Today we’ve heard a lot of noise about an “alarming” increase in bowel cancer in young people. But it’s been very difficult to make much sense of it all.
The ABC reported that:
Australian oncologists are warning that young people are developing bowel cancer at alarming rates. Figures collated by Bowel Cancer Australia over the past 20 years reveal the number of bowel cancer cases found in people aged between 20 to 34 has risen 64 per cent….
The Sydney Morning Herald reported that:
THE incidence of bowel cancer in young people has surged in the past decade, more than doubling in some age categories, but doctors have been unable to explain the increase. Analysis of data from the Australian Institute of Health and Welfare shows bowel cancer in men under 25 increased by more than 160 per cent in the five years to 1996 and the same period a decade later. In women of the same age, the increase was more than 75 per cent.….The same analysis, performed by Bowel Cancer Australia, shows an increase in bowel cancer rates of about 40 per cent for men and women between the ages of 25 and 34. Overall rates increased by about 12 per cent.
The Herald Sun reported that:
FATTY and processed food may be behind an alarming jump in the number of under 35s diagnosed with bowel cancer. The incidence of bowel cancer – and deaths – is rising more rapidly in younger people than in any other age group, Australian Institute of Health and Welfare data shows. Specialist surgeon Prof Graham Newstead said rates of the disease had increased 64 per cent in data collected from 2002-06 compared with data taken in 1992-96, the latest figures available. “This is more than five times the increase seen overall across all age groups,” the Bowel Cancer Australia spokesman said.
It is difficult for a casual reader to work out just how alarmed to be as a result of these alarming reports. No details were given about the absolute number of cases in young people, or the actual rates of diagnosis in young people – ie cases per 100,000 population, for example.
The Bowel Cancer Australia press release which seems to have generated all this coverage doesn’t help much. It doesn’t include such figures or provide any links to where they might be found.
You had to dig deep into the bowels of the release (sorry) to find a belated acknowledgement that bowel cancer – like most cancers – is most common in people over 50. There also was presumably the reason for the timing of the release: Bowel Cancer Awareness Week (5-11 June).
So my next stop was the Australian Institute of Health and Welfare, whose publications state that death from bowel cancer is “relatively rare” before age 50.
Its 2010 overview of cancer states:
The mortality rate of bowel cancer has decreased significantly in both males and females (Figure 3.4). The mortality rate for males decreased by 41% from 37 deaths per 100,000 males in 1982 to 22 deaths per 100,000 males in 2007. The mortality rate for females decreased by 47% from 27 deaths per 100,000 females in 1982 to 15 deaths per 100,000 females in 2007. The reasons for the continued decline in the death rate of bowel cancer are not clear, but may be due to earlier detection of precancerous polyps and improved treatment.
More details in the graphs below…
This still doesn’t answer my question about the absolute numbers and rates of bowel cancer in young people and how these have changed (Are we talking about five more cases in young adults every year? Or scores? Or hundreds?) But it does provide some broader context.
I sent Bowel Cancer Australia a tweet earlier today asking for clarification but have not yet heard back.
Of course, an increase in any cancer in young people is worrying. And the suggestion that this is an international trend sounds like an important story. But we can’t make sense of such concerns without better information about the extent of the increase.
If I hear more, I will update this post.
In the meantime, beware of headlines and experts who speak only in relative terms – whether they’re flogging products or sounding the alarm about a problem they want to highlight.
***
Update, 2 June
Meanwhile, Croakey readers have since helped provide some much needed context for this story…
Thanks to Mikaela (see comments below) who provided this data from AIHW:
Of 58,926 people who were diagnosed with bowel cancer between 1997-2001, 92 people were under 25 (0.16%) and 418 were aged 25-34 (0.71%).
Of 64,829 people who were diagnosed with bowel cancer between 2002-2006, 150 people were aged under 25 (0.23%) and 465 were aged 25-34 (0.72%).
***
Thanks to Perth GP Dr Brett Montgomery who provided this further analysis:
I too found it hard to interpret that press release and put all the figures in perspective. So, I followed Mikaela’s suggestion on your blog to look at the AIHW data.
I decided to make a graph from the data looking at the trend over time in the proportion of people with newly diagnosed bowel cancer who are under 35. Below is the data and the graph I came up with.
The graph seems to be at its lowest in approximately 1992-1997. The proportion of bowel cancers occurring in young people seems to have increased since then, but not to a level out of keeping with historical patterns. (I have included all the years for which data is available at ACIM – I haven’t cherry-picked a particular starting point.)
I don’t know if this graph demonstrates a true mid-90s dip in proportion of bowel cancers in younger people or if this appearance might all be due to chance. You would need someone more statistically savvy than myself to calculate that. However, I hope this graph might help put things in a useful perspective.
It is interesting that the authors of the press release chose the nadir of this graph as their baseline in their comparisons. If they had instead used the period 5 years earlier, 1987-1991, their press release may not have captured the same media attention.
***
PostScript from Croakey
For me, one of the main takes from this particular story is the value of an informed and engaged audience – and this is not only something of value for media organisations.
As previously noted at Croakey, Dr Nicholas Gruen, CEO of Lateral Economics, who chaired the Australian Government’s Web 2.0 taskforce, argues that enabling public participation has wide-ranging benefits.
AIWH has the absolute numbers in Australia in the ACIM books: http://www.aihw.gov.au/acim-books/
Of 58926 people who were diagnosed with bowel cancer between 1997-2001, 92 people were under 25 (0.16%) and 418 were aged 25-34 (0.71%).
Of 64829 people who were diagnosed with bowel cancer between 2002-2006, 150 people were aged under 25 (0.23%) and 465 were aged 25-34 (0.72%).
http://www.aihw.gov.au/acim-books/
Thanks Mikaela, much appreciated. Melissa
A data sheet containing the raw data (rates per 100,000) was provided to health reporters at metropolitan papers, AAP and any other news reporters who requested it.
The analysis of AIHW data was undertaken by two gastroenterologists, covered the three most recent datasets (five-year period each) with an average annual incidence and mortality calculated for each dataset. Age groups analysed were 20-24, 25-29 and 30-34.
Our release focussed on data from a 15 year period and the complete 20-34yr age bracket to get across the message that being young does not make you immune to bowel cancer. The release also stated Australians aged 50 and over are still the age group that need to be most vigilant as incidence and death from bowel cancer is still much higher in the over 50s than under 50s – a message communicated by our spokespeople too.
In my experience, clinicians aren’t wonderful at picking changes in disease incidence and prevalence. There are numerous examples of clinicians reporting alarming increases in disease incidence and prevalence, often in response to public health initiatives, without necessarily having the full story, statistically. For example, there have been reports of a massive rise in rickets attributed to skin cancer campaigns (increasing, but still relatively rare, and largely driven by an increase in dark-skinned immigrants) and eating disorders attributed to obesity campaigns (we know very little about the prevalence of eating disorders in the community, but estimates are in the order of 0.5-1.0% depending on diagnostic vigilance and criteria).
These observations shouldn’t be ignored – but they certainly require a bit more – often quite simple – investigation before we start panicking the population.
Melissa, thanks for your tenacity in parsing the original data on this. I always wonder about the veracity of these health ‘stories’ in the media. And Sleve above has highlighted an issue with the clinicians themselves. Croakey is THE ‘go to’ source for all things health!
To be fair to Bowel Cancer Australia, they did provide rates per 100,000 in the media information and some of us did report them in our stories, including the fact that in 2007, only about one per cent of bowel cancer cases in Australia were in the under 35s.
That said, I don’t think it’s wrong to highlight the fact that people under 35 can develop bowel cancer. I have interviewed people in their twenties with bowel cancer, and their families, who went undiagnosed for months because their doctors thought they were too young to have bowel cancer. If a story about young people with bowel cancer gets one person diagnosed sooner, or prompts someone to go to a doctor with their symptoms, I actually think it’s a good thing. The aim isn’t to create fear. It’s to be informative and get people thinking about bowel cancer symptoms.
These stories also need to be seen in context. We often write about bowel cancer in the over-50s and the need for a better screening program in Australia. I don’t think an occasional story about bowel cancer in young people negates that.
Thanks, Melissa, for valuing your “informed and engaged audience”. We value the fact that you are informed and engaged too, of course!
There is another “main take” from this story for me, and that is the value of open data. Without those AIHW data being openly available, we could never have looked at these allegedly “alarming” trends in context. We would have been left struggling to interpret a press release, and scratching our heads. So three cheers for open data!
Just for the record, I live in a fragile glass house. When critiquing media coverage of health stories, I’m always uncomfortably aware that I’ve made similar errors (whether of omission, fact or emphasis) over the years myself.
But there has been so much attention, in both the medical and journalism literature, to the problem of simply reporting relative risks (whether in promoting a new treatment or in highlighting a health problem) that these days I think most experts and journalists should be aware of the need to also include some absolute data, to give their audiences some context and perspective.
Mediadiagnosis, I quite agree that it is fine to highlight that younger people can get bowel cancer – so long as some sort of perspective or context is provided. I’m glad your report did this; but the ones that didn’t risk raising unnecessary alarm, and the potential for people to end up having unnecessary interventions which may cause harm. As the current “less is more” series in Archives of Internal Medicine shows, there is growing recognition within mainstream medicine of the potential harms of unnecessary tests and treatments. We know that media coverage of breast cancer in young women is associated with increases in medical presentations and testing of young women – where the potential harms/costs are likely to outweigh any potential benefits.
Brett, I absolutely agree about the importance of open data. In the future it would be good to see some collaborations between investigative journalism and open data crunchers. If you’ve any ideas that could be explored this way, don’t hesitate to get in touch. You might also like to check out a new initiative to enable community commissioning and funding of stories, YouCommNews: http://youcommnews.com/