Objective
In 2011, the American College of Rheumatology (ACR) and EULAR endorsed provisional criteria for remission in rheumatoid arthritis (RA), both Boolean‐ and index‐based. Based on recent studies indicating that a higher threshold for the patient global assessment (PtGA) may improve agreement between the 2 sets of criteria, our goals were to externally validate a revision of the Boolean remission criteria using a higher PtGA threshold and to validate the provisionally endorsed index‐based criteria.
Methods
We used data from 4 randomized trials comparing biologic disease‐modifying antirheumatic drugs to methotrexate or placebo. We tested the higher proposed PtGA threshold of 2 cm (Boolean2.0) (range 0–10 cm) compared to the original threshold of 1 cm (Boolean1.0). We analyzed agreement between the Boolean‐ and index‐based criteria (Simplified Disease Activity Index [SDAI] and Clinical Disease Activity Index [CDAI]) for remission and examined how well each remission definition predicted later good physical function (Health Assessment Questionnaire [HAQ] score ≤0.5) and radiographic nonprogression.
Results
Data from 2,048 trial participants, 1,101 with early RA and 947 with established RA, were included. The proportion of patients with disease in remission at 6 months after treatment initiation increased when using Boolean2.0 compared to Boolean1.0, from 14.8% to 20.6% in early RA and 4.2% to 6.0% in established RA. Agreement between Boolean2.0 and the SDAI or CDAI remission criteria was better than for Boolean1.0, particularly in early disease. Boolean2.0, SDAI, and CDAI remission criteria had similar positive likelihood ratios (LRs) to predict radiographic nonprogression and a HAQ score of ≤0.5 (positive LR 3.8–4.3). The omission of PtGA (BooleanX) worsened the prediction of good functional outcomes.
Conclusion
Using the Boolean 2.0 criteria classifies more patients as achieving remission and increases the agreement with index‐based remission criteria without jeopardizing predictive value for radiographic or functional outcomes. This revised Boolean definition and the previously provisionally endorsed index‐based criteria were endorsed by ACR and EULAR.