Health Promotion Hui: Cancer Society of New Zealand, 23 October, 2014
I have been asked to talk about how economic evaluation is used to determine the social costs of diseases and other health conditions. I could give a learned and technical lecture on the topic, but it strikes me that even were it a brilliant presentation most of you would lose the plot – it is not easy; even economists get confused. Instead I am going to give an example of economic evaluation to illustrate some of the issues, how it might be applied and how it can be helpful to other health professionals.
The example is the social costs of Fetal Alcohol Spectrum Disorders (FASD). I am aware that while FASD conditions appear to be associated with some cancers, the main damage occurs elsewhere. Even so, I hope you will value both its clarity and its conclusions.
I was the economist on a Canadian team, led by Lana Popova, concerned with FASD. I have no expertise on the medical condition. The rest of the team gave me key parameters which I used to do the evaluation. However, despite my lack of expertise, I have to say something about the nature of FASD. Here is my understanding of the disorder; remember I am not a clinician – although I have checked with clinicians about what I have to say.
Fetal alcohol spectrum disorders are a continuum of various permanent birth defects caused by the mother’s consumption of alcohol during pregnancy. (It includes its severest form, fetal alcohol syndrome (FAS) which has physical birth defects as well as neurological deficits.)
Diagnosing FASD involves medical and psychological evaluation so its clinical incidence is unknown – there is no other way to test for FASD so its prevalence in the general population is uncertain. We used the US estimate of 1 percent of children in the US having FASD, although I have seen much higher estimates. About one in ten of those with FASD are FAS. The uncertainty may not much affect the figures I am about to present; if it does, these estimates are on the conservative side.
As I have said, FASD is the consequence of alcohol in the mother’s bloodstream impacting on the development of the fetus. Even moderate amounts are able to cause significant alteration. There appears to be no safe amount of alcohol or safe time to drink alcohol during pregnancy.
To jump ahead to the policy conclusion, expert clinicians recommend a woman drink no alcohol at all if she is pregnant or could become pregnant. The corollary is that if she is drinking while pregnant, she should stop.
The symptoms are varied and the overt ones need not all be present. Underlying them is damage to the central nervous system. Prenatal alcohol exposure can damage the brain across a continuum of gross to subtle impairments, depending on the amount, timing, and frequency of the exposure as well as genetic predispositions of the fetus and mother. It can lead to neurological impairment and sometimes to growth retardation. The best known physical symptom is particular facial features associated with those with FAS. Many children with FASD don’t have the facial features but are nevertheless damaged.
All an economist need know is that an individual with FASD is likely to require greater health care while alive and that they may die earlier, perform more badly at school, have lower productivity at work and place greater demands on the justice, education, health and social services systems – not to forget the additional pressures on their families, their friends and their acquaintances. It is a very sorry condition, yet one which the affected individual had no control over; it is due to her or his mother drinking before they were even born. A mother drinking alcohol may not know the harm she is doing to her unborn baby; the damage can be much worse than that from tobacco, as pernicious as that is.
Each item on the list of effects which I have just gone through has economic consequences, such as higher costs to society in terms of less production or of resources which could be usefully deployed elsewhere. There are also non-market personal costs to those associated with FASD. Additionally a person with FASD can lead a less satisfactory life. We are not very good at quantifying these non-market costs but they are always in the back of our minds when we are doing an economic evaluation.
The Canadian team asked me to help them identify the costs of FASD from lost production because of lower productivity and earlier mortality. They gave me their estimates of how an individual with FASD might be affected, and wanted to know what was the total cost to society. There is not much point here going through the exact calculations – the methods are in the learned papers. But I need to set out the methodology.
A cost for an economist is almost always an opportunity cost – that is, if something was not occurring in the present situation what would be happening in an alternative one? This alternative is called the ‘counterfactual scenario’ in contrast with the actual fact. Choosing a counterfactual requires some skill; usually it depends upon what data is available but also one tends to choose the simplest from the complexity of realistic possibilities.
For this study we chose as our counterfactual scenario the assumption that there had never been any fetuses exposed to alcohol in the past, so that there was nobody with FASD today. We asked how much greater additional production would be if those with FASD did not suffer a productivity loss – either at work or being unemployed or because they died early.
The summary conclusion is that under the counter factual GDP – the measure of aggregate output by the market economy – would be higher by 0.2 to 0.5 percent if those with FASD had average productivity and employment for their circumstances. (I’ll skip the smaller mortality estimates because there is a complication when interpreting them because the population is larger.)
Your reaction may at first be that a loss of 0.2 to 0.5 percent of GDP is not much. But it is a loss from a mere one percent of the population. In New Zealand terms that represents a reduction of output of between $400m and $1b a year. The government often takes a lot of trouble to introduce policies which will generate far less than that.
The estimate does not include losses as a consequence of higher health costs, less effective educational services, higher judicial – police, court and corrections – costs, nor the emotional cost to families supporting their loved ones. The team looked at them; we should love to have been able to include them but we did not have statistics we could rely upon.
The sort of data we need are called ‘attributable fractions’. For example, we need to know what proportion of prisoners are in gaol because they have FASD, what proportion of any medical treatment occurs because of FASD. We have such figures for smokers and ex-smokers and for drinkers generally. But we don’t have them for those with FASD,
This means the figure of $400m as the social costs of FASD in New Zealand is very conservative, not only because it could be as high as $1b just for production losses. These other costs to the market economy are omitted and they are likely to be even larger. Nor does it cover the non-market costs.
Even so the $400m seems to be a large number. This year there will be born just under 600 New Zealand children with FASD and this year the economy is going to lose at least $400m, and probably much more, from the productivity losses of those with FASD.
What the figure is saying is that if in the past we had spent up to 0.2 percent of the GDP (perhaps even more) effectively preventing all FASD, the economy would have been ahead in output available for consumption, even after deducting the cost of the prevention program.
That does not mean that if we spend $400m this year on effective prevention programs the economy would benefit directly by that much this year. The $400m comes from all the earlier births – this year’s births will not impact on the labour force for two and more decades. (In the interim they will pressure health, educational, criminal justice and social services and their families and communities.) But if we carry out an effective prevention program long enough, there will be that sort of annual gain.
Of course a prevention program worth $400m plus targeted on preventing FASD is beyond the health promotion sector’s wildest dreams. Probably though, spending quite small sums would give great economic returns providing it was effective. Would you settle for $1m for a prevention program? If effective, the returns on the investment would be enormous.
I doubt any program would be 100 percent effective – alcohol is almost universally available, binge drinking is common and New Zealand has a relatively high rate of unplanned pregnancies. But suppose it prevented the most severe cases of FAS, reduced the number born with FASD and that those that remained had milder symptoms. Sound a good idea?
Now I have no expertise in health promotion either, but I am aware that on the whole we don’t do a very good job on FASD. Surveys suggest that many potentially or actually pregnant women are not aware of the damage that their drinking at any level can do to their baby’s health. The wider community is even less aware. There appears to be no systematic and coordinated program to inform us, which is surely the first step to prevention.
Among the resources available there is a non-government agency Alcohol Healthwatch. There is a FASD Project emerging at the Centre for Addiction Research in Auckland committed to advancing FASD research. As part of the Government’s response to the Health Select Committee Inquiry on Child Health, the Government is committed to an action plan to reduce harm from FASD which expects to be published by mid-2015. It is likely to involve boots-on-the-ground, as they say – people like you.
There is a problem which we might call ‘silos’. Does the Cancer Society have an interest in FASD? It does appears that individuals with FASD have an higher incidence of some cancers, but that is marginal, so let us leave this effect aside. It is also true that the Cancer Society has protocols about alcohol consumption because alcohol is associated with some kinds of cancer. Fair enough. But for the reasons I mentioned a few sentences ago they do not specifically mention FASD. And that is fair enough too.
However an approach in which each condition focuses only on its immediate concerns leads to a very fragmented approach to health promotion. What interested me, when I first came to FASD many years ago, was that the damage occurs at a time when the target audience is very vulnerable to common sense. Few pregnant women want anything other than the best for the babies they are carrying. I’m guessing that in such circumstances targeting may be very cost effective, especially as the pregnant women are already seeing medical professionals. But to be cost effective the targeting has to be holistic – not one agency concerned with alcohol, another with tobacco, a third with diet and so on.
It is not just a matter of targeting pregnant women. There is a period between conception and identification of pregnancy in which the consumption of alcohol is damaging even though the woman does not know she is pregnant. That means reaching out to women who can potentially get pregnant. (Allow me a side remark that if such a program was successful perhaps there would be fewer pregnancies conceived while the woman was drunk.)
By now I am well out of my depth, and I am relying on my mentor in this area, colleague Sally Casswell who directs SHORE, the Centre for Social and Health Outcomes Research and Evaluation, one of the research units with which I am associated. She says it is also a matter of the wider social culture and its attitudes to drinking, that we need to enhance our public policies: prices, purchase hours, marketing.
But we also need to target particularly damaging drinking, as we do with drink-driving. FASD is surely such a target. But what to do? (My small contribution is I wont drink alcohol in the presence of a pregnant women. I’m not a wowser; I enjoy a pinot with a meal. My attitude is one of solidarity.)
These are areas where you are much more proficient than I am. My economist contribution is that if you lot can set up a proposal for reasonably effective promotion program, I reckon an economist can almost certainly show that it will be cost effective, given the terrible damage that FASD does. We should even be able to get it past the Treasury, the keepers of the public purse, which is the gold standard.
To conclude more generally, I hope to have shown you that an economic evaluation – the estimation of the social costs of a disease or condition – is not just to generate Gee Whiz to impress everyone the disease is important. Done properly – sadly not all are – the estimate can be used to make better decisions about the deployment of health care services. In a resource-constrained world that is a very honourable objective, especially when it could lead to as beneficial an outcome as reducing FASD.