P47: Development of a VAD Specific Modified Early Warning Score

Background: The modified Early Warning – Sepsis Recognition Score (MEWS-SRS) is a widely utilized tool to identify patients who may be demonstrating early signs of systemic inflammatory response syndrome or sepsis. The components calculating MEWS-SRS include: temperature (T), heart rate (HR), respiratory rate (RR), systolic blood pressure (SBP), mental status and white blood count (WBC). A MEWS-SRS score greater than or equal to six, greater than or equal to a five if mental status change is excluded in scoring and/or a three in any single category prompts notification to a physician for assessment and evaluation of an infection. MEWS-SRS has subsequently been adapted to identify early signs of clinical compromise across disease entities, but the efficacy of this system in the ventricular assist device (VAD) specific population is unknown. Methods: We conducted a single-center retrospective analysis of all VAD patients who required in-patient admission between 1/1/15 – 12/31/20. Patients that had a diagnosis of sepsis on admission, were admitted for VAD implantation, or were younger than 18 years old were excluded.The MEWS-SRS score was calculated for all patients at time of admission and every 4 hours for 3 days. A VAD specific score (VAD-MEWS) was then calculated at the same interval for all patients, substituting the SBP with a Doppler return-to-flow (RTF). The two scores were then compared. Results: 92 patients met inclusion criteria. Mean age was 56.9 years, 27.2% were female and 41.3% were Non-White. Mean length of stay was 8.7 days. In-patient mortality was 5.5%. 10.9% of patients were noted to have a MEWS-SRS > 3, which increased to 19.6% when using the VAD-MEWS (p < 0.05, Fig 1). The mean change in score overall was 1.04. Among patients who did not survive the admission, the mean change from SRS-MEWS to the VAD-MEWS score was higher in comparison to those who survived (1.8 vs. 1.01, p=0.08). Conclusions: A VAD specific MEWS scoring system reclassifies a significant proportion of patients into a higher category, with a trend towards better identification of those who met in-patient mortality. Further enhancement of such nursing tools specific for VAD patients may lead to better patient outcomes.