Objective: Diagnostic testing is frequently overused in the intensive care unit. We devised guidelines to optimize blood tests utilization, and designed this study to quantify their efficacy over time, their safety, and their possible benefits.
Design: Laboratory testing guidelines were created by consensus and implemented through repeated staff education. The guidelines included: a) the tests to be obtained daily: complete blood count, serum electrolytes, urea nitrogen, creatinine, and blood glucose concentration; b) the need to discuss laboratory testing at daily patient's rounds; c) the need to provide a written order for all tests. The number of tests performed, corresponding physician orders, and various outcome measures were collected for two 6-month study periods, before and after the first day of implementation of the guidelines.
Setting: Twenty-bed surgical intensive care unit in a tertiary care teaching hospital.
Patients: All patients admitted during the two study periods.
Methods: Laboratory tests and related physician orders, demographics, blood products transfusion, and outcomes were collected from hospital databases. In prospectively defined subgroups, additional outcome measures were obtained by ad-hoc chart review.
Results: One thousand one hundred seventeen patients were enrolled. After the institution of the guidelines, the number of laboratory tests decreased by 37% (from 64,305 to 40,877), and the number of respective physician orders increased by 38% (from 20,940 to 35,472), p < 0.001. These results were manifest within 1 month, sustained through the study period, and still present at 1 yr. No changes in outcomes or in the rates of selected complications were detected. Red blood cells utilization correlated linearly (r .47) with the number of blood tests performed in both study periods.
Conclusions: Guidelines designed to optimize laboratory tests use in an intensive care unit can produce rapid and long-lasting effects, can be safe, and may affect the number of red blood cell units transfused.