There is a need for evidence-based usability principles to support the design of usable medication-related computerized CDS functions and systems. Such evidence requires establishing scientific relationships between usability principles, their violation in terms of usability flaws, issuing usage problems and their consequences or outcomes in the clinical work and patient care. This kind of evidence is not currently directly available in scientific evaluation studies of medication CDS functions. A possible proxy to seek evidence is systematic review of existing scientific evaluation reports. We rely on a four-stage framework describing the chain of consequences and inferences linking usability principles to clinical outcomes to design the systematic review methodology and interpretation principles. This paper describes the four-stage framework and the resulting consequences for the systematic review design.