Development of a hand-held real-time decision support aid for critical care nursing

Abstract

In the current health care environment, nurse clinicians must work "faster and smarter" making complex decisions on almost a continual basis. Evidencebased knowledge and standardized guides, such as clinical algorithms, can support clinical nursing decisions, however; effective real-time access is limited. This paper outlines research addressing this problem. In this research, current clinical knowledge is delivered to the clinician via an off-the-shelf handheld computer using wireless access to a central server and data repository. Innovative minimal-set database, data mining and knowledge discovery algorithms using a combination of case based and rule based learning with added confidence measures permitting bi-directional (forward and backwards) inferencing based on individual client data are developed and presented for the hand held device. The technology provides real-time decision support for the multiple cases and sequential decisions characterizing present critical care nursing practice. Nurses will be able to consider a full range of alternative explanations, determine additional data needs, find, isolate and examine patient case outliers for additional diagnostic data or verify the appropriateness of a selected strategy. Once fully developed the system will have the capacity to maintain a history of a series of decisions and outcomes thereby over time improving the case base and rule bases used for decision support. Outcomes of the real time decision support aid include more timely health care, less biased decisions, and improved patient outcomes.

A Tablet PC based System for Ubiquitous Patient Monitoring and Smart Alert Generation in an Intensive Care Unit

“…Here the vital signs (heart rate, blood pressure, blood oxygen saturation, and respiratory rate )from monitoring devices are fed to a classifier [7] module which can classify these signals into different labels like (Low, Normal ,High ,Very High etc.). The classifier output is fed to a decision support system [8] [11] which consists of a smart alarm generator [9] and a context aware rule set. The decision support system also collects inputs from electronic health records which can provide useful information regarding potential drug interactions and specific patient physiological parameters.…”