No one wants to prepare for the worst, but an offhand comment during a team meeting led a principal investigator (PI) to create a transition plan in case he suddenly died or became incapacitated. He suggests others take the time in the initial stages of a study to make such preparations, rather than making these decisions during the emotional aftermath of an event.

The management transition path began for study PI Brandon Brown, MPH, PhD, when he expressed gratitude for a team member’s work. Brown said the member’s “big shoes” would be hard to fill if he were hit by a bus. Already touched by a colleague’s recent sudden death, Brown then started to think: What happens if I get hit by a bus?

After the meeting, Brown, a health services researcher and an associate professor at the University of California, Riverside School of Medicine, began working with another investigator, Annie Lu Nguyen, PhD, MPH, an assistant professor at the University of Southern California Keck School of Medicine, to create a “just in case” transition plan.

Nguyen started with a literature search for such transition plans and came up empty. (Read more here.) Like writing a will, clinical researchers can pre-plan for potential transitions. “There are too many cases of unexpected death that occur, leaving more questions than answers,” Brown says.

Sometimes, the funds for a research project are returned to the sponsor when the PI dies. Other times, the institution designates a new PI. “But if the institution is not heavily involved in the research, they may not know who might fit, and many (perhaps most) institutions do not have a plan of how to handle studies in case of the death of a PI. I applaud those who have a plan,” Brown says.

When developing their plan, Brown ensured all study documents were backed up and accessible in case of his absence. He approached the financial administrators at his university and his grantor about what would happen to the funding if he died. The answers were unclear.

Brown then requested Nguyen lead the study if he no longer could. Assigning a proxy PI is a step away from the direction of complete confusion that has happened in many cases when a PI died, Brown says. Once the study PI selects a proxy PI, he or she should be informed of the selection and agree to it. “Then, the study team should be informed that this person is the proxy PI, and the proxy PI name should be included in the IRB application documents listed as proxy PI.” These documents include contracts, manuscripts, funding agreements, and research ethics applications.

Selecting a proxy PI would add little work to the IRB, Brown says. IRBs can add a single cell to applications to request the PI (at the time he or she submits the IRB application) list someone as the proxy PI in the application.

“This way, in case the PI leaves the institution or has an unplanned illness or death, or for whatever other reason, another PI can be contacted as the ‘next of kin’ in the research,” Brown says. “If the proxy PI must also sign the application, it means that they are more likely to read through and digest all of the written materials, providing them more training to take over the project when necessary.”

For many studies, the PI is the only person listed on the informed consent document. “This is administratively problematic since if the PI does not or cannot check their email or phone messages, any communication from participants will be lost, or at least unanswered,” Brown says. “If there is a ‘study email’ that is accessible by both the PI and proxy PI, then this should be less of an issue. All study data and documents should be accessible by both the PI and proxy PI. Data can be stored in a repository that is secured by the institution, but only accessible by the PI and proxy.”

Brown says he does not believe there should be any necessary documentation above and beyond listing the proxy PI in the IRB application. “The proxy PI should adhere to the same legal requirements as the PI,” he notes.

Naming a proxy PI can be a tricky and sensitive discussion with the study team, the sponsor, the IRB, the community advisory board, and other partners. To help with this, a space should be designated for a proxy PI in all new IRB applications to make it an automated process. “In this sense, there wouldn’t be a discussion, but the IRB is informed about the proxy at the earliest stage,” Brown says. “Since sponsors do not select PIs (the PI is self-selected as the lead for applications), they should not select proxy PIs. The PI of any study will likely have a much better gauge of who can lead in their absence than a study sponsor, IRB, or any administrator at their institution.”

Brown also assigned a percent effort of the research to each team member and how the funds would be distributed in his absence. “Payment to the proxy PI (in place of the PI) should be handled in the exact same way as it was proposed and handled for the PI,” he explains.

Sometimes, this depends on the progress of the study; other times, this is paid as a lump sum. “Most important is to remember the proxy PI is PI [when needed], so there may be some up-front training to ensure the proxy is prepared to take on the role and understand all aspects of the study similar to the current PI,” Brown says.

One major complication of distributing payment to the proxy PI is if he or she works at another institution. “It is often the case when a PI is the only investigator at their institution, and their collaborators are at external institutions,” Brown says. “This is a complication that I leave to those who have to think about the financial implications of losing funding to their institution.”

Researchers also should create a process for notifying a regulatory agency in case the PI dies or becomes incapacitated. “In some cases, IRBs or grants management offices at different institutions have no experience with a PI dying, so they have to figure it out themselves,” Brown explains. “The idea of a proxy PI is broader than an IRB issue. It must be something implemented/understood at various levels (IRB, sponsor, study team, grants management, administration, department) for it to work and be useful when the unexpected happens. Each of those levels may decide on the best way forward to record and be prepared for a change from PI to proxy PI, but there should be some harmony.”

Other questions to consider in the transition plan are instructions on whether the PI would like to be included as an author on publications, or at least acknowledged on them. Also, where should any royalties be directed?

The impetus for thinking about this topic was cemented by various deaths in academia due to the COVID-19 pandemic. “Death is a part of life, and we need to be prepared. Discussing death seems to be taboo in general, and perhaps spoken less about in research than our personal lives,” Brown says. “Unfortunately, many of us are spending more time or just as much time in our academic work as we spend on self-care and with our loved ones. This is yet another reason to plan ahead for the unexpected in the research that is so important to us and may define our lives."

“Contemplating and talking about our own deaths can be uncomfortable,” Brown and Nguyen wrote in their article. “But by planning ahead, we find peace in knowing that we are doing our best to help our life’s work continue beyond our physical presence.”