Objective: To construct percentile curve reference charts of physical function in a population of patients with rheumatoid arthritis (RA); to explore the relationship of age, gender, and disease duration on physical function using percentile curve reference charts; to explore the potential clinical applications of percentile curve reference charts.
Methods: We surveyed 358 patients with RA from a teaching hospital clinic and 4 rheumatological private practices. The study factors were age, sex, disease duration, and physical function (Health Assessment Questionnaire). The sample percentiles were derived empirically, using the weighted average method, and their distribution-free confidence limits were calculated. Cubic spline interpolation curves were used to smooth the percentile lines.
Results: We constructed percentile curve reference charts of physical function. The best time-dependent variable was increasing disease duration rather than increasing age. There was no overlap of the 95% confidence limits for the 10th, 50th, and 90th percentile curves.
Conclusion: Percentile curve reference charts could be used (1) to describe the distribution of health status in a defined population; (2) as an index of change, enabling clinicians to judge the success or failure of therapeutic interventions in terms of movement of their patients' values across percentiles of function; and (3) to set RA management objectives developed from population based norms.