It’s time for a model to report cancer news

Marc Beishon

Reporting science can be hard even for experienced journalists used to delving into the depths of papers published in journals, where findings are often expressed in statistical jargon and the science, especially in fields such as molecular biology, can be extremely complex.

Outside of the specialist outlets, much of the mainstream reporting on cancer tends to be about ‘breakthrough’ discoveries, most of which will never reach the clinic, and seemingly endless stories about ‘significant’ associations between foods and drinks (such as coffee) and cancer, either as a risk factor or a protective one.

The problem is that the actual impact or potential of the findings in the studies that led to the stories is often only given patchy or no analysis and balance, while other studies which are on genuine progress and which can have a big impact on patients, such as reducing the number of radiation doses for women after breast cancer surgery, are sidelined in the rush for the ‘cure or scare’ type items, especially in the popular press.

While many of the cancer stories about new drugs and the likes of the ‘coffee connection’ do not have a major public health implication, there are topics that do, such as the recent news items about the benefit of mammography screening, and items about sun and skin cancer and other proven risk factors such as smoking (and the rise of e-cigarettes).

Notably, earlier this year, a study by probably the world’s most eminent cancer researcher, Bert Vogelstein at John Hopkins, and Cristian Tomasetti, was widely reported and prompted fierce debate about the key finding that two-thirds of cancers are due to ‘bad luck’ owing to stem cell divisions that occur at varying rates in different tumours and happen independently of risk factors.

While the statistics, within the parameters of the study, are undoubtedly correct, there are limitations (it doesn’t include breast and prostate cancer) and is on US data, and there was an alarmed response from some, who were concerned that the ‘bad luck’ message would prompt people to lessen concern about proven risk factors such as smoking and sun exposure, and that overall nearly half of cancers can be prevented owing to different environmental and lifestyle factors around the world.

Writing a news item about studies in terms of local context, limitations, harms and costs is clearly very important but too many stories fail on many counts. A number of agencies, such as Cancer Research UK and the UK’s NHS do carry good, balanced interpretations of studies and announcements, and now one site,, has formalised this process by using a template that methodically assesses a news item and grades it with a mark out of ten. Further, the site is also assessing press releases put out by institutes and journals in the same way, as these releases are often picked up and used uncritically by news outlets.

While covering all of health, the site has so far included many on cancer as it tends to dominate news coverage. Take a paper in Nature on a discovery that a blood-based biomarker can pick up both late and early stage pancreatic cancer with ‘absolute accuracy’ – the reviewers found that several news stories in high-profile outlets failed to position the research in terms of harms, such as unnecessary surgery, and limitations, such as that finding early stage disease does not mean a major advance in life saving, and not least that it is a study that included only a small number of patients with early stage disease.

A screening test for pancreatic cancer would be a great boon but this is not near a clinical application – and achieving only say a half score in’s criteria does not convey the full implications of the study for readers if it does not grasp the quality of the evidence, commits ‘disease-mongering’ or omits harms and costs, among criteria used in the template.

Journalists are taught the ‘who what when where how’ model of reporting. A widely used process that works similarly through the merits of a health announcement would be a great addition to public understanding of science.


Leave a Comment

Your email address will not be published. Required fields are marked *

five + 10 =

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>