Tuesday, January 11, 2011

Private Rooms and Infection Control

Researchers at McGill University in Montreal published a study in Archives of Internal Medicine showing a 50% reduction in ICU infections in patients placed in private rooms compared to those in multiple-bedded rooms. Many media outlets [Science Daily, Eurekalert, Reuters] distributed the story. CBC ran it but somewhat conflated the ICU bed part with general hospital beds. Also, some prominent tweeters linked to #ptsafety failed to highlight that the study involved only ICU patients. [See photo.]
I am concerned that this might encourage consumers to ask for private rooms whenever they are hospitalized. In an ideal world, I certainly agree that every patient should have his own room, not only to reduce infection risk, but for privacy and comfort.
There are two major obstacles preventing us all from having private rooms. One, insurance generally will not pay for it. Two, most hospitals have a very limited number of private rooms. I spoke with Sara Beasley, of the American Hospital Association’s Resource Center. She told me that the AHA has no data on the number of private vs. semi-private rooms in US hospitals because that information is not currently being collected by any organization or agency she is aware of. She did point out that guidelines for new hospital construction do call for all private rooms. Most ICUs in the US have all private beds. But based on my own experience, I estimate that less than 10-15% of US hospital medical-surgical beds are private.

Until all beds are private, we will have to continue our efforts to practice good infection control and limit the over-utilization of antibiotics which is rampant in the country.


Anonymous said...

In my area, all are private rooms throughout, or quickly approaching that. As far as the study goes - no shit sherlock.

Skeptical Scalpel said...

Appreciate your comments, especially the last sentence which I thought but didn't say.

Post a Comment

Note: Only a member of this blog may post a comment.