Developing an Early Sepsis Alert Program :

An automated, real-time electronic medical record query and caregiver notification system was developed and examined for its utility in improving sepsis care. We hypothesize that the algorithm will increase the rate and timeliness of sampling of blood lactate and blood cultures, performance of chest radiography, and provision of antibiotics. A before-and-after, prospective study with consecutive enrollment examined an algorithm that automatically identified adult patients accumulating 2 or more systemic inflammatory response syndrome (SIRS) criteria and 2 or more blood pressure measurements less than or equal to 90 mm Hg during their emergency department (ED) stay. In phase 1, the system collected information but did not alert caregivers. In phase 2, caregivers were notified by alphanumeric paging and a text entry into the electronic medical record of the patients' potential illness and were provided with specific recommendations. Patients (33,460) were screened during 6 months; 398 patients activated the system, including 184 (46%) appropriately identified as severely septic. The algorithm had a 54% positive predictive value and 99% negative predictive value in detecting severe infection with acute organ dysfunction. The median time for patients to accumulate SIRS and blood pressure criteria was 152 minutes (interquartile range [IQR] 71 to 284 minutes), underscoring the dynamic nature of diagnosing critical illness in the emergency setting and the need for detection algorithms to repeatedly assess patients during their evaluation. After implementation, 2 interventions were performed more frequently, chest radiograph before admission (odds ratio 3.2; 95% confidence interval 1.1 to 9.5) and collection of blood cultures (odds ratio 2.9; 95% confidence interval 1.1 to 7.7). Only blood culture testing was performed significantly faster in the presence of decision support (median time to culture before intervention 86 minutes, IQR 31, 296 minutes; median time to culture after intervention 81 minutes, IQR 37, 245 minutes; P=.032 by Cox proportional hazards modeling). The predominant shortcoming of the strategy was failing to detect severely septic cases before caregivers. An automated algorithm for detecting potential sepsis increased the frequency and timeliness of some ED interventions for severe sepsis. Future efforts need to identify patient features present earlier in ED evaluation than SIRS and hypotension. Copyright © 2010 American College of Emergency Physicians. Published by Mosby, Inc. All rights reserved.

To determine whether automated identification with physician notification of the systemic inflammatory response syndrome in medical intensive care unit patients expedites early administration of new antibiotics or improvement of other patient outcomes in patients with sepsis.

We evaluate the impact that implementing an in-hospital protocol for the early detection of sepsis risk has on mortality from severe sepsis/septic shock. This was a prospective cohort study conducted in 2 phases at 2 general hospitals in Brazil. In phase I, patients with severe sepsis/septic shock were identified and treated in accordance with the Surviving Sepsis Campaign guidelines. Over the subsequent 12 months (phase II), patients with severe sepsis/septic shock were identified by means of active surveillance for signs of sepsis risk (SSR). We compared the 2 cohorts in terms of demographic variables, the time required for the identification of at least 2 SSRs, compliance with sepsis bundles (6- and 24-hour), and mortality rates. We identified 217 patients with severe sepsis/septic shock (102 during phase I and 115 during phase II). There were significant differences between phases I and II in terms of the time required for the identification of at least 2 SSRs (34 ± 48 vs 11 ± 17 hours; P < .001) and in terms of in-hospital mortality (61.7% vs 38.2%; P < .001). The early detection of sepsis promoted early treatment, reducing in-hospital mortality from severe sepsis/septic shock. Copyright © 2011 Elsevier Inc. All rights reserved.