(B-098) Factors linked to hospital-admission of chronic non-cancer pain patients prescribed long-term strong opioids

Background: Evidence indicates that long-term opioid use for chronic non-cancer pain is ineffective and linked to serious adverse effects. However, patterns of healthcare utilisation by patients prescribed long-term opioids are not well characterised in the UK.

Objectives: To assess the extent of healthcare utilisation of non-cancer pain patients in primary care treated with long-term strong opioids.

Methods: Design/Setting: this retrospective longitudinal cohort study used the Clinical Practice Research Datalink, linked to Hospital Episode Statistics data in England. Exposure: adults prescribed four strong opioids (morphine, buprenorphine, fentanyl, oxycodone) between 2000 and 2010 with no cancer diagnosis within 12-months to first prescription date (index date). Total day’s supply was calculated for each patient annually and cumulative >90 days per year defined long-term use. Main outcomes: annual primary care consultations and hospital admissions during strong opioid prescription use. Statistical analysis: baseline clinical and socio-demographics and time-dependent covariates linked to in-patient hospitalisation were assessed using generalised estimating equation model (binomial distribution with log-link function).

Results: A total of 26,703 non-cancer pain patients were prescribed strong opioids long-term. The median number of face-to-face GP consultations per user per year was 9 (IQR: 4,15) and the number of primary care consultations rose annually by 30.61% (95%CI: 29.70, 31.50, p< 0.001). Two-fifths of users had one or more in-patient hospitalisation, and musculoskeletal-related conditions were the main reason for in-patient admissions and re-admissions. Factors that increased risk of all-cause in-patient admissions amongst long-term strong opioid users included age >60 years (aRR: 1.20; 95%CI: 1.14, 1.26; p< 0.001), more GP visits per year (aRR: 1.77; 95%CI: 1.63, 1.92; p< 0.001), more pain diagnoses (aRR: 1.14; 95% CI: 1.07, 1.22; p< 0.001), comorbidities including psychiatric conditions (aRR: 1.53; 95%CI: 1.36, 1.73; p< 0.001), high dose of strong opioids (aRR: 1.20; 95%CI: 1.11, 1.29; p< 0.001) and co-prescriptions of psychoactive medicines. Of the long-term users only 6.86% (n=1,832) of patients had a first-time referral to a pain management specialist during the 11-years and the average OMED of referred patients was 110.16±116.65mg.

Conclusions: Comorbidities, greater GP visits, co-prescriptions of psychotropics and opioid dosage were linked to risk of hospitalisation. Recognition of these covariates would be useful in informing risk mitigation or prevention-strategies in clinical practice.