Initiation of Long-Acting Opioids Following Hospital Discharge Among Medicare Beneficiaries

Abstract

Background: Opioids with a longer duration of effect (“long-acting opioids”) are associated with risk of overdose and morbidity compared to short-acting opioids even at equivalent dosing. Guidelines recommend against initiating long-acting opioids during hospitalization. We evaluated the prescription of long-acting opioids following hospitalization among opioid-naïve older adults. Methods: We used a 20% random sample of fee-for-service Medicare beneficiaries with an acute care hospitalization in 2016 who were ≥65 years old, did not have cancer or hospice claims, and had not filled an opioid prescription in the 30 days prior to hospitalization. We used a multivariable generalized estimating equation model to identify characteristics associated with prescription of a long-acting opioid within 7 days of hospital discharge compared to those prescribed only short-acting opioids. Predictors included sociodemographics, chronic conditions and medication claims, markers of medical complexity and functional status, indication for hospitalization, and prior opioid use. Results: Of 300,715 hospitalizations in our cohort, 60,305 (20%) had a claim for short-acting opioids only and 1,651 (0.6%) had at least one claim for a long-acting opioid within 7 days of discharge. 64.9% of hospitalizations resulting in long-acting opioid prescription were surgical in nature. Compared to patients with receipt of short-acting opioids after discharge, patients with receipt of long-acting opioids had higher incidence of known risk factors for opioid-related adverse events, including history of mood and anxiety disorders, opioid use disorder, prior long-term high-dose opioid use, and co-administration of multiple sedating medications. Characteristics with the strongest positive associations with long-acting, compared to short-acting opioid receipt included previous high-dose long-term opioid use (aOR 34.99, 25.99-47.11), history of opioid use disorder (aOR 2.43, 1.87-3.15), medical DRG (aOR 2.94, 2.35-3.67), principal diagnosis of diseases of the musculoskeletal system and connective tissue (aOR 2.14, 1.63-2.80), and operations on the musculoskeletal system (aOR 2.68, 2.04-3.51). Conclusions: The majority of opioid-naïve individuals discharged with long-acting opioids after acute hospitalization are surgical patients, for whom pain is typically acute and self-limited. Moreover, patients started on long-acting opioids at discharge had higher incidence of known risk factors for opioid-related adverse events, highlighting the necessity to develop systems-based solutions to improve guideline-concordant prescribing of long-acting opioids.