[Comparison of different strategies for short-term death prediction in the infected older patient]

The aim of this study was to determine the utility of a post hoc lactate added to SIRS and qSOFA score to predict 30-day mortality in older non-severely dependent patients attended for infection in the Emergency Department (ED).

We performed an analytical, observational, prospective cohort study including patients of 75 years of age or older, without severe functional dependence, attended for an infectious disease in 69 Spanish ED for 2-day three seasonal periods. Demographic, clinical and analytical data were collected. The primary outcome was 30-day mortality after the index event.The antimicrobial susceptibility data and extended-spectrum beta-lactamase (ESBL) production in isolates recovered from intra-abdominal (IAI) (n=1,429) and urinary tract (UTI) (n=937) infections during the 2016- 2017 SMART study in 10 Spanish hospitals were analysed.

We included 739 patients with a mean age of 84.9 (SD 6.0) years; 375 (50.7%) were women. Ninety-one (12.3%) died within 30 days. The AUC was 0.637 (IC 95% 0.587-0.688; p<0.001) for SIRS ≥ 2 and 0.698 (IC 95% 0.635-0.761; p<0.001) for qSOFA ≥ 2. Comparing receiver operating characteristic (ROC) there was a better accuracy of qSOFA vs SIRS (p=0.041). Both scales improve the prognosis accuracy with lactate inclusion. The AUC was 0.705 (IC95% 0.652-0.758; p<0.001) for SIRS plus lactate and 0.755 (IC95% 0.696-0.814; p<0.001) for qSOFA plus lactate, showing a trend to statistical significance for the second strategy (p=0.0727). Charlson index not added prognosis accuracy to SIRS (p=0.2269) or qSOFA (p=0.2573).

Lactate added to SIRS and qSOFA score improve the accuracy of SIRS and qSOFA to predict short-term mortality in older non-severely dependent patients attended for infection. There is not effect in adding Charlson index.