Predictive value of platelet indices as early markers of mortality in patients with Crimean–Congo haemorrhagic fever: a retrospective observational study

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Background To evaluate the predictive value of admission platelet-related indices for in-hospital mortality in Crimean–Congo haemorrhagic fever (CCHF). Methods This retrospective single-centre study included adults with laboratory-confirmed CCHF between May 2022 and May 2023. The primary outcome was in-hospital mortality. Admission laboratory parameters and platelet-related indices were analysed using receiver operating characteristics analysis, DeLong comparisons, 2000-replicate bootstrap internal validation and reduced Firth logistic regression adjusted for age. Results A total of 167 patients (median age 47 y [interquartile range 34–62], 61.7% male) were included; 14 (8.4%) died. Non-survivors were older. Platelet count, platelet count multiplied by lymphocyte count (PLT×LYM) and plateletcrit were lower, whereas mean platelet volume:lymphocyte ratio, platelet distribution width:lymphocyte ratio and mean platelet volume:platelet count ratio were higher in non-survivors. Median PLT×LYM was lower in non-survivors than survivors (15.35 vs 49.56; p<0.001). PLT×LYM had the highest discriminatory performance (area under the curve 0.882, cut-off 39.33, sensitivity 100%, specificity 66.7%, positive predictive value 21.5%, negative predictive value 100.0%). In the logistic regression model, each 10-unit increase in PLT×LYM was associated with an adjusted odds ratio of 0.50 (95% confidence interval 0.30 to 0.73). Conclusion Admission PLT×LYM may serve as practical early marker for mortality risk stratification in CCHF; external validation is required.