2000 OPEN FORUM Abstracts
MEASURING PROCESS VARIATION IN BLOOD GASES AND VENTILATOR MANAGEMENT IN AN NICU
Kay Lockhart RRT. John W. Salyer, RRT, MBA, FAARC. Karen K. Burton RN, RRT. Respiratory Care Service and Outcomes Research Service, Primary Children's Medical Center, Salt Lake City, UT
Introduction: We sought to measure process variation in the use of blood gases in ventilator management by RCP's in our NICU. We suspected that there might be important delays in responding to blood gas results. The measures described below were made possible by the introduction of ventilator care plans in our NICU. These documents are intended to be prepared for all ventilated patients and act as both physician's orders and a plan of care for ventilator management. They describe conditions under which the patient should be weaned. Design: Copies of ventilator care plans were gathered for all ventilated patients between Dec 1999 and Apr 2000 inclusive. All blood gas results for these patients were reviewed to determine those gases that should have resulted in a change in ventilator settings according to the patients ventilator care plan. The time between the posting of the blood gas results and subsequent ventilator changes was computed in minutes for each result. Standard descriptive statistics were computed. Results were further categorized into day and night shift findings. Results: There were 600 blood gases analyzed from 39 different patients. Mean (± 1 SD) time from blood gas results posting to subsequent ventilator change was 72 ± 85 minutes, with a median of 43 minutes, and a range of 1 to 845 minutes. Thirty five percent of subsequent ventilator changes occurred in less than 30 minutes, 26% occurred between 30-60 minutes, 12% between 60-90 minutes, and 27% greater than 90 minutes. We found no statistically significant differences in these data between days and nights (Mann-Whitney U test, P = 0.14). Discussion: There is clearly considerable variation in responding to blood gases that should have resulted in ventilator adjustments in our NICU. We hypothesized that 30 minutes was a reasonable goal for having responded to all such blood gases. Using this standard it is obvious that we have considerable unwarranted variation in this process. There are a number of process "hand-off's" in the use of blood gases in our NICU. Nurses initiate the process of obtaining blood gases. Capillary gases are drawn by phlebotomists and arterial line sampling is done by the nursing staff. It is not uncommon for RCP's to be unaware that blood gases results have been posted. This could contribute significantly to delays in ventilator management. However, it is also possible that RCP's are not responding as quickly as they could. We intend to use continuous quality improvement methods to further analyze this process and to reduce unwarranted variation. One possible intervention is to make RCP's responsible for the entire process of obtaining, and responding to blood gases.