Ecological or recalled assessments in chronic musculoskeletal pain? A comparative study of prospective and recalled pain assessments in low back pain and lower limb painful osteoarthritis.

OBJECTIVES: The objective of this study was to compare ecological and recalled pain intensity assessments over 29 days in hip and knee osteoarthritis (O) and chronic low back pain (L). PATIENTS AND METHODS: Rheumatologists were asked to enroll patients with O and L, with pain intensity above 40 mm, in a prospective study for 29 days. Pain intensity was assessed with physicians on Days 1 and 29, and ecologically, over the intervening 28-day period, by random phone calls. RESULTS: We carried out correlation analyses for 353 (159 O, 194 L) patients: Overall recalled daily pain was strongly correlated with calculated 3-day mean pain assessments (r=0.96 [O]; 0.93 [L]) and evening pain (r=0.96 [O], 0.90 [L]). Correlations between ecological and recalled measures were stronger for recall over the last 7 days than for recall over the last 28 days in osteoarthritis patients (r=0.78, r=0.63), but were similar for both recall periods in low back pain patients (r=0.70, r=0.72). Correlations between assessments for the last 7 and 28 days were stronger for ecological (r=0.88 [O], 0.91 [L]) than for clinical (r=0.77 [O]; 0.86 [L]) assessments. After adjustment for current pain intensity, correlations remained significant for ecological assessments, but not for clinical assessments. Recalled pain assessments were more accurate when made after 24 hours (r=0.71 [O]; 0.70 [L]) than when made after 48 hours (r=0.63 [O]; 0.61 [L]). CONCLUSIONS: For both low back pain and osteoarthritis, overall daily pain recall is a reliable measurement correlated with daily ecological measurements, whereas a rapid decrease in recall occurs after 48 hours. The most reliable period for pain recall was 7 days, but the results obtained were influenced by current pain.