To seek or not to seek: A behavioral model of seeking mental health care

According to the National Institute of Mental Health 18.5% of the US adult population had a mental illness in 2013. The average annual private mental health expenditures per person amounted over $1,591 in 2006. From the classical economic theory stand point, rational individuals will seek care when ill until marginal utility derived from last unit of care equals marginal cost incurred for seeking care. Nevertheless, empirical evidence shows that individuals often do not seek care when needed or do not adhere to treatment as prescribed. Thus, more robust demand predictive models in healthcare are needed to identify significant predictors of the individual decision making process of seeking care. 

This study sought to develop a behavioral model to identify significant predictors of individual’s decision to seek care and thus, to forecast the demand of mental health services. The model builds upon Andersen’s behavioral model of health services use, considered the most comprehensive model up-to-date in terms of its structure, conceptual framework and components. Nevertheless, Andersen’s model fails to provide the mathematical framework and time dimension needed to evaluate alternative scenarios of changing health status and behaviors in dynamic economics.

In our model, the utility derived from seeking care in t is a function U_it of societal terminants, x_i,t (age, sex, race, zip code), predisposing, x_it^p (having a chronic disease) and enabling factors, x_it^e (income, employment), intrinsic characteristics of the health care system, h_ct (health insurance coverage -i_ct- and cost of healthcare –c_ht), and other healthcare needs, x_itⁿ (co-occurring health conditions -cc_it). Two scenarios are analyzed: the decision of seeking care for the first time in t and the individual’s adherence to treatment in future periods (t+j, where j≥1).

The proposed model illustrates the trade-offs between two effects: the positive returns of investment in healthy behaviors and the negative impacts of co-occurrence of other health conditions/mental health disorders that increases individual’s healthcare needs. The proposed model borrows from the Grossman model the depreciation of the stock of health concept due to aging and adds the effects of other co-occurring health conditions. The model shows that when the dominant effect on the health status (H_it) at time t is the impact of co-occurring health conditions (cc_it), the optimal decision is to seek care. Conversely, if the dominant effect is the return of the investment in healthy behaviors, the optimal decision is not to seek care, even if ill. The individual will seek care when ill if and only if the marginal utility derived from last unit of the individual’s behavioral health decisions in t is lower than the marginal cost of seeking care. Treatment adherence in t+1 depends on prior experience in twith regards to health outcomes and individual’s ability to pay for healthcare.

The proposed model will provide insights into the determinants of the individual’s demand for mental healthcare services and treatment adherence and thus, it will enable decision makers and stakeholders to design and implement health care policies aiming at improving efficient allocation of scarce mental health resources.