Aspartate aminotransferase as predictor of severity in SARSCoV-2 infection: linear regression model

Background and aim: Some patients with SARSCov-2 infection develop severe disease (SARS);however, the factors associated with severity are not yet fully understood Some reports indicate that liver injury may be a poor prognostic factor AIM: To identify the biochemical factors related to the development of SARS with mechanical ventilation (MV) requirement in patients with SARSCov-2 and COVID-19 Methods Type of study: Observational Cohort study Procedure: Data from COVID-19 patients were collected at admission time to a tertiary care center Differential factors were identified between seriously ill SARS + MV patients versus stable patients without MV Transformation to the natural logarithm of significant variables was performed and multiple linear regression was applied, then a predictive model of severity called AAD (Age-AST-D dimer) was constructed Results: 166 patients were included, 114(68 7%) men, mean age 50 6 ± 13 3 years-old, 27(16 3%) developed SARS + MV In the comparative analysis between those with SARS + MV versus stable patients without MV we found significant raises of ALT (225 4 ± 341 2 vs 41 3 ± 41 1;P = 0 003), AST 325 3 ± 382 4 vs 52 8 ± 47 1;P = 0 001), LDH (764 6 ± 401 9 vs 461 0 ± 185 6;P = 0 001), D dimer (7765 ± 9109 vs 1871 ± 4146;P = 0 003), age (58 6 ± 12 7 vs 49 1 ± 12 8;P = 0-001) The results of the regression are shown in the Table, where model 3 was the one that best explained the development of SARS + MV;with these variables was constructed the model called AAD, where: [AAD = 3 896 + ln(age)x-0 218 + ln(AST)x-0 185 + ln(DD)x0 070], where a value ≤ 2 75 had sensitivity = 0 797 and 1-specificity = 0 391, AUROC = 0 74 (95%CI: 0 62-0 86;P < 0 0001), to predict the risk of developing SARS + MV (OR = 5 8, 95%CI: 2 2-15 4;P = 0 001) Conclusions: Elevation of AST (probable marker of liver damage) is an important predictor of progression to SARS, together with elevation of D-dimer and age early (at admission) and efficiently predict which patients will potentially require MV Conflicts of interest: The authors have no conflicts of interest to declare [Formula presented] [Formula presented]