Individual and prescription level factors associated with overdose in opioid naïve older people

Abstract

Background

Opioid naïve older adults may be at risk of overdose after receiving an initial opioid prescription.

Methods

This population-based cohort study from a linked dataset of patients in Oregon, linking all payer claims data to other administrative datasets, aimed to assess the prescription- and patient-level characteristics associated with increased odds of opioid overdose after an initial opioid prescription. Included patients were ≥65 years old and received an index pain-formulation opioid prescription between 2016 and 2019. The primary outcome was an index nonfatal or fatal overdose within 6- or 12-months following index prescription. Patient characteristics included age, sex, insurance plan, number of medical comorbidities, and presence of psychiatric comorbidities. Prescription characteristics included opioid type, duration of action, and days' supply. A logistic regression model was used to determine the association with opioid overdose.

Results

There were 223,799 individuals included for analysis (58.6% 65–74 years old, 53.9% female). There were 183 fatal or nonfatal opioid overdoses in 6 months and 232 in 12 months following the index prescription. Adults aged ≥85 years were less likely to experience an overdose versus those 65–74 years (6-month adjusted odds ratio (aOR) 0.35, [95% confidence interval, 0.20–0.59]; 12-month aOR 0.38 [0.24–0.60]). Multiple factors were associated with increased odds, including dually enrolled in Medicare/Medicaid compared to commercial insurance (6-month aOR 5.99, [1.93–19.65]; 12-month aOR 3.53, [1.58–7.90]), three or more comorbidities compared to none: (6-month aOR 3.69, [1.91–8.13]; 12-month aOR 4.24, [2.32–7.74]), history of depression: (6-month aOR 1.94, [1.34–2.81]; 12-month aOR 2.20, [1.60–3.04]), received long-acting opioids (6-month aOR 5.76, [1.56–21.22]; 12-month aOR 4.0, [1.39–11.55]) compared to short-acting.

Conclusions

For older adults, there is an association between opioid overdose risk and factors including patient insurance type, patient comorbidities, and receiving a long-acting opioid prescription. Providers should be aware of the risks of opioids in this population.