A Model of Patient Choice with Mid-Therapy Information

For many treatments, a patient’s prognosis may be refined by conducting a mid-therapy assessment (MTA). When therapy lasts some time, an MTA predicts a patient’s ultimate outcome based on early signs of response. The availability and timing of such assessments also serve as a mechanism by which policy makers can influence patients’ treatment-initiation decisions. In this article, we examine how patients evaluate treatment options using MTAs of treatment effectiveness and discuss the policy implications of patient decision making in such contexts. To this end, we use indifference-curve analysis to characterize trade-offs between potential treatment success versus potential adverse effects in therapeutic situations where patients are encouraged to consider their own motivations for initiating treatment. Treatment initiation for chronic hepatitis C virus (HCV) infection is used for illustration, but the analysis can be easily adapted to a variety of clinical scenarios.

Analysis shows that the existence of MTAs influences patients’ initial treatment decisions by affecting the expectations of treatment success and adverse effects. Earlier MTAs have lower sensitivity and higher specificity, and the prior expectation of adverse effects is accordingly lower for two reasons. First, the lower chance for a false signal of treatment response and the higher chance for a false signal of no response make it more likely that the patient receives a signal leading to discontinuation. Second, if a signal to discontinue is received, fewer weeks of adverse effects have been endured before discontinuation occurs. Both factors make the expected burden of adverse effects lower when the MTA occurs earlier. Later MTAs have higher sensitivity and lower specificity, both of which serve to increase the probability of treatment completion, causing a higher prior likelihood of treatment success.

These findings indicate that treatment demand may be increased by changing the timing of MTAs, depending on the information content of the alternative mid-therapy signals and the nature of patient preferences. In HCV infection and other diseases, clinical practice guidelines for conducting MTAs might be modified to better achieve public health or other policy objectives by studying the economics of patient choice.