(085) Long-term use of prescription opioids and mortality: a population based, propensity-weighted cohort study with 767,299 patients initiating therapy
Background: Evidence on the association between patterns of opioid therapy initiation, long-term opioid use and increased mortality remains limited.
Objectives: To identify the patient factors and initial therapy characteristics associated with long-term opioid use, as well as the association between long-term use and death.
Methods: Retrospective cohort study combining several population-wide databases from the Valencia Health System Integrated Database (VID), covering a population of five million inhabitants. We included all patients aged 18 years old and over who were initiated opioid treatment from January 1, 2014, to December 31, 2018 for non-cancer pain. We used logistic regression models to identify factors associated with chronic opioid use. Then, to estimate the increased risk of mortality in long-term versus short-term users, we used the same set of covariates and carried out survival analyses using multivariable Cox regression modelling for all-cause mortality during follow-up. To adjust for potential confounding, we used inverse probability of treatment weighting (IPTW) using propensity scores, based on the probability of using opioids chronically.
Results: Individual factors associated with long-term use were older age (age 45 to 64, HR:1.863, 95%CI: 1.814, 1.913; age 65 to 74, HR: 2.56, 95%CI: 2.49, 2.64, and age 75 and over: HR: 3.35, 95%CI: 3.24,3.45, compared to age 18 to 44), female sex (male sex, HR: 0.88, 95%CI: 0.87, 0.90), low income level (compared to a yearly income of 18.00 euros or less, yearly income of 18.000-100.000, HR: 0.935, 95%CI: 0.91, 0.95; more the 100.000 euros/year, HR: 0.72, 95%CI:0.59, 0.87; and no resources, HR: 1.09, 95%CI: 1.06, 1.13), use of alcohol and tobacco (HR: 1.24, 95% CI:1.18, 1.30 and HR: 1.08, 95%CI: 1.06, 1.11). Initial therapy characteristics associated with higher risk for long-term use were initiating with short-acting (HR:1.24, 95%CI:1.14, 1.34) and long-acting opioids (2.49, 95%CI: 2.44, 2.54, when compared to tramadol), using higher daily doses (when compared to 50 MME or less, prescribing 50-89 daily MME, HR: 1.40, 95%CI: 1.32, 1.49; 90-119 daily MME, HR: 2.47, 95%CI: 2.03, 3.01; and more than 120 daily MME, HR: 1.81, 95%CI: 1.69, 1.94), and overlapping with benzodiazepines (HR: 1.32, 95%CI: 1.30, 1.35) and gabapentinoids (HR: 2.30, 95%CI: 2.24, 2.36). The weighted incidence of mortality was 21.56 per 10,000 patients for long-term users and 13.55 per 10,000 for short-term users. In Cox regression models using IPTW, chronic opioid use was associated with higher risk of all-cause mortality when compared to short-term use (HR: 1.37, 95%CI: 1.32, 1.42).
Conclusions: These findings may help healthcare managers to identify and address patients at higher risk of long-term use as well as riskier prescription patterns.
