Mike: Hello, I'm Mike Clarke, Podcast Editor for the Cochrane Library. In this podcast, review author Jan Verbakel talks with first author Liselore De Rop, a PhD candidate of the KU Leuven in Belgium who is involved in COVID-19 research, about their August 2024 review of the accuracy of routine laboratory tests to predict mortality and deterioration for patients with severe or critical COVID-19.

Jan: Hello Liselore.

Liselore: Hi Jan.

Jan: So, we’re sitting here today to discuss our review about the accuracy of routine laboratory tests to predict mortality and deterioration to severe or critical COVID-19 disease. First, could you tell us about routine laboratory tests, what are they and why do we do them?

Liselore: Yes of course. Routine laboratory tests are a set of commonly performed blood tests that provide information about a patient's health status. They can be used, for example, to identify disease or monitor health in a patient.

Jan: So why is it important to have a review about the accuracy of these types of tests to predict mortality and deterioration in people with SARS-CoV-2, the virus that causes COVID-19 disease?

Liselore: Many people with the virus have only mild disease but it’s crucial to identify those at high risk of developing severe COVID-19 or, among those who present to health with severe COVID-19, to identify those who might die because of the disease. This can help doctors to decide who needs to be admitted to hospital and we wanted to see if routine laboratory tests are sufficiently accurate to help with this. We wanted to know if the tests can predict mortality and deterioration, meaning worsening of the disease, in patients with confirmed SARS-CoV-2.

Jan: What types of evidence were we looking for and what did we find?

Liselore: Our outcomes of interest were, on the one hand, deaths in patients with SARS-CoV-2 and on the other hand deterioration from mild or moderate disease to severe or critical COVID-19. We had hoped to look at laboratory tests done within and outside hospital settings but found only studies that were done in hospitals. However, even then, we identified a total of 64 studies with data from more than 71,000 patients involving 53 different routine laboratory tests, which required a large amount of careful analysis given the sheer volume of the data.

Jan: And, what do those analyses tell us?

Liselore: We assessed the performance of the laboratory test by looking at 'sensitivity' and 'specificity’. The closer these are to 100%, the more reliable the test could be, and we found five tests with both sensitivity and specificity over 50%. Four of these, C-reactive protein, neutrophil-to-lymphocyte ratio, lymphocyte count, and lactate dehydrogenase, suggest an important inflammatory reaction during SARS-CoV-2 infection, which can reflect impaired immune response and tissue damage. The fifth test, d-dimer, reveals a state of increased blood clotting seen in COVID-19 patients.

Jan: What about safety. Might the tests give us a false sense of how well a patient might do?

Liselore: Yes, that’s important. When we want to safely rule out patients who will not die or deteriorate, we need a high sensitivity, for example above 90%. However, within our review, we did not find any tests that were reliable enough to single-handedly safely rule out a severe outcome.

Jan: Overall, what’s the take-home message about using routine laboratory tests to predict mortality and deterioration in COVID-19 patients?

Liselore: On the positive side, routine laboratory tests linked to inflammation and blood clotting in patients with COVID-19 disease, can be used to assess the risk that a patient will deteriorate towards an adverse outcome. However, none of the tests performed well enough to safely rule out progression to severe or deadly disease. Therefore, although using these tests might help assess the overall health status of a patient, predicting deterioration or mortality requires a more comprehensive assessment, including clinical signs and symptoms, radiological findings and patient's characteristics.

Jan: Thanks, Liselore. If people would like to read the review, how can they get hold of it?

Liselore: Thanks Jan. The review’s available online. If people go to Cochrane Library dot com and type 'routine laboratory tests to predict severe COVID‐19' in the search box, they will see a link to our review.