On the hidden conflicts of interest
The issue raised by Dr. Moynihan (1), of the conflict of interest
affecting many clinicians involved in the evaluation of efficacy and
safety of pharmacological agents, is very relevant. Undoubtedly, financial
links with pharmaceutical companies can produce a bias in favour of newer
and more expensive drugs. At the same time, in publicly-funded health
systems, such as those of most European countries, the majority of
clinicians receive their salaries by institutions which are interested in
reducing short-term direct costs, with a possible bias in favour of older,
cheaper drugs. The problem is aggravated by the opaqueness of many
disclosures of interests. For example, authors sometimes omit to declare
speaking fees for talks received from agencies which were ultimately
sponsored by pharmaceutical companies; in other cases, undeclared funds
from companies can be directed towards foundations or other no-profit
organizations controlled by, or sponsoring the activities of authors of
papers. Another typically undisclosed conflict of interest is that of
authors who receive financial bonuses by healthcare payers for reducing
short-term pharmaceutical expenditure, who could be biased toward cheaper
We should all be aware of the fact that most clinical research on drugs is
supported by pharmaceutical companies. Large-scale trials, although co-
authored by eminent scientists, are usually designed in conjunction with
drug manufacturers. Even spontaneous research by academic institutions is
often funded, at least partly or indirectly, by pharmaceutical companies.
The only possible, definitive solution to eliminate the interference of
companies in the clinical assessment of drugs is represented by the
organization of truly independent, entirely publicly sponsored trials. In
order to realize this approach, which requires large financial resources,
the scientific community should enhance its efforts to increase the
awareness of public health authorities toward the relevance of clinical
We also feel compelled to reply to the criticisms of Battaggia (2) to our
1) We confirm that the number of trials included is 164; Dr. Battaggia was
probably confused by the fact that some of those trials, having multiple
comparators, were used for more of one analysis. No duplication was made
for the global analysis.
2) There is no reliable method to assess heterogeneity when the number of
events in each trial is very small, so that the statement of homogeneity
is invalid. If heterogeneity cannot be assessed, fixed-effect models
cannot be used.
3) Incidence rates were not calculated simply multiplying the number of
patients and the duration of the trial; drop-outs and patients reaching
the trial endpoint were subtracted, assuming that the rate of events and
the rate of drop-out was constant throughout the trial. We recognized the
limitation of this assumption in the Discussion.
4) We agree with Dr. Battaggia that the short duration of many trials is a
limitation, when assessing typical long-term outcomes such as
cardiovascular disease; this is the reason for which meta-analyses such as
this one should never be used to assess efficacy, but only (short- and
medium-term) safety of drugs.
We are concerned as Dr. Battaggia for the long-term safety of drugs used
for treatment of diabetes. We feel that this issue deserves full, unbiased
and non-selective consideration, exploring thoroughly the safety of every
drug independent of its cost, date of approval, or manufacturer.
 Ray Moynihan: Rosiglitazone, marketing, and medical science. BMJ
 Alessandro Battaggia, et al: Misleading meta-analysis
bmj.com, 19 Apr 2010.
 Edoardo Mannucci , Matteo Monami, Mauro Di Bari, Caterina Lamanna,
Grancesca Gori, Gian Franco Gensini, Niccolò Marchionni Cardiac safety
profile of rosiglitazone A comprehensive meta-analysis of randomized
clinical trials Int J Cardiol (2009) doi:10.1016/J.iJcard.2009.01.064
Speaking fees from GSK and Takeda; research grants from Takeda.
Competing interests: No competing interests