Mental health comorbidities and cost/utilization outcomes in head and neck cancer patients

Purpose: Examine the relationship between mental health comorbidities and health services outcomes in non-elderly adults with head and neck cancer (HNC). Design: Retrospective, cross-sectional. Sample: Non-elderly adults with a primary diagnosis of HNC in U.S. Department of Defense (TRICARE) administrative claims data for fiscal years (FY) 2007–2014. Methods: Linear regression and generalized linear models were used to examine predictors of reimbursed cost and healthcare utilization, respectively. Findings: On average, there were 2944 HNC patients each year, the majority age 55–64, male, military retirees or family members of retirees, cared for in civilian facilities, and residing in the U.S. southern region. Between FY2007 and FY2014, there were slight increases in prevalence rates for diagnosed depression (12.4%–13.1%), anxiety (8.2%–11.9%), adjustment disorders (3.7%–5.8%), and drug use disorders (10.3%–19.4%), and a slight decrease in alcohol use disorders (12.3%–11.4%). In the cost regression model, depression and anxiety were the seventh and eighth strongest predictors (p <.001), behind hospice use, treatment modalities, chronic physical conditions, and tobacco use. In the utilization regression models, depression, adjustment disorder, and anxiety ranked seventh, ninth, and eleventh as the strongest predictors for the number of ambulatory visits; anxiety, depression and substance use disorder ranked fifth, sixth, and eighth in the model examining predictors of the number of annual hospitalizations; and anxiety and depression ranked fifth and sixth in the model examining predictors of the annual number of bed days. Conclusions: We found strong evidence that mental health comorbidities impact cost and utilization among HNC patients, independent of other factors. Implications for Psychosocial Providers or Policy: Addressing mental health comorbidities among HNC patients may reduce cost and improve resource efficiency.