Concern trolling 101

5 min read September 1, 2016 at 11:54pm on Health, Rant and Science

I've thus far avoided commenting beyond a small amount of twitter snark about the woes of Catalyst - once regarded as the ABC's flagship science program, but today I really feel like I have to join the chorus of condemnation.

In October 2013, Catalyst broadcast a two-part episode, entitled "Heart of the matter", which attempted to cast doubt on the efficacy of HMG-CoA reductase inhibitors - commonly known as "statins".  Two members of this class are in the top-ten most frequently prescribed medications in Australia, and estimates are that around 2.6 million people take them. The episode presented a contrary view to conventional medical wisdom - that statins were not as effective as trials made out, that the potential for adverse effects were greater, and that the available data was unduly influenced by industry.  An ABC internal review found that the second of the two episodes breached the ABC's Editorial guidelines by failing to present the conventional medical opinion.  

So what?  Well, a review of PBS prescribing data estimated that as many as 60,000 people may have stopped taking statins as a result of the episode.

Not content with resting on her laurels, the journalist responsible for Heart of the Matter (Maryanne Demasi - who read a PhD in medical science prior to moving into journalism) doubled down with "Wi-Fried" this year - a story about the potential "dangers" of radio-frequency radiation generated by phones, mobile internet and Wi-Fi routers. Wi-Fried has been taken down from the Catalyst website, received a caning on Media Watch, and Dr Demasi has been stood down from presenting on the show pending the results of a review due later this month.

So yesterday, Dr Demasi tweeted the following tweet:  

Tweet by Dr Demasi

which links to this story from the UK Telegraph. This story incredibly includes the following quote:

...research showed that cancer drugs are killing up to 50 per cent of patients in some hospitals. For the first time researchers looked at the numbers of cancer patients who died within 30 days of starting chemotherapy, which indicates that the medication is the cause of death, rather than the cancer.

I. Can. Not. Even.

This article, written by someone with the rather aspirational title of "Science Editor" is one of the most egregious examples of cherry-picking health data for a clickbait headline I can remember seeing in a long time.

So first, in the immortal sketching of Randall Monroe:


xkcd: Correlation - CC-BY-NC-2.5 - Original Link


Saying that all deaths within 30 days of chemotherapy are due to chemotherapy is exactly like saying that everyone who dies within a month of drinking a beer is a death due to alcohol. It's clearly ridiculous and does not stand up to even fairly passing scrutiny;  certainly something that I would expect a newspaper science editor to be able to parse. The article goes on to say:

In Milton Keynes the death rate for lung cancer treatment was 50.9 per cent, although it was based on a very small number of patients.

Well, I've got some news for all involved in this sorry tale, and it comes from this article. Lung Cancer Kaplan-Meier

This is a Kaplan-Meier survival curve;  you'll note that the y-axis is "cumulative survival".  You'll also notice that the group of patients treated with surgery (who all have early disease) have a survival fraction at 30 days of about 55%.  For the patients treated with chemotherapy or radiotherapy (who tend to have later-stage disease) it's less than 20%, and you're probably not able to pass a laser pointer between the curve for chemotherapy and palliative care. What does this mean for lay people? It means that lung cancer is really, really shit, and that your chance (based on this series) of being dead in 30 days if you receive palliative care for your lung cancer is around 90%. If you receive chemotherapy, it's about 80%.  These figures seem a bit grim even to me.  This American site quotes an overall 12-month survival of less than 50%, and for patients with advanced disease, a 5-year survival of only 4%.

The long and the short of it is that cancer is terrible. Chemotherapy is not an enormous amount of fun, as you would expect for a medication that kills cells which are basically your own, but to say that "chemo kills people" risks people declining life-saving treatment based on misinformation.

We have seen this in people who advocate natural therapies as an alternative to conventional medicine. We have good quality evidence that it occurred as a result of "Heart of the Matter". Of course, we should monitor death rates following chemo (and all the other things we do to people in hospital) - it can provide important signals about patient safety or systems issues that may lead to improved patient care.  

As with all health data, though, if it is being publicly reported -  as is increasingly the trend to do,  which is unequivocally a good thing - we must make sure it is reported into the public in a sensible, responsible way. In my opinion, science journalists - and doubly so for journalists who are scientists -  absolutely have a moral responsibility to report issues like this in socially responsible way that will not frighten the public away from medical care.

Catalyst is now clearly a repeat offender, and despite being on gardening leave, Dr Demasi continues to dump and run clickbait headlines without context, without explanation and without taking any responsibility for the unintended consequences.  Her responses to the ABC inquiries and criticism show no evidence of penitence, or more concerning, even any self-awareness.

Saying that "researchers say..." is no more a defence for providing contextless links to a story like this than it is for anti-vaccine crankery, chemtrails,  or alien abductions - all of which have allegedly serious research being done on them.

Sadly, Maryanne Demasi has form, and seems disinclined to change her ways. The public deserve better.