Whittle Rebecca, Royle Kara-Louise, Jordan Kelvin P, Riley Richard D, Mallen Christian D, Peat George
Arthritis Research UK Primary Care Centre, Research Institute for Primary Care & Health Sciences, Keele University, Keele, Staffordshire ST5 5BG UK.
Diagn Progn Res. 2017 Feb 8;1:1. doi: 10.1186/s41512-016-0006-6. eCollection 2017.
Prognosis research studies (e.g. those deriving prognostic models or examining potential predictors of outcome) often collect information on time-varying predictors their intended moment of use, sometimes using a measurement method different to that which would be used. We aimed to illustrate how estimates of predictor-outcome associations and prognostic model performance obtained from such studies may differ to those at the earlier, intended moment of use.
We analysed data from two primary care cohorts of patients consulting for non-inflammatory musculoskeletal conditions: the Prognostic Research Study (PROG-RES: = 296, aged >50 years) and the Primary care Osteoarthritis Screening Trial (POST: = 756, >45 years). Both cohorts had collected comparable information on a potentially important time-varying predictor (current pain intensity: 0-10 numerical rating scale), other predictors (age, gender, practice) and outcome (patient-perceived non-recovery at 6 months). Using logistic regression models, we compared the direction and magnitude of predictor-outcome associations and model performance measures under two scenarios: (i) current pain intensity ascertained by the treating general practitioner in the consultation (the intended moment of use) and (ii) current pain intensity ascertained by a questionnaire mailed several days after the consultation.
In both cohorts, the predictor-outcome association was substantially weaker for pain measured at the consultation (OR (95% CI): PROG-RES 1.06 (0.95, 1.18); POST 1.04 (0.96, 1.12)) than for pain measured in the questionnaire (PROG-RES 1.34 (1.20, 1.48); POST 1.26 (1.18, 1.34)). The -statistic of the multivariable model was lower when pain was measured at the consultation (-statistic (95% CI): PROG-RES 0.57 (0.51, 0.64); POST 0.66 (0.62, 0.70)) than when pain was measured in the questionnaire (PROG-RES 0.69 (0.63, 0.75); POST 0.72 (0.68, 0.76)), reflecting the lower OR for pain at the consultation.
Prognostic research studies ideally should measure time-varying predictors at their intended moment of use and using the intended measurement method. Otherwise, they may produce substantially different estimates of predictor-outcome associations and model performance. Researchers should report when, how and where predictors were measured and identify any significant departures from their intended use that may limit the applicability of findings in practice.
The protocol for the PROG-RES cohort data collection and primary analysis has been published in an open-access journal (Mallen et al., BMC Musculoskelet Disord 7:84, 2006). The POST trial was registered (ISRCTN40721988; date of registration: 21 June 2011; date of enrolment of the first participant: 3 October 2011) and had a pre-specified protocol covering primary analysis. There was no published protocol for the current secondary analyses presented in this manuscript.
预后研究（例如那些推导预后模型或检验潜在结局预测因素的研究）通常会在预测因素的预期使用时刻收集随时间变化的预测因素信息，有时使用的测量方法与预期使用的方法不同。我们旨在说明从此类研究中获得的预测因素与结局关联的估计值以及预后模型性能可能与早期预期使用时刻的情况有所不同。
我们分析了两个因非炎性肌肉骨骼疾病就诊的初级保健队列的数据：预后研究（PROG-RES：n = 296，年龄>50岁）和初级保健骨关节炎筛查试验（POST：n = 756，>45岁）。两个队列都收集了关于一个潜在重要的随时间变化的预测因素（当前疼痛强度：0-10数字评分量表）、其他预测因素（年龄、性别、医疗机构）和结局（6个月时患者自我感觉未康复）的可比信息。使用逻辑回归模型，我们在两种情况下比较了预测因素与结局关联的方向和大小以及模型性能指标：（i）由主治全科医生在会诊时确定的当前疼痛强度（预期使用时刻），以及（ii）会诊后几天通过邮寄问卷确定的当前疼痛强度。
在两个队列中，与通过问卷测量的疼痛相比，会诊时测量的疼痛的预测因素与结局关联明显较弱（比值比（95%置信区间）：PROG-RES为1.06（0.95，1.18）；POST为1.04（0.96，1.12））（PROG-RES为1.34（1.20，1.48）；POST为1.26（1.18，1.34））。当在会诊时测量疼痛时，多变量模型的卡方统计量较低（卡方统计量（95%置信区间）：PROG-RES为0.57（0.51，0.64）；POST为0.66（0.62，0.70）），而在问卷中测量疼痛时（PROG-RES为0.69（0.63，0.75）；POST为0.72（0.68，0.76）），这反映了会诊时疼痛的比值比较低。
预后研究理想情况下应在预测因素的预期使用时刻并使用预期测量方法来测量随时间变化的预测因素。否则，它们可能会得出预测因素与结局关联以及模型性能的显著不同估计值。研究人员应报告预测因素的测量时间、方式和地点，并识别与预期使用的任何重大偏差，这些偏差可能会限制研究结果在实践中的适用性。
PROG-RES队列数据收集和初步分析的方案已发表在一本开放获取期刊上（Mallen等人，《BMC肌肉骨骼疾病》7：84，2006年）。POST试验已注册（ISRCTN号：40721988；注册日期：2011年6月21日；第一名参与者入组日期：2011年10月3日），并有一份涵盖初步分析的预先指定方案。本手稿中呈现当前二次分析的方案未发表。
