Solutions Still Searching for a Problem: A Call for Relevant Data to Support Evergreening Allegations

For years pharmaceutical policymaking discussions have been revolving around allegations of supposed “evergreening” by pharmaceutical companies, and policymakers have considered a range of significant policy reforms — including to antitrust law and drug regulatory law — to address this purported problem. This paper evaluates empirical data offered to substantiate “evergreening” and explains that these data — though mostly accurate — do not support proposed policy changes.The “evergreening” claim is that by securing additional patents and FDA-related exclusivities after approval of their new drugs, brand drug companies enjoy a period of exclusivity in the market that is longer than the initial patent(s) and exclusivity on the drug would have provided and longer than acceptable as a normative matter. Policymakers have been invited to consider a database, hosted by the University of California Hastings College of Law, that counts patents and exclusivities associated with new drugs, identifies the earliest and latest expiring patent or exclusivity for each, and calculates the number of months between those dates. Our audit of more than 200 entries concludes that the underlying raw dataset can be a useful tool for policymakers, filling a gap that exists because early FDA publications have not been digitized. But our audit raises questions about inferences drawn in and from the secondary database that interprets the dataset.If the policymaking goal is to ensure that current patent and exclusivity policies do not prevent brand products from facing generic competition for “too long” — whatever “too long” might mean — the key questions are (1) when do brand products actually face this competition, and (2) what exactly drives the timing of this competition. For every new chemical entity we examined, a generic drug was commercially available before the date represented in the database as the “latest” expiry date, i.e., the date that — the database claims — reflects the “additional time for which a company may have limited generic competition and monopolized a drug product.” Indeed, within our dataset, generic competition launched on average 84 months (seven years) before the Hastings Database implies it would. The 79 new chemical entities in our dataset experienced generic competition on average 68 months (or more than five years) before the date the Hastings Database implies they would.Our claim, therefore, is that the latest expiration date of the various protections applicable to a specific new drug application is — QED — not the most relevant data point for policymaking that means to focus on ensuring timely generic competition with new drugs. Patients, healthcare providers, insurers, and the innovating and generic industries share an interest in evidence-based policymaking. But it is not enough for advocates of reform to offer data; the data must be not only accurate but also relevant. A study designed to produce relevant data would consider the market entry date of the first generic drug based on any brand product containing a particular new active ingredient and would actually determine the factors driving that market entry date. And if a more relevant dataset would more precisely document (or rule out, or add nuance to) a supposed problem that is said to justify reform, it is incumbent on supporters of reform to generate those data. Legislative change before relevant data are generated would be premature