Sandvall Brinkley, Okoroafor Ugochi C, Gerull William, Guattery Jason, Calfee Ryan P
Department of Orthopedic Surgery, Washington University School of Medicine, St. Louis MO.
Department of Orthopedic Surgery, Washington University School of Medicine, St. Louis MO.
J Hand Surg Am. 2019 Jun;44(6):454-459.e1. doi: 10.1016/j.jhsa.2019.02.015. Epub 2019 Apr 4.
This study was conducted to determine the minimal clinically important difference (MCID) of the Patient-Reported Outcomes Information System (PROMIS) Physical Function computer adaptive test (CAT) after distal radius fracture.
This study retrospectively analyzed data from 187 adults receiving nonsurgical care for a unilateral distal radius fracture at a single institution between February 2016 and November 2017. All patients completed the PROMIS Physical Function v1.2/2.0 CAT at each visit. At follow-up, patients also completed 2 multiple-choice clinical anchor questions querying their overall response to treatment. The MCID estimate was then calculated with an anchor-based method as the mean PROMIS Physical Function score change for the group reporting mild improvement and with a distribution-based method considering effect sizes of change and the minimum detectable change (MDC). The MCID estimate was examined for the influence of patient age, follow-up interval, and initial PROMIS score.
Change in PROMIS Physical Function scores between visits was significantly different between patients reporting no change, mild improvement, and much improvement on the anchor questions. The anchor-based MCID estimate for PROMIS Physical Function was 3.6 points (SD, 8.4). Among patients reporting mild improvement, individual changes in PROMIS Physical Function were not correlated with patient age or time between visits but were moderately negatively correlated with the initial absolute PROMIS Physical Function score. Applying the effect size parameters to our data when patients indicated minimal change, the distribution-based MCID estimate was 4.6 (SD, 1.8). Both the anchor-based and the distribution-based MCID estimates were judged sufficient because they exceeded the MDC value of 2.3.
The MCID value for PROMIS Physical Function is estimated between 3.6 and 4.6 in patients treated nonsurgically for distal radius fractures. Clinical improvement is associated with smaller magnitudes of change on PROMIS Physical Function when patients present with better reported function.
The MCID estimations are needed to determine the clinical relevance of changes in PROMIS scores and to more accurately calculate sample sizes needed for research incorporating PROMIS.
本研究旨在确定桡骨远端骨折后患者报告结局信息系统（PROMIS）身体功能计算机自适应测试（CAT）的最小临床重要差异（MCID）。
本研究回顾性分析了2016年2月至2017年11月期间在单一机构接受非手术治疗的187例单侧桡骨远端骨折成人的数据。所有患者在每次就诊时均完成PROMIS身体功能v1.2/2.0 CAT。在随访时，患者还完成了2个多项选择题临床锚定问题，询问他们对治疗的总体反应。然后采用基于锚定的方法计算MCID估计值，即报告轻度改善的组的PROMIS身体功能评分变化均值，并采用基于分布的方法，考虑变化的效应大小和最小可检测变化（MDC）。研究了MCID估计值受患者年龄、随访间隔和初始PROMIS评分的影响。
在锚定问题上报告无变化、轻度改善和显著改善的患者之间，就诊间PROMIS身体功能评分的变化有显著差异。PROMIS身体功能基于锚定的MCID估计值为3.6分（标准差，8.4）。在报告轻度改善的患者中，PROMIS身体功能的个体变化与患者年龄或就诊间隔时间无关，但与初始PROMIS身体功能绝对评分呈中度负相关。当患者表示变化最小时，将效应大小参数应用于我们的数据，基于分布的MCID估计值为4.6（标准差，1.8）。基于锚定和基于分布的MCID估计值均被判定为足够，因为它们超过了2.3的MDC值。
非手术治疗桡骨远端骨折患者的PROMIS身体功能MCID值估计在3.6至4.6之间。当患者报告的功能较好时，临床改善与PROMIS身体功能较小的变化幅度相关。
需要MCID估计值来确定PROMIS评分变化的临床相关性，并更准确地计算纳入PROMIS的研究所需的样本量。
