Information on Pharmaceutical Performance

There is an article in the Washington Post about the problems inherent in the increasing portion of drug studies funded by and controlled by the companies that would profit from successful marketing of the drugs.

Randomized case-control studies represent a major advance in the development of credible information on the efficacy of treatments, and can provide important information on unplanned side effects. Their value can be further enhanced by independent peer review done by expert peers in the craft of such studies. Multiple review, adding that of drug licensing agencies to that of journal reviewers further strengthens credibility. Of course, the credibility of the study and of the review depends on the credibility of the people and institutions doing the work -- unless their adherence to high ethical standards can be assured, it is hard to credit their published results.

That such studies are relatively credible does not mean that they are perfect, nor that they can not be improved, nor that they should not be complemented by other information products. Even when well done, the results are statistical in nature and there always remains a (small) probability that the stem from chance alone and not the efficacy of the drug. Unintended and unexpected side effects would be better studied by adding big data reviews of medical records after tested and introduced into medical practice.

Peer review is done by people who have less at stake than the people and organizations that conduct the research. They can and often do real service by calling attention to problems in the research, in the interpretation of the results, or the inferences as to their meaning. But peer reviews are less than perfect.

The article draws concern to two main issues:

• That companies rather than the government are increasingly funding the case-control studies, introducing a potential for bias, and
• That for-profit companies are increasingly conducting the case-control studies rather than academic institutions, introducing another and different potential for bias.

A key concern is bias. Bias may be deliberate and even malevolent. It certainly seems possible that company officials would deliberately make decisions to advance the sales of the company's product at the expense of patients and the public. However, I would suggest that unconscious bias in likely to be a greater problem. Indeed the line might be hard to define. Might an investigator word a finding differently if his work was funded by a company as opposed to funded by the government?

The article does not point out, but it is true that case-control studies are now more often conducted in poor countries. Studies done abroad introduce still another potential for bias.

I see several issues:

• How should clinical trials be funded?
• What kind of organization should carry out the clinical trials?
• What complementary studies should be standardized (in addition to preliminary lab and animal tests)?
• How could the entire system be better regulated?

Financing: The financing of drug trials can be distributed to the people who buy drugs (embodied in their prices) or by the general public. In the first case, presumably the cost would be borne by those who actually benefit from the drug, in the second case, by those who might in the future benefit from the specific drugs and from those who benefit from drug testing in general. In the first case, financing of the tests might be directly by the drug companies, but it might also be paid by a tax on the industry in general. Alternatively, the financing could come from the public sector, either from general taxes, or from specific taxes on the industry, or some combination of the two. Of course, if the United States Government were to finance tests of drugs used in many countries, then Americans might be paying more than their fair share of the cost.

Implementation of trials: Key players are prescribing physicians. However, specialized organizations are needed to plan trials, manage them, analyze the data, and report findings. In some cases, government agencies could do so, but the more likely alternatives are academic institutions or commercial ones. I am discounting the alternative of the pharmaceutical company itself managing the field trials, because of the potential conflict of interest, the question of the credence in the results such companies would report, and the fact that that is a different skill than drug discovery and drug manufacture. Academic organizations are perhaps more credible, since the principle investigators have responsibilities for medical and academic respectability. On the other hand, academics are notoriously challenged by keeping deadlines. 

Private firms may be either for profit or non-profit. In either case, a company that had the primary function of carrying out pharmaceutical field studies would be likely to be good at estimating costs and carrying out the studies expeditiously. I personally would lean towards non-profits to take more of the financial incentives out of play.

Complementary Studies: Now that we have digital health records I would suggest that we need a national system for drug review that would analyze the results of use of approved drugs in clinical practice.

Regulation: Perhaps we need laws to regulate this entire process. Clearly regulation should assure no fraud in the studies, and I would say complete reporting and availability of data. Perhaps there should be some kind of licensing for organizations conducting field studies.