The safety of newly marketed drugs is always provisional, based on the relatively limited population the drugs are used in prior to marketing. The number of subjects who take a drug in clinical trials is large enough to detect efficacy, but too small to detect rare, but serious, adverse events. In addition, the patients are necessarily not real world patients, warts and all. That’s not the only problem though, reporting of adverse events in clinical trials can less than optimal, and even if they are reported the data may not receive the prominence of other “more interesting” data. Ioannidis noted this in his excellent paper looking at the completeness of safety reporting in clinical trials. He looked at 192 randomized drug trials, and found:
- The severity adverse effects was adequately defined in only 39% of trial reports
- The frequency of specific reasons for discontinuation of study treatment due to toxicity was only stated in 46% of trials.
- Median space allocated to safety results was 0.3 page, comparable to space for contributor names and affiliations.
- Trials reporting statistically significant results for efficacy outcomes, provided statistically less data on safety.
Having full disclosure of safety data from a clinical trial is the only way of being able to judge the harm-benefit of the drug in the context of the trial. It is hard to understand how a trial can be considered a complete publication if the safety data is deficient. In this context, it is interesting to read a news report about dabigatran concerned with access to safety data from a recent trial published in the Lancet:
Pharmacist Aaron Tejani of the Therapeutics Initiative at the University of British Columbia, a group that independently evaluates clinical drug-trial data for doctors and pharmacists, is worried about a “signal of harm” in the data from the trial comparing dabigatran with warfarin.
“More people receiving dabigatran were experiencing myocardial infarctions or heart attacks,” Tejani said.
Those in the dabigatran group also had more cases of major stomach bleeding compared to those taking warfarin. It’s not clear if the problems were caused by the new drug.
The only way to find out would be to look at the full data of serious adverse events, such as heart attacks, strokes and problems requiring hospitalization, from the clinical trial, Tejani said.
So far, the dabigatran researchers have refused to share that information with Tejani and his colleagues to evaluate independently before the new approach is widely adopted.
In this case the reason put forward for not making all the safety data publicly available is interesting. One of the study co-authors, Dr. Stuart Connolly, director of cardiology at McMaster University in Hamilton, stated in an email that:
“The trial database won’t be released because researchers involved in the study want to publish their own papers on the findings”
This is hardly a convincing argument for delayed release of safety data. Safety and efficacy are intrinsically related. All prescribing decisions are, or rather should be, based on the balance of harms and benefits. Could there be a better example of how safety is in second place to efficacy, that only later publications by the study authors will explore safety, and that a desire for further academic papers appears to be being cited as reason for further delay of safety data.