Recall Bias in Low Back Pain Among Workers: Effects of Recall Period and Individual and Work-Related Factors

STUDY DESIGN. A prospective cohort study. OBJECTIVE. The aim of the present study was first to compare monthly measurements of low back pain (LBP) with quarterly and yearly retrospective measurements of LBP, and second to investigate possible bias effects for recall bias of LBP. SUMMARY OF BACKGROUND DATA. LBP is a subjective experience often measured by a single rating of recalled pain over a certain time interval. However, retrospectively reported pain may be subject to recall bias. METHODS. The agreement between monthly measurements of LBP and quarterly and yearly retrospective measurements of LBP was evaluated by Spearman rank order correlation coefficients, intraclass correlation coefficients, and Bland-Altman plots. Bias effects for recall bias were investigated by a linear regression model. RESULTS. There were no statistical significant differences in mean values of monthly measurements of LBP compared with quarterly and yearly retrospective measurements of LBP on a group level. However, the Bland-Altman plots revealed that within individuals, the difference between monthly measurements of LBP and quarterly and yearly retrospective measurements of LBP was highly variable. For both quarterly and yearly recall, social support from colleagues and average LBP days were significantly associated with the recall bias. CONCLUSION. The agreement of pain recall among workers in the current study seems to be good on a group level, but both between and within individuals, the difference between monthly and quarterly and yearly retrospective measurements was quite high. Factors that impacted the recall bias were social support from colleagues and average LBP days over the recall period. Level of Evidence: 3