RESUMO
OBJECTIVE: The objective of this study was to assess the usefulness of a software tool integrated into the medical electronic history at the time of emergency triage. The aim was the early detection of patients with severe sepsis, and the potential impact of this software tool on reducing the mortality rate in patients treated. METHOD: The study consisted of two comparative samples. Patient selection was performed retrospectively into two groups using ICD-9 codes from the hospital and emergency department discharge reports. The codes were 038.9, 995.9 and 995.92 for sepsis, and 785.52 for severe sepsis and septic shock. The sample called «alarms¼ consisted of patients studied after implementing the sepsis alarm system in the Emergency Department computer system. There were two types of alarms, a serious one and an alert one depending on the on vital signs defined. The historical sample called «no alarms¼ consisted of patients seen in the Emergency Department during the year before the introduction of the alarm system. RESULTS: The compliance rate of the sepsis treatment package was higher in the «alarms¼ sample, compared to the sample without alarms, with blood cultures, 96.3% versus 80.9% (P<.001), antibiotic treatment in less than one hour, 62.9% vs. 39.3% (P<.001), determination of lactic acid, 91.4% vs. 77.9% (P<.001), and applying appropriate volume, 57.7% vs 54.3% (P=.052), respectively. The hospital mortality was reduced in absolute terms from 25% in the sample without alarms to 13.6% in the sample with alarms. Survival at 30 days was higher in the sample with alarms (Log Rank=.004). CONCLUSIONS: There were no studies that evaluated the effectiveness of an alarm system in our literature search. An electronic identification system for patients with sepsis allows acting earlier, better compliance with basic measures, and a reduction in hospital stay and mortality.