(1574) Time-to-Antibiotics and Mortality in Suspected Sepsis Using NEWS2 vs Shock to Risk-Stratify Patients

Introduction: The Surviving Sepsis Campaign Guidelines suggest using the presence vs absence of shock to risk stratify patients with possible but unconfirmed sepsis and to inform the urgency of antibiotics. It is unknown whether physiologic scoring systems that include additional clinical parameters, such as the National Early Warning System 2 (NEWS2) score, may identify additional patients with possible sepsis who require immediate antibiotic therapy compared to using presence vs absence of shock alone.

Methods: We retrospectively identified all adults admitted with suspected infection (blood culture and IV antibiotic given within 6h of ED arrival) to 9 hospitals in Massachusetts, 2015-2022. We calculated each patient’s highest NEWS2 score within 1h of ED arrival using vital signs, supplemental oxygen use, and mental status assessments. We then compared hourly associations between time-to-first-antibiotic and hospital mortality when stratifying patients by presence vs absence of shock (persistent hypotension or lactate ≥4.0 mmol/L within 3 hours) versus stratifying patients into 4 NEWS2 score bins (0, 1-4, 5-6, ≥7). Associations were estimated using multivariable logistic regression models adjusted for patients’ demographics, comorbidities, vital signs, and lab tests.

Results: 104,003 patients met inclusion criteria (12.6% shock, 12.1% NEWS0, 52.0% NEWS1-4, 15.7% NEWS5-6, 20.3% NEWS≥7). On multivariable analysis, each hour until IV antibiotics was associated with increased mortality for patients with shock (OR 1.07, 95% CI 1.02-1.12) but not for patients without shock (OR 1.02, 95% CI 0.99-1.05). Similarly, each hour until antibiotics was associated with increased mortality in patients with NEWS2 scores 5-6 (OR 1.07, 95% CI 1.01-1.14) and ≥7 (OR 1.08, 95% CI 1.04-1.12) but not in those with lower NEWS2 scores. Each additional hour until antibiotics was associated with increased mortality amongst patients with NEWS2≥7 even in the absence of shock (OR 1.07, 95% CI 1.01-1.13); there was also a trend towards increased mortality for patients with shock but NEWS2 scores < 7 (OR 1.07, 95% CI 0.99-1.16).

Conclusions: NEWS2 scores may identify patients with suspected sepsis who do not have septic shock who also warrant immediate antibiotic therapy in addition to those with septic shock.