Added on June 27: Mahevas et al., from pre-print to publication to forced disclosure of conflicts of interest.
Physician reports on hundreds of successfully treated, high-risk patients are not anecdotal evidence, even if they are not published in peer reviewed papers. Treating such reports as scientific evidence is long overdue. When researchers publish on this topic, failure to properly cite these reports is bad scientific conduct. If issues or questions arise, researchers can contact the practicing physicians directly and ask for any additional information needed. Researchers may choose to dismiss these reports and explain their reasons for doing so, but they cannot simply ignore them. Stating that such reports do not exist is scientific fraud.
Top tier peer reviewed medical journals – The Lancet, NEJM, The BMJ, and JAMA Network – published and promoted low-quality or fraudulent anti-Hydroxychloroquine papers because of the political agenda of their editors. Other papers circulated as pre-prints.
May 22: Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis, Mehra et al., May 22, 2020, The Lancet.
This is an obvious fraud, using non-existent data from Surgisphere corporation, owned by the article’s co-author.
May 19: Lack of efficacy of hydroxychloroquine in covid-19 (sic!), Vinetz, Joseph M, May 19, 2020, BMJ.
This is a commissioned op-ed, using scientific terminology to conceal political agenda. One quote from it is enough:
“politicians with no expertise in science, medicine, or public health, supported by certain media, have picked up on the use of antimalarials as magic pills for the covid-19 pandemic. Rick Bright, former director of the US Department of Health and Human Services’ Biomedical Advanced Research and Development Authority, was moved from his post after resisting calls for widespread immediate dissemination of such drugs.”
Rick Bright was appointed to this position by Obama administration a few days after 2016 elections. He was removed because of incompetence and insubordination.
This op-ed is not peer reviewed, but the only place where it is mentioned is a sentence at the bottom of it. When cited, it looks like a peer reviewed paper.
May 14: Hydroxychloroquine in patients with mainly mild to moderate coronavirus disease 2019: open label, randomised controlled trial, Wei Tang et al., May 14, 2020, BMJ
Old news, new mistakes. HCQ should be given with AZM and/or Zn – not alone like in this study. Dosage (800-1,200 mg) and duration (14 days) were too high in this study. HCQ+ should be given early – not after 16 days like in this study (this is an old mistake). This study was conducted in China in February – early March, before good HCQ based treatment regimens were developed.
May 11: Association of Treatment With Hydroxychloroquine or Azithromycin With In-Hospital Mortality in Patients With COVID-19 in New York State, Rosenberg et al., May 11, 2020, JAMA Network. This one has 14 authors, including David Holtgrave. This alleged study was quoted by Andrew Cuomo. From the comments to the paper:
Comparing the Incomparable
Olga Goodman, MD | Rheumatology, Internal Medicine, Pediatrics
— Patients receiving hydroxychloroquine with or without azithromycin were overall sicker on presentation.
–The were significantly sicker given available information (table 1) and had multiple other risk factors …
But after receiving treatment much sicker patients had approximately the same mortality rates vs patients with a milder course of the disease and less risk factors. However, the authors conclude that “there are no significant benefits.”
Really? No benefits?
May 7: Observational Study of Hydroxychloroquine in Hospitalized Patients with Covid-19 (Geleris et al.), published in New England Journal of Medicine.
The paper “finds no evidence of benefit” of HCQ for COVID-19, AP claims. However, the data in this paper shows substantial benefits which have not been correctly reported. This observational, non-randomized study repeats the following errors found in similar previous studies.
Major disparities between the test & control groups:
- The patients who received HCQ were in much worse condition than those who did not.
- HCQ treatment was started in the late stages of the disease, when the patients were already in the emergency room.
- 22.5% of the non-HCQ group received Azythromycin.
Doctors chose to administer HCQ to the most severe patients:
- Practically ALL patients in the HCQ group had acute respiratory distress syndrome – ARDS – as defined by PaO2/FiO2 ratio of less than 300. Normal P/F is about 500. The average P/F ratio for the HCQ group was 223 compared with 360 for the non-HCQ. ARDS is the last stage of disease in most COVID-19 patients.
- 49% of the HCQ group had hypertension vs. less than 7% in the non-HCQ group
Statistical methods cannot eliminate such disparities. Even after torturing the data by Cox proportional-hazards regression models, propensity-score matching, multivariate logistic-regression model etc., a large difference of P/F ratios remained.
Further, the “study” combined death and intubation as the same bad outcome. But among persons with this combined outcome, there was a 30% lower percentage of deaths in the HCQ group than in the non-HCQ group. This study mentions that NY hospitals, where the treatment was conducted, stopped using HCQ in early April.
April 24: Effect of High vs Low Doses of Chloroquine Diphosphate as Adjunctive Therapy for Patients Hospitalized With Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection, Borba et al., April 24, 2020, JAMA Network
This is a good report of a bad clinical trial. Patients were given chloroquine (rather than hydroxychloroquine), and Azythromycin. Patients in the research group received high and known to be toxic doses of choloroquine. Many died. Hydroxychloroquine is much safer than chloroquine and is taken in lower doses. This study claims nothing about hydroxycholoroquine.
April 21: Outcomes of hydroxychloroquine usage in United States veterans hospitalized with Covid-19, Magagnoli et al., a pre-print on medRxiv.
It makes a retrospective statistical comparison of the outcome in COVID-19 patients, who received HCQ or HCQ+AZM treatment prior to April 11, in Veterans Affairs hospitals. In the Abstract, it claims that a larger percentage of HCQ treated patients died compared to untreated patients. This ignores the fact that HCQ or HCQ+AZM treatment was given only in the most desperate cases, frequently as compassionate care. Deep inside of the manuscript, it does acknowledge that initial conditions of the HCQ and HCQ+AZM groups was much worse than those of the untreated group, but then ignores it. The current version was published on April 23, hiding all the negative comments on the initial one. Casting even further doubt on the credibility of this study, one of the authors disclosed Gilead funding for another research. A month later, this piece has not been published in a peer reviewed publication.
There is commonality between these fraudulent studies: pre-print Magagnoli et al. and Geleris et al., published by the top tier New England Journal of Medicine on May 7. Both used data from patients hospitalized at the same time (March 9 to April 11 and March 7 to April 8, respectively)
Magagnoli is a retrospective analysis. Geleris et al. call their work an observational study, but it looks like a retrospective analysis. The authors have neither influenced nor observed the treatment, just gathered data after fact. Decisions to use or not to use HCQ were made by the treating physicians. The study had not been pre-registered (https://clinicaltrials.gov/ct2/manage-recs/how-register), although observational studies are frequently done without pre-registration is not frequently done for observational studies.
Magagnoli et al. had 7 authors. Geleris et al. has 12 authors – for something that looks like an exercise in statistical analysis, the type of work that could easily be done by one. Mahevas et al. has 30 authors! Such large numbers of authors in a low effort study frequently indicates that the authors know the weakness of their argument, and seek safety in their numbers.
Mahevas et al.
April 14: No evidence of clinical efficacy of hydroxychloroquine in patients hospitalized for COVID-19 infection with oxygen requirement: results of a study using routinely collected data to emulate a target trial, Mahevas et al., published as a pre-print on medRxiv.
This was a retrospective study on medical records, allowing the authors wide latitude to cherry pick data. It was incorrectly called “observational study.” It used terminology like primary and secondary outcomes, misleadingly suggesting that it was real time observation with pre-defined parameters.. The “primary endpoint” was composite: transfer to intensive care unit (ICU) within 7 days from inclusion and/or death from any cause. The authors made a “slight mistake” – included 8 patients that received HCQ into non-HCQ group. The patients were on oxygen, i.e., too late for viral effect of HCQ. Nevertheless, the paper reported lower death rate for the HCQ group compared with the control group – 2.8 vs. 4.1. Also, see Respone to Mahevas et al by P Brouqui, M Million & D Raoult.
The version of the “study”, published in The BMJ on May 14 under the title Clinical efficacy of hydroxychloroquine in patients with covid-19 pneumonia who require oxygen: observational comparative study using routine care data, was outright fraud. It claimed inaccurately that the decision whether to use or not to use HCQ was made before patients were admitted to hospital! The “primary endpoint” was changed from 7 days to 21 days. The eight initially mis-classified HCQ patients were removed from the study. Further, the Supplemental Data shows that >75% patients treated with HCQ were in one hospital (Mondor Hospital), while all but one patient not treated with HCQ were in other hospitals, with different care and population.
Five days later, The BMJ published an editorial Lack of efficacy of hydroxychloroquine in covid-19 based on this fraud. On June 18, the declaration of competing interests in Mahevas et al. was updated: “A reader queried the competing interests statement in the paper by Mahévas and colleagues … The whole author group has updated and expanded their competing interests statement on bmj.com, which now contains full details.” Originally, the authors claimed no competing interests. In the updated version, four authors reported personal fees from Gilead Sciences, and two other authors reported non-financial support from Gilead. It went unnoticed, and did not influence the conclusions and decisions made based on that article.
This post does not address all of them, but has been updated to include all the five references in the Surgisphere’s response of May 25.
So far, not a single clinical study put in doubt efficacy or safety of HCQ+AZM (with or w/out Zinc) treatment for COVID-19! But in the ocean of COVID-19 papers, the media has selected and extensively covered a few borderline fraudulent or misinterpreted studies, perceived negative toward HCQ-based treatment for COVID-19.
These attacks on Hydroxychloroquine are not science, but a Trump Derangement Syndrome by Proxy.
(Originally posted on May 8. Updated on May 11-12, 23-27; June 27)