Commentary: Risks of doing as the Romans doBMJ 2009; 338 doi: https://doi.org/10.1136/bmj.b1107 (Published 31 March 2009) Cite this as: BMJ 2009;338:b1107
Garattini and Chalmers (doi:10.1136/bmj.b1025) make four suggestions to improve the evaluation of drugs—involve patients in establishing research priorities, improve the transparency of drug evaluations and studies, fund independent drug evaluations, and require evidence that new drugs have added clinical value. Although controversial, many of their proposals would benefit patients. However, their suggestion to follow the Italian approach to funding independent clinical trials is unlikely to be a good idea.
In 2003, Italy decided to establish a tax on the drug industry of 5% of their medicines advertising budget. This money is used by the state medicines agency, AIFA, to fund clinical research on the efficacy of orphan drugs, comparisons of drugs for the same indication, observational outcome studies, and pharmacovigilance.1 2
We know that there are real problems with clinical trials as currently conducted, such as the use of surrogate end points, weak information on long term efficacy, and biased reporting of findings in industry sponsored clinical trials.3 But we also know that with the right public policy incentives drug companies will align their research and investment more closely with public priorities.4 The Italian approach does not take advantage of this knowledge.
Garattini and Chalmers take care to show that their proposed changes will also benefit the drug industry. This shows their understanding that clinical research sits within an interlinked complex system of industry, government, and universities, each with their own interests—a sort of mutually reinforcing “iron triangle” of relationships. The Italian approach dispenses with these relationships by singling out the industry part of this tightly interconnected system rather than by tackling the wider systemic problems. Funding more independent clinical trials does not on its own make the problem of bias, poor design, conflict of interest, and research misconduct go away; greater scrutiny and transparency (as the authors argue separately) might.
If we follow the money (the 5% tax), we see that the benefits and costs are being distributed through a transfer of money from the “haves” (industry) to the “have nots” (the research community), which will conduct the research. Apparently, the Italian government did not feel compelled to view funding clinical trials as a priority for general taxation. By creating a hypothecated (dedicated) tax, however, the policy is based on weak legs: AIFA’s ability to fund independent clinical trials now depends on a compulsory tax on discretionary advertising budgets, which can of course go down.
Furthermore, by creating apparent winners and losers, Italy has missed the opportunity to seek a wider consensus on how to achieve a better return on the public investment in medicines research—a point made by Garattini and Chalmers. Countries with a weak drug industry base tend to adopt stricter industry price controls and regulation than those with a strong industrial base, and this initiative from Italy is consistent with that view.5 Spain, another country with a comparatively weak drug industry has followed Italy with a similar tax, but on sales volume.6
The worry is that this policy will have unintended consequences. The Italian government’s objective to improve research productivity may suffer if investment in drug research is moved to countries with more receptive commercial environments. In addition, the Italian research community may not have the capacity to design and conduct appropriate independent clinical trials if researchers also leave Italy. The industry might also reassess Italy as a congenial jurisdiction in which to do medicines research in the first place, although this may already be the case.
The conclusion is that if you want to build a strong medicines research community that is more likely to act in the wider public interest, copying Italy may not be a good idea.
Cite this as: BMJ 2009;338:b1107
Competing interests: None declared.