• Juan González Del Castillo, Luis Escobar-Curbelo, Mikel Martínez-Ortíz de Zárate, Ferrán Llopis-Roca, Jorge García-Lamberechts, Álvaro Moreno-Cuervo, Cristina Fernández, Francisco Javier Martín-Sánchez, and Representing the Infectious Disease Group of Spanish Emergency Medicine Society.
• aEmergency Department bResearch Unit, Preventive Medicine Department cHealth Research Institute, Hospital Universitario Clínico San Carlos, Madrid dEmergency Department, Hospital Universitario de Basurto, Bilbao eEmergency Department, Hospital Universitario de Bellvitge, Barcelona, Spain.
• Eur J Emerg Med. 2017 Jun 1; 24 (3): 183-188.

ObjectiveTo determine the validity of the classic sepsis criteria or systemic inflammatory response syndrome (heart rate, respiratory rate, temperature, and leukocyte count) and the modified sepsis criteria (systemic inflammatory response syndrome criteria plus glycemia and altered mental status), and the validity of each of these variables individually to predict 30-day mortality, as well as develop a predictive model of 30-day mortality in elderly patients attended for infection in emergency departments (ED).MethodsA prospective cohort study including patients at least 75 years old attended in three Spanish university ED for infection during 2013 was carried out. Demographic variables and data on comorbidities, functional status, hemodynamic sepsis diagnosis variables, site of infection, and 30-day mortality were collected.ResultsA total of 293 patients were finally included, mean age 84.0 (SD 5.5) years, and 158 (53.9%) were men. Overall, 185 patients (64%) fulfilled the classic sepsis criteria and 224 patients (76.5%) fulfilled the modified sepsis criteria. The all-cause 30-day mortality was 13.0%. The area under the curve of the classic sepsis criteria was 0.585 [95% confidence interval (CI) 0.488-0.681; P=0.106], 0.594 for modified sepsis criteria (95% CI: 0.502-0.685; P=0.075), and 0.751 (95% CI: 0.660-0.841; P<0.001) for the GYM score (Glasgow <15; tachYpnea>20 bpm; Morbidity-Charlson index ≥3) to predict 30-day mortality, with statistically significant differences (P=0.004 and P<0.001, respectively). The GYM score showed good calibration after bootstrap correction, with an area under the curve of 0.710 (95% CI: 0.605-0.815).ConclusionThe GYM score showed better capacity than the classic and the modified sepsis criteria to predict 30-day mortality in elderly patients attended for infection in the ED.

16 keywords...

hide…