Empirical Studies of the Development and Diffusion of Medical Technologies

Technological change is a central feature of health care markets, and over the past few decades has revolutionized the treatment of many medical conditions. Although breakthrough scientific advances are an important driver of the pace of medical innovation, market incentives – such as those provided by the patent system – are often a critical determinant of which potential technologies successfully make the transition ‘from
the lab to the market.’ The key contribution of this project is to develop new data (Aims 1 and 3) and new empirical methods (Aims 2 and 4) to investigate how patents shape the development and diffusion of medical innovations. Specifically, we apply these new data and new methods to estimate two relevant parameters: the extent to which patents provide incentives for the development of new technologies, and the extent to which
patents on existing technologies hinder subsequent innovation. The more effective patents are in inducing research investments, the stronger the case for longer or broader patents. On the other hand, the larger the costs of patents in terms of hindering subsequent innovation, the weaker is this case. Evaluating these costs and benefits of patents is a key input into optimal policy design. To estimate the first parameter – how the prospect of stronger patent protection affects research investments – we develop a new empirical method (Aim 2) based on the following idea: because pharmaceutical firms file patents prior to starting clinical trials, shorter clinical trials grant firms longer effective patent terms. Hence, if longer patent terms encourage more research investments, we should see higher levels of research investments on treatments for patient groups that require shorter clinical trials. We apply this method to a newly developed data set measuring research investments on treatments for cancer patients (developed in Aim 1). To estimate the second parameter – how patents on existing technologies affect subsequent research investments – we develop a new empirical method (Aim 4).
The key idea behind our approach is to take advantage of two facts: first, patent applications are quasi randomly assigned to patent examiners at the US Patent and Trademark Office; and second, patent examiners differ in their likelihood of awarding a patent to any given application. We apply this method to a newly developed data set measuring research investments related to the human genome (developed in Aim 3) and
quantify the extent to which patents on human genes have hindered subsequent medical innovation. The new empirical methods (Aims 2 and 4) will be applied to two particular classes of medical technologies – cancer treatments and gene-based diagnostics and treatments – but each method can be applied in other contexts. In addition, the new data construction methods (Aims 1 and 3) – text-based matching with clinical trial enrollment lists, and MeSH-ICD matching of research investments to the relevant groups of patients – can also be applied in other contexts.