Prediction of unfavorable outcomes in community-acquired bacteremia by SIRS, SOFA and qSOFA scores.

BACKGROUND

Sepsis diagnostic and prognostic scoring systems have changed over time. It remains uncertain which scoring system is the best predictor of unfavorable outcomes. We aimed to evaluate prediction of community-acquired bacteremia (CAB) outcomes using on-admission systemic inflammatory response syndrome (SIRS), sequential organ failure assessment (SOFA) and quick sequential organ failure assessment (qSOFA).

METHODS

We present a retrospective observational cohort study of consecutive adult patients hospitalized with CAB over ten years. SIRS, qSOFA and SOFA scores calculated on admission were dichotomized as ≥2 or 0-1. Raw and adjusted incidence of a composite unfavorable outcome (death, septic shock, invasive mechanical ventilation, extra-corporeal membrane oxygenation, renal replacement therapy) over 35 days were compared.

RESULTS

Among 1930 patients, 1221 (63.3%) had SIRS, 196 (10.2%) had qSOFA, and 1117 (57.9%) had SOFA≥2. Respective raw and adjusted probabilities of the outcome were similar. Incidence for qSOFA≥2 was high (41.3%) and still considerable for qSOFA 0-1 (5.4%). SOFA≥2 indicated higher risk than SIRS≥2 (14.7% vs. 12.4%), while SOFA 0-1 indicated lower risk than SIRS 0-1 (1.2% vs. 3.1%). This relationship between SOFA and SIRS was also observed in patients with qSOFA 0-1.

CONCLUSIONS

qSOFA≥2 was associated with highest probability of unfavorable outcome, but dichotomized SOFA was more precise at high vs. low-risk distinction. Consecutive use of dichotomized qSOFA and SOFA on admission of adults with CAB enables fast and reliable identification of patients at high (qSOFA≥2, risk ~≥35%), moderate (qSOFA 0-1, SOFA≥2, risk ~10%), and low risk (qSOFA 0-1, SOFA 0-1, risk 1-2%) of subsequent unfavorable events.