Optimal D-Dimer Cutoff for Prediction of Ards in Covid-19 Infection

ABSTRACT

INTRODUCTION: Dynamic monitoring of D-dimer levels is a prognostic tool used for patients with COVID-19 and higher levels are associated with increased mortality. However, optimal D-dimer cutoff models for ARDS have not been previously evaluated. We aimed to determine the optimal D-dimer level for the prediction of ARDS in patients admitted with COVID-19. METHOD(S) We conducted a two-center retrospective study of 502 adult patients hospitalized between 2020 and 2021 with confirmed COVID-19 infection. The D-dimer on admission and peak value during hospitalization were obtained. Differences between groups were determined by one-way ANOVA, Wilcoxon rank-sum, or Fisher exact test. The cutoff D-dimer level and the C-index were obtained using the receiver operating curve (ROC). Univariate and multivariate regression models were used. Statistical analyses were performed using SPSS 28. RESULT(S) The mean age was 65+/-15 and 58.4% (294/502) were males. ARDS developed in 51.3% (258/502) of the patients, and 30.5% (153/502) died. Elevated D-Dimer (>0.5 ug/mL) was present on admission in 84% (423/502) of all patients. Patients who developed ARDS had higher peak (median 3.54 ug/mL, p< 0.001) and admission (median 1.5 ug/mL, p=0.005) D-dimer levels. D-dimer level of >3.05 ug/ ml predicted ARDS with a sensitivity of 71%, specificity of 67% and C-index 0.723 (p=0.001, CI 0.68-0.77). The C-index for D-dimer on admission was 0.59 (p=0.001, CI 0.54-64). Using the peak D-dimer level of 3.05 ug/mL, unadjusted logistic regression models showed a statistically significant effect on ARDS (p< 0.001, OR 4.62, CI 3.12-6.83). This effect persisted after adjusting for other variables (p< 0.001, OR 3.81, CI 2.51-5.79). CONCLUSION(S) Peak D-dimer level during hospitalization with a cutoff of 3.05 ug/mL is a useful tool to predict ARDS in patients admitted with COVID-19. Dynamic monitoring of D-dimer is an adequate and objective measurement during hospitalization for assessment of deterioration. Further studies with better-controlled monitoring on the timeframe of obtaining D-dimer are needed to further evaluate this threshold.