Section for Sports Traumatology M51, Bispebjerg and Frederiksberg Hospital, Copenhagen, Denmark.
The Research Unit for General Practice and Section of General Practice, Department of Public Health, University of Copenhagen, Copenhagen, Denmark.
Scand J Med Sci Sports. 2021 May;31(5):982-990. doi: 10.1111/sms.13880. Epub 2020 Dec 14.
Deviations from adequate use and reporting of PROMs may be problematic and misleading. The aim of this study was to investigate the extent of such problems in randomized clinical trials (RCTs). RCTs involving sports medicine research that used PROMs as primary outcomes were identified in 13 preselected journals. The articles were reviewed for nine potential problems related to how the PROM was used and how the data had been reported. The potential problems were as follows: aggregating subscale scores; combining patient-reported scores with physical, clinical, or para-clinical measures; using a PROM to diagnose or evaluate the individual patient; using a PROM for one leg or arm; selectively excluding domains or items; constructing a PROM for the specific occasion; mixing PROM formats (ie, digital, paper, telephone, e-mail, in person); ambiguous instructions for how the PROM should be completed; and recall bias. As covariates, we registered journal impact factor, year of publication, and existence of a registered protocol. In 29 (53.7%) of 54 identified RCTs, at least one potential problem was identified, the most common being aggregation of domain scores. This was not different with a published protocol or dependent on journal rankings, except for exclusion of domains, which was most common in high-ranking journals. Aggregation of domain scores was significantly less common in recently published articles compared with older articles (P = .03). Potential problematic use of PROMs and reporting of PROM data are common in RCTs, also in high-ranking journals, but less so in more recent articles.
在使用和报告 PROMs 时出现偏差可能会产生问题和误导。本研究旨在调查随机对照试验 (RCTs) 中存在此类问题的程度。在 13 种预先选定的期刊中确定了涉及运动医学研究并将 PROM 作为主要结果的 RCTs。对这些文章进行了九个与 PROM 使用方式和数据报告方式相关的潜在问题的审查。这些潜在问题如下：聚合子量表分数；将患者报告的分数与身体、临床或临床相关措施相结合；使用 PROM 进行诊断或评估个体患者；使用 PROM 评估一条腿或一只手臂；选择性排除特定领域或项目；为特定场合构建 PROM；混合 PROM 格式（即数字、纸质、电话、电子邮件、亲自）；如何完成 PROM 的说明不明确；以及回忆偏差。作为协变量，我们登记了期刊影响因子、出版年份以及是否存在注册方案。在 54 项确定的 RCT 中，有 29 项（53.7%）至少发现了一个潜在问题，最常见的是领域分数的聚合。这与已发表的方案或期刊排名无关，除了排除领域，这在高排名期刊中最为常见。与较旧的文章相比，最近发表的文章中聚合领域分数的情况明显较少（P =.03）。在 RCT 中，尤其是在高排名期刊中，PROM 的潜在问题使用和 PROM 数据的报告很常见，但在最近的文章中则较少。
