Work in pharma versus work in academia (created 2010-10-23).

This page is moving to a new website.

Dear Professor Mean, How does working as a statistician in the pharmaceutical industry (pharma) differ than working as a statistician in academia?

At 54, I'm old enough to be a senior statistician, I suppose. I have never been employed full time with pharma, but I have served as an independent consultant on a few projects. I've worked in academia way back in the 1980s and have recently come back, part-time, to an academic position, to supplement my income as an independent consultant. Most of my work has been with the government (9 years) and with a private hospital (12 years). So my perspectives may not be as well informed as others here. These are all general tendencies, and there are many exceptions that could be noted. In general, the work is not too different. The biggest difference is that the statistics used in academia are far more heterogeneous. Although there are some exceptions, most of the work in pharma is prospective. In academia, the work could just as easily be retrospective as prospective. In academia, I have worked with problems in business, law, psychology, sociology, and other areas. Pharma, of course, is restricted to health care problems, and even in that arena, it is limited to interventions that can be packaged and sold (drugs and devices). So the work is much narrower in scope. In academia, much of the work involves writing grants, which is a high-risk, high-reward situation. Pharma projects are all internally funded, which means that you are closer to the people that you need to influence to get funding for your work. Also in pharma, the research is reviewed by an outside agency (the FDA in the United States, and EMEA in Europe). This leads to a higher quality of research, as Stephen Senn has pointed out. In academia, there is an ideal which is never quite obtained, but which is almost always striven for, the "disinterested researcher." This is a researcher who is equally glad to see positive or negative results, as either one contributes to the knowledge base. Very few researchers meet this ideal, but many come close. In pharma, there is an asymmetry, in that positive findings usually lead to profits and negative findings usually lead to losses. This leads to a temptation to possibly skew the results in a favorable light. A smart company will not do this, as false positive findings can lead to bad publicity, drug recalls, and inefficiency in the long run. But there is a temptation, at least, even if it is not acted on, and a statistician in pharma is often called on to act as a gatekeeper to prevent distortion of research findings. This is not to say that academic researchers are not sometimes tempted to distort results, especially if a positive finding is like to lead to more grant funding. But there are more settings in academia where economic success is not tied to a positive result than there are in pharma. In academia, you often get to choose which researchers you work with. Research teams are formed informally in academia, and if a certain project does not suit your temperament, you won't be forced to work on it. In pharma, you can find different projects to work on, but only by moving from one department to another. Finally, work in pharma is heavily deadline dependent. There are a few deadlines in academia, especially with grants, that can cause you to put in a lot of extra time, but generally, the pace is less hectic. Time is money in pharma because the profits in pharma are much higher during the time that patent protection holds. Serious delays in statistical analysis are not tolerated in pharma. I hope this helps.