How the COVID-19 Pandemic has Affected Clinical Trials

Throughout this blog series, we will be discussing how the COVID-19 pandemic situation has affected clinical trials, focussing on the pertinent statistical issues and associated key recommendations. Rather than providing an exhaustive summary of all potential issues and possible solutions, the aim is to concentrate on the key areas and questions that statisticians on project teams should be asking and thinking about. Whilst there are important aspects regarding safety data, the focus of this series is on the considerations for efficacy data and so the main points covered are estimands and missing data.

The World Health Organisation (WHO) declared COVID-19 a global pandemic on 11th March 2020 and the subsequent restrictions on everyday life caused significant disruption to clinical trials. The full extent of the impact of COVID-19 will not be known for some time but the effect on clinical trials will most certainly be significant. Although sponsors have been quick to make changes to clinical trials there will undoubtably still be many issues to address. The range of disruptions of COVID-19 to clinical trials are expected to be so diverse that no single solution will be appropriate for all trials. It is likely that each trial will need to be assessed on a case-by-case basis, although there will be many common issues. It is critical for sponsors and regulators to work together and make the best use of clinical trials affected by COVID-19. Collaboration and sharing of learnings during the pandemic are going to be vitally important for a long time to come.

The COVID-19 pandemic brought about global quarantines and travel restrictions which resulted in site closures and interruptions to investigational product (IP) supply. These all affected trial conduct, data integrity, and interpretation of the results in numerous ways, including protocol deviations, timing of and missed assessments, collection of trial endpoints, treatment discontinuations or interruptions, participant withdrawal, and use of alternative medications. In view of this, the use of different methods and/or assessments for collection of endpoints needs to be considered. Additional data, such as reasons for missed/delayed visits/assessments, will also need to be collected in the clinical database as it is important to capture all COVID-19-related data in current trials. In terms of safety data, occurrences of COVID-19 need to be reported as adverse events. The Medical Dictionary for Regulatory Activities (MedDRA) coding dictionary Version 23 has now been updated with a coding term for COVID-19.

A review of recently published documents from the FDA and EMA relating to the current statistical issues faced in clinical trials during the COVID-19 pandemic confirms that the various information and guidelines are consistent with each other. The main statistical issues are:

  • missed visits/assessments,
  • discontinuations (treatment and study),
  • treatment interruptions,
  • and altered types of visits (virtual instead of on-site)

Our standard processes are well-equipped for us to deal with these issues appropriately, but care should be taken not to introduce any bias when making amendments or modifications to trials. Any decisions taken regarding modification to the trial and/planned statistical analyses should be based on blinded data only and in some cases, it may be necessary to consider using an Independent Data Monitoring Committee (IDMC).

It may also be important to assess the impact of stopping a trial early on the sample size and power. Additionally, it will also be important to consider the impact that the COVID-19 vaccines may have on participants in clinical trials, in particular additional adverse events and determining the relatedness of these adverse events to study drug/vaccine, interactions with study drug, etc. Further, consultation with regulatory agencies is highly recommended.

The general process is to first assess the impact of the pandemic on the trial, then define the risks, so that once we understand the risks, we can then look at the contingency measures/mitigation whilst ensuring that data integrity remains of key importance. We will be looking at these in more detail in our next blog.