select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

expert reaction to observational study looking at chloroquine and hydroxychloroquine in COVID-19 patients

A study, published in the Lancet, reports on chloroquine and hydroxychloroquine in COVID-19 patients.


Dr Stephen Griffin, Associate Professor in the School of Medicine, University of Leeds, said:

“This is potentially a landmark study for COVID19 therapy.  Whilst not a placebo controlled trial, an observational study on this scale undertaken with stringent and meticulous analysis is powerful indeed.

“This is the first study on this scale assessing the potential benefits of chloroquine, or hydroxychloroquine either alone, or in combination with a macrolide antibiotic (usually azithromycin) in patients suffering with COVID19.  These treatments have been used widely in the clinical setting both within and outside of clinical trial conditions.  However, large scale placebo controlled studies are yet to be completed.  Smaller studies reveal conflicting evidence of potential benefits, despite some laboratory findings supporting antiviral efficacy.

“The key finding of this study is that neither single, nor combination therapy with either chloroquine or hydroxychloroquine appear to provide any patient benefit in large numbers of SARS-COV2 infected patients.  Moreover, and worryingly, such therapy actually may be harmful as treatment was associated with cardiovascular problems relating to heart rhythm.  This seems worse for drug combinations, but is concerning under all conditions.

“The need for effective COVID19 therapies as the world awaits an effective vaccine is paramount, and several widely used drug regimens have been retargeted in this regard.  Nevertheless, despite large numbers of clinical studies, accurate data from phase three, placebo controlled trials is lacking for the majority of compounds currently under assessment.  The scale of this study enables sound preliminary judgements to be made regarding safety and efficacy, although it can’t 100% rule out confounding factors that may influence outcomes.

“Thus, whilst controlled trials will be required for confident affirmation, the indications are that these drugs certainly ought not to be used outside of a trial setting where patients can be monitored for complications.  One might also question whether the resources necessary for large studies might be better diverted to other causes, at the same time allowing the drugs themselves to be used for their original purpose.  Furthermore, it is clear that high profile endorsements of taking these drugs without clinical oversight is both misguided and irresponsible.”


Prof Stephen Evans, Professor of Pharmacoepidemiology, The London School of Hygiene and Tropical Medicine, said:

“This is an interesting study based on co-operation across six continents (mainly North America and Europe) in 671 hospitals.  It was originally based on a data collection scheme for surgery based in the US, but this collaboration recorded patients admitted to hospital with a positive test for SARS Cov-2.

“They included over 96,000 patients.  They excluded the small number of patients treated with remdesivir (0.3%) and rather more who started treatment with hydroxychloroquine or chloroquine more than 48 hours after diagnosis (0.9%) or when on mechanical ventilation (1.1%).  These exclusions are reasonable, since starting treatment late in the course of disease may not give an opportunity for the drug to work.  They could not apply similar exclusion criteria to the control group who did not receive hydroxychloroquine or chloroquine.  This could lead to a slight bias against finding possible benefit of hydroxychloroquine or chloroquine.

“They also classified those who had hydroxychloroquine or chloroquine into those who received a macrolide antibiotic, specifically either azithromycin or clarithromycin.  All these drugs have been suggested as being efficacious for the treatment of Covid-19 but the evidence suggesting that is very weak and no randomised trials have yet shown such benefit, though there are trials running which will give the answers before too long.

“This study, apart from the slight potential bias noted above, is one of the better ones, both in its size and its general conduct, to assess the potential benefit or harm of hydroxychloroquine or chloroquine combined or not with a macrolide antibiotic.  There are some questions that could be raised about the study; it is not clear whether the main results are based on propensity score matching or on covariate adjustment.  It is not clear whether adjustment at least for continent, if not hospital or country, was made.  The contribution of anti-virals to effects on mortality does not seem to have been reported though it does seem to have been taken into account.  They may or may not have benefits on mortality but do not seem to affect the risk of arrhythmia.

“Those (14,888) who were in one of the four treatment groups had a number of differences to those in the control group; the treatment groups had a tendency to be at higher risk than the control groups, though the authors say such differences were not statistically significant.  The authors performed analysis adjusted for the measured differences in risk factors.  They found effects of risk factors that have generally been found elsewhere, such as increasing age, being male, diabetes and other cardiovascular risks.  Surprisingly they found Asians did better than others as did those on ACE inhibitors (a blood pressure lowering type of drug).  It is possible that South Asians/Indian subcontinent background does worse than those from Eastern Asian countries (Chinese background etc) who were probably more frequent in this study than in those from UK studies.

“The main findings are definitely not good for any of these four drug treatment groups, as the authors say “Each of these drug regimens was associated with decreased in-hospital survival and an increased frequency of ventricular arrhythmias when used for treatment of COVID-19.”

“The arrythmias are not surprising since these are well-known adverse effects of each of the drugs and while they are not a frequent occurrence, to have a 2.5 to 5 times increase is a major concern.

“The increases in mortality are found to be about 33% to 45% after adjustment.

“This is an observational study and it is possible that inadequate adjustment was made for the risk factors for mortality (some possibly not measured) and the bias noted above could explain a little of the difference, but it can be stated with some confidence that it is unlikely that the randomised trials will find substantial benefit to these drugs, and the excess in heart rhythm problems is consistent with everything we know about them.  A definitive answer still awaits the results of the randomised trials, but it is clear that the drugs should not be given for treatment of Covid-19 other than in the context of a randomised trial.

“It might even be said that to go on giving them other than in a trial is unethical, given this evidence that is not yet contradicted by other available evidence.”


Prof Babak Javid, Principal Investigator, Tsinghua University School of Medicine, Beijing, and Consultant in Infectious Diseases at Cambridge University Hospitals, said:

“This study, published in The Lancet, performs a retrospective analysis of a large number (>18000) patients treated with chloroquine (CQ), hydroxychloroquine (HCQ) alone or in combination with azithromycin compared with not receiving any of the agents.  There was absolutely no benefit seen with either HCQ or CQ +/- azithromycin, indeed, use of these agents was associated with increased risk of death in hospitalized COVID-19 patients.  It was also independently associated with increased risk of abnormal heart rhythms.  This was a retrospective analysis, not a prospective randomised controlled trial.  Therefore one can’t formally state that HCQ or CQ are not associated with benefit, but it certainly casts a great deal of doubt whether these agents are effective in the setting in which these drugs are currently being used in COVID-19 patients: i.e. severely ill patients in hospital.

“It should be noted that no data were available with regards to when drug treatment was initiated with respect to symptom onset.  We know that anti-virals for COVID-19 are more likely to be effective given early in the disease process, since that is the time of maximum viral replication.  Since many of the patients were severely ill, it’s possible that treatment was started fairly late.  Also, as the authors themselves state, it does not evaluate the potential role of either agent in earlier disease before patients are admitted to hospital, or prevention of infection.  Studies examining potential benefits in these specific circumstances are underway.”


‘Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis’ by Mandeep R Mehra et al was published in the Lancet on Friday 22 May 2020.

DOI: 10.1016/S0140-6736(20)31180-6


All our previous output on this subject can be seen at this weblink:


Declared interests

Dr Stephen Griffin: “No conflict.”

Prof Stephen Evans: “No conflicts of interest.  I am funded (1 day/week) by LSHTM.  They get funding from various companies, including Astra Zeneca and GSK but I am not funded by them, I have no involvement in obtaining funding from them and I am not an investigator or any grants obtained from them.  I am the statistician to the “meta-Data Safety and Monitoring Board” for CEPI [].  I will probably be paid for my attendance at meetings and expenses for travel.”

Prof Babak Javid: “None.”

in this section

filter RoundUps by year

search by tag