RESUMEN
OBJECTIVE: Several scoring systems have been proposed for EDs to identify patients at increased risk of mortality from sepsis. The modified Sequential Organ Failure Assessment (mSOFA) score, proposed in 2019, demonstrated a high negative predictive value. We aimed to validate mSOFA and compare its accuracy for predicting 30-day mortality to the simple bedside score, quick SOFA (qSOFA). METHODS: Over 1 month in 2018, consecutive patients with suspected sepsis were prospectively identified. A retrospective chart review was conducted to calculate both the mSOFA and qSOFA scores for these patients. The primary outcome was 30-day mortality. RESULTS: There were 252 patients with suspected sepsis identified over the study period. Thirty-day mortality was 13/39 (33.3%) for those with a positive mSOFA and 15/211 (7.1%) for those with a negative mSOFA score. Sensitivity was 46.4% (95% confidence interval [CI] 27.5-66.1%), specificity 88.3% (95% CI 83.3-92.2%), positive likelihood ratio 3.96 (95% CI 2.32-6.78), negative likelihood ratio 0.61 (95% CI 0.43-0.86). The area under the curve (AUC) was 0.74 (95% CI 0.64-0.85). qSOFA sensitivity was 39.3% (95% CI 21.5-59.4%), specificity 91.9% (95% CI 87.5-95.1%), positive likelihood ratio 4.85 (95% CI 2.56-9.18) and negative likelihood ratio 0.66 (95% CI 0.49-0.89). The AUC for qSOFA was 0.81 (95% CI 0.73-0.88). The difference in the AUC was -0.07 (95% CI -0.18 to 0.05), P = 0.25. CONCLUSIONS: In the present study, neither mSOFA nor qSOFA was adequately sensitive for predicting 30-day mortality, although both scores were highly specific and their overall accuracy was similar. The added complexity of the mSOFA without a significant increase in discriminative ability makes it unlikely to replace qSOFA in the ED setting.