Drug companies sponsor clinical research so that they can influence the results to their advantage. There is no incentive to be truthful and there is no penalty for dishonesty. Should a drug have serious side effects resulting in lawsuits then the discovery process might reveal some manipulation of the research results. But this happens far too rarely and even if dishonesty is revealed there are no effective penalties. Usually problems take years to come to light and by that time the drug company has already made millions or billions in profits. There is no way to force the drug company to disgorge the profits.
Authorship of a clinical study: Most clinical studies are published in medical journals as “articles” with a principal author and often several co-authors. The articles are judged on the prestige of the medical journal where they are published and the prestige of the principal author. The authors tend to be on teaching faculties of various medical schools. However, frequently the articles are written by drug company employees and some professor of medicine simply signs off on it. In one case it was found that the drug company completed the clinical trial and then went looking for some academic doctor to sign in as the principal author (1). So the authors may have no knowledge of the study until after it is over.
Multi-center Studies: Many clinical studies are carried out in medical centers spread across the world. The principal investigator does not have the time or the resources to monitor the study or to analyze the data from other centers. The drug company employees are happy to do that. All of the data is collected by the drug company employees and analyzed by company consultants. This data is treated as confidential information. The investigators may be prohibited by contract to divulge this data to anyone. The principal investigator may not even get the data from the different centers (2). The results of the trial, as written by the company employees, are presented to the investigators for their signatures.
Small Companies: There is trend toward small companies carrying out clinical trials. Each small company is engaged in the development of only one drug. The employees of the company are given stocks and stock options in that company. If the drug is not shown to be effective these stocks and stock options are worthless. These same employees are then given the job of running the clinical trials for that drug. They are gathering the data and doing the statistical analysis. If they err on the side beneficial to their company, they stand to gain financially. There is no downside to this for them. If “mistakes” are uncovered it could be years later and by that time they would have sold their stocks. There is no civil or criminal liability for such mistakes.
Patient Selection: There are criteria for inclusion and exclusion of patients for clinical trials. The company sets these criteria and usually excludes those patients that may be harmed by the drug. The investigators may take this a step further and restrict inclusion to relatively healthy individuals. Patients who get the drug after it is approved generally have a very different profile than the study patients.
After the drug is approved by the FDA there is no mention of the fact that some particular types of patients were excluded from the study.
Adverse Effects: Drug companies use various methods to minimize the reporting and statistical impact of adverse effects. Some investigators simply ignore a side effect and do not record it at all. Others use various cut-off dates to remove the adverse effects from the analysis. For example they might say that the data collection was stopped at a certain date, excluding adverse effects that happened after that date (3). At other times the company may remove the adverse effect if it happened immediately following a treatment, noting that it occurred during the normal recovery period. Only the data in the post-treatment period is analyzed (4).
If there are more deaths in the treatment group than the placebo group the difference is not analyzed statistically. Usually a statement is made saying the total number was too small or that the deaths were unrelated (5).
Statistical Analysis: This is where drug companies influence the outcomes of the trials most easily. Usually the patients in a study are evaluated in several different ways. Patients are often asked to keep a diary of their symptoms which provides a “symptom score.” Doctors examining the patients keep detailed notes. There is data from blood tests, x-rays, and other clinical tests. All this data is collected and analyzed by the drug company employees. Wherever the drug seems to produce a statistically significant benefit that particular finding becomes the center point of the clinical trial report. Areas where the patients on the study drug did worse than the patients on a placebo are not reported.
When the effects between the study drug and the placebo are small the data is analyzed in various ways to find some place where there is a statistically significant improvement for the study drug. For example, they may focus on the total number of patients who got better on the study drug, or on the degree of improvement from the drug, or on the percentage of patients who had a particular degree of improvement from the drug. The data that shows benefit, no matter how insignificant, becomes the selling point. The data that shows the placebo was equal to or better than the study drug gets ignored.
Publication of results of clinical study: If the clinical study shows beneficial results then it is published and publicized extensively. One study may be published in several journals in slightly different versions, giving the impression that there were multiple studies. If the study shows no benefit it may not get published, be delayed or be published in some obscure journal.
1. Merck designed and conducted study on its drug Vioxx. Then it went looking for authors in academic field. The study was published as ADVANTAGE study in Annals of Internal Medicine. 2003; 139(7):539-546. The lead author did not know of the study till after its completion.
Ross JS, Hill KP, Egilman DS, Krumholz HM. Guest authorship and ghostwriting in publications related to rofecoxib: a case study of industry documents from rofecoxib litigation. JAMA. 2008;299(15):1800-1812.
2. The Pharmaceutical Research and Manufacturers of America, the trade association of the industry, justified withholding data in this way: “As owners of the study database, sponsors have discretion to determine who will have access to the database.”
Steinbrook R. Gag clauses in clinical-trial agreements. N Engl J Med. 2005; 352(21):2160-2162.
3. Merck was testing Vioxx for prevention and/or treatment of Alzheimer’s disease. The company monitored the patients for one year after the discontinuation of drugs for possible improvement but did not report the excess heart attacks in the patients who had taken Vioxx. There were four times as many heart attacks on patients on Vioxx as compared to patients on placebo. The company did not notify the FDA, the volunteers or the institutes who carried out similar studies in subsequent years.
Reines SA, Block GA, Morris JC, et al. Rofecoxib: no effect on Alzheimer’s disease in a 1-year, randomized, blinded, controlled study. Neurology. 2004; 62(1):66-71. See the analysis of this study by Psaty et.al. in JAMA. 2008;299(15):1813-1817.
4. A medical device company, Asthmarx®, conducted studies on asthma patients. The patients were subject to a surgery called Bronchial Thermal Cautery (promoters call it Bronchial Thermoplasty). A large number of patients developed severe asthma immediately after surgery. The total number of asthma episodes was more in treatment group as compared to the placebo group, in the one year study period. However the company reported that asthma attacks decreased after the surgery. They excluded all asthma attacks in the post-operative period, calling it “treatment phase” and counted only the ones in “post treatment phase.”
Castro M, Rubin AS, Laviolette M, for the AIR2 Trial Study Group, et al: Effectiveness and Safety of Bronchial Thermoplasty in the Treatment of Severe Asthma: A Multicenter, Randomized, Double-blind, Sham controlled Clinical Trial. Am J Respir Crit Care Med 2010, 181:116-124.
5. In two studies on Alzheimer’s patients there were 57 deaths in 1069 patients on Vioxx and 29 deaths in 1074 patients on placebo (p=<0.001). However no statistical analysis was given in the articles. The articles concluded that “Vioxx is generally well tolerated by elderly patients.”
Psaty BM, Kronmal RA. Reporting mortality findings in trials of rofecoxib for Alzheimer disease or cognitive impairment: a case study based on documents from rofecoxib litigation. JAMA. 2008;299(15):1813-1817.