Objectives Patients, as well as healthy people, often fail to remember prior symptoms, illnesses, or treatments. The authors investigated how well people in a workplace recalled their prior musculoskeletal pain and which factors influenced recall accuracy.

Methods In this prospective study (1992–1998), among a cohort of automobile manufacturing workers (N=464), musculoskeletal pain reported at baseline was compared with recalled pain at follow-up. Two outcomes (ie, forgetting and over-recalling) were examined in relation to several personal and occupational characteristics.

Results Of those who had experienced pain or discomfort in the upper extremities at baseline, 72% did not recall it 6 years later. Symptoms at the time of recall strongly influenced pain recall; forgetting approached 100% among those with no current or recent symptoms. However, forgetting was considerable even among those with current symptoms (45%). Prior pain status was over-recalled by 37% of those with upper-extremity pain at the time of recall, but only by 3% of those without symptoms. Women, those with history of an upper-extremity disorder or systemic disease, who were clinical cases or had more anatomical areas in pain at baseline forgot less often. Over-recalling was related to age, having current symptoms, or being a clinical case. The results were similar for low-back and knee pain.

Conclusions Prior musculoskeletal symptoms are poorly remembered after some years, and the recall is strongly influenced by current symptoms. Recalled information should not be relied upon when an epidemiologic case definition is being constructed or when an attempt is being made to describe the natural history of a disorder.

Key terms epidemiology; memory; musculoskeletal pain; musculoskeletal symptom; recall; recall bias; self-assessment; self-report; working population