Quality indicators for the care of osteoarthritis in general practice: identification, synthesis, and implementation

Abstract

Background
Previous studies have demonstrated suboptimal management of care for osteoarthritis (OA). The objectives of this study were to (i) identify indicators of quality of care for OA in general practice, (ii) measure quality of care using routine general practice records and through an enhanced recording template (iii) estimate the effect of the template introduction on quality of care, and (iv) assess the feasibility of quality indicators as trial outcome measures.

Methods
A systematic review and narrative synthesis of quality indicators was undertaken. An iterative process of development resulted in an electronic template to record management of OA in consultations, based on identified quality indicators. This was triggered by a case definition of clinical OA derived through consensus. An assessment of coding, diagnostic misclassification using consultation narrative, and baseline recorded quality of care before template installation in eight practices was undertaken. Measurement after template installation facilitated a before-and-after comparison of care. The indicators were used as secondary outcomes in a cluster-randomised trial of a model OA consultation.

Results
There were fifteen valid, feasible quality indicators. Consultation prevalence of clinical OA was comparable to other estimates but up to one-third of cases may not represent true OA. Prescribing and referral data were well-captured in the routine record; assessment and core treatment indicators (such as education and advice) were not and so were included in the recording template. The template had small-to-moderate effects on weight recording, and paracetamol and topical anti-inflammatory prescription.
Assessment of the effect of the model consultation was limited by high baseline quality achievement and variation between trial arms, practices and clinicians.

Conclusion
Assessment of quality of care for OA in general practice through quality indicators is feasible but comprehensive assessment requires enhanced recording approaches. Inter-clinician variability requires further understanding and reduction, and triangulation with patient-experienced quality is needed.