Rational decision making in medicine: Implications for overuse and underuse

Abstract

In spite of substantial spending and resource utilization, today's health care remains characterized by poor outcomes, largely due to overuse (overtesting/overtreatment) or underuse (undertesting/undertreatment) of health services. To a significant extent, this is a consequence of low-quality decision making that appears to violate various rationality criteria. Such suboptimal decision making is considered a leading cause of death and is responsible for more than 80% of health expenses. In this paper, we address the issue of overuse or underuse of health care interventions from the perspective of rational choice theory. We show that what is considered rational under one decision theory may not be considered rational under a different theory. We posit that the questions and concerns regarding both underuse and overuse have to be addressed within a specific theoretical framework. The applicable rationality criterion, and thus the "appropriateness" of health care delivery choices, depends on theory selection that is appropriate to specific clinical situations. We provide a number of illustrations showing how the choice of theoretical framework influences both our policy and individual decision making. We also highlight the practical implications of our analysis for the current efforts to measure the quality of care and link such measurements to the financing of health care services.

Keywords: clinical decision making; health policy; overtreatment; overuse; practice; rationality; undertreatment; underuse.

Figure 1

Threshold model of decision making. A, The model states that the most rational decision is to prescribe treatment when the expected treatment benefit outweighs its expected harms at given probability of disease or clinical outcome. The horizontal line indicates the probability at which physicians should treat the patient with suspected tuberculosis (2.7%). B, Actual threshold for treating a patient suspected of having tuberculosis (based on Basinga et al50; graph: Courtesy of Dr Jef Van den Ende (see Table 1 and text for details)