As the U.S. population ages a greater proportion have multiple chronic conditions (MCCs). Understanding the order and the rate of accumulation of chronic conditions is important to planning for effective care management for patients. Existing work has attempted to assign weights to chronic conditions or focused on a target diagnosis and assigned a count of other conditions. These counts are often used to explain variations in treatment or outcomes for the target disease, but miss the unique effects of specific conditions. This research draws on 2008- 2011 linked Medicare and Medicaid claim files for 289,000 dually eligible beneficiaries in the State of Michigan. We test the model that argues that the accumulation of chronic conditions in subsequent years will depend on;the pattern of chronic conditions in the baseline year which will be moderated by the age, sex and race of the patient and mediated by their use of outpatient and specialty services. We then investigate the health care utilization (admissions/readmissions, emergency visits, nursing home placements) associated with specific combinations of chronic conditions. Our analytical approach incorporates the theory of stochastic processes and will present an advance in the methodology of the analysis of longitudinal data by describing the specific accumulation of MCCs within patients over time. The combinations of specific conditions will be treated as the state space for a stochastic process X(t) with discrete time t=2008, 2009, 2010, 2011. Data from multiple patients will be treated as multiple independent realizations of the random sequence, one for each patient. This approach overcomes the limitation of existing methodology that incorporates only a count or a weighted count of comorbid conditions without looking at the combinations of specific conditions and their accumulation over time. While a Markov property will be employed, we are not going to make an arbitrary assumption that Markov property holds but will test this assumption. This is important because for some patients with certain constellation of conditions the Markov property may hold. For others, however, current constellations may be unrelated to prior states. This analytical approach will identify how MCCs lead to subsequent conditions, and how the role of age, sex, and race and contact with the health care system to differentially explain accumulation of conditions over time. It will further describe the health care utilization associated with specific combinations of conditions that may support health care planning activities.
As the U.S. population ages, more people are diagnosed with multiple chronic conditions. Over 50% of Medicare beneficiaries have at least two conditions but care may not always be coordinated. This project will identify how specific chronic conditions accumulate, factors that are associated with new conditions and the public health implications for management of multiple chronic conditions.