Predictors of urine drug testing for patients with chronic pain: Results from a national cohort of U.S. veterans

Background: Urine drug testing (UDT) is recommended for all patients who initiate chronic opioid therapy (COT) for the treatment of chronic pain; however, it is infrequently utilized. Some prior research has identified factors that may predict UDT, but studies have been limited. The purpose of this study is to examine the rate and predictors of UDT among a national sample of patients with chronic pain who had new initiations of COT. Methods: Administrative data were examined for all veterans receiving medical care at Department of Veterans Affairs medical facilities who had new initiations of chronic opioid therapy (COT) during fiscal year 2011. Results: Nineteen percent of patients who had new initiations of COT for chronic noncancer pain received UDT within 90 days of starting opioids. In adjusted analyses, patient-level factors that predicted increased likelihood of UDT included male gender (risk ratio [RR] = 1.23, 95% confidence interval [CI] = 1.02-1.49), Black race (RR = 1.20, 95% CI = 1.06-1.37), divorced/separated marital status (RR = 1.13, 95% CI = 1.02-1.25), higher pain intensity (RR = 1.03, 95% CI = 1.01-1.05), comorbid substance use disorder (RR = 1.42, 95% CI = 1.27-1.60), posttraumatic stress disorder (PTSD) (RR = 1.14, 95% CI = 1.01-1.29), bipolar disorder or schizophrenia (RR = 1.29, 95% CI = 1.08-1.53), having received UDT prior to initiating opioid therapy (RR = 1.43, 95% CI = 1.26-1.62), and a higher baseline opioid dose (RR = 1.38-1.81, 95% CIs = 1.20-1.58, 1.57-2.09). Age was also associated with UDT, in a nonlinear manner. Several factors were associated with lower likelihood of UDT, including living in a highly rural setting (RR = 0.62, 95% CI = 0.29-0.99), having a VA service-connected disability (RR = 0.85-0.89, 95% CIs = 0.75-0.97, 0.79-0.99), and having a nurse practitioner or physician assistant as one's primary care clinician (RR = 0.72, 95% CI = 0.61-0.85). Conclusions: Urine drug testing was conducted with 19% of patients who had new initiations of COT. Factors that predicted UDT were multifaceted and included patient and clinician variables. Multidimensional system-level interventions may be needed to facilitate widespread implementation of UDT.