Estimating the cost and epidemiology of mild to severe chronic pain associated with osteoarthritis in England: a retrospective analysis of linked primary and secondary care data

OBJECTIVE: Osteoarthritis (OA) affects 10% of adults in the UK. Despite over one-third of people with OA experiencing chronic pain, few studies have examined the population-level impact of chronic pain associated with OA. We compared resource-use and epidemiological outcomes in patients with mild, moderate and severe chronic OA-associated pain and matched controls without known OA. DESIGN: Retrospective, longitudinal, observational cohort study (July 2008 to June 2019). SETTING: Electronic records extracted from Clinical Practice Research Datalink GOLD primary care linked to Hospital Episode Statistics (HES). PARTICIPANTS: Patients (cases; n=23 016) aged ≥18 years with chronic OA-associated pain. Controls (n=23 016) without OA or chronic pain matched on age, sex, comorbidity burden, general practitioner practice and available HES data. INTERVENTIONS: None. PRIMARY AND SECONDARY OUTCOME MEASURES: Total healthcare resource use (HCRU), direct healthcare costs in 0–12, 12–24 and 24–36 months postindex. Secondary outcomes included incidence and prevalence of chronic OA-associated pain and pharmacological management. RESULTS: HCRU was consistently greater in cases versus controls for all resource categories during preindex and postindex periods. Across follow-up periods, resource use was greatest in patients with severe pain. In the first 12 months postindexing, mean total costs incurred by cases were four times higher versus matched controls (£256 vs £62); costs were approximately twice as high in cases vs controls for months 12–24 (£166 vs £86) and 24–36 (£150 vs £81; all p<0.0001). The incidence of new cases of chronic pain associated with OA was 2.64 per 1000 person-years; the prevalence was 1.4%. CONCLUSIONS: This study highlights the real-world cost of chronic pain associated with OA in cases versus matched controls. We included patients with mild, moderate and severe pain associated with OA, and showed HCRU in discrete 1-year time frames. The true economic burden of pain associated with OA is likely to be considerably higher when indirect costs are considered.