PROJECT SUMMARY/ABSTRACT
Both chronic pain and opioid use disorder (OUD) are linked to opioid-related morbidity and mortality. People with
comorbid chronic pain and OUD are vulnerable due to their potential under-treatment of pain and OUD, and
elevated risk of adverse outcomes such as transition to illicit drugs and overdose. Despite prominent clinical
practice guidelines that emphasize opioid alternatives for pain management, little is known about the use of non-
opioid and non-drug pain treatments nationally among chronic pain patients. Even less is known within the
context of co-occurring chronic pain and OUD, and racial/ethnic minorities. Using Medicare data, the proposed
research aims to: (1) evaluate the relationship between OUD status and type of chronic pain management
(prescription opioid, non-opioid medication [e.g., gabapentinoids], or non-drug strategies [e.g., physical therapy),
and (2) assess if receipt of non-opioid and non-drug treatments is associated with opioid prescribing and drug-
related overdoses and acute care use among patients with and without OUD. Key innovations of the proposed
research include its focus on overlapping of chronic pain and OUD, and analysis of newly-available OUD
indicators in the Medicare Chronic Condition Data Warehouse. The study will also offer new insight into the use
of opioid alternatives and outcomes among disabled and older adults who have a high prevalence of chronic
pain, are often prescribed opioids, and have rising OUD and overdose rates. We will analyze a 2016-2018
nationally representative 20% random sample of Medicare beneficiaries using enrollment, inpatient, outpatient,
and pharmacy claims data. In Aim 1 the main independent variable is OUD status and the primary outcome is
the type of pain treatment received defined as a categorical variable. In a secondary analysis, we will evaluate
the intensity of each type of pain treatment as measured by opioid morphine milligram equivalents, and count of
prescriptions and visits. In Aim 2, the main independent variable is type of pain treatment and the outcomes are
long-term high dose opioid use, drug overdoses, and drug-related hospitalizations and emergency department
visits. For each aim, we will assess race and ethnicity as a key potential effect modifier. The expected results
are that Black and Hispanic beneficiaries with OUD will be less likely to receive any pain treatment compared to
non-Hispanic Whites with or without OUD. We further hypothesize that non-opioid treatment is associated with
fewer opioid-related adverse outcomes and lower duration and dose of opioid use. Minorities with OUD will have
worse outcomes than Whites. The findings have the potential to identify and help address disparities in chronic
pain management by providing evidence to inform the development of future practice guidelines that consider
the nuanced challenges faced by marginalized groups such as people with OUD and racial minorities.