Expert Insights

Expert Insights

Impact of COVID-19 on Biostatistical Options

Impact of COVID-19 on Biostatistical Options
01 May 2020

With most standard clinical trial activity slowing down, or even stopping because of the impact of COVID-19, there are still meaningful things that can be done to continue clinical development. Within biostatistics there are the obvious, minimum steps we can take such as updating the SAP with regards to Protocol Deviations due to COVID-19 per FDA guidance or helping with CRF updates; but these don’t exhaust our options for action.

As the FDA guidance indicates, there will almost certainly be issues with patient compliance to trial conduct due to COVID-19. This may ultimately result in a reduction in your “per protocol” population. But that’s not the only way in which your analysis populations, or more generally trial sample size, could be impacted.

Some patients will end their trial conduct early, not wishing to expose themselves to additional risks by visiting a medical facility. Some of these patients may even end their conduct prior to the collection point of key efficacy data, thus ending their ability to contribute to the primary analysis. In other cases, patients who might have otherwise joined a trial will have their participation, at a minimum, delayed, and some will be completely discouraged for months to come (or permanently).

How can statisticians respond to COVID’s impact?

First, consider how we got to our study’s target sample size to begin with. While in some cases it may have been arbitrarily set, especially in later phases, it was almost certainly based on a concept of statistical power.

Statistical power is a measure of probability for a hypothesis test to be able to detect an effect, given that one actually exists for a particular sample size and set of product performance assumptions. Statistical power allows us to quantify a key aspect of trial design and ultimately aids in critical trial design decisions.

Using statistical power, we can gauge the impact of an increase (or decrease) in our estimated sample size, which in turn lets us make decisions on trial design characteristics. As we go through this optimization problem — balancing information, cost, and time — we come up with an estimated sample size that allows us to meet our clinical goals within our operational constraints.

That said, while there is a mathematical rigor behind the sample size estimate we have selected, there are many scenarios that would have been appropriate and simply weren’t chosen. In other words, the sample size specified in the protocol isn’t necessarily the only sample size that will allow you a successful trial and valid analysis.

How does lower enrollment impact sample size?

In some cases, the simplest solution available may be to merely wait out the delay that is causing lack of enrollment, and gradually enroll additional patients until you reach the necessary nominal or effective sample size.

But that doesn’t have to be your only solution. Knowing that a trial’s sample size estimate is one of many that may be sufficient, we can begin to probe at the information we have and make critical decisions from there. Options include:

  • Consider the data collected and determine the effective statistical power based on pretrial assumptions. You may find that in cases where the patient deficit is minimal (perhaps less than 5% of the total target) that a sufficient power is available to consider ending the trial early. Or, more generally, you can assess potential scenarios of the completion rate of active patients to determine a threshold of concern. This can be a tool to determine whether you need to enroll additional patients.
  • A more complicated option could be to consider a blinded sample size reassessment. This is a statistical tool where, based on the data as collected, the operating characteristics of the trial are assessed, and a new sample size estimate can be determined to meet the necessary statistical power. Sometimes this can mean enrolling additional patients beyond the originally planned sample size, but it can also let you know that the patients you have may be sufficient.
  • For trials very early in their progress, a statistical analysis of futility could be considered. A futility analysis is an early estimate that compares the likelihood of a successful trial given the data collected against a prespecified threshold; if it the likelihood is less than that threshold then the trial is stopped. While none of us hope for this outcome, knowing this early and prior to an extended hold can allow a team to use that same time to plan the next trial instead of waiting for one unlikely to succeed.

One important word of caution regarding taking any of these steps which may alter a trial’s conduct outside of the protocol guidelines: Don’t do anything without taking the necessary steps to review and assess.

At a minimum, this includes consulting your statistical, regulatory, clinical, and medical advisors and likely also includes the relevant regulatory agencies. Further, these scenarios are only a few of the potential avenues you have available to you; the right plans of action are specific to your trial and the best course of action may be to hold the line. Even in times where action may seem difficult, there are options if we are simply willing to step back and ask.

x

Submit RFI / RFP

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

Required

x

Contact Synteract

Tell us how to stay in touch with you:

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

Required

x

 

x

 

x

Download Resource

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

*Required

x

To complete your download, please choose an option in the Internet Explorer pop-up window at the bottom of this browser.

x

Download Resource

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

*Required

x

To complete your download, please choose an option in the Internet Explorer pop-up window at the bottom of this browser.

x

Opt In

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

*Required