The website The Conversation recently published an article titled, "FactCheck: is domestic violence the leading preventable cause of death and illness for women aged 18 to 44?".
Overall the article is a decent explanation of the difference between burden of disease and risk factors. We hope it will serve to prevent false and misleading claims such as Tanya Plibersek’s statement that “Domestic violence claims more Australian women under 45 than any other health risk, including cancer.”
Unfortunately however the team at The Conversation appear to have gone out of their way to try and make it look like a statistic published on One in Three’s website was (at least partially) inaccurate or incomplete, going so far as to take the claim out of context. It is also possible they may not have followed their own Fact Check process.
The Conversation's article also unfortunately contained two errors of fact, one of which has thankfully been corrected.
Lastly, while the article’s finding that, “although intimate partner violence is not a leading cause of death, injury and illness among Australian women aged 18-44, it does appear to be a leading contributor” is technically correct, it is still incomplete and misleading. It is an advocacy statistic designed to amplify the risk intimate partner violence poses to the health and wellbeing of women, and is regularly misquoted by media and government alike (hence the need for the "Fact Check").
An equally correct and less misleading way of presenting the same data would be to say:
Intimate partner violence (physical and sexual violence plus emotional abuse) is not a leading cause of death, injury and illness among Australian women aged 18-44. As one of several known risk factors, it contributes 5.1% toward the burden of disease in Australian women aged 18-44. The vast majority of this burden of disease is anxiety and depression - death and injury contribute very little. Among all Australian women, intimate partner violence contributes an estimated 2.2% to the burden of disease. The leading cause of death for Australian women 15-44 years (2016) was intentional self-harm, with 368 deaths per annum, while there were 50 intimate partner homicides of women of all ages per year in 2012-14. The leading cause of injury for all Australian women was falls (51%), with assaults (of which intimate partner assault is a sub-category) making up 3.2%.
While well-intentioned, efforts to reduce family violence against women that use incomplete, incorrect or misleading statistics unfairly stigmatise men and boys as violent and abusive, while simultaneously denying or downplaying the existence of male victims of violence.
All victims of violence and abuse, whether male or female, deserve policies based upon up-to-date accurate data. Flawed data can only lead to flawed policies and actions, and many children continue to be exposed to violence because of these misinterpretations.
The One in Three Campaign is fully supportive of all genuine programs designed to protect women and children from violence. We are simply seeking similar protection for men and boys and asking that the vast majority who are not violent are no longer tarred with the brush of "violent males." You can read more at oneinthree.com.au/misinformation.
For purposes of transparency, the following is a complete email transcript of One in Three’s correspondence with The Conversation, followed by a video clip outlining The Conversation’s Fact Check methodology. Readers can decide for themselves whether The Conversation followed its own methodology when dealing with One in Three.
From: Sophie Heizer, Commissioning Editor, Education, The Conversation
To: Greg Andresen, Senior Research, One in Three Campaign
Date: 27/3/18 at 3:41pm
Hi Greg,
Sophie Heizer from The Conversation here.
I'm emailing you on behalf of our FactCheck editor, Lucinda Beaman.
We're in the process of putting together a FactFile on domestic violence. In this FactFile, we are exploring two statements about domestic violence, one of which comes from the One in Three website:
The latest available data shows that the top five causes of death, disability and illness combined for Australian women aged 15-44 years are anxiety and depression, migraine, type 2 diabetes, asthma and schizophrenia. Violence (let alone the subset of family violence) doesn't make the list.
As per our FactCheck protocol, I'm writing to ask, on the record and for quoting:
1. Could you please provide a link to a source, or sources, to support the statement that "the top five causes of death, disability and illness combined for Australian women aged 15-44 years are anxiety and depression, migraine, type 2 diabetes, asthma and schizophrenia. Violence (let alone the subset of family violence) doesn't make the list."
2. The website states that these figures are according to "the latest available data". Could you please provide a link to the data on which this statement is based?
3. Is there any other comment One in Three would like to include?
Many thanks for your time. If you would like to get in touch with Lucinda by phone, you can reach her on xxxx xxx xxx or by email at xxxx@theconversation.edu.au. Additionally, I can be reached on my mobile.
Thanks again and best regards,
Sophie Heizer
From: Greg Andresen, Senior Researcher, One in Three Campaign
To: Sophie Heizer, Commissioning Editor, Education, The Conversation
Date: 27/3/18 at 10:06pm
Hi Sophie,
Many thanks for your email - I’d be very happy to assist with your inquiry.
I have looked into the relevant data source and have discovered that since we posted the statement in question on our website, more recent data has been published by the AIHW. As we are a small group of volunteers, unfortunately we lack the time and resources to keep sources such as this up-to-date unless new data is brought to our attention, as you have just done (thank you!). We will revise our website very soon. The original statement in question comes from the AIHW Burden of Disease study 2003 (published in 2007). The most recent data comes from the AIHW Burden of Disease study 2011 (published in 2016).
Here is the information you requested:
1. Could you please provide a link to a source, or sources, to support the statement that "the top five causes of death, disability and illness combined for Australian women aged 15-44 years are anxiety and depression, migraine, type 2 diabetes, asthma and schizophrenia. Violence (let alone the subset of family violence) doesn't make the list.”
The revised statistic (soon to be updated on our website) is as follows:
The top five causes of death, disability and illness combined (i.e. Disability-Adjusted Life Year, or DALY) for Australian women aged 15-44 years are:
Anxiety disorders Depressive disorders Back pain and problems Asthma Other musculoskeletal conditions
Homicide and violence (of which family violence is a subset) ranks 26th on the list of causes.
The source data is as follows:
Australian Institute of Health and Welfare (2016). Australian Burden of Disease Study: impact and causes of illness and death in Australia 2011. https://www.aihw.gov.au/reports/burden-of-disease/abds-impact-and-causes-of-illness-death-2011/data.
2. The website states that these figures are according to "the latest available data". Could you please provide a link to the data on which this statement is based?
Here is a link to the Excel Spreadsheet containing the latest 2011 data from the AIHW. You will need to refer to Sheet 5: "S13.5: DALY by disease, sex and broad age group, Australia, 2011". The relevant data in in Column O ("F 15–44”).
https://www.aihw.gov.au/getmedia/27216d6d-cd27-46ff-8776-aaf44df97c1f/abod-2011-aus-2017.xls.aspx
I hope this helps! Please feel free to call or email if you require any further clarification.
Kind regards,
Greg
From: Sophie Heizer, Commissioning Editor, Education, The Conversation
To: Greg Andresen, Senior Researcher, One in Three Campaign
Date: 28/3/18 at 9:58am
Hi Greg,
Thanks very much for getting back to me so quickly and sending that information through, it's a real help.
I'll let you know if we need anything else, but that should be all good.
Thanks again and all the best,
Sophie Heizer
From: Greg Andresen, Senior Researcher, One in Three Campaign
To: Sophie Heizer, Commissioning Editor, Education, The Conversation
Date: 28/3/18 at 11:23pm
No problem at all Sophie!
May I ask what the second statement about domestic violence that you are exploring in your fact file is? We may wish to take up your offer of further comment, but would need a bit more context before doing so.
Warm regards,
Greg
From: From: Sophie Heizer, Commissioning Editor, Education, The Conversation
To: Greg Andresen, Senior Researcher, One in Three Campaign
Date: 29/3/18 at 2:50pm
Hi Greg,
Thanks for your question. Unfortunately, that's not how this process works, so I can't give you that information at this time.
We really appreciate your interest and your help so far.
All the best,
Sophie Heizer
From: Greg Andresen, Senior Researcher, One in Three Campaign
To: Sophie Heizer, Commissioning Editor, Education, The Conversation
CC: Lucinda Beaman, FactCheck Editor, The Conversation
CC: Charis Palmer, Deputy Editor/Chief of Staff, The Conversation
Date: 17/4/18 at 11:15am
Dear Sophie,
I see that you have published your Fact Check article - great stuff! It’s wonderful to have such a thorough explanation of the difference between burden of disease and risk factors to refer people to in the future.
I would like to make a couple of quick comments, if you don’t mind.
Firstly, I would like to voice One in Three’s concerns about the way you have taken our use of the statistic in question out of context in order to claim it is (at least partially) inaccurate or incomplete. You say,
The One in Three Campaign has correctly quoted the AIHW Australian Burden of Disease Study 2003. But it is not accurate to say the 2003 report is the “latest available data”, as the Australian Burden of Disease Study 2011 is the most recent data...
...If One in Three used the words “previous and current data shows”, and updated its reference, the claim would be correct.
You correctly link to the Misinformation page of our website as the source of the statistic in question. If you read this page carefully, you will see that it has been set up to log our attempts at correcting false and misleading information about family violence published in the media.
For example, the news article containing the misleading statistic used by ABC Radio National AM was last updated on 7/1/13. Therefore the ABC would not have had access to the latest 2011 AIHW data, as it wasn’t published until three years later in 2016. Likewise our other uses of the statistic in question on the Misinformation page (correcting publications from 25/11/2008 and 21/05/2010 respectively). Therefore, in context, our use of the statistic in question to correct the published errors was perfectly correct. To put it another way, our use of the phrase “the latest available data” was the language we used when we contacted the respective media outlets and government departments, prior to 2016, to request that they amend their articles and websites. It was never intended to be taken out of context, as you appear to have done.
Ironically, our initial attempt last month to update our website to reflect the latest available data (because of your inquiry) actually made our own website less accurate! We have now amended it again to be as accurate and transparent as possible, including both sets of data.
Secondly, would you consider amending a couple of words in your article for reasons of clarity and accuracy? Normally these changes would be pedantic and I wouldn’t make such a request of you. However because your article is a “Fact Check” and because it rests entirely upon the use of precise health terminology I hope you may humour my request to help ensure that your article is as clear and precise as possible.
You say,
If The Age editorial used the words “contributor to”, it would be correct, i.e. “leading preventable contributor to death and illness for women aged 18 to 44”.
A more accurate sentence might read,
If The Age editorial used the words “contributor to”, and the full phrase “death, disability and illness” it would be correct, i.e. “leading preventable contributor to death, disability and illness for women aged 18 to 44”.
As you know, burden of disease terminology uses the phrase “death, disability and illness” to refer to the DALY metric. Leaving out any of these three terms (e.g. referring to “death and illness” or “injury and illness” or just “death”) no longer refers to the burden of disease measurement, but to something else entirely.
You also used the following heading for your graph of the AIHW data:
Selected contributors to the burden of death, disability and illness for Australian women aged 15-44 years, 2011.
May I suggest the following re-wording?
Burden of death, disability and illness by selected disease for Australian women aged 15-44 years, 2011.
As the whole point of your article is to distinguish between causes and contributors, using the word contributor in a table of causes may serve only to confuse your readers!
Many thanks once again for the excellent in-depth article, and for contacting us for comment.
Warm regards,
Greg
From: Fron Jackson-Webb, Senior Health + Medicine Editor/Deputy Chief of Staff, The Conversation
To: Greg Andresen, Senior Researcher, One in Three Campaign
Date: 17/4/18 at 12:38pm
Hi Greg,
Thanks for this. We will be updating the chart to reflect that we are indeed talking about the causes of, rather than contributors to, death, disability and injury.
However I don't agree with the other suggestions.
Regards, Fron
From: Greg Andresen, Senior Researcher, One in Three Campaign
To: Fron Jackson-Webb, Senior Health + Medicine Editor/Deputy Chief of Staff, The Conversation
Date: 17/4/18 at 12:45pm
Dear Fron,
Thanks for your email. I’m very glad you have taken one of my suggestions on board.
Do you think I might expect a reply from either Sophie Heizer, Lucinda Beaman or Charis Palmer, or is your email the complete response from The Conversation?
Kind regards,
Greg
From: Fron Jackson-Webb, Senior Health + Medicine Editor/Deputy Chief of Staff, The Conversation
To: Greg Andresen, Senior Researcher, One in Three Campaign
Date: 17/4/18 at 12:46pm
Hi Greg,
I'm responding on behalf of the team.
Regards,
Fron
From: Greg Andresen, Senior Researcher, One in Three Campaign
To: Misha Ketchell, Editor, The Conversation
CC: Samara McPhedran, Author, The Conversation
CC: Terry Goldsworthy, Reviewer, The Conversation
Date: 17/4/18 at 1:20pm
Dear Misha,
I am writing to you to voice our concerns about yesterday’s article "FactCheck: is domestic violence the leading preventable cause of death and illness for women aged 18 to 44?”
Overall, we are very happy with the article. It is a thorough and long-overdue explanation of the difference between burden of disease and risk factors.
As you can see from the correspondence below, one of our concerns Is that the article takes our use of the statistic in question out of context in order to claim it is (at least partially) inaccurate or incomplete. Our other concern is that the correction to the statistic used by The Age also contains a significant factual error.
If this was a regular article I wouldn't bother voicing our concerns. However this is a “Fact Check” article that will be probably be accessed considerably as a reference over the coming months and years. It is important that it is factually correct.
I would appreciate hearing your thoughts.
Kind regards,
Greg
From: Samara McPhedran, Author, The Conversation
To: Greg Andresen, Senior Researcher, One in Three Campaign
Date: 17/4/18 at 1:45pm
Hi Greg,
Thanks for your email. I support The Conversation's decision on the matters you've raised.
Regards,
Samara
From: Greg Andresen, Senior Researcher, One in Three Campaign
To: Samara McPhedran, Author, The Conversation
Date: 17/4/18 at 2:39pm
Thanks for letting me know Samara.
Kind regards,
Greg
Follow-up emails were sent to Misha Ketchell and Terry Goldsworthy on 20/4/18 but neither responded.