Hydroxychloroquine study not all that it seems

An article on MailOnline and several Facebook posts have claimed that a new study shows that hydroxychloroquine “can increase COVID survival rates by 200%”.

This is not as promising as it sounds, because the study in question, which has not yet been peer-reviewed, is flawed.

The paper looked at 255 Covid-19 patients in a hospital in New Jersey, who were at some point given invasive mechanical ventilation, where a tube is inserted into the patient’s airway in order to get air into their lungs. 

Dr Penny Ward, Visiting Professor in Pharmaceutical Medicine at King’s College London explains, via the Science Media Centre, that 88% of the 255 people received at least some hydroxychloroquine, and a majority also were given azithromycin. From that, the researchers assessed how much both contributed to their survival according to how much of these drugs the patients were given overall.

Out of the 255, 37 patients were given a combination of the highest cumulative doses of hydroxychloroquine (that is, over the course of their whole treatment) and an antibiotic called azithromycin. Of this group, 18 survived, which is just under half.

Dr Ward says: “This is a flawed analysis...because the small number of survivors are likely to have received treatment for a longer period than individuals dying rapidly despite [invasive mechanical ventilation] and [it] thereby potentially favours survivors.

“This analysis is flawed as longer survival is likely to be associated with greater cumulative doses of any treatments given and in addition there is a major imbalance between the numbers of individuals in each group compared.”

In other words, patients may have been given more of these drugs simply because they lived longer, so the study doesn’t necessarily mean that the drugs prolonged their lives.

This study does not prove that hydroxychloroquine works

Dr Jeffrey Aronson, Consultant Physician and Clinical Pharmacologist at the University of Oxford told Full Fact via the Science Media Centre: “This was a small, non-randomized, retrospective, observational study, of a kind that would not be taken as good evidence, even if we did not already have strong evidence from large, well-designed, randomized studies that hydroxychloroquine is not only not beneficial in Covid-19, but may even be harmful, especially if combined with azithromycin.”

As Dr Aronson says, this paper is observational, meaning that the researchers didn’t intervene (i.e. purposely give one group a treatment and not the other).

The best evidence for proving that a treatment works comes from randomised double blind trials, where neither the patients nor the doctors know whether they are getting a potential medication, or are in the ‘control’ group without it. Patients in each group are also chosen so each group is similar as possible, to prevent other factors affecting their outcomes.

Last week, the Medicines and Healthcare products Regulatory Authority (MHRA) suspended recruitment to any clinical trials that were investigating whether hydroxychloroquine might treat or prevent Covid-19.

This followed results from a large-scale UK trial, which found “no beneficial effect of hydroxychloroquine in patients hospitalised with COVID-19”.

Where did the 200% come from?

Of the 37 patients who were given the most hydroxychloroquine and azithromycin, just under half survived. In comparison, 16.5% of the 218 other patients who didn’t receive as much of the drugs survived.

The paper reports that the relative difference in survival between these groups was 198%.

This doesn’t mean any Covid-19 patient is 200% more (or three times as) likely to survive given this combination of medication. It means the survival rate in the very small group given this treatment was quite a lot higher than in the rest of the patients in this study.