Projected death toll from drug-resistant superbugs ‘based on incomplete data and flawed assumptions’

This investigation appears together with “The Independent“ and in German with „Spiegel online“

An official estimate of deaths expected to be caused annually by antimicrobial resistance – cited by the United Nations and world leaders – is „unreliable“ and undermines the fight against superbugs, a group scientists has warned.

Antimicrobial resistance (AMR) arises when micro-organisms that cause an infection survive medicine that is intended to kill them or stop their growth.

A figure of „10 million“ deaths a year has been widely used to illustrate the dangers of not taking action on AMR, but leading scientists now say it is based on incomplete data and flawed assumptions.

The number first appeared in 2014, in the initial report of the Review on Antimicrobial Resistance – a body set up by former Prime Minister David Cameron and headed by the economist Lord Jim O’Neill.

The aim of the AMR Review was to analyse the potential impact to human health of growing resistance to antibiotics and antivirals.

It warned that without urgent action, by 2050 a further 10 million people around the world would lose their lives to drug-resistant infections – a figure that has been cited for the past two years by media organisations, politicians, the UN, the European Commission, the World Health Organisation, the G7, and many others. 

The number also has prompted headlines warning of an “antibiotic apocalypse“, in which superbugs would “kill more than cancer“ – and a world that’s on the cusp of a “post-antibiotic era“.

Now scientists, led by Marlieke de Kraker at the HUG (Geneva University Hospitals), say that while the dangers of the overuse of antibiotics are undisputed, the 2014 death toll estimate does not stand up to scientific scrutiny.

In a peer-reviewed article, published in the scientific journal PLOS Medicine, the researchers acknowledge that action is required to fight antibiotic resistance, but recommend that „estimates for such an important, ‘hot’ topic should undergo scrutiny by independent experts before being made publicly available“.

“We contend that unreliable global estimates like those provided in the AMR Review potentially undermine, rather than support, the fight against a post-antibiotic era“, the study’s authors conclude.

Speaking to correctiv.org, Ms de Kraker said her article had been well-received by the scientific community – many had long doubted the figure and are reportedly glad it has finally been disputed.

Petra Gastmeier, director of the Institute of Hygiene and Environmental Medicine at the Charité University Hospital in Berlin – not involved in the study – agreed with its conclusions, saying „the authors have shown the mistakes of the British study step-by-step“.

But neither scientist would provide their own estimates, saying there were too many uncertainties and not enough data.

Lord O’Neill, who was the first minister to resign from Theresa May’s new Government – two days after presenting the Review’s findings to a major summit of the UN in New York – said the AMR Review no longer existed in a formal capacity, but was keen to comment on the new findings.

He said: „Two specialist forecasting organisations put these numbers together for us about the possible future consequences of inaction, and this work has helped spur huge momentum on AMR in the UK and internationally in the last two years, which continues.

“One of my Review’s 10 specific calls was for improved surveillance to get better data to continue to assess this huge health and economic threat so in that sense we agree on the need for better data.“

Consulting giant KPMG was one of the two forecasting organisations mentioned by Lord O’Neill. A KPMG spokesperson said: „Our own analysis was based on assumptions and the best data available at the time. Since then more data has become available and we are looking to update our estimates next year.“

When David Cameron commissioned the review in 2014, he warned that antibiotic resistance could lead to medical „dark ages“. He said it was one of the gravest crises the world faced.

Lord O’Neill, a former senior Goldman Sachs banker with a worldwide reputation as a top economist, was asked to head the review – and in 2015 he issued a call for urgent action.

But according to the PLOS Medicine study, Lord O’Neill’s AMR Review contained three main errors.

The scientists led by Marlieke de Kraker found the Review’s global deaths estimate not credible because it was based on incomplete data from the European Antimicrobial Resistance Network (EARS-Net), which records instances of infections in 895 European hospitals.

The new study argues this data is not representative of infections across Europe as a whole. This is because a third of the hospitals that report to EARS-Net are large specialist hospitals, whose patients will naturally be carrying more infections than patients in, for example, a smaller hospital or a GP’s surgery.

The AMR Review had extrapolated infections data from the 895 European hospitals to calculate the global number of infections, which the new study calls „a crude approach“.

Fears over antibiotic usage leading to lethal infections were discussed by the G8 in 2013

Suppose that 2,000 patients die from a resistant strain from a bug such as MRSA, and 1,000 patients with a treatable, non-resistant strain also die. It would be reasonable to assume the deadliness of the resistant bug is twice as high. It was precisely this assumption the Review made – according to Ms de Kraker’s team who contest it.

Charité University Hospital’s Petra Gastmeier explained that the two groups of patients were different. She said the MRSA deaths will be, on average, older – and therefore at a bigger risk of serious illness and death than those who are infected with the non-resistant bug.

The older people are, the more antibiotics they are likely to have taken – increasing the likelihood of them carrying resistant bugs, she added.

While infections with resistant bugs can result in death, the risk is probably lower than what the Review had assumed.

Based on these assumptions, the Review used a figure of 700,000 deaths a year from antimicrobial resistance as its baseline. And then, „based on [these] already uncertain… estimates“, the PLOS Medicine study suggests that the Review’s final error was assumptions about increasing infection rates in the future.

According to the study, there is already evidence that fewer people are dying from infections because of better medical care. The 10 million figure was calculated based on the AMR Review’s assumption that resistant bugs will become significantly more common and twice as infectious, and that the proportion of people who die from these infections will not change – even though, as the study points out, improvements in public health systems, especially in middle-income countries, mean that even now the number of people who die from such infections is getting smaller.  

While the study acknowledges that the Review did call for better data and better surveillance of infections, it also says that the Review did not clearly report „how existing uncertainties in each of the applied steps could affect their estimates.“

The scientists said: „The Review estimated that antimicrobial resistance could cause 10 million deaths a year by 2050. This estimate has become a familiar refrain. It has been quoted repeatedly by lay media, experts, and public health agencies. Frequently, only this specific, frightening conclusion is reproduced from the report, unaccompanied by caveats or confidence intervals… We contend that unreliable global estimates like those provided in the AMR Review potentially undermine, rather than support, the fight against a post-antibiotic era.“

A Department of Health spokesperson said: „The figures in Lord O’Neill’s report are predictions based on a number of assumptions about current and future trends. They serve as a warning as to what might happen if we don’t act now – and there is certainly consensus from experts across the world that the threat from antimicrobial resistance is very real.“