Prognostic accuracy of severity grading score and severity scoring index for predicting severe outcomes in Crimean-Congo hemorrhagic fever: a systematic review and meta-analysis

Purpose Early risk stratification is critical in Crimean-Congo hemorrhagic fever. We compared the prognostic accuracy of the Severity Grading Score and the Severity Scoring Index for predicting severe outcomes and mortality in laboratory-confirmed Crimean-Congo hemorrhagic fever virus. Methods We conducted a systematic review and meta-analysis of prognostic accuracy studies according to PRISMA 2020 guidelines. MEDLINE, Embase, Scopus and Web of Science were searched from inception to November 7, 2025. Observational studies evaluating Severity Grading Score and/or Severity Scoring Index and reporting extractable prognostic accuracy data were included. Pooled sensitivity, specificity, likelihood ratios and diagnostic odds ratios (DOR) were estimated using bivariate random-effects models at prespecified cutoffs for mortality prediction (Severity Grading Score ≥ 9; Severity Scoring Index ≥10). Risk of bias was assessed using QUADAS-2. Results Fourteen studies met inclusion criteria. Ten studies evaluating Severity Grading Score (n = 1521) and four evaluating Severity Scoring Index (n = 478) were included in quantitative synthesis for mortality. Severity Grading Score showed very high specificity (0.99, 95% CI 0.95–1.00) but limited sensitivity (0.47, 95% CI 0.27–0.69), yielding a pooled DOR of 94.8. Severity Scoring Index demonstrated higher sensitivity (0.78, 95% CI 0.63–0.88) with high specificity (0.97, 95% CI 0.91–0.99) and a pooled DOR of 100.7. Using an observed mortality prevalence of ~9%, Severity Grading Score provided strong rule-in capacity, whereas Severity Scoring Index offered superior rule-out performance. Conclusions Both scores are clinically useful prognostic tools in Crimean-Congo hemorrhagic fever, but with distinct roles. Severity Grading Score is better suited for confirming high-risk status, while Severity Scoring Index performs better as an early triage tool to identify patients at risk of death. Prospective head-to-head validation with standardized outcomes is warranted to guide their incorporation into clinical triage protocols and management guidelines.