Risk of adverse events in patients prescribed long‐term opioids: A cohort study in the UK Clinical Practice Research Datalink

Abstract

Background Long‐term opioid prescribing for musculoskeletal pain is controversial due to uncertainty regarding effectiveness and safety. This study examined the risks of a range of adverse events in a large cohort of patients prescribed long‐term opioids using the UK Clinical Practice Research Datalink. Methods Patients with musculoskeletal conditions starting a new long‐term opioid episode (defined as ≥3 opioid prescriptions within 90 days) between 2002 and 2012 were included. Primary outcomes: major trauma and intentional overdose (any). Secondary outcomes: addiction (any), falls, accidental poisoning, attempted suicide/self‐harm, gastrointestinal pathology and bleeding, and iron deficiency anaemia. “Control” outcomes (unrelated to opioid use): incident eczema and psoriasis. Results A total of 98,140 new long‐term opioids users (median age 61, 41% male) were followed for (median) 3.4 years. Major trauma risk increased from 285 per 10,000 person‐years without long‐term opioids to 369/10,000 for a long‐term opioid episode (<20 mg MED), 382/10,000 (20–50 mg MED), and 424/10,000 (≥50 mg MED). Adjusted hazard ratios were 1.09 (95% CI; 1.04, 1.14 for <20 mg MED vs. not being in an episode of long‐term prescribing), 1.24 (95% CI; 1.16, 1.32: 20–50 mg MED) and 1.34 (95% CI; 1.20, 1.50: ≥50 mg MED). Significant dose‐dependent increases in the risk of overdose (any type), addiction, falls, accidental poisoning, gastrointestinal pathology, and iron deficiency anaemia were also found. Conclusions Patients prescribed long‐term opioids are vulnerable to dose‐dependent serious adverse events. Opioid prescribing should be reviewed before long‐term use becomes established, and periodically thereafter to ensure that patients are not being exposed to increased risk of harm, which is not balanced by therapeutic benefit. Significance Long‐term opioid use is associated with serious adverse events such as major trauma, addiction and overdose. The risk increases with higher opioid doses. Opioid prescribing should be reviewed before long‐term use becomes established, and periodically thereafter to assess ongoing effectiveness.

Safety Events Associated with Tramadol Use Among Older Adults with Osteoarthritis

“…Falls/hip fractures are a consistent safety event reported in the literature for tramadol and other opioids use. 13,14,[16][17][18][19] Over the 33-month time period about 25% of tramadol or other opioid users experienced a fall/hip fracture compared to about 16% for nonusers. In the present study, the risks for new users of tramadol and other opioids were 60% and 70% increased, respectively.…”

“…11,15 Most research studies have focused on risks associated with overall opioid use. [16][17][18][19] Furthermore, despite indications of potential risks, many primary care providers have assumed this drug to be relatively safe compared to other pharmaceutical options for patients vulnerable to adverse events, especially older adults. 15 Diagnostic documentation of some adverse events, such as seizures, respiratory distress, non-injurious falls, or other adverse drug-drug interactions, are difficult to track in administrative databases because of insufficient coding detail.…”

“…In another population-based study including those ages ‡18 years, tramadol was associated with a 52% increased risk of hospitalization for hypoglycemia with the highest risk evident at initiation. 10 Other sources of safety event hospitalizations potentially associated with opioids/ tramadol include respiratory distress, 8 pneumonia, 20 fractures, 14,[16][17][18][19] and seizures. 8,[21][22][23] The risk for injurious falls/fractures is consistently identified as an adverse event associated with overall opioid use, likely prompted by dizziness, fatigue, somnolence, and the other CNS effects of opioids.…”

“…13,14 Fewer studies focus exclusively on the association of tramadol with injurious falls/fractures but, generally, the level of risk associated with tramadol (eg, odds ratio [OR] 1.54) 14 compared to other opioids (eg, ORs 1.36-1.59) 14 was similar. 13,[17][18][19] The association between opioids and cardiovascular (CVD) outcomes has not been widely reported and results have been mixed. 2,16,24,25 In a recent study among those ages ‡45 years using subgroups of opioids, no significant association was evident between opioid medication use and coronary artery disease.…”

“…27 Because this population may differ from general older adult and/or overall Medicare populations, it was of interest to estimate the risk for longerterm safety events, such as ER visits, hospitalizations, falls/fractures, and/or mortality, associated with chronic tramadol use compared to other opioids or opioid nonuse. In addition, because most opioid studies focus on opioid initiation or new users, 10,11,13,16,17,19 the present study included a comparison of new and continuing users to document the assumption that risks associated with tramadol or other opioid use would dissipate over time. 10,13,14,18 The expectation was that tramadol would be associated with excess risk for safety events compared to no opioid use, and that the risks associated with tramadol would be less than those associated with other more potent opioids.…”