Consider Data Needs Carefully When Altering Trial Protocols in Response to COVID-19
To get through the COVID-19 crisis with their trials intact, sponsors and sites should take a new look at their data and data collection methods, including evaluating whether a trial will have enough reliable data to support a positive decision, finding ways to collect more data or using data monitoring committees (DMCs) to help assess and modify studies.
Different types of trials need different solutions, says Jonathan Seltzer, chief scientific officer of WCG Clinical. Some trials may have enough data to close out the study early, others will have to decide whether data still to be collected is important enough to be worth putting people at risk or find other ways to get needed data, Seltzer says.
“The DMCs may have some really good input about what data can be collected during this time,” he said at a WCG Clinical webinar last week on coping with the pandemic, “but you should really consult with the DMCs to discuss the safety of the participants and the maintenance of the trial’s integrity.”
Seltzer also said that DMCs and Data Safety Monitoring Boards (DSMBs) can play a role in amending protocols to mitigate the risk of adverse events (AEs) and serious adverse events (SAEs) related to the virus. Typically, DMCs/DSMBs are unblinded and they may have already reviewed efficacy data.
Guidance from the FDA allows for consultation with DMCs/DSMBs about study modification. Guidance from the European Medical Association recommends using DMCs/DSMBs to help:
- Re-start usual trial operations and additional measures when completing the trial after the pandemic (e.g., validation of outcomes that were measured differently);
- Determine the need to adjust the trial sample size and additional analyses to investigate the treatment effect of the pandemic before, during and after the trial; and
- Deal with any identified potential sources of bias, such as missing values, newly identified intercurrent events or other unforeseeable required changes to trial elements.
Sponsors may consider forming DMCs/DSMBs if their trial doesn’t already have one, said Janet Wittes, founder and president of WCG Statistics Collaborative. The DMCs/DSMBs may help sponsors clearly define AEs of special interest and help identify whether the trial should continue. In terms of COVID-19, the DMCs/DSMBs may also be able to help sponsors define the impact of the pandemic on the trial itself.
But, Wittes warned, sponsors should avoid giving DMCs/DSMBs too many responsibilities, which could compromise the integrity of the trial and may ultimately render its data invalid. “While it may be tempting to go to them to seek recommendations on what to do with a trial,” she added, “an open-ended question like that can affect the validity of the trial.”
When capturing data on COVID-19-related AEs, “Let’s think about the most important safety information,” Seltzer said on navigating trials during the pandemic. AEs can be classified as severe, serious and of special interest and Seltzer offered guidelines for how to classify COVID-19-related AEs when reporting them to sponsors and regulators.
He provided the following classifications:
- AEs – A trial participant who becomes infected with the virus but is asymptomatic.
- SAEs – A patient who is hospitalized with COVID-19 and/or dies, including patients treated in non-hospital environments used to handle the overflow of COVID-19 patients.
- AEs of special interest – events specific to a certain trial, patient or COVID-19 that might otherwise not be considered reportable.
Wittes also cautioned webinar attendees about making protocol adjustments to a trial that has an FDA-approved Special Protocol Assessment (SPA) plan. To protect the integrity of the SPA, Wittes recommended:
- Writing a supplement to the trial’s statistical analysis plan;
- Considering carefully whether a change in primary outcome or primary method of analysis will impact the SPA; and
- Clearly documenting and justifying changes or additions being made.
In addition to working with DMCs/DSMBs to guide protocol changes and AE reporting, Wittes said sponsors of trials that are not directly studying COVID-19 should think about whether their trial offers enough clinically meaningful benefit to warrant continuation. Additionally, she suggested studies in their early phases should hit the pause button. “If your trial is still in the design phase or in the screening phase,” she said, “my advice would be to halt recruiting and take this time to make sure your protocol, case report forms and databases are clean before you can start up your trial.”
If trials are in the phase where the last visit has occurred and it’s time to collect the queries, Wittes suggests sponsors should think a little harder about how to collect those queries. “Remember that the [sites] are going to be extremely busy with COVID-19 cases,” she said, “so some of the queries that really aren’t essential to the analysis of the data or interpretation of the study don’t necessarily need to be collected.” She emphasized the importance of only collecting those data that are central to the interpretation of the trial. These data include important safety and efficacy endpoints as well as the primary outcome and important secondary outcomes.
Seltzer added that the possible silver lining of this pandemic will be that the industry will develop some best practices for clinical trials that will help prevent possible future infections. These best practices will probably be developed in relation to remote assessments and virtual visits. “We all know that the reason people don’t enroll in clinical trials is the burden of coming in,” he added, “so we may be able to improve the entire process [with virtual trials] and potentially protect the safety of our patients by keeping them out of hotbeds of infection.”
The COVID-19 crisis may also help the industry learn how to perform trials in a simpler way that’s less visit- and data-intensive, Wittes added. “The clinical trials community should do a real self-reflection about whether we are collecting too much data and if we are demanding too much of our participants,” she said. In addition, Wittes said she hopes that this unprecedented time may inspire sponsors to consider thinking about other pandemics or natural disasters when writing future protocols. “We should consider putting in a section in the protocol on what centers should do when these things happen,” she said, “so a combination of protocols that identify that these things may happen and what centers should do will be really important.”
To listen to a recording of the webinar, click here: https://bit.ly/2X2I0KC.