The Herald has a fascinating example of how statistics can be filtered through an agenda to say, more or less, what you want. The latest set of statistics from the Children’s Social Health Monitor have been interpreted by Prof Cindy Kiro and the current Children’s Commissioner, John Angus, to mean that the recession is having a deleterious effect on the health of children living in poverty.
I confess that, as soon as I saw Prof. Kiro’s name, I immediately suspected disinformation and re-interpretation. For years she has produced a remarkably constant message that all children’s problems will disappear if only we redistribute someone else’s money to them. She has done this regardless of the data that comes her way. She has done it with complete disregard to the fact that the only proven way of lifting children out of poverty is for their parents to find some work. It is therefore unsurprising that she sings the same message here, despite the data saying otherwise. It is a little alarming that her successor to the Children’s Commission appears to be equally unable to objectively interpret data.
It is good to see that Dr Elizabeth Craig, director of the New Zealand Child and Youth Epidemiology Service, is far more level-headed about the applicability of the data
However because of the anonymous nature of the data used, it is not possible to prove any direct causal links between rising unemployment, increases in the number of children reliant on benefit recipients, and increasing hospital admissions for socioeconomically sensitive medical conditions.
If you ignore Dr Craig’s warnings then, at first glance , some of the data would superficially appear to support Kiro’s thesis:
Conditions with a social gradient (sensitive to socioeconomic status), such as asthma, show a noticeable increase in hospital admissions in 2008 and 2009, supporting Kiro’s proposition. However, I’m certain all of my readers immediately noticed that the same graph also shows a sharp rise in hospital admissions in 2000 – 2002 which cannot, in any way, be explained by a recession. In fact, this coincides with the first few years of the last Labour government. Some of this rise will be from the meningitis epidemic which peaked in 2001, but the numbers involved in the epidemic were relatively small compared to, say, asthma. One is forced to the conclusion that there are likely to be bigger forces at work than a recession. One gets some inkling of this when looking at ethnic data.
The rise in admissions is mainly in Pacific Islanders and Maori. At this point, most would point out the PI and Maori are overly represented in poverty statistics, explaining the disparity. Unfortunately for this theory, the following graph clearly shows that the recent increase in admissions is not related to deprivation…
As one would expect, the higher the deprivation index, the higher the rate of admission. No-one disputes the well-established link between poverty and hospitalisation. But the graph for all indices is relatively flat, meaning that the rise in recent admissions is much more related to ethnicity than it is to poverty.
Let me be clear here. I do not, for a moment, think that Maori and Pacific Island parents are less willing to pay for their children to have medical care. What is happening here is that Maori and Pacific Islanders are accessing emergency departments more often. If you preferentially present to an emergency department, you stand a significantly greater chance of being admitted than you would if you had presented to a GP with the same problem. This is because the expectation of care from a hospital is significantly greater than the expectation of care from a GP.
To illustrate: A 2 year old presents to a GP with a harsh cough and wheezy chest. The GP diagnoses bronchiolitis on purely clinical ground and gives the child some Ventolin with good result. Child goes home 30 minutes later with some steroid syrup and an inhaler. Same child arrives at emergency department and is not seen for nearly two hours. The house surgeon then orders a chest X-ray and some blood tests and gives the child Ventolin. It takes a further 2 hours before the test results are back by which time the child is a bit wheezy again. The house surgeon does not want to send the child home because he is “unstable” and so admits the child.
Repeat this conservative approach over the gamut of child presentations and you will see an obvious trend towards admissions. Maori and Pacific Islanders access Emergency departments in disproportionately large numbers. This might be exacerbated in times of recession. Thus, the increase in admissions over the past two years is more likely to be an artifact due to increase Maori and Pacific Islander presentations.
In addition, there are other, more likely, causes of increased admissions in 2008 -2009 , the most obvious being Swine Flu in 2009. Extremely hot, dry summers are another.
Instead of knee-jerk reaction based on our prejudices and agendas, it would be helpful if we could be more objective in our analysis. It should be fairly obvious that simply throwing more money into welfare is unlikely to make the requisite difference in child health that we are looking for. It is possible that extending the free child subsidy to age 14 might improve things but I personally have my doubts about that. There does not seem to be particularly good evidence that children’s health is dependent on child health subsidies. The only objective data we have would suggest that we put all of our efforts into helping parents into employment, as this has been shown to have the most pronounced effect on overall child health. Figure 4 illustrates this well. It does not take much to go from Decile 10 to Decile 8 and yet this has significant effects on child admissions.
If you really want to do something “for the sake of the children” then get their parents employed.