RAPS is closely monitoring developments in the Coronavirus (COVID-19) outbreak. See our public safety page for the latest updates.

Regulatory Focus™ > News Articles > 2020 > 6 > FDA offers statistical guidance for trials impacted by COVID-19

FDA offers statistical guidance for trials impacted by COVID-19

Posted 17 June 2020 | By Michael Mezher 

FDA offers statistical guidance for trials impacted by COVID-19

The US Food and Drug Administration (FDA) on Wednesday issued immediately effective guidance offering statistical advice to clinical trial sponsors with the aim of maintaining trial integrity and mitigating the effects of coronavirus disease (COVID-19) public health emergency on clinical trials.
The new document follows the agency’s guidance on the conduct of clinical trials amid the pandemic, which has been updated several times to include additional questions and answers since being issued in March. (RELATED: FDA unveils guidance for trials impacted by COVID-19, Regulatory Focus 18 March 2020; FDA updates guidance on clinical trials amid COVID-19, Regulatory Focus 13 May 2020).
While the agency’s clinical trial conduct guidance has focused on ways to protect patients and mitigate the effects of COVID-19, the new statistical considerations guidance is meant to help sponsors with statistical considerations for changes to a trial.
“To help ensure that the trial will provide interpretable findings with correct statistical quantification of uncertainty, this guidance addresses statistical considerations for proposed changes to trial conduct due to the COVID-19 pandemic that may impact the analysis and interpretation of the primary or key secondary endpoints in the trial,” FDA writes, adding that sponsors should discuss any proposed protocol changes and changes to their statistical analysis plan with the relevant review division.
To ensure trial integrity, FDA says sponsors should “proactively plan to address the impact of COVID-19 on the ability to meet the trial objectives.”
Changes that would impact the ascertainment of primary or key secondary endpoints should be submitted in a protocol amendment or an investigational device exemption (IDE) supplement, while modifications to primary or key secondary endpoint analyses “should be reflected in an updated statistical analysis plan before locking the database and before any modification to unblinded interim analyses.”
FDA also says that sponsors should avoid any modifications to the trial that may introduce bias into the interpretation of the results of the study.
“Generally, for a blinded trial, modifications based on information that reveals the magnitude of the treatment effect or information presented by treatment arm have the potential to introduce bias,” FDA writes.
The guidance provides a list of data to consider when making modifications to a trial to address COVID-19, including summaries pooled over treatment arms, participant treatment discontinuation or interruptions, participant trial withdrawal and endpoints. Site closures and supply disruptions may also be appropriate to consider.
Additionally, the guidance provides several design and analysis strategies for evaluating primary and key secondary endpoints in trials affected by COVID-19.
“As discussed in the Conduct Guidance, it is important to capture specific information at the participant level, describing the context and/or reasons for post-baseline events as they relate to COVID-19, such as discontinuation of treatment, withdrawal from the trial, use of alternative or rescue treatments, missed endpoint ascertainment, and the use of alternative endpoint ascertainment methods,” FDA writes, noting that such information may be useful to address potential biases or for sensitivity analyses.
Other strategies include how to handle loss of statistical power due to smaller sample sizes or shorter follow-up durations; how to approach statistical analyses in trials stopped early or the addition of interim analyses to trials; considerations for increasing enrollment after the pandemic or extending follow-up; how to handle missing endpoint ascertainment or investigational product supply disruptions; and where modifications to the definition and ascertainment of endpoints might be warranted.

Regulatory Focus newsletters

All the biggest regulatory news and happenings.