ABSTRACT
Introduction D-dimer levels are elevated in COVID 19 and they correlate to the levels of other inflammatory markers such us ferritin, fibrinogen and C-reactive protein. It may be possible to correct D-dimer value in function of inflammatory markers, thus identifying patients at higher risk of venous thromboembolism (VTE). Our objectives are estimating a corrected value of plasma D-dimer as a linear function of ferritin, C-reactive protein and fibrinogen and stablishing a cut-off point of high probability of VTE. Patients and methods Age and sex matched case-control study of all patients diagnosed with COVID 19 and VTE between March and May 2020 in a tertiary hospital in Madrid (Spain). Using linear regression, the best predictive model will be estimated and residual D-dimer values will be obtained and analyzed using ROC curves to determine its discriminative performance. Results Thirty-eight cases and seventy-six controls were included. There was 63.2% of men and mean age was 68.2. D-dimer was best predicted by a linear model including fibrinogen, ferritin and C-reactive protein. Using residual values, the optimal cutoff point was 2165 ng/mL, with a sensitivity of 57.9% and specificity of 98.7%. Conclusion It is possible to estimate a D-dimer corrected value in function of ferritin, C-reactive protein and fibrinogen. Using the observed and estimated value we can obtain a residual value that performs well as a screening method to detect patients who would benefit for further VTE diagnostic testing.