Overdose deaths in people prescribed opioids for chronic non-cancer pain: Systematic review and meta-analysis

Abstract

Background: Opioid prescribing for chronic non-cancer pain has increased dramatically in the United States and other high-income countries. Pharmaceutical opioids pose a risk of fatal overdose, even if used as prescribed. Guidelines for opioid prescribing in chronic pain emphasise the need to discuss risks and benefits of opioid use with patients, but there is a lack of data to inform these discussions. We completed the first systematic review to estimate crude mortality rates (CMRs) and standardised mortality ratios (SMRs) of people prescribed opioids for chronic non-cancer pain, with a focus on drug-related deaths.

Method: We searched Medline, Embase, and PsycInfo for cohort studies reporting crude mortality rates (CMR) and standardised mortality ratios (SMR) among people prescribed opioids. We included studies of people prescribed opioids for chronic non-cancer pain, including studies using registry data (e.g. prescribing within a health insurance plan or universal health system). SMRs were rarely reported so we used CMRs, information about the cohort (e.g. sex and age distribution) and mortality rates in the relevant country to estimate the all-cause and cause-specific SMRs.

Results: We identified seven studies meeting the inclusion criteria, including 409,341 people and 504,865 person-years (py) of observation. All were registry studies and all but one were undertaken in the United States. The pooled all-cause CMR was 2.66 per 100 person-years (py; 7 studies; 95% CI 1.45, 4.89) and participants died at twice the rate of their age- and sex-matched peers (7 studies; pooled SMR 1.94; 95% CI: 1.01, 3.72). There was variation in definitions used for drug-related deaths, but these were a relatively small proportion of all deaths observed in these cohorts (5 studies; pooled drug-related CMR 0.09 per 100 py; 95% CI 0.03, 0.26). Significant excess drug-related mortality was observed, with participants 8 times as likely as their peers to die from drug-related causes (5 studies; pooled drug-related SMR 8.19; 95% CI: 2.70, 24.88). By contrast, the pooled SMR for accidental injury deaths was 3.13 (2 studies; 95% CI: 1.79; 5.45) and for cardiovascular disease was 1.79 (3 studies; 95% CI: 0.95, 3.38).

Conclusion: The drug-related mortality rate in people prescribed opioids was low relative to other causes of death, but drug-related deaths occurred at 8 times the rate as in the age- and sex-matched general population, suggesting that overdose is a significant contributor to excess mortality in this population. As only registry studies were available for analysis, there was insufficient information to determine if participants in these studies were taking opioids as prescribed. There is an urgent need for prospective cohort studies of people prescribed opioids for chronic pain to better understand risks, including mortality.