This composite index includes seven domains (pain, function, fatigue, physical and psychological wellbeing, sleep disturbance and coping). Each domain is evaluated using a single question answered by a 0 to 10 numerical rating scale. Each domain also has a specific weight assigned by a patient survey. The RAID score is a continuous variable ranging from 0 (best) to 10 (worst)(Table ​(worst)(Table11).

In general, the results of clinical studies and trials are reported at group level, for example by the mean change from baseline and this makes determination of the relevance of the results for an individual patient challenging. In order to make interpretation easier, data may be presented at individual level by considering the proportion of patients with an improvement above a threshold of an important change from baseline. Moreover, apart from the concept of improvement ('feeling better), the concept of status ('feeling good') has become increasingly important [13]. In order to assess these individual outcomes, continuous outcome measures (absolute value or change in RAID score) for each patient must be converted into a dichotomous variable (that is, change from baseline above a clinically relevant cutoff defining an important improvement from the patient's perspective, or absolute value below a clinically relevant cutoff defining an acceptable or good condition from the patient's perspective). These cutoffs have been called Minimal Clinically Important Improvement (MCII) for improvement and Patient Acceptable Symptom State (PASS) for status [14-17].

Three types of technique have been proposed to determine thresholds. According to the first simple empirical method, an absolute change of at least 1 or 2 points on a 0 to 10 scale [18,19] or a relative change of at least 20, 30 or 50% [20,21] have been proposed as thresholds for several patient-reported outcomes in rheumatic disorders. The second technique considers a change to be relevant when it exceeds the measurement error [22]. The third technique uses a gold standard anchor (usually the patient's global assessment) to determine the threshold from the best ratio between sensitivity and specificity using the receiver operating characteristic (ROC) curve [23], correct classification probabilities [24] or the 75^th percentile [14,15].

To our knowledge, despite the recognition of the validity of the RAID questionnaire, no formal threshold has been proposed to present results at the individual level. Furthermore, although several techniques have been used to establish threshold values as detailed above, no comparison has been performed between these techniques.

This study has some weaknesses, but also several strengths. The very wide range of threshold values proposed using different methodologies raises the question of the optimal way to address this issue. All the techniques used in this study have been previously adopted, though no consensus has been reached in the field of clinical epidemiology [10-21]. This can be easily explained by the different rationales of each technique: the empirical technique involves asking physicians to propose relevant thresholds based on the simplicity of their proposal or their experience [20]. The aim of another technique is to avoid proposing a value below the measurement error of the outcome measure, as any interpretation of results using a threshold below the noise due to this measurement error is hazardous [22,23]. Finally, the techniques using an external anchor are also very relevant [14,15]. Although the validity of this external anchor may be questioned (here we used the previously reported gold standard MCII and PASS questions, which might raise the issue of a circular reasoning), these techniques make it possible to select the optimal threshold based on the arguments for and against, favoring sensitivity (for example, 75^th percentile technique [14,15]), sensitivity and specificity (for example, ROC curve and correct probability technique [23,24]). In this study therefore, we decided to use all the different techniques in a uniform group of patients (for example, active definite RA requiring a TNF blocker) receiving the same TNF blocker (etanercept). Despite this fact, we observed a very wide variability in the thresholds proposed by these analyses. From our point of view, such variability justifies a systematic evaluation of the validity of any proposed threshold and the main question is to define the optimal methodology for evaluating such validity. In this study, we approached this question by calculating the capacity of a proposed threshold to adequately classify a patient by considering previously validated external anchors from both a patient's perspective and a physician's perspective. The MCII and PASS questions were considered to be a gold standard anchor for the patient's perspective [14,15]. Because we also used the MCII and PASS questions for the elaboration of such thresholds, one might be concerned by the potential circular reasoning of this approach. This is why we decided to use not only this external anchor but also another one (the DAS28), which is considered a physician's perspective [26], while evaluating a patient. We then calculated the positive LR. This approach, using two different external anchors resulted in quite good concordance between the two analyses for each proposed threshold, strengthening our finding. This agrees with the results of a previous study suggesting that the PASS corresponds to moderate disease activity [29]. The data presented in the figures suggest also that the most stringent thresholds are also the most valid, at least with regard to our definition of external validity.