How Clinical Trial Data Impact Issues in Litigation

Clinical trials can provide the most robust and comprehensive evidence that a drug or medical device is safe and effective. The Department of Health and Human Services recognizes that “clinical trials produce the best data available for healthcare decision making” and “are important because they advance medical knowledge and help improve patient care.”[1] It is therefore not surprising that the data from clinical trials and a drug manufacturer’s interpretation of such data often take center stage in product liability cases.

A well-designed and conducted clinical study has many benefits: it may be the most powerful evidence to dismantle the opinions of a plaintiff’s expert (e.g., one who selectively relies upon negative aspects of the study and disregards data that do not support the expert’s opinions), thus disproving allegations of design defect or inadequate warnings. A well-designed study can also be effective proof of a company’s good faith in developing and appropriately marketing a product based on sound study results.

On the other hand, any clinical trial is always subject to retrospective criticism. Such criticisms may include that the study was under-powered and therefore cannot provide a statistically supported conclusion, or that the trial was plagued with poor patient selection, lacked sufficient follow-up, or failed for lack of blinding. Beyond these structural problems, plaintiffs often fault the study’s ultimate conclusions that the product is safe and effective due to the improper use of liberal success standards and large confidence intervals to inflate results.

Given the current state of the mass tort market, clinical trial data are—and will be—the focus of important litigation issues. A few major topics, including discovery, witness selection and evidentiary issues, are discussed below.


Clinical trial data may be held by a trial investigator, an institution or a university. Such information may be subject to subpoena, which risks confidentiality of the data as well as presenting a time-consuming obligation to respond for the investigator. This is particularly true if release of the information could impact the sponsor’s ultimate goal of product approval by leaking useful information to competitors. In such a situation, the sponsor may seek a protective order or otherwise intervene to quash the subpoena.

In addition, such data may be considered a business record of the trial sponsor: a compilation of events near the time of experience that the sponsor would maintain in the regular course of business. This designation not only makes the data discoverable from the sponsor, but paves the way for wholesale admission of the information at trial.

The clinical trial data may even be considered an admission of the sponsor, particularly where the information was provided to the FDA as part of an approval process. Of course, one can presume that if the information was submitted to the FDA as part of the approval process, then it would support the sponsor’s position as to the safety and efficacy of the drug or device. This is more likely to become an issue if later studies refute the originally submitted information or otherwise undercut the propriety of the sponsor’s reliance on it. Thus, even where clinical trial data are not within the control of the product manufacturer, any completed protocol forms or data compilations provided to the manufacturer from the study can be cast as business documents or potentially as admissions.


Clinical trial data and the patterns discerned from such studies will be a likely focus in any products liability trial. Clinical studies are often sponsored to some degree by a manufacturer, especially when a company is looking to validate a drug or device’s safety and efficacy before that product reaches the market. Plaintiffs will undoubtedly seek to develop a narrative to cast doubt on the results—such as strongly suggesting (if not saying outright) that any positive results from the study are tainted and cannot be trusted because the results are bought and paid for. Knowing this tactic is in plaintiffs’ playbook, manufacturers must develop a more accurate narrative, such as explaining how clinical trial data demonstrate the manufacturer’s careful study of the product and development of adequate warnings.

No matter the vantage point, one thing is clear: clinical trial data are complicated and highly technical material. Despite their utility in showing product safety and labeling adequacy, clinical data are often rife with dense statistical analyses of complicated information that may confuse jurors—or perhaps even worse, lull them to sleep. Judges, particularly those undertaking a new docket of pharmaceutical product liability cases, may have never been exposed to the complex science behind clinical trials, and as a result may be completely unfamiliar with the methodology supporting the interpretation of clinical trial data. Biostatistics (statistics related to biological data) must be broken down into manageable information to establish either the safety of the product or to show that any reported adverse events are not statistically significant related to product safety.

Given that a lay jury (or judges without scientific expertise) may be called upon to scrutinize clinical trial data, a critical issue is identifying the best witness for introduction of this proof. A biostatistician can delve deeply into the minutiae of the data, but unless such a deep dive into the particulars is critical, such information may be of limited value due to its complexity.

Often the best witness would be the clinical trial investigator who can explain the details of the study by reason of first-hand participation. An investigator knows exactly how the trial was conducted, how data were compiled and interpreted and how such information was reported to the sponsor. But this witness may also be subject to impeachment on the basis of bias, either by virtue of being directly compensated by the sponsor or by virtue of working at the university or institution that is paid by the sponsor.

And a thorough company-sponsored clinical trial can itself be used to undercut such claims of bias. This is particularly true where the study identifies:

  • Shortcomings or limitations of a drug or device
  • New complications associated with a drug or device
  • Scientific or medical controversies surrounding a drug, device, or surgical procedure
  • Adverse events, including statistically supported percentages and durations
  • Recurrence rates and/or the need for surgical reintervention
  • Definitions of surgical or treatment success
  • Ideas to make future studies more robust
  • Identification of areas in which a drug or device can be improved as medical science evolves.

Openly discussing the complications that arise in a clinical trial illustrates that the sponsor carefully considered and was transparent about complications. This can undercut a common plaintiff theme: that the manufacturer was motivated to downplay negative results in a rush to get the product to market.


Not to be overlooked is the usefulness of clinical trial data in exposing the bias of a plaintiff’s expert. If the expert failed to consider all of the data or dismissed the entire trial as company-sponsored, this provides a basis to challenge the plaintiff’s expert’s methodology because of cherry-picking data and focusing only on information helpful to the plaintiff’s case. Such cherry-picking renders these opinions unreliable.

In addition, plaintiff’s experts who testify on clinical trial data may be tempted to go beyond “just the facts” and offer opinions that venture into corporate ethics, conduct or state of mind. Experts may attempt to use any negative data from the clinical trial to opine that the product manufacturer “knew better” or put profit over public safety. Courts have routinely excluded such testimony as speculative and misleading.


There are important determinants to evaluate the strength or weakness of a clinical trial when evaluating the data in a litigation setting.[2] Such factors should be thoroughly vetted with the product manufacturer’s witnesses related to clinical trials. These factors should also be evaluated in preparing to examine plaintiff’s experts and to discount any study that undercuts the safety and efficacy of the subject product.

One key factor concerns the population of patients in the study. Whether the selected participants are proper ones to test the ultimate hypothesis is a central question. It is also critical to determine if the participants were told of the potential effects of treatment, both positive and negative. Planting the “seed” of the ultimate conclusion may introduce bias into the trial by stimulating the placebo effect.

In addition, the sample size must be sufficient to yield the statistical power necessary to identify and define clinically meaningful differences. This may require a power analysis to determine an appropriate sample size that will permit the investigator to identify the meaningful differences between treatments. Any such power analysis should be conducted in advance of the study. So, too, the participants must be sufficiently diverse to mirror the age, gender and race distribution of the target population so that the results can be generalized to additional treatment settings. Finally, the site(s) of the clinical trial can impact the eventual success/failure rates of the trial. For example, if the trial is conducted at a single site with exceptionally skilled physicians, the results may be subject to additional scrutiny due to the inability to recreate the results in the general population.

The outcome of any clinical trial should be scrutinized to determine if it was clinically supported. This may include evaluation of both objective and subjective outcomes and an analysis of the clinical relevance of the supporting data. Part of the examination of the outcome will include whether there was an open-phase of the trial where patients were no longer blinded to treatment. This again may create an issue of bias/placebo effect and compromise the conclusions because of the lack of a control group. So, too, whether clinical trial follow-up has been conducted may be paramount. The follow-up period should be sufficient for the outcomes being evaluated to manifest, and a failure to follow-up may compromise the reliability of the study.

While clinical trial data, their interpretation and their analysis may be complex and are likely something never encountered by the ordinary juror, dissection of the information plays a key role in both bolstering a product manufacturer’s claims of product safety and efficacy and in challenging a plaintiff’s expert’s opinions to the contrary. Presentation of such important data at trial requires a deft hand and a knowledgeable witness.


[2] Hannan EL. Randomized Clinical Trials and Observational Studies: Guidelines for Assessing the Respective Strengths and Limitations, J. Am Coll. Cardiol. Intv., 2008;1:211-217.