Factors associated with primary care prescription of opioids for joint pain


  • Funding sources

    Daniel Green was funded by the National Institute for Health Research (NIHR). This report presents independent research commissioned by the National Institute of Health Research. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. The study was supported by Medical Research Council, UK (grant code: G9900220).

  • Conflicts of interest

    The authors declare no conflict of interest.


Daniel Green

E-mail: d.j.green@cphc.keele.ac.uk



Opioids are commonly prescribed in primary care and can offer pain relief but may also have adverse effects. Little is known about the characteristics of people likely to receive an opioid prescription in primary care. The aim is to identify what factors are associated with primary care prescribing of high-strength analgesics in a community sample of older people with joint pain.


A prospective two-stage postal survey completed at baseline and 3-year follow-up in a population aged 50 and over registered with eight general practitioner (GP) practices in North Staffordshire (North Staffordshire Osteoarthritis Project cohorts) linked with data from medical records. Participants were selected who reported joint pain in one or more joints at baseline. Outcome measures were the number of prescriptions for high-strength pain medication (opioids) in the following 3 years. Socio-demographic and health status factors associated with prescription were assessed using a zero-inflated Poisson model.


873 (19%) people were prescribed opioids (out of 4652 providing complete data) ranging from 1 to 76 prescriptions over 3 years. Baseline factors significantly associated with increased rates of prescription were younger age group [65–74 group: incidence rate ratio (IRR) = 1.26 (1.18–1.35)], male gender [IRR = 1.17 (1.12–1.23)], severe joint pain [IRR = 1.19 (1.12–1.26)] poor physical function [IRR = 0.99 (0.99–0.99)] and lower frequency of alcohol consumption [once/twice a year: IRR = 1.13 (1.06–1.21), never: IRR = 1.14 (1.06–1.22)]. Restricting the analysis to those without prior prescriptions for strong opioids showed similar results.


Poor physical function and participation restrictions were strongly associated with prescriptions of stronger opioids in addition to several socio-demographic and lifestyle factors. Given the uncertainties over the effectiveness and risks of opioid use, future research could investigate decision making of GPs, exploring reasons for prescribing them.