Are foreign students at high risk of death?

I’d never seen any data on the deaths of international students while in Australia, so I was interested in this story in The Age this morning reporting 54 deaths in the year to November 2008 (though annoyed at the beat-up elements – claiming the information was ‘suppressed’ by the coroner, when there is no evidence of anything other than reluctance to publish possibly unreliable data).

Obviously 54 deaths is 54 too many, but so far as I can work out this a death rate below that of the general population. Though there are statistical problems in working out the base population for overseas students (because the number of overseas students who will be in Australia at some time during a year will give a too-high number, due to short courses, mid-year starts and finishes etc), my estimate is that this gives a death rate of about .02%.

For a local comparison, I looked at deaths of 20-somethings in Australia. That works out at around .04% of the base population, or around double the death rate of overseas students. On the other hand, perhaps the relevant comparison group is Australian students – if we assume that the local death rate is increased by including the kinds of risk-taking and underlying illness that is under-represented in the student population. (The death rate of Australian students is not ‘suppressed’, it is just not recorded.)

Indians appear to be over-represented among the deaths, so perhaps another comparison point is the death rate of young Indians of similar backgrounds in India. I would have thought that the risk of death from accidents or disease was much lower here.

4 July update: Coroner to improve statistics on international student deaths.

6 Responses to “Are foreign students at high risk of death?

  • 1
    Son of the Ratpack
    July 1st, 2009 15:10

    The Indian students might be dieing not in their capacity as students but in their capacity as taxi drivers.

  • 2
    Rajat Sood
    July 1st, 2009 16:34

    I suspect Indian students are more inclined to commit suicide than students of other nationalities, perhaps because of the pressure they are under to succeed and make a return on their parents’ investment, as well as the fact that suicide has historically been an acceptable way of dealing with loss or failure in India. Another point is that Indian students are basically all boys (who are usually more successful at killing themselves), whereas maybe other countries send some girls over.

  • 3
    melaleuca
    July 2nd, 2009 00:06

    Oh I love this. Son of the Ratpack shows himself to be a language pedant then writes:

    “The Indian students might be dieing not in their capacity as students but in their capacity as taxi drivers.”

    The Urban dictionary says:

    1. Dieing

    A common misspelling of dying that that only dumbasses use.

    http://www.urbandictionary.com/define.php?term=dieing&defid=3096307

    Sorry but I couldn’t resist. As you were.

  • 4
    John
    July 2nd, 2009 10:07

    Yikes, this is a bad use of statistics. First off, the type of death is important. If 90% of those deaths were murders, then you’d have to compare MURDER RATES for the student population against that of the general AUSTRALIAN population. You’d also need to do SUBGROUP ANALYSIS, stratifying by country of origin.

    Incidentally, the Age reported that Australian coroners noted there were NO deaths due to suicides, while the Age noted 3 deaths. Hence, any anomalous reporting is being done by the Government of Australia.

    A comparison of Indian students against the death rates / murder rates in India would not form a valid comparison since they are in different social climates and circumstances. Take this example: say a gang is targeting ethnic group X, but ethnic group X comes from a country where there is a high DEATH rate (not murder rate) for a number of reasons (social, poverty, etc). All the gang would have to do is keep the murder rate slightly lower than the death rate in country X to, using your line of reasoning, make things look “okay” when in fact they are being targeted at a higher rate than other groups.

    You might want to pick up your high school stats textbook and look up base comparisons, subgroup analysis, and controlling for confounders.

  • 5
    Son of the Ratpack
    July 2nd, 2009 11:04

    I am a grammar pedant, not a spelling pedant.

  • 6
    Andrew Norton
    July 2nd, 2009 12:02

    John – The data to do a proper analysis is not available. I just did a quick examination based on the numbers that are available, total populations and total deaths, noting that there were problems with these numbers. This is a sensible starting point in the absence of detailed data.

    There is no ‘anomalous reporting’. If you read The Age article carefully, the coroner did not say there were no suicides. They did not supply any information at all because there is no requirement to record the nationality or occupation of the deceased, and therefore they did not believe they could provide meaningful totals.

    Nor did the government say there were no suicides. They were relying on Department of Immigration statistics which did not require cause of death to be reported. Suicide was only on the list because one visa holder who died overseas was reported as a suicide. They clearly state that they don’t know how most students died. The Age has been in contact with other organisations who have explained how some of these students died.

    The India comparison depends what question is being asked. If the question is ‘will my kid be at increased danger if he/she comes to Australia?’ the India comparison is relevant. If the question is being asked by Australian authorities ‘are Indian students at particular risk in Australia compared to other students or other Australians’ then the comparison you suggest is appropriate.

    There does seem to be a need to pay closer attention to what is happening to international students, and possibly improved collection of deaths statistics is part of that. This will have to come via the coroners, as educational institutions do not have the capacity in most cases to give a conclusive cause of death.