Measuring chronic pain, non-drug pain therapy, and opioid-related mortality

Abstract

Chronic pain comprises three of the top five causes of physical disability in the United States (U.S.) and affects over 60% of U.S. military veterans seen in Veterans Affairs primary care clinics. The opioid epidemic stems in part from health systems’ efforts to treat chronic pain pharmacologically, leading to opioid addiction and opioid-related mortality. Many non-drug therapies are effective for chronic pain but are clinically underused. Managing the widespread problems of chronic pain and opioid-related mortality requires valid and reliable measurement approaches. This dissertation addresses measurement challenges in assessment of chronic pain, non-drug pain therapy use, and opioid-related mortality. The first manuscript examines relationships between physical performance measures and patient-reported outcome measures of pain-related functional interference and pain severity among U.S. military veterans in a 12-month randomized clinical trial of opioid vs. non-opioid medication therapy for chronic musculoskeletal pain. The second manuscript explores patterns of use of non-drug pain therapies among a national cohort of U.S. military veterans prescribed long-term opioid therapy for chronic pain. The third manuscript adapts expert elicitation approaches from Bayesian statistics to incorporate physician opinion into a quantitative bias analysis of potential age-related misclassification in opioid-related mortality data from death certificates in the U.S.