Fishy

Not taking fish oil apparently costs Australia $4 billion per year.

The Deloitte study was commissioned by the Complementary Healthcare Council of Australia (CHC), a body which appears to be funded by the likes of Pfizer, Swisse Vitamins and Blackmores.
The press release which accompanied the research, and dutifully reported by the press, quoted CHC executive director, Dr Wendy Morrow, thus: "The report has found that the Australian community could be saving up to $4.19 billion, in terms of costs related to disease burden and premature life loss, through the preventive use of fish oils in Australians with heart disease. This research is of major significance to the industry, government and to the health and wellbeing of the Australian population."

Deloittes reckons that giving fish oil to every person with coronary heart disease would cost about $40 million per year. But a reasonable portion of the consequent benefits are in disability years avoided rather than costs avoided by the health system; it looks on a first cut that the measure passes cost-benefit on the basis of private benefits accruing to heart disease patients rather than from savings to the health system.

If the analysis hinges on benefits accruing to heart disease patients, then it's harder to make the case for that failure to take fish oil costs the country substantially. First, the people most likely to enjoy benefits from taking the supplement are likely already being advised to do so by their doctors; extending treatment to more people reduces the average efficacy of treatment. Second, while we can build a case for provision based on the private gains in monetized statistical life years, we also have to wonder why those who could benefit aren't already taking supplements. Is it because they can't afford it, or because doctors are less convinced than are Deloittes about the potential health benefits? If the former, don't we need to be sure that giving money to poor heart patients in form of subsidized fish oil does more good than giving money to other people in other forms?

I'm always a bit skeptical about studies claiming benefit to cost ratios of 39:1 - their base case result where net benefits are a billion dollars rather than the headlined $4 billion.

But if the Australian government had twenty seven million dollars to spend (the gross tallied cost in the baseline case), it probably does better buying fish oil on competitive markets and giving it to sick people rather than, purely hypothetically, throwing it at a company best known for making the kind of cars that increase Australia's need for a carbon tax.* So long as the fish is harvested sustainably I suppose.

* Update: It's actually unclear whether they do; cars that are more expensive to drive because they're worse on gas will emit more CO2 per mile but will be driven fewer miles. Net results then are ambiguous.