“John McCain’s health plan would reduce the ranks of the uninsured by about 21.1 million people if fully put in place by 2010, while Barack Obama’s would reduce the number by 26.6 million, an analysis predicts.” That was the lede of an AP story that circulated last week. Those numbers come from a study done by the Lewin Group, a prominent health care consulting firm that advises governments, insurance carriers, drug companies, and health care providers.

The Lewin analysis is important not for the new estimates it throws on the table, but because those estimates are likely to spur another game of “My Plan Covers More People Than Yours,” played by Hillary Clinton and Obama during the primary campaign. Remember, Clinton said 15 million people would not be covered under Obama’s plan; Obama said 10 million wouldn’t be covered under hers.

Lewin’s report adds to the growing number of studies offering different estimates of how many would receive health care under the candidates’ proposals. In September, the health policy journal Health Affairs published a study that indicated McCain’s plans would have little effect on the uninsured. The number indicated that initially giving people a tax credit and taxing some portion of employer-sponsored benefits would have little impact on the number of uninsured. Within five years, the number of uninsured would increase.

Last May, the Urban Institute/ Brookings Institution Tax Policy Institute said that Obama’s plan would cover nearly half of America’s uninsured over a ten year period, while McCain’s plan would cover less than five percent.

All of this is just so much crystal-balling, since the numbers reflect the inputs that go into the statistical models, and sometimes those inputs are subjective. That’s not to say, however, that such numbers shouldn’t be reported, especially if the candidates start their “My Plan Covers More People Than Yours” game again. But they should be reported with context, clarity, and completeness. The AP story which was picked up by news outlets around the country fell short on these measures.

For starters, the story didn’t contextualize the projected number of newly insured individuals in terms of the total number currently without coverage. A vital question, no? The latest census data show that 45.7 million people were uninsured last year, so it would appear that McCain’s plan would leave almost 25 million people without coverage; Obama’s would leave 19 million uncovered. Whether that’s good or bad depends on whether you want universal health coverage or don’t mind that some people will be left out. Having the AP actually report those numbers, along with the 45 million stat it did use, would have made it much easier for readers to see what they mean—without having to fool with the math themselves.

As I read through the AP story, I saw a quote from Kenneth Thorpe, identified by the AP as an Emory University professor “who conducts similar analysis and advice for Democratic candidates’ health plans.” Uh, oh, I thought, some Republican or conservative outfit must have paid for the study, especially since Thorpe said: “They pulled the same overestimates for Bush last time with a similar bill.” It turns out that the study was an independent analysis conducted by Lewin.

“We didn’t do it for anybody. We have complete control over our own work,” John Sheils, Lewin’s senior vice president, told me. Fine, but why didn’t the AP tell readers it was Lewin’s own report? The release sent out on the PR Newswire noted that it was an independent study. How difficult could it have been for the AP to clue its readers into an important fact that spoke to the study’s credibility?

Trudy Lieberman is a fellow at the Center for Advancing Health and a longtime contributing editor to the Columbia Journalism Review. She is the lead writer for The Second Opinion, CJR’s healthcare desk, which is part of our United States Project on the coverage of politics and policy. Follow her on Twitter @Trudy_Lieberman.