Towards effective data utilization in congenital cardiac critical care

Abstract

Critical care is among the most data intensive fields in health care, with multiple sources of physiologic measurements that are tracked both continuously and intermittently for the purpose of guiding ongoing treatment. Clinicians have a limited capacity to convert this data into actionable information, and thus there is an ongoing effort to develop sophisticated analytic support systems. The immediate technical issues of aggregating this data for analysis are significant but manageable. Analytical models may be generally categorized based on their abstraction of underlying physical principles. Models may be derived from experimental data through statistical processing (black box), from first physiologic principles (white box), or some combination of the two (grey box). Ultimately, successful analytic technologies will distill and reduce data and present the resultant information in a centralized, intuitive, and efficient manner.