Introduction:

The reference standard for the diagnosis of COVID-19 remains reverse transcriptase polymerase chain reaction (RT-PCR)-positive, while plain chest computed tomography (CT) has been more sensitive and specific in severity classification. In defining the severity of SARS-CoV-2, clinical, imaging and laboratory criteria have been specified, mentioning several markers, including lactate dehydrogenase (LDH). Historically, LHD has been of interest as a biomarker associated with lung damage, with a greater peak during the influenza A H1N1 pandemic, and taken as a marker in the definition of severity.

Objective:

To determine the diagnostic performance of serum LHD levels on admission to the ICU as a marker of severity in patients with SARS-CoV-2 infection.

Material and methods:

Observational, analytical, longitudinal and prospective study, carried out in two stages, the first (April 1, 2020 to September 30, 2020) consisted of a correlation study between serum LHD levels at Intensive Care Unit (ICU) admission and simple chest CT, a moderate correlation was obtained between serum LHD levels and severity score by pulmonary tomography, with an r = 0.5 and p < 0.0001; a severity-associated serum LHD cut-off point of 396 U/dL was obtained (p < 0.001, sensitivity 86.02%, specificity 75%). The second stage was extended until December 31, 2020, in which analysis is performed for diagnostic test performance, in inferential statistics correlation tests are performed, a multiple linear regression model is built, analysis was performed for diagnostic tests by building two-by-two contingency Tables, determining sensitivity, specificity, positive predictive value, negative predictive value, likelihood ratios and Bayesian analysis.

Results:

155 patients were included in the study to determine the diagnostic performance of serum LHD levels in patients with severe SARS-CoV-2 infection; by multiple linear regression analysis, it was determined that LHD levels on admission to the ICU is a predictor of severity in patients with SARS-CoV-2. A sensitivity of 0.92, specificity 0.41, with PPV 0.854, NPV 0.57 and an accuracy of 0.83, an LR + 1.57 and LR- 0.63 were obtained. In the post-test probability results, patients with DHL > 396 U/dL have a 86.3% probability of developing severe COVID-19. DHL levels have a direct correlation with ICU days of stay and days of invasive mechanical ventilation with statistical significance (p = 0.001, p = 0.033 respectively).

Conclusion:

Serum LHD levels have a high sensitivity to diagnose COVID-19 severity; and in the absence of simple chest CT, it can predict severe lung damage. It constitutes a test with high screening power in severe cases of COVID-19; with a probability of almost double (LR + 1.6) that of predicting a severe case.