Confounding in epidemiological studies: why "independent" effects may not be all they seem.

passed a child health insurance bill, and we had to stop enrolling children because we ran out of money. It was only at the very last moment, at this last special legislative session, through extraordinary efforts by some really committed folks in the governor's office, that a deal was struck to expand the programme so we could reopen enrolment. The programme provides subsidised health insurance to children who do not qualify for Medicaid for primary ambulatory care-not hospital care, no medications, no high-tech procedures, just basic routine care, getting the infection treated before it turns into a major problem. It's a very narrow, circumscribed programme. But we need to expand it to all eligible children. The legislators chose children because frankly, from a political standpoint, children are easiest to sell. But as we discussed, there's so much need for expanded access to programmes like this, even before we get to comprehensive reform. It can be effective, I think, in targeting the most needy popu- lations and solving some of the biggest inequities in the system. So I think we can push ahead with things like insurance reform as we did in this legislative session- community rating, eliminating pre-existing conditions. We can push ahead with patchwork programmes like child health. Any amount of resources we can commit to that while we're trying to deal with the larger issues-how we're going to control costs in the major programmes and expand access and make sure that quality doesn't suffer at the same time. We have to do both.