Session P92.7

Protocol Adherence on the Cardiac Intensive Care Unit: Glucose Control as a Model for Implementation of Clinical Decision Support

JA Lipton*, RJ Barendse, SP Nelwan, TB van Dam,
MJB van Ettinger, N van der Putten

Erasmus Medical Center
Rotterdam, Netherlands

Introduction: Glucose regulation in acute cardiac disease is difficult to achieve and may improve patient outcome, however the effect of tight glucose regulation on these patients is not well known. In 709 patients admitted to the Cardiac Intensive Care Unit (CCU) over a 12 month period mean glucose levels were above protocol advised levels (4.5-7.0 mmol/l) in 62%. Subsequent glucose measurements were done within the protocol advised timeframe (and a 10% margin) in 36%. Limited availability and complexity of the paper protocol, no notification of new lab result, and difficulty to achieve the time for next measurement may have limited adherence to the glucose management protocol. Automatisation is associated with increased adherence to protocols, and may improve patient outcomes.
Methods: We built a lightweight, AJAX driven decision support program (DSP), for easy implementation and integration with the different applications on the various displays on the CCU. The hospital information system transmits laboratory values in HL7 to the patient data monitoring system (PDMS). The glucose DSP checks for new laboratory results every 10 seconds, and simultaneously retrieves insulin infusion pump settings from the PDMS. When a new glucose value is retrieved, this is entered into the DSP, which generates an audible alert and a pop-up window with the protocol advised action regarding change in insulin infusion rate and the time for the next glucose measurement. A pop-up window and alert are also generated at the time when the next measurement is to be done. A graphic display of the temporal glucose trends and insulin infusion rate serves as an additional aid in the glucose management.
Evaluation of Methods: Times of glucose results and insulin infusion pump change are stored in a separate database to evaluate protocol compliance. Additionally glucose ranges and patient outcomes will be compared before and after implementation of the DSP. User feedback will be used to optimize the DSP interface.

(Abstract Control Number: 305)