Questions raised over research

A group of researchers, statisticians and clinicians, including doctors and pharmacists, have called for an independent audit and review of The Lancet’s large hydroxychloroquine study

Last month, a large observational study published in The Lancet linked hydroxychloroquine and chloroquine use to increased rates of mortality and heart arrhythmias in hospitalised COVID-19 patients.

In response to the findings, the World Health Organization paused the hydroxychloroquine arm of its SOLIDARITY trial, an international hydroxychloroquine trial led by the University of Oxford in the UK was also paused, and the AustralaSian COVID-19 Trial (ASCOT) is reportedly under review.

The Australian Department of Health has now changed its advice to “strongly discourage” hydroxychloroquine use in treating COVID-19 in hospitalised patients unless the patient is enrolled in a clinical trial.

Meanwhile a group of clinicians, including doctors and pharmacists, as well as researchers, epidemiologists and statisticians, have shared methodological and data integrity concerns regarding the Lancet study.

There were initial concerns about the Australian numbers cited in the article, which the authors subsequently noted was an error in their data.

They admitted that one hospital had self-designated as belonging to the Australasia continental designation when it should have been assigned to the Asian continental designation. While the data was amended, no changes were made to the findings of the paper.

Meanwhile there are further concerns about the integrity of the data, including how it was sourced, with questions raised over the role of the company Surgisphere which provided the data for the paper.

Surgisphere is run by Dr Sapan Desai, one of the co-authors of the paper.

In an open letter addressed to Richard Horton, editor in chief of The Lancet, over one hundred signatories call for more information.

They say the error with the Australian numbers “indicates the need for further error checking throughout the database”, noting that the results of the study have had “a considerable impact on public health practice and research and therefore require further scrutiny”.

The signatories share concerns about:

  1. Inadequate adjustment for confounders.
  2. Authors not releasing their code or data.
  3. Lack of ethics review.
  4. No mention of the countries or hospitals that contributed to the data source and no acknowledgments to their contributions. A request to the authors for information on the contributing centres was denied.
  5. A classification error in the way some data were reported as Australia numbers. (This error has been corrected in the paper)
  6. “Unlikely” data from Africa.
  7. Unusually small reported variances in baseline variables, interventions and outcomes between continents.
  8. Mean daily doses of hydroxychloroquine that are 100 mg higher than FDA recommendations, whereas 66% of the data are from North American hospitals.
  9. Implausible ratios of chloroquine to hydroxychloroquine use in some continents.
  10. “Unlikely” 95% confidence intervals reported for hazard ratios.

In response to a request for data by the authors of the open letter, lead researcher behind the study Dr Mandeep Mehra, who is a Professor at Harvard Medical School, reportedly replied: “Our data sharing agreements with the various governments, countries and hospitals do not allow us to share data unfortunately.”

The letter calls for provision of details on data provenance—at the very minimum, sharing of aggregated patient data at the hospital level; independent validation of the analysis; and open access to all the data sharing agreements cited.

It also calls on The Lancet to make openly available the peer review comments that led to the manuscript being accepted for publication.

The list of signatories comprises clinicians and researchers from all over the world, including several Australians.

One of these, Professor Allen Cheng from Monash University, told AJP : “Letters to have been sent not only to The Lancet but also to NEJM about their prior paper”, referring to a study published in the New England Journal of Medicine in May which also used data provided by Surgisphere.

“The authors state that they are not able to release the data, or even the names of participating hospitals,” he said.

On Tuesday, The Lancet editors published an expression of concern over the study.

“Important scientific questions have been raised about data reported in the paper,” they said.

“Although an independent audit of the provenance and validity of the data has been commissioned by the authors not affiliated with Surgisphere and is ongoing, with results expected very shortly, we are issuing an Expression of Concern to alert readers to the fact that serious scientific questions have been brought to our attention.

“We will update this notice as soon as we have further information.”

This article was updated at 10.20am to include new comment from The Lancet.

See the full letter here.

Previous New advice on combinations
Next ‘Historians and epidemiologists will write whole books on what’s happened here.’

NOTICE: It can sometimes take awhile for comment submissions to go through, please be patient.

No Comment

Leave a reply