Extracting Evidence from Health IT Studies to Populate Logic Models

Abstract

Background: Logic models graphically present the socio-technical components of a variety of 'programs' such as educational programs. They show the underlying logic and assumptions of how a program is supposed to work. We suggest that they can be used to describe the mechanisms of complex socio-technical health IT interventions. Objective: To assess the suitability of logic models to describe cause-effect chains of health IT. Results: We are currently conducting an integrative review of the impact of patient portals on patient outcomes. We extracted the following elements of logic models from the found publications: resources, activities, output, outcome, and impact. These factors are then used to populate the logic model and form a structured graphical representation of the evidence. Until now, all the evidence we found could be fit into the logic model. The logic model was able to accommodate diverse types of evidence. Conclusion: Logic models seem to be suitable for representing evidence on the impact of health IT.