The goal of pharmacovigilance is to detect the subtle signal of a causal effect otherwise buried amongst the noise of all the maladies which by coincidence alone happened to manifest after a drug was taken. The FDA Adverse Event Reporting System (FAERS) gathers reports of these adverse events and presents them in a database that can be analyzed for just such signals. Since previously unknown causal associations are what are being sought there is no requirement that the reported adverse effect be demonstrably causal. To work properly the system merely requires accurate and unbiased reporting of as much adverse event data as possible.

You'd think health care professionals would be responsible for most of the reported adverse events and most of the time you'd be right. But zealous advocates (aka lawyers) can submit adverse reports too and when they do the resulting signal amplification is amazing (or appalling, depending on your perspective). Take the case of metoclopramide and tardive dyskinesia (TD) examined in:  "The Metoclopramide Black Box Warning for Tardive Dyskinesia: Effect on Clinical Practice, Adverse Event Reporting, and Prescription Drug Lawsuits".

For years the rate of TD per metoclopramide (Reglan) prescription bumped along at the bottom of the graph never rising above one half of one tenth of one percent. It is, or was, commonly prescribed in those suffering from gastroparesis. Then came a black box warning advising of the very rare but very serious side effect TD. Suddenly the TD rate took off like the Blue Angels at an airshow. What was behind it? "In the FAERS, only 3.6% of all the reports are from lawyers, whereas 70% of TD reports related to metoclopramide came from lawyers, resulting in a significant magnification of TD reports disproportionate to all other adverse reports." Interestingly, the authors using published legal opinions regarding TD/metoclopramide as a proxy for filings found that a wave of claims took off just before the wave of reports of TD. 

Distorting the metoclopramide/TD risk signal caused doctors and patients to overestimate the risk posed by metoclopramide. Often they chose instead domperidone which has not been approved by the FDA; unfortunately that "[puts] patients receiving this drug at potential risk for serious or even fatal arrhythmias."

Perhaps even more unsettling is the fact that as fewer and fewer prescriptions for metoclopramide were written the increase in the rate of TD got steeper and steeper. One hypothesis that readily emerges is that lawsuits drove the increase in TD cases rather than the other way around. If so, it points to another chronic problem in mass tort litigation - expert witnesses whose primary contribution is to dramatically over-diagnose ailments like TD.

The authors note that theirs is not the first study to find that lawyers are able to modulate the signals relied on for pharmacovigilance. It'll be interesting to see what if anything the FDA does about it.