We read with great interest the paper ‘A Systematic Review and Meta-analysis of Radiofrequency Procedures on Innervation to the Shoulder Joint for relieving Chronic Pain’ by Pushparaj et al. (2021). This is an excellent and timely systematic assessment of predominately pulsed radiofrequency neurotomy (pRF) of the suprascapular nerve and we commend the authors for their robust methodology and willingness to hold up a commonly accepted pain clinic treatment to appropriate scientific scrutiny.

The authors concluded from a meta-analysis of seven randomized controlled trials that there was no analgesic benefit from pRF treatment over conventional medical management (CMM). This brings into question the appropriateness of current ongoing clinical use of this therapy.

We note that the 95% confidence interval for the mean difference ranged from −1.83 (significantly effective and in favour of radiofrequency treatment) to +0.05 (very minimally in favour of CMM). We also note that all but the study of Yan and Zhang (2019) involved small numbers of participants and have lower observed differences than the differences needed to be seen between the two means of the procedures if the power of the test is to be as high as 90%. Thus, the rationale for a meta-analysis against the comparator.

Our interest was piqued by reviewing the study of Eyigor et al. (2010), which, in contrast to the other six studies, showed 95% confident intervals all in favour of CMM. This seems to be an outlier with respect to the other studies, which were in favour of pRF at least on one end of the 95% confidence interval.

We extracted the data of what constituted CMM and, not surprisingly, it was heterogeneous. In the Eyigor study (2010), it consisted of a triple injection of local anaesthetic and steroid to the acromioclavicular joint, the subacromial bursa, and the glenohumeral joint. This provided slightly superior efficacy at 12 weeks versus. pRF of the suprascapular nerve. In the remaining studies, CMM consisted of transcutaneous electrical nerve stimulation, suprascapular nerve injection of 2 ml of 1% lidocaine, physical therapy, sham transcutaneous pRF, photobiomodulation and suprascapular nerve needle placement followed by sham radiofrequency treatment.

We note the inclusion criteria for the Eyigor study (2010) consisted of chronic shoulder pain, and that patients who had recently received a shoulder injection were excluded. It is highly likely that a heterogeneous group of patients with acromioclavicular joint pain, chronic subacromial bursitis and glenohumeral joint osteoarthritis were recruited, and it is possible that many of these patients were naive to local anaesthetic and steroid injection and had not failed that therapy. Thus, it is not surprising to see efficacy of a triple injection of local anaesthetic and corticosteroid occurred. We hypothesize that pRF of the suprascapular nerve would optimally treat the subgroup that had glenohumeral joint osteoarthritis but may be insufficient by itself to block pain arising from the acromioclavicular joint and the subacromial bursa/rotator cuff tendons. We agree with Pushparaj and colleagues when they state that additional neural targets may be required for more structures supplied by anterior nerves. In discussion with colleagues worldwide, it appears that pRF of the suprascapular nerve is mostly reserved for treatment of refractory glenohumeral joint osteoarthritis where the patient has already failed injections of local anaesthetic and steroid to relevant structures. We suspect that applying further such injections in a patient who had already failed them would not yield a highly clinically significant result.

If one excludes the Eyigor study (2010), it would appear that the meta-analysis falls significantly in favour of pRF. We kindly ask whether it is possible for the authors to rerun the meta-analysis with the Eyigor study excluded? Therefore, we propose that a more nuanced interpretation of the findings would be that pRF of the suprascapular nerve is more efficacious than a heterogeneous mix of CMM approaches, other than a comprehensive multi-target injection of local anaesthetic and steroid at 12 weeks.

Of course, the area of greatest interest is in the longer-term results and, ultimately, if the clinical benefit of pRF of the suprascapular nerve will succeed or fail based on 6 or 12-month results. It would perhaps be an opportune extension of the work of Pushparaj and colleagues to consider a meta-analysis of these and future studies with long term results. Given that the typical clinical duration of pRF appears to vary between 9 and 12 months, there may be a clearer diversion of efficacy between pRF and CMM over a longer time period.

We entirely agree with the authors’ call for high-quality anatomical dissections, large-scale registries and clinical studies evaluating different imaging techniques for radiofrequency electrode guidance. The authors have clearly outlined appropriate next steps, which we fully endorse and support. We thank the authors for the comprehensiveness of their paper and believe that their approach serves as an excellent model for assessing interventional pain techniques.

我们饶有兴趣地阅读了 Pushparaj 等人的论文“对肩关节神经支配的射频手术治疗慢性疼痛的系统评价和荟萃分析”。（2021 年）。这是对肩胛上神经主要脉冲射频神经切断术 (pRF) 的出色而及时的系统评估，我们赞扬作者稳健的方法论和愿意将普遍接受的疼痛诊所治疗置于适当的科学审查之下。

作者从七项随机对照试验的荟萃分析中得出结论，与传统医疗管理 (CMM) 相比，pRF 治疗没有镇痛益处。这使人们对目前正在进行的这种疗法的临床使用的适当性产生了疑问。

我们注意到平均差异的 95% 置信区间的范围从 -1.83（显着有效且有利于射频治疗）到 +0.05（极少有利于 CMM）。我们还注意到，除了 Yan 和 Zhang（2019 年）的研究外，所有研究都涉及少量参与者，并且观察到的差异低于如果测试的功效要如此高，则需要在两种程序之间看到的差异。作为 90%。因此，对比较者进行荟萃分析的基本原理。

回顾 Eyigor 等人的研究激起了我们的兴趣。( 2010 )，与其他六项研究相比，显示 95% 的置信区间都支持 CMM。与其他研究相比，这似乎是一个异常值，这些研究至少在 95% 置信区间的一端支持 pRF。

我们提取了构成 CMM 的数据，毫不奇怪，它是异质的。在 Eyigor 研究 (2010) 中，它包括向肩锁关节、肩峰下滑囊和盂肱关节注射局部麻醉剂和类固醇三次。与 12 周相比，这提供了略微优越的疗效。肩胛上神经的 pRF。在其余研究中，CMM 包括经皮神经电刺激、肩胛上神经注射 2 ml 1% 利多卡因、物理治疗、假经皮 pRF、光生物调节和肩胛上神经针放置，然后是假射频治疗。

我们注意到 Eyigor 研究（2010 年）的纳入标准包括慢性肩痛，并且排除了最近接受肩部注射的患者。很可能招募了一组患有肩锁关节痛、慢性肩峰下滑囊炎和盂肱关节骨关节炎的异质性患者，并且这些患者中的许多人可能对局部麻醉剂和类固醇注射不熟悉，并且没有使治疗失败。因此，看到局部麻醉剂和皮质类固醇三重注射的疗效并不令人惊讶。我们假设肩胛上神经的 pRF 将最佳治疗患有盂肱关节骨关节炎的亚组，但其本身可能不足以阻止由肩锁关节和肩峰下囊/肩袖肌腱引起的疼痛。我们同意 Pushparaj 及其同事的观点，他们指出前神经提供的更多结构可能需要额外的神经目标。在与世界各地同事的讨论中，似乎肩胛上神经的 pRF 主要用于治疗难治性盂肱关节骨关节炎，患者已经无法向相关结构注射局部麻醉剂和类固醇。我们怀疑对已经失败的患者进行进一步的此类注射不会产生具有高度临床意义的结果。我们同意 Pushparaj 及其同事的观点，他们指出前神经提供的更多结构可能需要额外的神经目标。在与世界各地同事的讨论中，似乎肩胛上神经的 pRF 主要用于治疗难治性盂肱关节骨关节炎，其中患者已经未能向相关结构注射局部麻醉剂和类固醇。我们怀疑对已经失败的患者进行进一步的此类注射不会产生具有高度临床意义的结果。我们同意 Pushparaj 及其同事的观点，他们指出前神经提供的更多结构可能需要额外的神经目标。在与世界各地同事的讨论中，似乎肩胛上神经的 pRF 主要用于治疗难治性盂肱关节骨关节炎，其中患者已经未能向相关结构注射局部麻醉剂和类固醇。我们怀疑对已经失败的患者进行进一步的此类注射不会产生具有高度临床意义的结果。似乎肩胛上神经的 pRF 主要用于治疗难治性盂肱关节骨关节炎，其中患者已经向相关结构注射局部麻醉剂和类固醇失败。我们怀疑对已经失败的患者进行进一步的此类注射不会产生具有高度临床意义的结果。似乎肩胛上神经的 pRF 主要用于治疗难治性盂肱关节骨关节炎，其中患者已经向相关结构注射局部麻醉剂和类固醇失败。我们怀疑对已经失败的患者进行进一步的此类注射不会产生具有高度临床意义的结果。

如果排除 Eyigor 研究（2010 年），则荟萃分析似乎明显偏向于 pRF。我们恳请作者是否有可能在排除 Eyigor 研究的情况下重新进行荟萃分析？因此，我们建议对研究结果进行更细微的解释，即肩胛上神经的 pRF 比 CMM 方法的异质混合更有效，除了在 12 周时全面多靶点注射局部麻醉剂和类固醇。

当然，最感兴趣的领域是长期结果，最终，根据 6 个月或 12 个月的结果，肩胛上神经 pRF 的临床益处是成功还是失败。Pushparaj 和他的同事考虑对这些和未来研究进行荟萃分析，并获得长期结果，这也许是一个适当的扩展。鉴于 pRF 的典型临床持续时间似乎在 9 到 12 个月之间变化，因此在更长的时间内，pRF 和 CMM 之间的疗效可能会有更明显的转移。

我们完全同意作者对高质量解剖解剖、大规模注册和临床研究的呼吁，以评估用于射频电极引导的不同成像技术。作者清楚地概述了适当的后续步骤，我们完全赞同和支持。我们感谢作者的论文的全面性，并相信他们的方法可以作为评估介入疼痛技术的优秀模型。