Monday, August 31, 2015

Bringing Clear Thinking to Evidence Based Medicine and Evidence Based Policy

Two very recent articles, each discussed in a separate blog, discuss how to bring more lucid and objective insights to two common policy problems.

The question of whether the FDA is "too strict" or "too lax" is perennial for decades.  In a new open access paper, MIT authors Montazaerhodjat and Lo frame this as a "type I" and "type II" error problem, and show an analytical approach to decision-making.  The paper is here - "Is the FDA Too Conservative or Too Aggressive?  A Bayesian Decision Analysis of Clinical Trial Design."  For a summary you can read on your smartphone, see the blog at marginalrevolution.com by Alex Tabarrok (here).  [*]  Tabarrok's blog includes an active discussion chain.

Another very important and perennial question is:  "How Do You Know Which Health Care Effectiveness Research You Can Trust?"  Soumerai, Starr, and Majumdar (from Harvard, Boston University, and University of Alberta) discuss this at the CDC journal Preventing Chronic Disease, here (open access).  Soumerai and Koppel overview the study at Health Affairs Blogs, here.

[*]
Footnote.
Companion piece to Montazerhojdat and Lo.  In July 2015, BIO released a white paper called, "Life Cycle Perspective for Structured Risk Benefit Design."  Here.