Objective: To test the hypothesis that implementation of a practice guideline for blood gas measurement would decrease numbers and increase appropriateness of tests (according to criteria in the guideline) for up to 1 yr after introduction of the guideline.
Design: Numbers of tests and appropriateness of each test were measured retrospectively during each of five periods: two baseline periods 2 yrs and 1 yr before introduction of the guideline and three follow-up periods 2 to 3 months, 6 to 7 months, and 12 to 13 months after introduction of the guideline.
Setting: A ten-bed multidisciplinary intensive care unit (ICU) within a 500-bed tertiary teaching hospital.
Patients: A random sample of 30 patients admitted to the ICU during each of the periods specified above.
Interventions: The nominal group process was used to develop a new guideline and a multipronged educational approach was used to facilitate implementation of the guideline.
Measurements and main results: At 2 to 3 months, test numbers decreased from 4.9 +/- 1.6 to 3.1 +/- 1.8 (SD) tests/patient/day and to 2.4 +/- 1.2 tests/patient/day at 12 to 13 months. Appropriateness increased from a mean of 44% at baseline to 78% at 2 to 3 months and 79% at 12 to 13 months. There were no differences in Acute Physiology and Chronic Health Evaluation scores or ICU mortality among the patient groups and no differences in number of ventilator days or time to wean from ventilation. Cost-minimization analysis showed that the incremental cost-saving 1 yr after introduction of the guideline was $19.18 per patient per day.
Conclusions: Implementation of this guideline for arterial blood gas measurement increases efficiency of test utilization without prolonging mechanical ventilation or affecting outcome.