How do initial signals of quality influence the diffusion of new medical products? The case of new cancer drug treatments.

Abstract

Purpose: Objective measures of a new treatment's expected ability to improve patients' health are presumed to be significant factors influencing physicians' treatment decisions. Physicians' behavior may also be influenced by their patients' disease severity and insurance reimbursement policies, firm promotional activities and public media reports. This chapter examines how objective evidence of the incremental effectiveness of novel drugs to treat cancer ("chemotherapies") impacts the rate at which physicians' adopt these treatments into practice, holding constant other factors.

Design/methodology: The novelty of the analysis resides in the dataset and estimation strategy employed. Data is derived from a United States population-based chemotherapy order entry system, IntrinsiQ Intellidose. Quality/price endogeneity is overcome by employing sample selection methods and an estimation strategy that exploits quality variation at the molecule-indication level. Pooled diffusion rates across molecule-indication pairs are estimated using nonparametric hazard models.

Findings: Results suggest incremental effectiveness is negatively and nonsignificantly associated with the diffusion of new chemotherapies; faster rates of diffusion are positively and significantly related to low five-year survival probabilities and measures of perceived clinical significance. Results are robust to numerous specification checks, including a measure of alternative therapeutic availability. We discuss the magnitude and potential direction of bias introduced by several threats to internal validity. Evidence of incremental effectiveness does not appear to motivate the rate of specialty physician diffusion of new medical treatment; in all models high risk of disease mortality and perceptions of therapeutic quality are significant drivers of physician use of novel chemotherapies.

Value/originality: Understanding the rate of technological advance across different clinical settings, as well as the product-, provider-, and patient-level determinants of this rate, is an important subject for future research.