What happens if the tailoring variable that is needed to inform the next intervention could not be collected, or its value is missing, for a specific patient?
This should be planned for as part of the adaptive intervention design. High-quality adaptive interventions include contingency plans for events or circumstances that are not uncommon but cannot be predicted with great certainty, such as missing data on a tailoring variable. Note that missing data on a tailoring variable is not a statistical challenge; rather, it is a clinical challenge.
Adaptive interventions should include components that aim to prevent missing data on tailoring variables; and they should guide how to make intervention decisions when a tailoring variable is missing. When a tailoring variable (value) is missing, intervention options available to the clinician include (i) using the last known values of the tailoring variable to decide the patient’s current status, (ii) assuming the patient is a non-responder (or a responder) until the tailoring variable can be collected, (iii) specifying a window of time during which the tailoring variable can be collected, and/or (iv) .