Objective: To determine the reliability of self-assessment of disease by patients with rheumatoid arthritis (RA), with a particular emphasis on the assessment of articular swelling.
Methods: A questionnaire was developed using components from validated instruments. Information was obtained on global function, global joint tenderness/swelling, and joint pain (10 cm visual analog scales), duration of musculoskeletal morning stiffness (grade 1-6), ACR functional score (grade 1-4), tender joint count (0-20), tender joint score (0-60), swollen joint count (0-20), and swollen joint score (0-60). Data were collected prospectively on 61 patients with RA in a teaching clinic or office practice by 4 staff rheumatologists. Patient questionnaires were completed within 24 h before physician assessments. Followup assessments were carried out on 27 patients after a mean interval of 3 months (range 0.5-6).
Results: At the initial assessment there was a significant correlation between patient and physician assessments for global function (r = 0.83; p = 0.01) [intraclass correlation coefficient (ICC) = 0.83; p < 0.01], global joint tenderness/swelling (r = 0.83; p < 0.01) (ICC = 0.83; p < 0.01), global joint pain (r = 0.83; p < 0.01) (ICC = 0.81; p < 0.01), duration of morning stiffness (r = 0.83; p < 0.01) (ICC = 0.85; p < 0.01), ACR functional score (r = 0.61; p < 0.01) (ICC = 0.62; p < 0.01), tender joint count (r = 0.57; p < 0.01) (ICC = 0.31; p < 0.01), and tender joint score (r = 0.60; p < 0.01) (ICC = 0.35; p < 0.01). However, there was poor correlation between patient and physician assessments for both swollen joint count (r = 0.16; p > 0.05) (ICC = -0.02; p > 0.05) and swollen joint score (r = 0.24; p > 0.05) (ICC = 0.12; p > 0.05). Longitudinal analysis indicated significant correlation between changes in patient and physician assessments in all variables except swollen joint count and score.
Conclusion: Although there was good correlation between most variables for patient and physician assessments of disease activity in RA, there were substantial differences in the assessment of joint swelling. This objective determinant of disease activity cannot be ascertained in self-report measures of health status.