Candidate for the Lee B. Lusted Student Prize Competition
Purpose: Current guidelines recommend treatment for rheumatoid arthritis (RA) be informed by a measure of disease activity such as the Disease Activity Score-28 (DAS28-CRP), which includes physician assessment of tenderness and swelling in joints. We explored the impact of basing treatment recommendations for RA patients on patient joint assessments rather than physician assessments.
Methods: The analysis used data for 364 patients and physicians enrolled in the Rheumatoid Arthritis Comparative Effectiveness Research (RACER) registry, who completed joint exams during routine clinic visits. We modeled discrepancies in the DAS28-CRP when based on patient vs. physician assessments, as well as discrepancies in disease activity categories and treatment recommendations. Treatment recommendations were compared with actual treatment decisions.
Results: Mean DAS28-CRP score based on patient assessment was higher than when based on physician assessment (mean ±SD, 3.49 ±1.50 vs. 3.02 ±1.25; p<0.0001). In 56% of 364 cases with DAS28-scores available, the patient and physician-based DAS28-CRP led to categorization in the same disease severity level; in 44% of cases the patient and physician-based scores led to discrepant categorizations. Following American College of Rheumatology (ACR) guidelines, in 84% of cases the physician and patient-based scores would yield the same treatment recommendation, but in 16% of cases they would yield different treatment recommendations. When evaluating actual decisions against treatment recommendations based on ACR guidelines (n=342), only 8.77% of patients actually augmented therapy, compared to 33.33% and 39.18% recommended to augment therapy according to physician and patient DAS28-CRP, respectively (Table 1; recommendations both significantly different from actual decisions at p<0.0001, but not significantly different from each other at p=0.1307). Actual rates of therapy augmentation are low among those recommended to augment therapy according to physician DAS28-CRP (16.7%) as well as patient DAS28-CRP (14.9%).
Conclusions: Although discrepancies in disease severity categorization occur frequently (44% of the time), discrepancies in treatment recommendations arise less frequently (16% of the time) as a result of differences in patient and physician joint exams. Regardless of whether treatment recommendations are based on patient or physician joint exams, actual treatment decisions often appear to be inconsistent with disease severity, primarily because therapy is often not augmented despite indications of moderate to high disease activity. Further research will explore disease and health-related outcomes associated with miscategorizing disease severity or under-/over-treating patients.
See more of: The 35th Annual Meeting of the Society for Medical Decision Making