Why the REMAP-CAP rebuild mattered
The rebuild of REMAP-CAP was not only in response to growth, the addition of more global sites, or additional treatment options. The protocol itself had fundamentally changed.
REMAP-CAP evolved from an earlier model focused on severe community-acquired pneumonia in the ICU, 90-day mortality, and regimen-level randomisation, to a broader adaptive structure with a new composite endpoint, more sophisticated patient categorisation, domain-level randomisation, and greater flexibility to add, pause, or adapt interventions over time.
For software, those changes are fundamental. Randomisation is not a step added at the end of enrolment. It depends on eligibility, domain availability, intervention constraints, allocation probabilities, site settings, and the current state of the protocol. When those rules change, the safest approach is often not to keep layering exceptions onto an older system, but to rebuild the underlying logic so it reflects how the trial operates in reality.
Spiral and the REMAP-CAP team recognised that a new foundation would better support the trial's future direction, rebuilding the database to reflect the protocol as it exists today, rather than continuing to add workarounds to a system designed for an earlier version of the trial.
REMAP-CAP
One example of that evolution is PANDA, a framework used within REMAP-CAP to more clearly define categories for patient, pathogen, disease, and illness states. Some of those characteristics are fixed at enrolment, while others can change throughout a patient's illness. Those categories may then influence eligibility, randomisation, intervention availability, and later analysis.
In practical terms, this moves the trial beyond simply asking whether a patient is enrolled. It allows the protocol to define how different groups of patients are categorised and how those categories influence the subsequent treatment pathways and analyses.
"The newer system makes it much easier to verify what is happening and why. When you're working with allocation updates or protocol changes, you can see the logic more clearly and be confident that what you're applying reflects the protocol as intended."
That clarity extends to the enrolment process itself. Rather than progressing through multiple steps only to discover an issue at the end, users receive clearer feedback throughout the workflow, helping them identify what needs attention before moving forward. In a critical care environment, where clinicians are balancing patient care alongside research activity, that matters a great deal.
"ICU trials don't happen in a quiet room. The system needs to support fast decisions and clear workflows without expecting users to hold complex protocol logic in their heads."
The rebuild provided REMAP-CAP with a stronger foundation for the future. It reduced ambiguity around enrolment and allocation, made protocol logic easier to verify, and created a more flexible framework for supporting future domains, interventions, and adaptations as the trial continues to evolve.
