Patterns of routine primary care for osteoarthritis in the UK : a cross-sectional electronic health records study

Jackson, Holly and Barnett, Lauren and Jordan, Kelvin and Dziedzic, Krysia and Cottrell, Elizabeth and Finney, Andrew and Paskins, Zoe and Edwards, John (2017) Patterns of routine primary care for osteoarthritis in the UK : a cross-sectional electronic health records study. BMJ Open, 7 (12): e019694. ISSN 2044-6055

Abstract

Objective To determine common patterns of recorded primary care for osteoarthritis (OA), and patient and provider characteristics associated with the quality of recorded care. Design An observational study nested within a cluster-randomised controlled trial. Setting Eight UK general practices who were part of the Management of Osteoarthritis in Consultations study. Participants Patients recorded as consulting within the eight general practices for clinical OA. Primary outcomes Achievement of seven quality indicators of care (pain/function assessment, information provision, exercise/weight advice, analgesics, physiotherapy), recorded through an electronic template or routinely recorded in the electronic healthcare records, was identified for patients aged ≥45 years consulting over a 6-month period with clinical OA. Latent class analysis was used to cluster patients based on care received. Clusters were compared on patient and clinician-level characteristics. Results 1724 patients (median by practice 183) consulted with clinical OA. Common patterns of recorded quality care were: cluster 1 (38%, High) received most quality indicators of care; cluster 2 (11%, Moderate) had pain and function assessment, and received or were considered for other indicators; cluster 3 (17%, Low) had pain and function assessment, and received or were considered for paracetamol or topical non-steroidal anti-inflammatory drugs; cluster 4 (35%, None) had no recorded quality indicators. Patients with higher levels of recorded care consulted a clinician who saw more patients with OA, consulted multiple times and had less morbidity. Those in the High cluster were more likely to have recorded diagnosed OA and have knee/hip OA. Conclusions Patterns of recorded care for OA fell into four natural clusters. Appropriate delivery of core interventions and relatively safe pharmacological options for OA are still not consistently recorded as provided in primary care. Further research to understand clinical recording behaviours and determine potential barriers to quality care alongside effective training for clinicians is needed. Trial registration number ISRCTN06984617; Results.