There is constant research on products and procedures to improve health and health outcomes. How does all that knowledge get out into medical practice? That is the question studied in a new paper at the National Bureau of Economics. (NBER Paper) The authors looked at the experience regarding new information on use of statin drugs, which lower lipid levels. They studied how this knowledge appeared to have diffused into everyday practice in Italy, using data from 2003 to 2014. As in most countries, physicians in Italy are required to do continuing medical education and have other information sources which update them on medical practice recommendations. In addition to physicians, patients may also become aware of medical developments and may request changes to their treatment. The specific piece of medical knowledge in this case related to a potential liver damage side effect from statin use; a concern which likely inhibited some prescribing of statins and which led to recommendations of regular liver function testing for patients on statins. In 2010, the results of a large randomized trial were published which showed that statins did not cause liver damage. This led to a discontinuation of the recommendation for widespread liver function tests in the US and some other countries. Although the Italian recommendation didn’t change, Italian physicians were aware of the trial results because, among other things, of pharmaceutical sales representatives presenting the information to them.
So the authors looked for changes in prescribing patterns before people first began questioning the statin/liver damage link, during the period when concerns about the validity of that link were more widespread and after the trial results were released. They used data from Italy that linked the prescriber, patients and prescriptions. They considered a large number of variables, including the number of patients with high lipid levels that the physician treated, the lipid level before and during treatment, and possible side effects. An initial finding was that in fact patients who are more adherent to statin treatment have larger cholesterol reductions than those who are less so. A second finding was that doctors appeared to use liver function testing less than the guidelines suggested during the period in which there were concerns about liver damage. There was substantially more testing of patients with higher doses of statins, suggesting doctors were aware of the concern, but doctors with more patients did less testing for patients at all doses, indicating that experience lessened their concern regarding the side effect. Across all subgroups, there was a decline in testing and an uptick in statin use following the publication of the trial results. The results suggest that this new information diffused rapidly through the system. But one thing you can always count on is that drug companies will be really sure that people have any data that might get them to prescribe more of the companies’ products.