Abstract
BACKGROUND: Prognostic research in low back pain (LBP) is essential for understanding and managing the condition. This study aimed to, (1) describe the proportions with mild-moderate and severe pain and disability at baseline, 1-year and 4-year follow-up, and (2) investigate prognostic factors for improvement in pain and disability over 4 years in a cohort of secondary care LBP patients. METHODS: This was a secondary analysis of a cohort of patients with LBP aged 18-40 years recruited from a non-surgical outpatient spine clinic between March 2011 and October 2013 (n = 1037). Questionnaires were collected at baseline, 1-year, and 4-year follow-up. Disability was assessed using the Roland Morris Disability Questionnaire (RMDQ 0-100) and pain intensity using the Numeric Rating Scale (NRS 0-10). 'Mild-moderate pain' was defined as NRS < 7 and 'severe pain' as NRS ≥ 7. Likewise, 'mild-moderate disability' was defined as RMDQ < 58.3, and 'severe disability' was RMDQ ≥ 58.3. In the prognostic analysis, improvement in pain and disability over 4 years was defined as meeting both criteria: decrease of ≥ 2 on the NRS and of ≥ 20.8 on the RMDQ. Sixteen candidate prognostic factors were assessed by multivariate logistic regression. RESULTS: Among patients with information available at all three time points (n = 241), 54%/48% had persistent mild-moderate pain/disability, while only 7%/15% had persistent severe pain/disability. Of patients included in the multivariate prognostic analysis regarding improvement over 4 years (n = 498), 32% had improved in pain and disability after 4 years. Positive associations were found for pain intensity (OR 1.34 [95%CI: 1.17-1.54]), disability (OR 1.01 [1.00-1.02]), and regular employment or studying (OR 1.67 [1.06-2.64]), and negative associations for episode duration (OR 0.99 [0.99-1.00]) and risk of persistent pain (OR 0.58 [0.38-0.88]). CONCLUSION: Patients with persistent LBP in secondary care had mostly mild-moderate pain and disability consistently at all three time points, with few having consistently severe symptoms over 4 years. Moreover, approximately half of the included patients improved in pain and disability. We found that pain intensity, disability, episode duration, regular employment or studying, and risk of persistent pain predicted a long-term improvement. However, the limited availability of complete follow-up data may affect generalisability.