How bad is bad – rating cancers

How bad is bad – rating cancers

Around 125,000 Australians will be diagnosed with cancer sometime this year. That’s just over 5 new cases per 1,000 Australians. Chances are, someone I know knows some one being diagnosed this year, and they’ll tell me about it. Cancer is never good news. But they’re also not all the same. While some are hard to beat, others have very decent survival rates.

So how bad is my mate’s mate’s cancer?

(This post focuses purely on the likelihood of death as a measure of ‘badness’.)


Some are deadlier than others

One method of comparing a condition’s “deadliness” is the mortality-to-incidence ratio (MIR). The MIR denotes the number of people who die of a particular cancer in a given year, to the number of people diagnosed with the same cancer in the same year. The ratio ranges from 0 to 1, and the lower the value the longer one is expected to survive.  A MIR of 0 means no one dies of that particular cancer.

The 10 most common cancers in 2012, in terms of incidence, accounted for 71% of new cases. These cancers, listed below, have MIRs ranging from 0.13 to 0.90. That’s to say, some common cancers are 8 times as deadly as other common cancers.

Cancers aint cancers 1


In short, lung and pancreatic cancers have a much worse outlook than prostate, breast or melanomas.


Visualising MIR’s results

The MIR has a huge impact on how many people die from a particular cancer, compared to how many are diagnosed with it. For example, even though prostate cancer impacts 8 times as many people as pancreatic cancer (20,637 vs 2,383), both claimed roughly the same number of lives in 2012 (3,173 vs 2,437). The graph below shows the incidence and mortality of Australia’s 21 most common cancers.

Cancers aint cancers 2


How bad is bad: not as bad as it used to be

Huge improvements in survival rates are being made across most cancers. Over the past 30 years, 8 of the top 10 cancers saw large drops in mortality ratios. The two most common cancers, prostate and breast, are now less than half as deadly as they were in the early 1980s. Unfortunately, progress has been less effective for bladder cancer, which has in fact gone backwards, by 39%.

Cancers aint cancers 3

* 1982 MIRs are age-adjusted based on the 2012 population, to make the figures more comparable.

** Care must be taken when comparing colon and rectal cancers  over time, as it is likely that the figures are disturbed by coding changes, thus may not reflect real changes in survival rates


Neither me nor my mate’s mate get a say on which cancer they have, but it does help to know that treatments and support are improving every year.





All data used sourced from the AIHW’s Australian Cancer Incidence and Mortality (ACIM) books.


Is a breast worth 15 lungs

Is a breast worth 15 lungs

Lung cancer is by far the biggest killing cancer in Australia. In 2014 it claimed the life of over 8,200 people. That’s almost as many as the next three cancers combined (prostate 3 102 + breast 2 844 + pancreas 2 547 = 8 493).


In popstats format, that’s one Australian death every hour.

Fortunately, much like pop, lung cancer’s mortality rate peaked in the early 80s, and has been declining steadily since.


Women catching up on the wrong race

This decrease, however, has been entirely gender lopsided.

While the anti-smoking initiatives have helped halve the mortality rate of men’s lung cancer since 1981, women’s has increased by 60% in the same period.



The increase in women dying of lung cancer has been so drastic that it has overtaken breast cancer as the biggest killer of women among all cancers. Back in the 1970s, breast cancer killed 4 times as many women as lung cancer.



Yet, lung cancer seems to be largely ignored (relatively speaking).

Research conducted by Cancer Australia, shows that even though lung cancer kills about 3 times as many as breast cancer, it receives less than a fifth of the research funding. Similar comparisons can be made with prostate and other cancers.  The graph below from their 2016 Cancer Research Review[1] provides a great representation of the inequality of research funding distribution currently in the field.



Lungs don’t sell

The communities’ disdain for lung cancer is also clear in the organizations we support. The Australian Charity and Not-for-profit Commission’s register includes 18 organisations mentioning “Breast cancer” by name, and another 15 mentioning “prostate cancer”.  Yet not one combined the words “lung” and “cancer” in their name[2].

This is not to say that there aren’t any organisations working in the area, but rather suggests that highlighting their cause is not considered a draw card.


Who’s to blame

Many suggest the community ignores lung cancer sufferers because a many of them are somewhat responsible for their condition. After all, smoking is linked with about 80%-90% of lung cancer sufferers[3]. But since when have we been so spiteful?

We help countless who have had a hand in their demise.

When the injured arrive at Emergency, triage forms don’t cover culpability.

We help those who drove too fast for the unexpected just as much as careful drivers who became their victims.

We help young men who go clubbing in Sydney, even if they threw the first punch.

People take all sorts of risks. Yet help is at hand when things don’t work out the way they hoped.

If James Dean does it


Not to mention that around 4 out of 5 sufferers took up smoking before the Vietnam War[4]; smoking warnings were not even a thing[5], and ads were the epitome of cool.


Not to mention the other 15%

That’s all without even thinking of the roughly 1,500 sufferers who never touched a smoke!


Heal the world

This, by the way, is a global phenomena. Lung cancer killed 1.6 million people worldwide in 2014 [7], yet similar under funding occurs across the major economies (or at least the ones I could find on a quick google search). So, any impact local research has in Australia could potentially help millions across the world.


So, why not?

Why not indeed.

In this age of cost-benefit analysis, we sometimes forget to put it into practice where it matters most. Lung cancer might be decreasing, but it sure isn’t going away. Smoking rates may have decreased, but they still haunt half as many as they did in the 80s[6]. At this rate, lunch cancer will still be the biggest killer for generations to come.

It’s time to stop victim blaming smokers, and put some money where are lungs are.




The author is a reformed smoker… the worst kind.


Feature pic by hey_paul:

Human Lung Embroidery Wall Decor




[2] Based on their 2014 data.


[4] Based on their age – over 65s in 2013.  And research showing 90% of smokers pick up the habit before the age of 20 (United States Department of Health and Human Services. Substance Abuse and Mental Health Services Administration. Center for Behavioral Health Statistics and Quality. National Survey on Drug Use and Health, 2014. ICPSR36361-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2016-03-22.





The death of dying

The death of dying

Life is getting longer.

Life expectancy in Australia grew from 50 years in the late 1800s, to 70 in the 1960s, to 82 or so today. And we’re nowhere near finished.  But while many are aware of this, not many appreciate the magnitude of this achievement, nor its continuous impacts.

Perhaps there’s a lack of newspaper headline moments claiming “1,000 people did not die yesterday”.  Or perhaps it’s because people think lives are merely being elongated when we’re at our worst, forcing us into a dependent cycle of medical attention from which the doctors gain more than the patients.

Yet, neither of these is entirely true.

Making headlines every day

While 154,000 Australians died in 2014 (latest data available), the figure would have been closer to 176,000 (22,000 more deaths) had the mortality rate not improved in the last 10 years. (And 10 years is hardly a long time. In today’s currency, it’s barely 5 Prime Minsters ago.)  Improvements since 2004 are saving an extra 2,200 people each year, on average.  To put it in perspective, around 230 people are murdered in Australia each year. So mortality improvements, which rarely rate a mention, are saving 10 times the number murdered more every year than the last!

Seeing as murders are front page material, mortality improvements should have their own weekend section.


Saving lives at all stages

Unlike popular opinion, extending life expectancy is not about delaying death while holding people to ransom, feeding them mashed cauliflower through transparent tubes.  Death rates for teenagers and kids under 5 year-olds have both decreased by a third.  Basically, we’ve saved 1 in 3 of the teenagers (and babies) who would have died had the improvements not occurred.

Had we saved a third of the price of electricity Josh Frydenberg would have declared it a national holiday.

All ages decreasing


These decreases were not a once off, nor were they specific to one group.  Improvements across the younger years have been pretty steady over the past decade.

constant decreast young


Improvements in mortality rates translate into less people dying than would have under the previous rates.

Deacreading death is people


As expected, the majority the decrease in deaths occurs at the latter stages of life.  Though surprisingly, the only cohort with an increasing mortality rate are those over 95 years of age.  And yet, the improvements are so large that even though the younger cohorts make up smaller percentages, they are still newsworthy. Without the improvements of the last 10 years, 2014 would have seen the death of around:

  • 376 more kids under 5
  • 124 more teenagers
  • 426 more people in their 20s
  • around 3000 more working aged folk (15-65 year-olds)

Instead, our attention is focused on gruesome anecdotes of villains and victims.  A hero cowardly king-hit, a young family destroyed by a murder-suicide, three teenagers overdosing on ice behind the local servo, or a joy ride gone awfully wrong.  Without balancing these out with the constant excitement of decreasing mortality and longer lives, we fall into a pessimistic spiral of despair.  We fail to recognise the huge progress made and foster a sense of nostalgia for times which were in fact significantly worse.

Shit’s never been this good. We’re just too busy reading the headlines to realise it.



Death figures sourced from : ABS Deaths Publication, 2014

Converted to rates using ERP by age

Death Tree

Death Tree

Not quite as cool as the Death Star, this Death Tree breaks down the 153,580 deaths which occurred in Australia in 2014, by cause:

Interpreting the Tree
Size of the box shows relative number of deaths, compared to all other deaths.
The colour represents the sex divide.
The bluer boxes represent diseases which kill more men than women. The yellow boxes kill more women than men. The legend on the top right provides a guide as to the sex representation.

Navigating the Death Tree
Left clicking drills down into finer level causes (e.g.: Cancer breaks down into Lung cancer, colon cancer, breast cancer, etc.).
Right clicking drills back up.

All causes are classified using the International Classification of Diseases (ICD).

The Death Tree provides a bigger, more user-friendly representation.

In the meantime, here is a smaller embedded version:

Everyday people, everyday deaths

Everyday people, everyday deaths

You may not have read it in the newspapers this morning, nor on social media, but 421 Australians died yesterday. And the day before. And most likely today too. Roughly speaking of course, averaging out the 153,580 who died over the course of 2014¹. (Coincidentally, 153,000 is roughly how many people die worldwide per day².)

The media may focus on the half a person murdered per day, or the 3.8 people tragically killed on the roads, but to use a common idiom, the majority of people who died in 2014 (55%) ‘had a good innings’. (A good innings, in my eyes, equates to living past 80.) But this has not long been the case. Until 2005 less than half of Australians who died each year reached 80. In fact, the percentage of people who died before reaching 80 was 44% in 2000, 34% in 1990, and 29% in 1980³.

The graph below shows the percentage of deaths by age group, since 1910.Age of death 1910-2013

The gains achieved over the past 115 years are huge. Much of the improvement, in particular in the first half of the century, was a result of decreased infant (under 1) and child (1 to 4 years) mortality. Infant deaths decreased from 81 deaths per 1,000 live births in 1910 to 25 per 1,000 in 1950, to just 3 in 2014 [4].

The improvements over the last 50 years, however, have also been driven by decreased rates of death in older Australians. This, in no small part, is a result of the improved ways in which we deal with heart disease. Heart disease is “the largest single cause of death in Australia” [5], but if current trends continue it won’t be for long. While heart disease (ICD-10 I20–I25) accounted for 30% of all deaths in 1970, by 2014 this had decreased to 13%.

Over the past 18 years heart attacks (which account for about half of all heart disease deaths) have dropped from 13% of all deaths to 6%. Dementia on the other hand, has increased from 1% to 5%. The changes have been so pronounced that dementia now kills more women than heart attacks do.

Heart attcks vs dementia

Since 1997, the number of heart disease deaths has reduced by an average of 505 people per year. That’s twice the average number of murder victims per year (which itself is decreasing).

Society is not getting more violent, or more dangerous. But improvements in the way we treat health conditions have dramatically improved and extended our lives.

If the amount of attention on scientific and medical improvements, as well as sensationalised but unlikely scenarios, more accurately reflected reality, perhaps we would have a very different perception of our state of affairs.

Shit’s getting better. Way better.


To better understand what people die of these days, the Death Tree Map linked here shows all Deaths in Australia in 2014 by cause. Clicking each category drills into finer categories in greater detail (i.e. Cancer breaks down into Lung cancer, colon cancer, breast cancer, etc.). Right clicking drills up.
The colour represents the sex divide. The bluer boxes represent diseases which kill men more than women. The yellow boxes kill more women than men. The legend on the top right guides the sex representation.
(below is a smaller representation of the Death Tree Map)




The roads know no date

The roads know no date

By the time the Christmas Holiday Road Toll counter begins, the worst is all but over. For many, unfortunately, the reminders of the roads’ dangers come too late.

Every year the Australian media and Police and Roads Departments produce an hour by hour, day by day counter of the road fatalities occurring during the festive season. News bulletins focus on yet another untimely death, and end of year best wishes are accompanied by a “stay safe on the roads” message.

It has almost become a new Hallmark theme, alongside “It’s a boy”, “Get Well soon”, and “I might not want to personally join such an outdated and discriminative institution, but I respect your right to do so, even if many aren’t yet afforded that right”¹.

Whilst road fatalities are a serious issue deserving of media attention, the Christmas period is not news worthy. In fact, the focus might incorrectly imply that the rest of the year is less dangerous, and therefore less deserving of our vigilance.

Road Toll Graph

Based on figures from 1989-2014 from the Bureau of Infrastructure, Transport and Regional Economics’ (BITRE), the Xmas Toll period (which this year runs from the 23rd December to the 3rd January) has a lower road fatality average than the rest of the year².

The graph above compares each day in December through to mid-January to the average death rate for the entire year. These figures are based on data in the 1989-2014 period, for which the average is 4.7 road fatalities per day.  The red areas denote a day where there are more deaths on the road than average, and grey areas denote smaller number of casualties.

Other than a small spike on Boxing and New Year’s Day, the rest of the Xmas Road Toll Period is remarkably below average.

Unlike the Festive period, the weeks just prior to it do seem more dangerous than average. On average, 110 people will die on the roads in the first 3 weeks of December every year. This is more than in any other 3 week continuous period during the year.

Newspapers are not entirely unaware of the numbers. In fact they usually publish an article or two every year showing how misleading these tolls are (example ³). Yet they continue to misguide the community and create panic by perpetuating the myth.

The question is how we focus our educational and advertising campaigns to have the greatest impact.  I doubt Christmas is the answer. It rarely is.


[1] This comment is source-less, and potentially misleading.



Time for a road trip!

Time for a road trip!

Today is the fifth day of the 2011-12 Christmas holiday road toll period¹ and it is hard to open a newspaper or news website without being informed that so far 18 people have lost their lives on Australian roads since last Friday. According to the World Health Organisation (WHO), road traffic accidents amount to over 1.3 million deaths per year, and injure in excess of 50 million people. They are the biggest killer of people aged 15 to 39 worldwide, and ranked around 9th overall². Worst of all, these figures are expected to increase over the coming years, mostly due to the growing access to vehicles in the developing world*. These facts make road safety a worthwhile focus of media attention. What makes less sense is the misleading obsession which the media has on the holiday periods.

By focusing so overtly on the Easter³ and Christmas periods, it is assumed by the public that they claim more lives than other times. Headlines speak of the climbing toll, morbidly focusing on every death recorded during the period. They even include a state by (vs.) state tally, and each police jurisdiction either congratulates their constituents, or shows their disappointment, depending on how their stats is faring. What they rarely report, however, is that the holiday periods are nothing special. If anything, the Easter and Christmas periods claim less lives than similar periods throughout the year.

Based on data published by the Department of Infrastructure and Transport 4 (Commonwealth level), the Christmas/New Year holiday period claims an average of 67 lives per year (since 1989, which is the earliest freely available data). The average for a similar period throughout the year is 70. Similarly, the Easter period claims an average of 22 lives per year, while 24 die on similarly lengthed periods the rest of the year. Christmas day itself falls in the lowest 5% of days in terms of deaths, in the period 1989-2010.

The month of December does appear to be slightly more deadly than the rest of the year, but surprisingly it is the period before Christmas which makes it so. By the time we first hear about the Christmas road toll, the worst has passed. Similarly, March is worse than April, meaning the Easter reminder comes too late.

The calendar below shows the average number of road fatalities for each day of the year, for the period 1989-2010. The cells are colour coded, with green having the smallest number of deaths and red being the deadliest.

Unfortunately, the road toll is devastating every week, not just when the media decides to focus on it. During the last two decades, an average of 34 people died on Australian roads every week. Most of the accidents which claimed lives occurred during the weekends. Saturdays claim 63% more lives than Mondays and Tuesdays. Fridays and Sundays aren’t far behind. If the media (and law enforcement agencies) want to focus their attention on a particular period (double demerit points, increased breathalysers, etc. etc.) then perhaps every weekend is as good as the others. This would send the message that the roads are dangerous throughout the year, not just on holidays.

NB: As the numbers represent the average from 1989 to 2010, the day of the week they land on changes from year to year. For presentation purposes, this calendar is based on the 2010 layout, with the 29th February added for completeness.

Dataset available upon request.

* Conversely, one could argue that since the trend is due to expanding access to motorised vehicles, a growing road tally in the developing world is a horrible side effect of great progress, bringing with it increased freedom of movement, greater access to trading routes, and generally higher quality of life.

1. Christmas/New Year holiday period begins at 00:01 on the last Friday before December 25, and ends at 23:59 on the first Friday after December 31.
3. Easter holiday period begins at 00:01 on Easter Thursday and ends at 23:59 on Easter Monday.