A contemporary and comprehensive analysis of the costs of stroke in the United States.

Background: While several studies have determined direct costs associated with stroke there is a paucity of research involving indirect costs, especially in younger patients.

Aim: Perform a cost of illness analysis for ischemic and non-traumatic hemorrhagic stroke in the US involving both direct and indirect costs in all age groups.

Methods: Nationally representative data was obtained from the Medical Expenditure Panel Survey (2003-2014). Subjects were dichotomized based on reported history of stroke. Two-part econometric models were used to estimate the adjusted incremental direct expenditure for patients with stroke. We used generalized linear model with family gamma, log link for the adjusted analysis of annual wage, a negative binomial regression model for the adjusted analysis of missed-work day, and a logistic regression model to estimate the probability of full-year employment. Loss of productivity due to premature death was computed using Present-Value of one life and annual number of deaths from 2014 National Vital Statistics. All costs are represented in 2016 US-dollar values.

Results: Out of 253,235,052 participants, 8,101,159 (3.2%) reported history of stroke. Weighted samples of 10,155 stroke participants and 314,694 control group were compared. Adjusted annual direct costs for each stroke participant was $4317 (95% CI: $3828-$4807) greater than control resulting in a net $38 billion incremental expenditure. Based on salary difference, missed workdays, and mortality, indirect cost from under-employment was $38.1 billion and from premature mortality was $30.4 billion.

Conclusion: Total aggregate of $103.5 billion expenditure was incurred with 66% being from indirect costs based on 2016 US-dollar values.