Method
Pain drawings were collected from thirty-four patients with various types of musculoskeletal pain. On the first clinic visit, patients were instructed to draw in their pain areas on a diagram depicting the body (P1). The examining physiatrist, blinded to the patients drawing, also drew in the pain areas on a separate diagram based on thorough history taking (D). After 2 to 9 days without treatment, patients repeated their pain drawings (P2). A second physiatrist input each of the pain drawings into the Pain Chart System (PCS), for comparison of pain distribution sites and areas among the three drawings. The weighted-sum of comparison algorithms was calculated for similarity determinations between drawings.

Results
The similarity values between P1 and P2, and P1 and D were 0.63±0.18 (mean±S.D.) and 0.62±0.18 (mean±S.D.), respectively, reflecting a high reliability of pain drawings.

Conclusion
High test-retest and inter-individual reliability of successive pain drawings suggest that pain drawings may be a useful tool to describe the painful sites. (J Korean Acad Rehab Med 2003; 27: 404-409)