Long-term opioid prescribing and the risk of adverse events in patients with musculoskeletal pain: a cohort study in the CPRD

Talk Code: 
Ying Chen
Ying Chen, Richard A Hayward, Julie Ashworth, Kate Walters, Kate Dunn, Kelvin Jordan, John Bedson
Author institutions: 
Keele University, University College London


One in seven new consulters for musculoskeletal pain in the UK is prescribed an opioid. A recent study showed a large increase in the number prescribed long-term opioids (>90 days) from 2002-2013, with an increase in more potent long-term controlled opioids. In the USA, long-term opioid use has been associated with adverse events including substance abuse, self-poisoning and bone fractures. However, in the UK examination of potential adverse events has so far been limited and may not be the same as in the USA due to differences in prescribing patterns. As 20% of adult patients consult UK primary care each year for a musculoskeletal problem, the increasing use of opioids, including long-term potent controlled opioids, presents primary care with an as yet unquantified level of risk. Our aim was to assess whether using long-term opioid analgesics for musculoskeletal pain is associated with increased risk of adverse events.


This was a cohort study of adults aged ≥ 18 years starting on long-term opioids (≥ 3 opioid prescriptions within 90 days) for musculoskeletal pain, undertaken in the UK Clinical Practice Research Datalink (190 practices) from 2002-2013. Start and end dates of each long-term opioid episode were determined for each patient. Cox proportional hazards models were used to compare the risks for adverse events between the periods on and off opioids, with adjustment for confounding factors including smoking, NSAID use, ethnicity, geographical region, deprivation level and number of comorbid conditions. The adverse events examined were major trauma, falls, gastric and non-gastric bleeding, accidental and non-accidental self-poisoning, suicide/self-harm, incident depression, anaemia, osteoporosis, and opioid addiction. We also examined two “control” outcomes (eczema and psoriasis) which a priori were assumed should not be related to long-term opioid use.


There were 103,297 adults newly prescribed long-term opioids (41% males; median age 60 (IQR 46,72) at baseline. Periods of long-term opioid episode use were associated with increased risk of adverse events, compared with opioid free periods. For example, major trauma (hazard ratio (HR) 1.12, 95% CI 1.08, 1.17; n(events) = 11,307), falls (HR 1.22, 1.17, 1.26; n(events) = 14,731), gastric bleeding (HR 1.15, 1.05, 1.27; n(events) =1,895), accidental poisoning (HR 2.27, 1.51, 3.40; n(events) = 109), incident depression (HR 1.27, 1.21, 1.32; n(events) = 9,762), incident osteoporosis (HR 1.22, 1.15, 1.30; n(events) = 4,511), incident opioids addiction (HR 2.87, 2.16, 3.82; n(events) = 236). There was no association of long-term opioid use with either of the “control” outcomes.


There was a significant association between long-term opioid use and adverse events in the UK. Doctors need to understand the risks of long term opioid use and adopt a policy of regular review to monitor efficacy, side-effects and potential complications when initiating and continuing opioid therapy.

Submitted by: 
John Bedson
Funding acknowledgement: