Two weeks ago, a study in The Lancet , perhaps the most influential medical journal in the world, found no benefit from the use of hydroxychloroquine (HCQ) , a well-known antimalarial, to treat sick COVID-19 patients. Today, that study stands retracted . As it had relied on a huge dataset of about 96,000 patients sourced from 671 hospitals in six continents, the World Health Organization, citing a ‘do no harm’ principle, suspended drug trials pending a safety review. This led to some countries in Europe withdrawing the drug from their own trials. Another study involving some of the same authors and relying on the same data published in The New England Journal of Medicine , which sought to answer questions on the associations between cardiovascular disease, COVID-19 and drugs that target the enzymes that play a role in facilitating the virus in attacking a host, has also been retracted. The Lancet study triggered a backlash from scientists who found problems with the methodology and, more importantly, the dataset. It emerged that mortality attributed to the disease in Australia did not match with the country’s own estimates; there was no way to tally patient records and the hospitals they were sourced from; and there were problems with the statistics deployed and the conclusions about the potential risk from the drug.
The bigger concern was that the data was supplied by Surgisphere Corporation, which had a handful of employees with limited scientific expertise, and claimed to have aggregated its numbers by compiling electronic health records in less than two months. Experienced clinical trial specialists said this was a labour-intensive process. Moreover, when aspersions about the data started to swirl, the company, citing client confidentiality, said it was unable to share its data sources for independent assessment. In their retractions, the journals have blamed Surgisphere for being opaque with its primary data. So far, neither journal has introspected on the peer-review process that led to these studies being published in the first place. In hindsight it seems obvious that a disinterested analysis would have raised eyebrows regarding data sourcing, but the post-COVID world is a panic-driven one that has left no institution or appraisal process untouched. The average peer-review takes weeks and the clinical trial process months, but now the expectation is that science delivers its results like magic. For years now, questions have been raised on the effectiveness of the traditional, time-consuming peer-review process and this has launched a welcome culture of papers being uploaded as preprints for review. In the present instance of the HCQ imbroglio, it is the independent effort by external scientists that brought the blight to light. The key lesson is that it is a mistake to assume the scientific process as one divorced from the influence of power, privilege, finance and politics. The means and methods to a scientific result matter more than results — only achieved through global scrutiny. Openness, more than blame game, is what the post-COVID world needs now.
