Background 

The American Academy of Orthopaedic Surgeons recently proposed quality measures for the initial surgical treatment of carpal tunnel syndrome (CTS). One measure addressed avoidance of adjunctive surgical procedures during carpal tunnel release; and a second measure addressed avoidance of routine use of clinic-based occupational and/or physical therapy (OT/PT) after carpal tunnel release. However, for quality measures to serve their intended purposes, they must be tested in real-world data to establish that gaps in quality exist and that the measures yield reliable performance information.

Questions/purposes 

(1) Is there an important quality gap in clinical practice for avoidance of adjunctive surgical procedures during carpal tunnel release? (2) Is there an important quality gap in avoiding routine use of clinic-based occupational and/or physical therapy after carpal tunnel release? (3) Do these two quality measures have adequate beta-binomial signal-to-noise ratio (SNR) and split-sample reliability (SSR)?

Methods 

This retrospective comparative study used a large national private insurance claims database, the 2018 Optum Clinformatics® Data Mart. Ideally, healthcare quality measures are tested within data reflective of the providers and payors to which the measures will be applied. We previously tested these measures in a large public healthcare system and a single academic medical center. In this study, we sought to test the measures in the broader context of patients and providers using private insurance. For both measures, we included the first carpal tunnel release from 28,083 patients performed by one of 7236 surgeons, irrespective of surgical specialty (including, orthopaedic, plastic, neuro-, and general surgery). To calculate surgeon-level descriptive and reliability statistics, analyses were focused on the 66% (18,622 of 28,083) of patients who received their procedure from one of the 24% (1740 of 7236) of surgeons with at least five carpal tunnel releases in the database. No other inclusion/exclusion criteria were applied. To determine whether the measures reveal important gaps in treatment quality (avoidance of adjunctive procedures and routine therapy), we calculated descriptive statistics (median and interquartile range) of the performance distribution stratified by surgeon-level annual volume of carpal tunnel releases in the database (5+, 10+, 15+, 20+, 25+, and 30+). Like the Centers for Medicare & Medicaid Services (CMS), we considered a measure “topped out” if median performance was greater than 95%, meaning the opportunity for further quality improvement is low. We calculated the surgeon-level beta-binomial SNR and SSR for each measure, each stratified by the number of carpal tunnel releases performed by each surgeon in the database. These are standard measures of reliability in health care quality measurement science. The SNR quantifies the proportion of variance that is between rather than within surgeons, and the SSR is the correlation of performance scores when each surgeons’ patients are split into two random samples and then corrected for sample size.

Results 

We found that 2% (308 of 18,622) of carpal tunnel releases involved an adjunctive procedure. The results showed that avoidance of adjunctive surgical procedures during carpal tunnel release had a median (IQR) performance of 100% (100% to 100%) at all case volumes. Only 8% (144 of 1740) of surgeons with at least five cases in the database had less than 100% performance, and only 5% (84 of 1740) had less than 90% performance. This means adjunctive procedures were rarely performed and an important quality gap does not exist based on the CMS criterion. Regarding the avoidance of routine therapy, there was a larger quality gap: For surgeons with at least five cases in the database, median performance was 89% (75% to 100%), and 25% (435 of 1740) of these surgeons had less than 75% performance. This signifies that the measure is not topped out and may reveal an important quality gap. Most patients receiving clinic-based OT/PT had only one visit in the 6 weeks after surgery. Median (IQR) SNRs of the first measure, which addressed avoidance of adjunctive surgical procedures, and the second measure, which addresses avoidance of routine use clinic-based OT/PT, were 1.00 (1.00 to 1.00) and 0.86 (0.67 to 1.00), respectively. The SSR for these measures were 0.87 (95% CI 0.85 to 0.88) and 0.75 (95% CI 0.73 to 0.77), respectively. All of these reliability statistics exceed National Quality Forum’s emerging minimum standard of 0.60.

Conclusion 

The first measure, the avoidance of adjunctive surgical procedures during carpal tunnel release, lacked an important quality gap suggesting it is unlikely to be useful in driving improvements. The second measure, avoidance of routine use of clinic-based OT/PT, revealed a larger quality gap and had very good reliability, suggesting it may be useful for quality monitoring and improvement purposes.

Clinical Relevance 

As healthcare systems and payors use the second measure, avoidance of routine use of clinic-based OT/PT, to encourage adherence to clinical practice guidelines (such as provider profiling, public reporting, and payment policies), it will be critically important to consider what proportion of patients receiving OT/PT should be considered routine practice and therefore inconsistent with guidelines. The value or potential harm of this measure depends on this judgement.

中文翻译：

---

拟议的腕管释放质量措施是否揭示了重要的质量差距以及它们是否可靠？

 背景

美国骨科医师学会最近提出了腕管综合征（CTS）初始手术治疗的质量措施。一项措施解决了在腕管松解期间避免辅助手术的问题；第二项措施涉及避免在腕管松解术后常规使用基于临床的职业和/或物理治疗（OT/PT）。然而，为了使质量措施达到其预期目的，必须在现实世界的数据中对其进行测试，以确定质量方面是否存在差距，并且这些措施可以产生可靠的绩效信息。

 问题/目的

(1) 在临床实践中，对于避免腕管松解期间的辅助手术程序，是否存在重要的质量差距？ (2) 腕管松解术后避免常规使用基于临床的职业和/或物理治疗是否存在重要的质量差距？ (3) 这两个质量指标是否具有足够的 β 二项式信噪比 (SNR) 和分割样本可靠性 (SSR)？

 方法

这项回顾性比较研究使用了大型全国私人保险索赔数据库，即 2018 Optum Clinformatics® 数据集市。理想情况下，医疗保健质量措施是在反映这些措施所适用的提供者和付款人的数据中进行测试的。我们之前在大型公共医疗保健系统和单个学术医疗中心测试了这些措施。在这项研究中，我们试图在使用私人保险的患者和提供者的更广泛背景下测试这些措施。对于这两项测量，我们纳入了由 7236 名外科医生中的一名对 28,083 名患者进行的首次腕管松解术，无论手术专业如何（包括整形外科、整形外科、神经外科和普通外科）。为了计算外科医生水平的描述性和可靠性统计数据，分析的重点是 66%（28,083 名中的 18,622 名）患者接受了 24%（7236 名中的 1740 名）外科医生之一的手术，该外科医生在手术期间至少进行过 5 次腕管松解术。数据库。没有应用其他纳入/排除标准。为了确定这些措施是否揭示了治疗质量方面的重要差距（避免辅助手术和常规治疗），我们计算了数据库中按外科医生水平每年腕管释放量分层的性能分布的描述性统计数据（中位数和四分位数范围）（ 5+、10+、15+、20+、25+ 和 30+）。与 Medicare 和 Medicaid 服务中心 (CMS) 一样，如果中值绩效高于 95%，我们认为该衡量标准已“达到顶峰”，这意味着进一步提高质量的机会很小。我们计算了每项测量的外科医生水平 β 二项式 SNR 和 SSR，每个测量均按数据库中每位外科医生执行的腕管松解次数进行分层。 这些是医疗保健质量测量科学中可靠性的标准测量。 SNR 量化了外科医生之间而不是内部的方差比例，SSR 是当每个外科医生的患者被分为两个随机样本，然后根据样本大小进行校正时，绩效得分的相关性。

 结果

我们发现 2%（18,622 例中的 308 例）腕管松解术涉及辅助手术。结果显示，在所有病例量中，在腕管松解期间避免辅助手术的中位 (IQR) 表现均为 100%（100% 至 100%）。在数据库中至少有 5 个病例的外科医生中，只有 8%（1740 名中的 144 名）的绩效低于 100%，只有 5%（1740 名中的 84 名）的绩效低于 90%。这意味着很少执行辅助程序，并且根据 CMS 标准，不存在重要的质量差距。关于避免常规治疗，存在较大的质量差距：对于数据库中至少有 5 个病例的外科医生，中位绩效为 89%（75% 至 100%），其中 25%（1740 名中的 435 名）有性能低于 75%。这意味着该措施尚未达到顶峰，并可能揭示出重要的质量差距。大多数接受临床 OT/PT 的患者在术后 6 周内仅就诊过一次。第一项措施（解决避免辅助外科手术）和第二项措施（解决常规临床 OT/PT）的中位 (IQR) SNR 分别为 1.00（1.00 至 1.00）和 0.86（0.67 至 1.00） ， 分别。这些措施的 SSR 分别为 0.87（95% CI 0.85 至 0.88）和 0.75（95% CI 0.73 至 0.77）。所有这些可靠性统计数据都超过了国家质量论坛新兴的 0.60 最低标准。

 结论

第一项措施是在腕管松解期间避免辅助手术，缺乏重要的质量差距，表明它不太可能有助于推动改进。第二项措施是避免常规使用临床 OT/PT，显示出较大的质量差距，并且具有非常好的可靠性，表明它可能有助于质量监测和改进目的。

 临床相关性

由于医疗保健系统和付款人使用第二种措施，即避免常规使用基于临床的 OT/PT，以鼓励遵守临床实践指南（例如提供者分析、公开报告和付款政策），因此考虑这一点至关重要有多少比例的患者接受 OT/PT 应被视为常规做法，因此不符合指南。这一措施的价值或潜在危害取决于这种判断。