22,000 or 45,000?

I've seen this story a few times today:

WASHINGTON (Reuters) - Nearly 45,000 people die in the United States each year -- one every 12 minutes -- in large part because they lack health insurance and can not get good care, Harvard Medical School researchers found in an analysis released on Thursday.

But what I haven't seen is an explanation of why this is double the estimate of 22,000 annual excess deaths that has been sort of the standard number that gets quoted.

Studies estimate that the number of excess deaths among uninsured adults age 25-64 is in the range of 22,000 a year.

I'll keep looking. I've felt that, at a gut level, the latter number seemed low, but I would still like to understand why the former number should be accepted as accurate.

I also winced a little at this passage:

The National Center for Policy Analysis, a Washington think tank that backs a free-market approach to health care, said researchers overstated the death risk and did not track how long subjects were uninsured.

Uh... unless you're "lucky" enough to be 65 or older, the "free-market approach" is exactly what's killing either 23,000 or 45, 000 people per year.

Followup, a few minutes later: The Left Coaster has a good take on the news...

What would this nation do if some foreign power killed 45,000 people in this country in a single year? What should we do to the insurance industry?