the predefined study period and cross-checking each entry against the electronic medical records using unique hospital identifiers. Recorded time windows were prespecified as 1 h (± 15 min), 4 weeks (± 7 days), 12 weeks (± 14 days), and 6 months (± 30 days), and were applied consistently throughout eligibility assessment, data extraction, and analysis. The study was approved by the Necmettin Erbakan University Ethics Committee (2025/6137). It was conducted in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its subsequent amendments. Due to the retrospective design and the use of de-identified data, the requirement for informed consent was waived. Inclusion criteria Clinical records were included if the patient met the following criteria:Adults aged 18 to 65 years.Clinical diagnosis of unilateral cervicothoracic MPS was based on established criteria [1], including the presence of at least one active myofascial trigger point identified by palpation as a hyperirritable spot within a taut band of skeletal muscle associated with localized tenderness and reproduction of the patient’s typical referred pain pattern upon compression. All clinical examinations were performed by a board-certified pain medicine specialist with experience in musculoskeletal pain disorders. The diagnosis was primarily based on clinical examination, whereas ultrasound was used to support anatomical localization of the target region and procedural guidance rather than as a stand-alone diagnostic tool.Persistent unilateral cervicothoracic pain lasting a minimum of three months.Documentation of a single SPSIPB procedure performed under ultrasound guidance as part of their routine pain management plan. Adults aged 18 to 65 years. Adults aged 18 to 65 years. Clinical diagnosis of unilateral cervicothoracic MPS was based on established criteria [1], including the presence of at least one active myofascial trigger point identified by palpation as a hyperirritable spot within a taut band of skeletal muscle associated with localized tenderness and reproduction of the patient’s typical referred pain pattern upon compression. All clinical examinations were performed by a board-certified pain medicine specialist with experience in musculoskeletal pain disorders. The diagnosis was primarily based on clinical examination, whereas ultrasound was used to support anatomical localization of the target region and procedural guidance rather than as a stand-alone diagnostic tool. Clinical diagnosis of unilateral cervicothoracic MPS was based on established criteria [1], including the presence of at least one active myofascial trigger point identified by palpation as a hyperirritable spot within a taut band of skeletal muscle associated with localized tenderness and reproduction of the patient’s typical referred pain pattern upon compression. All clinical examinations were performed by a board-certified pain medicine specialist with experience in musculoskeletal pain disorders. The diagnosis was primarily based on clinical examination, whereas ultrasound was used to support anatomical localization of the target region and procedural guidance rather than as a stand-alone diagnostic tool. Persistent unilateral cervicothoracic pain lasting a minimum of three months. Persistent unilateral cervicothoracic pain lasting a minimum of three months. Documentation of a single SPSIPB procedure performed under ultrasound guidance as part of their routine pain management plan. Documentation of a single SPSIPB procedure performed under ultrasound guidance as part of their routine pain management plan. Exclusion criteria History of interventional procedures in the affected region within the previous three months.Coagulopathy or known hypersensitivity to local anesthetics. History of interventional procedures in the affected region within the previous three months. History of interventional procedures in the affected region within the previous three months. Coagulopathy or known hypersensitivity to local anesthetics. Coagulopathy or known hypersensitivity to local anesthetics. Sample selection All eligible cases were identified through systematic screening of routine clinical records and included in the analysis. Given the retrospective design, the sample size was determined by the fixed availability of existing records at the time of study initiation; therefore, no a priori sample size calculation was performed. A total of 75 patients who underwent ultrasound-guided SPSIPB for unilateral cervicothoracic MPS met the predefined inclusion and exclusion criteria and were retained in the cohort. Availability of complete outcome data across all recorded time points was not required for inclusion; instead, missing data were addressed at the level of individual outcomes and time points without excluding entire patient records. Post-baseline co-interventions, including physiotherapy, repeat injections, and patient-initiated home exercise programs, were documented based on patient reports and medical records. Their timing and clinical indications were recorded for each patient. Potential sources of bias inherent to the retrospective design were considered, including information bias related to medical record abstraction, selection bias due to differential availability of outcome data across recorded time points, and time-varying confounding arising from post-baseline co-interventions. To mitigate these issues, only the index procedure was analyzed, prespecified time windows were applied consistently throughout eligibility assessment, data extraction, and analysis, and all data were collected using a standardized case-report form. Participant flow and the handling of missing follow-up and post-baseline co-interventions are shown in Fig. 1. Fig. 1Flow diagram of record identification, cohort inclusion, post-baseline co-interventions, and outcome data availability. Footnote: A total of 75 eligible patients who underwent ultrasound-guided SPSIPB for unilateral cervicothoracic MPS were included in the retrospective cohort. Post-baseline co-interventions were documented rather than used as exclusion criteria and were incorporated into the primary analysis as a time-varying covariate. NRS analyses included all available observations from the full cohort, whereas NDI and EQ-5D-5L analyses were performed in the predefined subset of patients with available data. NRS, Numerical Rating Scale; NDI, Neck Disability Index; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; SPSIPB, Serratus Posterior Superior Intercostal Plane Block; MPS, Myofascial Pain Syndrome Flow diagram of record identification, cohort inclusion, post-baseline co-interventions, and outcome data availability. Footnote: A total of 75 eligible patients who underwent ultrasound-guided SPSIPB for unilateral cervicothoracic MPS were included in the retrospective cohort. Post-baseline co-interventions were documented rather than used as exclusion criteria and were incorporated into the primary analysis as a time-varying covariate. NRS analyses included all available observations from the full cohort, whereas NDI and EQ-5D-5L analyses were performed in the predefined subset of patients with available data. NRS, Numerical Rating Scale; NDI, Neck Disability Index; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; SPSIPB, Serratus Posterior Superior Intercostal Plane Block; MPS, Myofascial Pain Syndrome Intervention All procedures were performed under ultrasound guidance using an ultrasound system (Esaote MyLab Six, Genova, Italy) equipped with a high-frequency linear transducer (B-mode, 4–13 MHz) by a single pain physician with over seven years of experience in ultrasound-guided interventional pain procedures. Patients were placed in the prone position with the head turned contralaterally. To facilitate lateralization of the scapula, patients were instructed either to grasp the contralateral shoulder with the ipsilateral hand or to place the affected arm in adduction and internal rotation. Subsequently, anatomical landmarks—including the spinous processes of C7 and upper thoracic vertebrae, the superomedial and inferior angles of the scapula and spina scapula—were identified and marked. The procedure was performed unilaterally on the symptomatic side. No procedural sedation was used. Standard monitoring, including non-invasive blood pressure, heart rate, and peripheral oxygen saturation, was applied to all patients. Intravenous access was established prior to the procedure. The skin was prepared using antiseptic solution and sterile draping was applied in accordance with routine aseptic technique. The transducer was placed longitudinally along the medial border of the scapula, and active myofascial trigger points in the cervicothoracic region were confirmed through palpation before skin preparation, with ultrasound used to identify the relevant fascial and muscular planes. SPSIPB The transducer was positioned over the medial scapular border at the level of the second to fourth thoracic vertebrae to visualize the SPS muscle and the underlying intercostal muscles. After skin antisepsis, a 22-gauge, 50 mm echogenic block needle (Stimuplex Ultra 360, B. Braun, Melsungen, Germany) was inserted in-plane from 45° caudal to cephalad direction, targeting the fascial plane between the SPS and intercostal muscles in accordance with the ultrasound orientation shown in Fig. 2. Once correct needle placement was confirmed by hydrodissection with 2 mL of normal saline, negative aspiration was performed, and a total volume of 20 mL, consisting of 0.25% bupivacaine, 2 mL dexamethasone (8 mg), with normal saline added to achieve the final volume was slowly injected under real-time visualization to ensure appropriate spread along the interfascial plane [17]. Care was taken to avoid intramuscular spread or pleural contact and patients were monitored for at least 30 min post-injection for potential complications. Fig. 2Ultrasound-guided Serratus posterior superior intercostal plane block. Legend : (A) Longitudinal paraspinal ultrasound image demonstrating the serratus posterior superior (SPS) muscle overlying the third thoracic ribs. The needle is advanced in-plane from caudal to cephalad toward the fascial plane deep to the SPS muscle (red arrows). B Parasagittal view showing needle tip positioning beneath the SPS muscle with the distribution of local anesthetic spreading (between the two blue lines) within the serratus posterior superior–intercostal fascial plane. SPS — serratus posterior superior; Rhom — rhomboid; Trap — Trapezius Ultrasound-guided Serratus posterior superior intercostal plane block. Legend : (A) Longitudinal paraspinal ultrasound image demonstrating the serratus posterior superior (SPS) muscle overlying the third thoracic ribs. The needle is advanced in-plane from caudal to cephalad toward the fascial plane deep to the SPS muscle (red arrows). B Parasagittal view showing needle tip positioning beneath the SPS muscle with the distribution of local anesthetic spreading (between the two blue lines) within the serratus posterior superior–intercostal fascial plane. SPS — serratus posterior superior; Rhom — rhomboid; Trap — Trapezius Outcome measures and statistical analysis Demographic and clinical characteristics of all participants, including age, sex, symptom duration, and outcome measures (NRS, NDI, and EQ-5D-5 L), were routinely documented as part of standard clinical care and retrieved from electronic medical records. Procedure-related adverse events were predefined as secondary outcomes and systematically evaluated through retrospective review of these records, including pneumothorax, vascular puncture, local anesthetic systemic toxicity, infection, vasovagal reactions, and prolonged sensory or motor deficits. Baseline data were obtained from the pre-procedural clinical assessment, while post-procedural outcome data were extracted from recorded clinical encounters corresponding to prespecified time windows (1 h, 4 weeks, 12 weeks, and 6 months). Primary outcome The primary outcome was the change in pain intensity over time, assessed using the Numerical Rating Scale (NRS). The NRS ranges from 0 (no pain) to 10 (worst imaginable pain) and was recorded at baseline, 1 h, 4 weeks, 12 weeks, and 6 months. The 4-week time point was prespecified as the primary endpoint for clinical interpretation. Clinically meaningful improvement was additionally defined as a ≥ 50% reduction in NRS score from baseline at each follow-up time point [26]. Secondary outcomes Functional disability related to the cervical region was evaluated using the NDI [27].The NDI is a validated, self-reported questionnaire consisting of 10 items, each assessing different domains of neck-related disability (e.g., pain intensity, personal care, lifting, reading, headaches, concentration, work, driving, sleeping, and recreation) each scored from 0 to 5, with a total score ranging from 0 to 50, where higher scores indicate greater disability. The validated Turkish version of the NDI [28] was used in this study. NDI data were available at baseline and at recorded time points corresponding to 4 weeks, 12 weeks, and 6 months. Health-related quality of life was assessed using the EQ-5D-5 L Questionnaire developed by the EuroQol Group [29]. It includes five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension has five levels: no problems, slight problems, moderate problems, severe problems, and extreme problems. Utility index values were calculated using the England EQ-5D-5 L value set (Devlin et al., 2018; hybrid model) [30], as a country-specific Turkish EQ-5D-5 L value set was not available. Index scores were derived according to the standard tariff algorithm, with higher values indicating better health status. EQ-5D-5 L was completed at baseline, 4-week, 12-week and 6 month post-intervention. Statistical analysis Statistical analyses were performed using IBM SPSS Statistics for Windows, version 22.0 (IBM Corp., Armonk, NY, USA). Continuous variables were summarized as mean ± standard deviation (SD) or median (interquartile range [IQR]) depending on distributional characteristics, while categorical variables were presented as counts (percentages). The primary analysis was conducted using a linear mixed-effects model including all available observations from the full cohort (n = 75), regardless of post-baseline co-interventions, to reflect real-world clinical practice. Time was modeled as a categorical fixed effect (baseline, 1 h, 4 weeks, 12 weeks, and 6 months), and a random intercept was included at the patient level to account for repeated measurements within individuals. Post-baseline co-interventions (including repeat injections, physiotherapy, and patient-initiated home exercise programs) were incorporated into the model as a time-varying covariate. This variable was coded as 0 for observations obtained prior to any co-intervention or in patients without co-intervention, and 1 for observations obtained after the first documented co-intervention. The model was adjusted for age, sex, and symptom duration. Adjusted mean changes from baseline with 95% confidence intervals (CIs) were reported. Given the retrospective design and decreasing follow-up completeness over time, missing outcome data were expected. The primary analysis relied on likelihood-based estimation within the mixed-effects modeling framework, which provides valid inference under the missing-at-random (MAR) assumption and allows inclusion of all available observations without listwise deletion. To assess the robustness of the findings, several prespecified sensitivity analyses were conducted. A per-protocol (while-on-treatment) analysis was performed by excluding observations obtained after post-baseline co-interventions. In addition, a complete-case analysis was conducted including only patients with available NRS data at all follow-up time points. An all-inclusive sensitivity analysis without adjustment for co-interventions was also performed, incorporating all observations regardless of treatment changes. Furthermore, to evaluate potential departures from the MAR assumption, an additional missing data sensitivity analysis was conducted under a missing-not-at-random (MNAR) framework. In this analysis, a delta-adjustment approach was applied, whereby missing follow-up NRS values were imputed under a conservative assumption of worse outcomes (i.e., higher NRS scores) compared with observed values. The primary model was then re-estimated under these alternative assumptions to assess the robustness of the findings. In addition, a tipping-point sensitivity analysis was conducted to further assess the robustness of the results under potential MNAR mechanisms. Missing 6-month NRS values were systematically shifted by a range of delta (δ) values (from − 2 to + 3 NRS points) relative to model-based estimates, and the corresponding changes in treatment effect were re-evaluated to determine the magnitude of deviation required to alter the clinical interpretation. Secondary outcomes (NDI and EQ-5D-5 L index scores) were analyzed using separate linear mixed-effects models in a predefined subset of patients with complete data across all scheduled follow-up time points. All statistical tests were two-tailed, and a p-value < 0.05 was considered statistically significant. Results Patient demographics and baseline characteristics A total of 75 patients were included in the analysis. The mean age was 42.0 ± 12.5 years, and the majority of participants were female (80%). The median symptom duration was 24 months (IQR: 12–60). Baseline pain intensity was 6.7 ± 1.7 on the NRS and functional disability, assessed using the NDI, was 26.0 ± 9.4. The mean EQ-5D-5 L index score was 0.768 ± 0.087. Of the 75 patients, 60 did not receive any post-baseline co-intervention; among these, 30 had complete outcome data across all time points, whereas 30 had incomplete outcome data. The remaining 15 patients received post-baseline co-interventions, including repeat injections, physiotherapy, or patient-initiated home exercise programs. Among the 15 patients who received post-baseline co-interventions, 5 were treated due to insufficient pain relief, 7 due to persistence or recurrence of symptoms, and 3 based on patient preference or referral for additional therapy. Baseline characteristics stratified by outcome data availability and post-baseline co-intervention status are presented in Table 1. Table 1Baseline characteristics stratified by outcome data availability and post-baseline co-intervention statusVariableComplete outcome datano co-intervention(n = 30)Incomplete outcome datano co-intervention(n = 30)Co-intervention (n = 15)Age, years(mean ± SD)43.1 ± 11.841.5 ± 13.242.2 ± 13.0Female, n (%)24 (80%)24 (80%)12 (80%)Baseline NRS (mean ± SD)7.2 ± 1.66.0 ± 1.55.9 ± 0.8Duration, months (median [IQR])24 (12–60)18 (12–72)36 (12–96)NDI available, n (%)30 (100%)0 (0%)0 (0%)EQ-5D-5 L available, n (%)30 (100%)0 (0%)0 (0%)NRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire, SD standard deviation Table 1Baseline characteristics stratified by outcome data availability and post-baseline co-intervention statusVariableComplete outcome datano co-intervention(n = 30)Incomplete outcome datano co-intervention(n = 30)Co-intervention (n = 15)Age, years(mean ± SD)43.1 ± 11.841.5 ± 13.242.2 ± 13.0Female, n (%)24 (80%)24 (80%)12 (80%)Baseline NRS (mean ± SD)7.2 ± 1.66.0 ± 1.55.9 ± 0.8Duration, months (median [IQR])24 (12–60)18 (12–72)36 (12–96)NDI available, n (%)30 (100%)0 (0%)0 (0%)EQ-5D-5 L available, n (%)30 (100%)0 (0%)0 (0%)NRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire, SD standard deviation Baseline characteristics stratified by outcome data availability and post-baseline co-intervention status Age, years (mean ± SD) NRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire, SD standard deviation All patients contributed baseline and 1-hour NRS data, whereas the number of available observations decreased at later time points. NDI and EQ-5D-5 L were available only for patients followed under a structured clinical follow-up protocol during routine care. Because these measures were not collected for the entire cohort, they were analyzed as exploratory secondary outcomes in this predefined complete-data subset. The availability of outcome data across prespecified time points is summarized in Table 2. Table 2Availability of outcome measurements across time points in the retrospective cohortOutcomeBaseline (n = 75)1 h (n = 75)4 weeks (n = 51)12 weeks (n = 41)6 months (n = 36)NRS, n (%)75 (100)75 (100)51 (68)41 (55)36 (48)NDI, n (%)30 (40)Not scheduled30 (40)30 (40)30 (40)EQ-5D-5 L, n (%)30 (40)Not scheduled30 (40)30 (40)30 (40)Values represent the number of patients with available outcome data at each prespecified time point. Percentages are calculated using the full cohort (n = 75) as the denominator. No patients were excluded due to missing data, and analyses were performed using all available observations. NDI and EQ-5D-5 L outcomes were collected in a predefined subset of patients and were not scheduled at the 1-hour assessmentNRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire Table 2Availability of outcome measurements across time points in the retrospective cohortOutcomeBaseline (n = 75)1 h (n = 75)4 weeks (n = 51)12 weeks (n = 41)6 months (n = 36)NRS, n (%)75 (100)75 (100)51 (68)41 (55)36 (48)NDI, n (%)30 (40)Not scheduled30 (40)30 (40)30 (40)EQ-5D-5 L, n (%)30 (40)Not scheduled30 (40)30 (40)30 (40)Values represent the number of patients with available outcome data at each prespecified time point. Percentages are calculated using the full cohort (n = 75) as the denominator. No patients were excluded due to missing data, and analyses were performed using all available observations. NDI and EQ-5D-5 L outcomes were collected in a predefined subset of patients and were not scheduled at the 1-hour assessmentNRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire Availability of outcome measurements across time points in the retrospective cohort Values represent the number of patients with available outcome data at each prespecified time point. Percentages are calculated using the full cohort (n = 75) as the denominator. No patients were excluded due to missing data, and analyses were performed using all available observations. NDI and EQ-5D-5 L outcomes were collected in a predefined subset of patients and were not scheduled at the 1-hour assessment NRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire Primary outcome This all-inclusive modeling approach ensured that all patients contributed to the primary analysis, including observations obtained after post-baseline co-interventions. Pain severity (NRS scores) In the primary all-inclusive linear mixed-effects model including all available NRS observations from the full cohort, NRS scores decreased significantly over time following the index procedure. The greatest reduction was observed at the early post-procedural time point, with a gradual attenuation of the effect across subsequent recorded time points. Detailed estimates of the longitudinal changes in NRS are presented in Table 3. The time-varying post-baseline co-intervention variable was not significantly associated with NRS scores (β = −0.32, 95% CI − 1.00 to 0.36, p = 0.361). Table 3Longitudinal changes in pain severity, functional disability and health-related quality of life outcomesOutcomeTime point n Mean ± SDAdjusted mean change vs. baseline (95% CI)p-valueNRSBaseline756.7 ± 1.7—1 h752.5 ± 1.3−4.27 (− 4.62 to − 3.92)< 0.0014 weeks513.7 ± 1.5−3.01 (− 3.42 to − 2.60)< 0.00112 weeks413.5 ± 1.3−3.23 (− 3.68 to − 2.78)< 0.0016 months364.0 ± 1.3−2.73 (− 3.18 to − 2.28)< 0.001NDIBaseline3026.0 ± 9.4—4 weeks3018.1 ± 7.0−7.83 (− 9.34 to − 6.33)< 0.00112 weeks3018.6 ± 7.5−7.37 (− 8.87 to − 5.86)< 0.0016 months3019.3 ± 7.5−6.67 (− 8.17 to − 5.16)< 0.001EQ-5D-5 LBaseline300.768 ± 0.087—4 weeks300.793 ± 0.082+ 0.025 (0.012 to 0.038)< 0.00112 weeks300.799 ± 0.075+ 0.031 (0.018 to 0.044)< 0.0016 months300.798 ± 0.064+ 0.030 (0.017 to 0.043)< 0.001NRS was analyzed in the full cohort using an all-inclusive linear mixed-effects model including all available observations and a time-varying co-intervention covariate. NDI and EQ-5D-5 L were analyzed in the predefined subset of 30 patients with available dataNRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire, CI confidence interval, SD standard deviation Table 3Longitudinal changes in pain severity, functional disability and health-related quality of life outcomesOutcomeTime point n Mean ± SDAdjusted mean change vs. baseline (95% CI)p-valueNRSBaseline756.7 ± 1.7—1 h752.5 ± 1.3−4.27 (− 4.62 to − 3.92)< 0.0014 weeks513.7 ± 1.5−3.01 (− 3.42 to − 2.60)< 0.00112 weeks413.5 ± 1.3−3.23 (− 3.68 to − 2.78)< 0.0016 months364.0 ± 1.3−2.73 (− 3.18 to − 2.28)< 0.001NDIBaseline3026.0 ± 9.4—4 weeks3018.1 ± 7.0−7.83 (− 9.34 to − 6.33)< 0.00112 weeks3018.6 ± 7.5−7.37 (− 8.87 to − 5.86)< 0.0016 months3019.3 ± 7.5−6.67 (− 8.17 to − 5.16)< 0.001EQ-5D-5 LBaseline300.768 ± 0.087—4 weeks300.793 ± 0.082+ 0.025 (0.012 to 0.038)< 0.00112 weeks300.799 ± 0.075+ 0.031 (0.018 to 0.044)< 0.0016 months300.798 ± 0.064+ 0.030 (0.017 to 0.043)< 0.001NRS was analyzed in the full cohort using an all-inclusive linear mixed-effects model including all available observations and a time-varying co-intervention covariate. NDI and EQ-5D-5 L were analyzed in the predefined subset of 30 patients with available dataNRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire, CI confidence interval, SD standard deviation Longitudinal changes in pain severity, functional disability and health-related quality of life outcomes NRS was analyzed in the full cohort using an all-inclusive linear mixed-effects model including all available observations and a time-varying co-intervention covariate. NDI and EQ-5D-5 L were analyzed in the predefined subset of 30 patients with available data NRS Numerical Rating Scale, NDI Neck Disability Index, EQ-5D-5 L EuroQol 5-Dimension 5-Level questionnaire, CI confidence interval, SD standard deviation When clinically meaningful improvement was defined as a ≥ 50% reduction in NRS score from baseline, responder rates were 90.7% at 1 h, decreasing to 34.7% at 4 weeks, 33.3% at 12 weeks, and 28.0% at 6 months, based on all available observations. Secondary outcomes In the predefined subset of patients with available data, both functional disability and health-related quality of life improved significantly over time following the intervention. NDI scores demonstrated a consistent reduction across subsequent recorded time points, while EQ-5D-5 L index scores showed corresponding improvements. Detailed results for secondary outcomes are presented in Table 3. No major procedure-related complications (pneumothorax, local anesthetic systemic toxicity, motor block, or persistent neurological deficit) were documented in the available medical records. Sensitivity analyses Sensitivity analyses were consistent with the primary analysis. Comparable reductions in NRS scores at 6 months were observed in the all-inclusive model without co-intervention adjustment, the per-protocol (while-on-treatment) analysis, and the complete-case analysis, with somewhat larger effect estimates in the latter. Conservative MNAR sensitivity analyses, assuming worse outcomes for missing follow-up data, also demonstrated persistent significant improvement, although with attenuated effect sizes. Detailed results are presented in Table 4. Additional tipping-point analyses demonstrated that the observed reduction in NRS remained clinically meaningful across a wide range of MNAR assumptions. The treatment effect remained above the minimal clinically important difference threshold unless extreme deviations (approximately δ ≥ +3 NRS points) were assumed (Supplement 1). Table 4Primary and sensitivity analyses for change in pain intensity at 6 monthsAnalysis / ModelEffect estimate, change from baseline95% CIp-valuePrimary analysis, LMM adjusted for age, sex, symptom duration, and time-varying co-intervention−2.73−3.18 to − 2.28< 0.001All-inclusive model without co-intervention adjustment−2.75−3.20 to − 2.30< 0.001Per-protocol / while-on-treatment analysis−3.11−3.65 to − 2.57< 0.001Complete-case analysis−3.73−4.24 to − 3.23< 0.001Conservative MNAR sensitivity analysis, delta + 1 NRS point−2.19−2.55 to − 1.83< 0.001Conservative MNAR sensitivity analysis, delta + 2 NRS points−1.71−2.13 to − 1.30< 0.001Effect estimates represent adjusted mean changes in NRS scores at 6 months compared with baseline. Estimates and 95% confidence intervals were obtained from linear mixed-effects models with time as a categorical fixed effect and a random intercept at the patient level. The primary model was adjusted for age, sex, symptom duration, and post-baseline co-intervention as a time-varying covariate. The all-inclusive model omitted the co-intervention covariate. The per-protocol analysis included only observations obtained before post-baseline co-interventions. The complete-case analysis included patients with available NRS data at all follow-up time points. MNAR sensitivity analyses imputed missing follow-up NRS values under conservative assumptions of one- and two-point worse outcomesNRS Numerical Rating Scale, LMM linear mixed-effects model, CI confidence interval, MNAR missing not at random Table 4Primary and sensitivity analyses for change in pain intensity at 6 monthsAnalysis / ModelEffect estimate, change from baseline95% CIp-valuePrimary analysis, LMM adjusted for age, sex, symptom duration, and time-varying co-intervention−2.73−3.18 to − 2.28< 0.001All-inclusive model without co-intervention adjustment−2.75−3.20 to − 2.30< 0.001Per-protocol / while-on-treatment analysis−3.11−3.65 to − 2.57< 0.001Complete-case analysis−3.73−4.24 to − 3.23< 0.001Conservative MNAR sensitivity analysis, delta + 1 NRS point−2.19−2.55 to − 1.83< 0.001Conservative MNAR sensitivity analysis, delta + 2 NRS points−1.71−2.13 to − 1.30< 0.001Effect estimates represent adjusted mean changes in NRS scores at 6 months compared with baseline. Estimates and 95% confidence intervals were obtained from linear mixed-effects models with time as a categorical fixed effect and a random intercept at the patient level. The primary model was adjusted for age, sex, symptom duration, and post-baseline co-intervention as a time-varying covariate. The all-inclusive model omitted the co-intervention covariate. The per-protocol analysis included only observations obtained before post-baseline co-interventions. The complete-case analysis included patients with available NRS data at all follow-up time points. MNAR sensitivity analyses imputed missing follow-up NRS values under conservative assumptions of one- and two-point worse outcomesNRS Numerical Rating Scale, LMM linear mixed-effects model, CI confidence interval, MNAR missing not at random Primary and sensitivity analyses for change in pain intensity at 6 months Effect estimates represent adjusted mean changes in NRS scores at 6 months compared with baseline. Estimates and 95% confidence intervals were obtained from linear mixed-effects models with time as a categorical fixed effect and a random intercept at the patient level. The primary model was adjusted for age, sex, symptom duration, and post-baseline co-intervention as a time-varying covariate. The all-inclusive model omitted the co-intervention covariate. The per-protocol analysis included only observations obtained before post-baseline co-interventions. The complete-case analysis included patients with available NRS data at all follow-up time points. MNAR sensitivity analyses imputed missing follow-up NRS values under conservative assumptions of one- and two-point worse outcomes NRS Numerical Rating Scale, LMM linear mixed-effects model, CI confidence interval, MNAR missing not at random Discussion The findings of this study suggest that ultrasound-guided SPSIPB is associated with a reduction in pain intensity, with the greatest decrease observed in the early post-procedural period and more modest improvements during follow-up. When clinically meaningful improvement was defined as a ≥ 50% reduction in NRS score, reduced pain severity was observed in approximately 91%, 35%, and 28% of patients at 1 h, 4 weeks, and 6 months, respectively. Disability scores also improved, with reductions of 30% at 4 weeks and 26% at 6 months, while quality-of-life measures showed improvement at 4 weeks that remained evident at 6 months. The cervicothoracic region is particularly vulnerable to myofascial pain due to its high biomechanical demand and postural strain [3]. SPS muscle, in this context, has long been recognized as a key pain generator. Travell and Simons originally described “painful shoulder syndrome,” highlighting trigger points within the SPS as frequent sources of referred pain to the shoulder and arm [4, 5]. Fourie later reinforced these observations by demonstrating that tenderness near the superior scapular angle often originated from the SPS and that both steroid injection and surgical release could yield lasting symptom relief in refractory cases [6]. From a functional standpoint, the SPS is thought to act primarily as a proprioceptive and stabilizing muscle at the cervicothoracic junction. Studies by Peck [31] and Voss [32] demonstrated its high muscle-spindle density, supporting a sensory rather than respiratory function. Consequently, chronic overload, fascial densification, or altered proprioceptive feedback within this region may perpetuate nociceptive input and contribute to persistent cervicothoracic pain. Several small series have reported short-term benefits following trigger point injections into the SPS; however, these effects are typically transient [33, 34]. In contrast, fascial plane–directed interventions such as the SPSIPB may produce broader and more durable analgesia by modulating the richly innervated interfascial plane [7, 9] and interrupting nociceptive signaling across the cervicothoracic fascia, rather than merely treating focal intramuscular loci. Anatomically, the SPSIPB targets the plane between the SPS and intercostal muscles, an area innervated by Aδ and C-fiber afferents from the dorsal rami of spinal nerves—providing a rational basis for its extensive sensory coverage [17]. Tulgar et al. first described SPSIPB in cadaveric dissections and a limited clinical series, suggesting coverage of the upper thoracic and posterior cervical dermatomes [17, 20]. Sawada et al. subsequently demonstrated its feasibility in three patients with cervicoscapular MPS, reporting 2–3 weeks of pain relief using 30 mL of 0.25% ropivacaine or physiological saline [35]. In contrast, the present study extends these preliminary findings by providing longitudinal data in a larger cohort, demonstrating reductions in pain intensity of approximately 28% at 6 months. Unlike prior reports, our study used a lower injectate volume with corticosteroid and additionally assessed functional disability and quality of life, which may contribute to the observed longer-term clinical effects. The therapeutic mechanism of SPSIPB likely involves a combination of mechanical and neurophysiological pathways. Chronic musculoskeletal pain has been attributed to reduced fascial mobility and persistent mechanical stress that sensitizes nociceptors within the fascia [36]. A systematic review by Laimi et al. emphasized the importance of restoring fascial gliding in relieving chronic pain [37], while Ichikawa et al. [38] and Arıcı et al. [39] demonstrated that interfascial injections restore sliding between tissue layers and suppress nociceptive afferent activity. Beyond local effects, blocking continuous nociceptive drive may also reduce central sensitization by attenuating dorsal horn hyperexcitability and restoring descending inhibitory control. The addition of dexamethasone could further prolong this effect by reducing perineural inflammation and stabilizing neuronal membranes, accounting for the extended analgesic response observed up to six months in our cohort. The present findings align with emerging evidence supporting the role of ultrasound-guided interfascial plane blocks in the management of MPS [14, 15]. Previous studies investigating blocks such as the ESPB and the RIPB have consistently demonstrated meaningful reductions in pain and functional disability in patients with chronic dorsal or cervicothoracic MPS [21–24]. In the one-year follow-up case series by Taskiran et al. [23], RIB was associated with sustained improvements in pain severity, disability, and quality of life, supporting the potential long-term role of interfascial plane blocks in MPS. However, their protocol used 20 mL of 0.25% bupivacaine without corticosteroid, whereas the present study achieved sustained clinical improvement using combined with 0.25% bupivacaine and 8 mg dexamethasone. In contrast, Köse et al. [24] evaluated RIB in 30 patients and reported a reduction in median NRS from 5 to 2 at 6 weeks, with improvement in disability but no statistically significant change in quality of life. Their technique differed from ours by using 15 mL of 0.25% bupivacaine with 8 mg dexamethasone. In a prospective randomized trial comparing RIB and ESPB in patients with MPS [21], both techniques demonstrated comparable improvements in pain, disability, and quality of life, with no significant difference between groups at any time point. Given these comparable clinical outcomes, differences in anatomical spread may be relevant when considering the potential advantages of SPSIPB. ESPB provides broad craniocaudal spread, although its ventral diffusion and engagement of the dorsal rami may be variable [40, 41]. In contrast, RIB offers more predictable dorsal coverage but has limited cephalad extension beyond the second rib, which may restrict its applicability in cervicoscapular pain [21]. By targeting the fascial plane between the SPS and intercostal muscles, SPSIPB may facilitate coverage of the upper thoracic and lower cervical dermatomes, potentially addressing some of the anatomical and technical limitations of these approaches. However, these assumptions require confirmation in well-designed comparative studies. Additional interfascial approaches have similarly demonstrated favorable short-term analgesic effects. Domingo et al. [42] demonstrated that ultrasound-guided interfascial injection to the trapezius muscle using 10 mL of 0.125% bupivacaine resulted in a marked immediate reduction in pain intensity (from 7.6 to 1.6 within 10 min); however, only the immediate analgesic effect was evaluated, and no short- or long-term follow-up data were reported. Recent evidence has also explored the trapezius–rhomboid plane block (TRB) [43] as a novel interfascial approach for the management of MPS. In a case series including six patients with dorsal MPS, TRB was performed using 10 mL of 0.25% bupivacaine combined with 40 mg methylprednisolone, resulting in a substantial reduction in pain intensity of approximately 70–80% during short-term follow-up (up to 8–10 weeks). These findings indicate that while various interfascial techniques consistently provide robust short-term analgesia, differences in injectate composition, volume, anatomical target, and follow-up duration may substantially influence the persistence of clinical benefit. In this context, the present study adds to the existing literature by showing reductions in pain intensity alongside improvements in functional disability and quality of life over a 6-month follow-up period, suggesting that SPSIPB may represent a potentially useful interfascial approach in cervicothoracic MPS. Future prospective studies evaluating different injectate volumes and compositions are warranted to determine the optimal protocol and to better define their impact on fascial spread, analgesic duration, and safety. This study has several limitations. First, its retrospective design limits causal inference, and the absence of a control or comparator group precludes definitive attribution of the observed effects to the block itself. A placebo effect or regression-to-the-mean phenomenon cannot be fully excluded in the absence of a sham or alternative intervention group. Although post-baseline co-interventions were systematically documented and incorporated into the primary analysis as a time-varying covariate, their indications and timing were not standardized due to the retrospective nature of the data. Consequently, residual time-varying confounding cannot be completely excluded. Furthermore, potential attrition-related bias must be acknowledged, as patients with complete follow-up exhibited higher baseline NRS scores than those with incomplete follow-up or post-baseline co-interventions. This baseline disparity may have influenced the magnitude of observed improvement, potentially overestimating the treatment effect. The absence of a prospective sample size calculation inherent to the retrospective design, may limit statistical precision and the generalizability of the results. In addition, the relatively wide age range of the study population may introduce heterogeneity, as age-related degenerative changes could have coexisted and influenced pain perception despite standardized diagnostic criteria. Finally, the lack of objective quantitative measures, such as pressure pain threshold, limits the ability to evaluate changes in local nociceptive sensitivity beyond patient-reported outcomes. Nevertheless, the inclusion of all eligible patients and the use of an all-inclusive analytical approach incorporating all available observations enhance the real-world applicability of the findings. Despite these limitations, to our knowledge, this is the first longitudinal real-world cohort evaluating SPSIPB in patients with cervicothoracic MPS The present study adds clinically relevant follow-up data on its use in this population, extending prior reports focused mainly on technical feasibility and short-term outcomes by showing improvements in pain, function, and quality of life over a six-month follow-up. The use of an all-inclusive modeling strategy, incorporating post-baseline co-interventions as a time-varying covariate, may enhance clinical interpretability by reflecting routine care pathways. Given its targeted anatomical approach and observed safety profile, SPSIPB may be considered a feasible minimally invasive option for patients with refractory cervicothoracic myofascial pain. However, these findings should be interpreted cautiously. Well-designed comparative trials are needed to clarify relative effectiveness versus ESPB, RIB, and trigger-point injections, to optimize injectate composition, and to better understand underlying mechanisms. The integration of quantitative imaging methods, such as ultrasound elastography, may further help elucidate fascial changes and their relationship with clinical outcomes. Conclusion Ultrasound-guided SPSIPB may be a feasible interventional option for cervicothoracic MPS, showing improvements in pain, function, and quality of life over six months. Further prospective controlled studies are needed to confirm these findings and clarify its role in multimodal pain management. Supplementary Information Supplementary Material 1. Supplementary Material 1.