LAUSR

Insomnia and other sleep problems commonly occur in patients with chronic pain. While behavioral approaches are recommended first-line treatments for conditions such as insomnia, patients with sleep complaints often receive sleep medications. Thus, we aim to understand patterns of prescription sleep medication use in Veterans with chronic pain receiving care at the Veterans Health Administration (VHA). We reviewed electronic medical records to identify patients with moderate/severe chronic pain receiving care at a VA hospital between FY2019—2021. Multivariate logistic regressions examined associations between 14 patient demographic and clinical variables on sleep medication prescription. Short-term prescriptions associated with upcoming procedures were excluded. Of the 1,706,214 moderate/severe patients with chronic pain (88% male, 66% White), 14% had comorbid diagnosis of insomnia, 30% sleep disordered breathing, and 19% were prescribed sleep medications. Trazodone (40%), melatonin (23%), and benzodiazepines (14%) were most commonly prescribed. Although 14% of patients without insomnia or sleep apnea diagnostic codes were prescribed medication, patients with insomnia were four times more likely to use sleep medication (odds ratio (OR) = 4.05, 95% CI [4.01-4.09]). Strongest predictors of medication use were insomnia (adjusted OR = 2.84, 95% CI [2.81-2.86]), PTSD (adjusted OR = 2.77, 95% CI [2.75-2.79]), depression (adjusted OR = 1.89, 95% CI [1.87-1.91]), bipolar disorder (adjusted OR = 1.84, 95% CI [1.81-1.87]), and ongoing opioid use (adjusted OR = 1.82, 95% CI [1.80-1.84]). Prevalence of medication use was similar across gender, age, and racial/ethnic groups. Sleep medication use is common in Veterans with chronic pain, with highest rates among patients with insomnia diagnoses. Additionally, a large number of patients without a sleep disorder diagnosis also receive medications. Findings highlight need to connect chronic pain patients with behavioral sleep interventions. Notably, we observed a wide variety of medications without strong patterns across demographic characteristics. This work was conducted as a non-research quality improvement effort supported by Quality Enhancement Research Initiative (PEC 13-001) and the VA Office of Patient Centered Care and Cultural Transformation. YIL is supported by a VA MIRT Fellowship. JAC is supported by a VA HSR&D CDA (IK2 HX002866). LMD is supported by a VA HSR&D CDA (CDA 18-187).