Factors associated with long-term opioid therapy discontinuation for people with chronic non-cancer pain in UK primary care: a population-based retrospective cohort study

ObjectivesTo identify factors associated with long-term opioid therapy (L-TOT) discontinuation in people with chronic non-cancer pain (CNCP). Design and settingPopulation-based retrospective cohort study using UK Clinical Practice Research Datalink Aurum data between 01/01/2000-31/12/2020. PopulationThe study cohort comprised adults ([≥]18 years) with CNCP who were using L-TOT, defined as [≥]3 opioid prescriptions within 90 days, or a [≥]90 days of supply within the first year, excluding the first 30 days. Main outcome measuresDiscontinuation was defined as opioid-free for [≥]180 days following an episode of L-TOT. Mixed-effects logistic models with a random intercept for general practice were used to identify sociodemographic, comorbidities, lifestyle and pharmacological factors associated with L-TOT discontinuation. ResultsAmong 573,639 L-TOT users, 5.2% (n=29,589) discontinued (mean age 56.45{+/-}18.41, female 60.04%) within the first year. Factors significantly associated with a higher likelihood of L-TOT discontinuation included being Asian (adjusted odds ratio: 1.31, 95% confidence interval: 1.23 to 1.39) or Black (1.21, 1.13 to 1.31), a non-smoker (1.09, 1.05 to 1.12) or a former smoker (1.05, 1.02 to 1.09), living in the least deprived area (1.09, 1.04 to 1.14), using weak and short-acting opioids (1.50, 1.39 to 1.61), coexisting osteoarthritis (1.06, 1.02 to 1.11) or anxiety (1.04, 1.01 to 1.07), and concurrently using non-steroidal anti-inflammatory drugs (1.06, 1.03 to 1.09) or benzodiazepines (1.08, 1.05 to 1.11). Those younger (0.990, 0.988 to 0.992), with lower daily dose (0.981, 0.980 to 0.982), substance use disorder (0.77, 0.72 to 0.83), and taking gabapentinoids (0.83, 0.69 to 1.00) or antidepressants (0.89, 0.87 to 0.92) were less likely to discontinue L-TOT. ConclusionsThis study identified key characteristics of people with CNCP who are more likely to discontinue L-TOT in UK primary care. Such information can help to guide the development of targeted, personalised interventions to support safe and effective opioid discontinuation. What is already known on this topicO_LIExisting evidence on L-TOT discontinuation primarily comes from US-based studies, revealing several common associated factors, including younger age, mental health conditions, and substance use disorders. C_LIO_LIHowever, conclusions vary across studies due to differences in study populations and L-TOT definitions. Some studies link discontinuations to lower opioid doses, while others associate it with higher doses, concurrent use of benzodiazepine, or being unmarried. C_LI What this study addsO_LIOur study identified additional factors associated with a greater likelihood of L-TOT discontinuation that have not been previously reported, including being Asian or Black, a former or a non-smoker, living in less deprived areas, co-existing osteoarthritis or anxiety, and concurrent NSAIDs use. C_LIO_LIL-TOT discontinuation rates varied across regions, with the likelihood decreasing in more northern regions, compared to London. Addressing these geographical inequities through a fairer distribution of healthcare resources, particularly in underserved areas, could improve support for patients attempting to discontinue L-TOT. C_LIO_LIDemographics, comorbidities, medication profiles, and lifestyle factors should be considered when devising L-TOT discontinuation strategies among people with CNCP in UK primary care. C_LI