Health and Access Effects of New Drugs: Combining Experimental and Non-Experimental Data

We propose to combine clinical trial and estimates of behavioral responses in the population to quantify the value of new drug innovations when such values cannot be obtained by randomized experiments alone. New drugs are seen as having two distinct effects on patients. First, they can provide better outcomes for patients currently under treatment, due to better clinical efficacy. Second, they can also provide treatment access to more patients, perhaps by reducing side effects or expanding treatment. We compare these “clinical” and “access” effects using claims data, data on the arrival rate of new drugs, and the clinical trials literature on the effectiveness of these drugs. We find that the effect of new drug introductions on the number of patients treated accounts for a substantial majority of the value created by new drugs.