Glucose meter standardization across a large academic hospital system.

• A strong multidisciplinary organizational structure is necessary to drive consensus in selection and successfully implement point of care glucose meters across a large multi-hospital system. • Beyond the standard analytical laboratory evaluation, meter performance should be evaluated at the point of use, employing consensus error grids and defined acceptability criteria. • Structured end-user evaluation is as an essential element of the meter selection process which commonly includes the evaluation of cost, connectivity, indications for use, and ease of implementation. To select and standardize point-of-care (POC) glucose meters across a multi-hospital system. We formed a multidisciplinary POC glucose standardization working group including key stakeholders from each site. A set of selection criteria: usability, clinical and laboratory performance, indications for use, interface connectivity, ease of implementation and ongoing operational costs were used to develop a scoring schema to facilitate a consensus-driven selection process. Method comparison and consensus error grid evaluation against the clinically validated reference methods demonstrated that the analytical performance for all candidate meters was comparable for both the laboratory and clinical evaluation. However, Meter 1 ranked highest in usability evaluations, implementation and streamlined interface connectivity. The meter selection process and implementation were staggered across sites due to complexity of transitioning to a new manufacturer's meter and limitations in vendor support for training and ongoing troubleshooting of interface connectivity. Standardization of POC glucose meters in a large multi-hospital system is a complex undertaking requiring robust, multidisciplinary organizational structure both system-wide and locally, development of consensus-driven selection tools, usability evaluation by end-users, laboratory and clinical evaluation of the analytical performance, and a strong vendor-laboratory partnership during the implementation process. [ABSTRACT FROM AUTHOR]