New newspaper series on HA-MRSA

The Seattle Times this morning launched an three-day investigative project on incidence of HA-MRSA in Washington State that is worth reading.

As readers here already know, MRSA is not a reportable disease, and there are no diagnosis codes that directly correspond to MSRA that make infection or death easily trackable through hospital records or death certificates. The Times' team came up with some innovative data-drilling techniques and apparently did a massive amount of number-crunching to come up with the incidence estimates that underpin their reporting. They use those to challenge hospitals' reluctance to undertake surveillance and treatment that would wipe out MRSA on colonized patients and thus reduce the likelihood of MRSA infecting those patients or spreading to others via healthcare workers who neglect infection control. (NB, Michael Berens, the series' co-author, did a huge project on nosocomial infections when he was at the Chicago Tribune a number of years ago.)

I am puzzled by one thing I am seeing on the story's web page — one of the items in the break-out box that sums the story up very quickly to attract eyeballs to it. It says: "About 85 percent of people infected with MRSA get the germ at a hospital or other health-care facility. " That figure doesn't make sense to me; it sounds as though it is a mis-translation of the CDC finding a year ago (in the Klevens JAMA paper) that approximately 85% of invasive cases of MRSA have hospital-associated risk factors. Constant readers will remember that estimate has been challenged by researchers on community MRSA, who believe that CA-MRSA accounts for a much larger proportion of the current epidemic than has been acknowledged, and think that the wide spread of the community strain is the actual driver of the overall epidemic. I can't see where in the text the Times team has done the math to support that assertion, so if anyone else spots it, or knows the reference it comes from, please let me know.