Demographic, Social, Behavioral, and Clinical Characteristics Associated with Long-Term Opioid Therapy and Any Opioid Prescription in High-Risk VA Patients.

BACKGROUND

Social risks (individual social and economic conditions) have been implicated as playing a major role in the opioid epidemic and may be more prevalent in the most medically vulnerable patients. However, the extent to which specific social risks and other patient factors are associated with opioid use among high-risk patients has not been comprehensively assessed.

OBJECTIVE

To identify patient-reported and electronic health record (EHR)-derived demographic, social, behavioral/psychological, and clinical characteristics associated with opioid use in Veterans Affairs (VA) patients at high risk for hospitalization or death.

DESIGN

We used generalized estimating equations to calculate the probability of long-term opioid therapy (LTOT) and the probability of filling any opioid prescription (regardless of duration) over five intervals during a 4-year period (12/2016-12/2020).

PARTICIPANTS

Prospective cohort of 4121 medically high-risk VA patients not receiving palliative or end-of-life care, and who responded to a survey mailed to a nationally representative sample of 10,000 high-risk VA patients.

MAIN MEASURES: Patient-reported demographic, social risk, behavioral/psychological, and clinical measures, and linked EHR-derived data.

KEY

RESULTS

The average age was 69.8 years, 6.7% were female, and 17.5% were Non-Hispanic Black race/ethnicity. The majority had diagnosed chronic pain (76.1%). LTOT and any opioid prescription were positively associated with the following: younger age, non-Hispanic White race/ethnicity (compared to non-Hispanic Black race/ethnicity), male sex assigned at birth (LTOT only), not being currently employed, current tobacco use, no alcohol use, higher grit (any opioid prescription only), functional limitations, diagnosed chronic pain, lower comorbidity burden (LTOT only), obesity class I or class II/III (any opioid prescription only), undergoing surgery (any opioid prescription only), and diagnosed cancer (any opioid prescription only).

CONCLUSIONS

Multifactor screening could help identify individuals at elevated risk for adverse opioid-related outcomes and augment current multifaceted initiatives, as several social risks and patient characteristics were predictors of LTOT and any opioid prescription.