LAUSR

Aim To determine whether statistically distinct classes of smokers exist according to mental health (MH) diagnoses within primary care and to evaluate whether class membership is associated with healthcare utilization. Subject and methods Data were obtained from the US Department of Veterans Affairs (VA) electronic medical record for encounters between January 1, 2014, and December 31, 2014, from the New York/New Jersey region. Data from a subset of 25,713 smokers from a larger sample of 111,333 primary care patients were used. An exploratory latent class analysis (LCA) was conducted to determine the presence of distinct groups of smokers based on observed MH diagnosis, while a confirmatory LCA assessed for reliability of those distinct groups. Using group as a predictor, we then utilized regression to determine whether group membership was associated with elevated numbers of primary care, integrated primary care, and specialty care encounters. Diagnosis was based on primary care encounter codes. Results Three independent groups of current smokers were established: those with no MH diagnoses (77.8%), those with depressive diagnoses (16.7%), and those with alcohol/substance use diagnoses (5.5%). Most smokers were in the non-MH group, which utilized the least healthcare services (e.g., 2.7 primary care visits per year). The depressed group had 3.6 primary care visits per year and were most likely to use integrated primary care. The alcohol/substance-use group utilized primary and specialty services significantly more often than the other two groups, with an average of 5.5 primary care visits per year. Conclusion Distinct classes of smokers can be identified in primary care settings.