Rob Sutton is an incoming junior doctor in Wales and a former Parliamentary staffer. He is a recent graduate of the University of Oxford Medical School.

The official Covid-19 death figures produced by Public Health England (PHE) may have been inflated. The revelation that their daily count included all fatalities following a positive test, regardless of the time interval since the result or whether it was the most likely cause of death, risks undermining one of the key metrics steering the Government’s response.

Judgement has been swift, severe and bipartisan. The Government has been trapped in a pincer movement of criticism from both sides of the political spectrum. Some on the right argue inflated statistics have led to economic disruption and the restriction of individual freedoms. Leftists point to it as another indication of a disorganised government response. Publication of daily figures has been halted and Matt Hancock has ordered an urgent review.

The criticism might seem fair. Why should there be any difficulty in counting deaths from the Coronavirus? The political fault lines exposed by the pandemic have brought scepticism to every aspect of the scientific community’s role in policymaking. Every figure and calculation, from test accuracies to the danger of asymptomatic patients to the effectiveness of masks, has been questioned. We might hope the death toll is one thing which can be agreed upon.

Yet the calculation is not trivial. Difficulties arise from numerous technical, clinical and political considerations, and these considerations have shifted during the pandemic. Deaths can be counted based on the confirmation of a positive test result or the judgement of a medical professional in cases where testing has not been performed.

Different types of tests might have different interpretations. Polymerase chain reaction (PCR) is the gold standard and is more likely to indicate a current infection, but it requires expensive equipment and technical expertise. Antibody tests are relatively cheap and easy to perform, but a positive result might indicate infection weeks or months prior. Their accuracy also varies.

Where testing is not available or has not been performed, the clinical challenges of defining a Covid-19 death are more pronounced. It must be decided on a balance of probabilities whether a given patient was infected. Many who die from the Coronavirus have comorbidities, and it can be difficult to decide whether the virus was the main cause of death or a contributing factor.

Public health officials, policymakers and statisticians must consider these factors. Should they only record those who have tested positive or should they include cases where there was clinical suspicion? Should we include positive results from PCR, antibody tests, or both? Should recorded deaths be those where it was the main cause, a contributing factor, or (as in the case of PHE) in all cases where there is a positive test result on file?

Deepening the problem is a lack of international consensus on best practices for different countries. Different countries have different levels of healthcare infrastructure and testing capacity, a one-size-fits-all approach is unpractical and would risk distorting true figures. This should have been established early on in the pandemic.

The World Health Organisation (WHO) has been busy fighting its own battles. There has been little opportunity to develop an international consensus on how to count deaths while battling on multiple controversies. Its ability to provide global leadership has thus been compromised.

In the vacuum of international medical leadership, political considerations have been added to the technical and clinical difficulties of counting deaths. The focus of our government should be simple: to obtain the best available data and the knowledge to effectively use it to save lives while protecting the economy.

Instead, perverse incentives have made an already challenging technical exercise into a political minefield. No government wants to look bad when compared to international peers, and overcounting risks inflating the statistics and leading to embarrassment on the world stage. Having the most comprehensive testing program sounds less appealing when it also means counting the most deaths.

Our policymakers are therefore pulled in different directions. We now face a situation in which the limited resources of PHE are being further stretched to run a recount in the hope that we might deflate our disturbingly high death toll. One can imagine better ways to invest these resources.

Feigning shock and pointing fingers is profoundly unhelpful at this point. There are clear actions that should be taken promptly. Amended figures should be released and anonymised data provided for scrutiny. Methodology for calculation should be explained unambiguously and a consensus formed as to how to record data moving forward.

If a major discrepancy is found in the new figures, policies which were based on those numbers should be reviewed and changed accordingly. Instead of bouncing the blame between PHE, SAGE, the Department of Health and Social Care, and the Cabinet, we need to acknowledge the difficulties inherent to the problem and act to ensure our decisions are guided by the best available data.