Real Time Clinical Alerting with Decision Support

Team
Medical Informatics

Real Time Clinical Alerting of Critical Laboratory Values: Improving the safety of patients with critical laboratory values with automated alerts and real time decision support

Patients with critical laboratory values and other signs of deterioration require prompt and correct clinical care, often with evidence based resources to inform an appropriate response. Done jointly with Sunnybrook Hospital, the University Health Network (UHN) is evaluating a system to deliver clinical alerts with associated decision support.

Abstract

Patients with critical illness, physiologic derangement and corresponding critical laboratory values require prompt and correct clinical care. Unfortunately, delivery of timely and appropriate care is often delayed. Lack of access and delayed knowledge of critical laboratory values are significant contributors to potentially preventable adverse events. While many patients may have laboratory evidence of deterioration, the appropriate clinical response is delayed because the clinician may not become aware of the abnormalities in a timely and consistent manner. Additionally, clinicians may require additional resources or information to respond promptly and appropriately. Our goal is to integrate critical laboratory data with evidence based decision support, and automatically deliver this information to the clinician responsible for the patient. The University Health Network and Sunnybrook Health Sciences Centre has developed and pilot tested an automated system that detects and communicates critical laboratory values to clinicians at the point of care. In a pilot study, this system reduced the time to response, defined as the time from the abnormality being reported by the laboratory to the time of a physician order. However, it was found that even though the clinicians received this information in a timely manner, they often did not provide the recommended or appropriate care. Providing prompt but inappropriate care may be a much greater risk to patient safety than providing delayed or no care.

Evidence-based medicine is the application of knowledge gained from clinical trials and studies to clinical practice, in an effort to promote patient safety and quality of care. In an environment where there is great variation in treatment, even within the same organization, standardization of best-practice treatment is important to ensure consistent best practice care to patients. The development of evidence-based guidelines can help to reduce treatment variation but asking a busy clinician to actively retrieve this information during the flow of care is very ineffective. An approach that presents this information in a succinct format within the flow of care may be a more effective tool towards evidence-based care delivery.

Integrating evidence-based medicine with clinical decision supports systems has the potential to substantially improve patient safety and quality of care. Decision support systems that communicate critical laboratory values to clinicians at the point of care allow them to respond immediately to the problem. Incorporating succinct, evidence-based guidelines with this communication ensures that treatment is appropriate for the given problem. Supplementing this communication with electronic access to patient specific data would also promote patient-specific decision support at the point of care, greatly reducing adverse clinical events and increasing patient safety.

Funding Agency

Canadian Patient Safety Institute

Project Sponsors

Dr. Ed Etchells

Dr Peter G. Rossos

Dr. Howard Abrams

N.K.J. Adhikari

R. Fowler

M. Cheung

Sherman Quan

Dr. Robert Wu

Organizations

University Health Network

Sunnybrook Health Sciences Centre

Project Team

Tim Tripp

Melinda Hamill

Svetolik Kovacevi