The performance of screening tools for predicting mortality across multi-site international sepsis cohorts

BackgroundDirect comparisons of sepsis screening tools for prognostication have largely been limited to single-centre or high-income countries despite a disproportionately high burden of sepsis in low- and middle-income countries (LMICs). We evaluated the performance of commonly used sepsis screening tools across prospective sepsis cohorts in the United States, Cambodia, and Ghana. MethodsFrom 2014 to 2021, participants with 2 or more SIRS (Systemic Inflammatory Response Syndrome) criteria and suspected infection were enrolled in emergency departments and medical wards at hospitals in the Cambodia and Ghana and hospitalized participants with suspected infection were enrolled in the United States. Cox proportional hazards regression was performed, and Harrells C-statistic calculated to determine 28-day mortality prediction performance of the qSOFA score [&ge;]2, SIRS score [&ge;]3, NEWS [&ge;]5, MEWS [&ge;]5, or UVA score [&ge;]2, Screening tools were compared to baseline risk (age and sex) with the Wald test. ResultsThe cohorts included 567 participants (42.9% female) including 187 participants from Kumasi, Ghana, 200 participants from Takeo, Cambodia, and 180 participants from Durham, North Carolina in the United States. The pooled mortality was 16.4% at 28-days. The mortality prediction accuracy increased from baseline risk with the MEWS (C-statistic: 0.63, 95% CI: 0.58, 0.68; p=0.002), NEWS (C-statistic: 0.68; 95% confidence interval [CI]: 0.64, 0.73; p<0.001), qSOFA (C-statistic: 0.70, 95% CI: 0.64, 0.75; p<0.001), UVA score (C-statistic: 0.73, 95% CI: 0.69, 0.78; p<0.001), but not with SIRS (0.60; 95% CI: 0.54, 0.65; p=0.13). Within individual cohorts, only the UVA score in Ghana performed better than baseline risk (C-statistic: 0.77; 95% CI: 0.71, 0.83; p<0.001). ConclusionsAmong the cohorts, MEWS, NEWS, qSOFA, and UVA scores performed better than baseline risk, largely driven by accuracy improvements in Ghana, while SIRS scores did not improve prognostication accuracy. Prognostication scores should be validated within the target population prior to clinical use. Key questionsWhat is already known on this topic - While single-centre cohorts and retrospective analyses have been performed, the optimal sepsis screening tool for prognostication in low- and middle-income countries is unknown. What this study adds - The MEWS, NEWS, qSOFA scores, but not SIRS, were additive over baseline risk for prognostication in prospective hospitalized infection cohorts, but with variable additive performance within each cohort. How this study might affect research, practice or policy - Prognostication scores should be validated within the target population prior to clinical use.