An evaluation method for the evaluation of big data based streaming analytic clinical decision support systems

Abstract

This thesis presents a methodology for evaluating a scalable clinical decision support systems (CDSS) that uses high frequency streaming physiological data using a holistic approach that includes the presence of population health indicators. The plan applies concepts and uses indicators suggested in the HOT-Fit framework, while applying the evaluation template developed by Public Health Ontario and uses an indicator structure described in York Region Public Health’s Monitoring and Evaluation Framework. The methodology is applied within the research to the implementation of the Artemis Platform at the McMaster Children’s Hospital (MCH) Neonatal Intensive Care Unit (NICU). NICUs, have specific requirements relating to the use of clinical data and the implementation of new IT infrastructure. These requirements predicate the need for informative documentation that describes the utilization of the CDSS including a Privacy Impact Assessment (PIA), Threat and Risk Assessment (TRA), and a research and ethics proposal.