The preliminary study, published by its authors on Tuesday before being evaluated by the reading committee of a medical journal, covers 368 network patients in public hospitals for American veterans who either died or were discharged from the hospital before April 00.
The authors analyzed the medical records of these patients a posteriori, which they grouped into three groups to compare: those treated with hydroxychloroquine (HC) only; those who had the hydroxychloroquine-azithromycin cocktail (an antibiotic) promoted in particular by the French doctor Didier Raoult; and those who have never received hydroxychloroquine.
The proportion of patients who died was in the hydroxychloroquine group alone (28%) compared to the cocktail group ( highest.%) And in the group without HC (11%).
However, this conclusion can be misleading because the group of patients receiving only hydroxychloroquine was initially ill and more at risk than the other two groups: it contained more smokers and people with diabetes or a history of cardiovascular and lung diseases.
The authors statistically corrected this initial imbalance and found that “the increased risk of mortality persisted in the hydroxychloroquine group alone”.
The specificity of the patients treated must also lead to caution when generalizing to an entire population. The patients examined were all men, the majority black, a population most affected by the epidemic in the United States. The average age was advanced: over 65 years.
Hydroxychloroquine is one of the treatments used in multiple countries in emergencies with severe COVID – 19, but there is controversy about its effects.
Scientifically speaking, it is ideal to conduct a randomized clinical trial in which groups of comparable patients randomly perform different treatments.
Such large scale trials are underway, particularly the European Discovery trial, but their results are not yet known.
In the meantime, doctors are experimenting with the molecules and treatments. Researchers can look at the results after the fact, but without harmonized protocols on duration, doses, intervention times, and severity of the cases at the beginning, it is difficult to draw reliable conclusions.