Wednesday, April 11, 2007

Central Line Infections, both better and worse

Here are our latest figures for central line infections, measured in cases per thousand ICU patient days. The average over the last several months remains better than for the previous year, but the rate for February comes from two actual cases, worse than January and with 100 fewer patient days. As always, we treat them as sentinel events and try to learn what went wrong and why.

Our folks are really serious about this and, in my opinion, deserve a lot of credit. A friend of mine was recently in the hospital and had one of these lines put in his chest for delivery of an anti-cancer drug. His wife, a medical professional, watched the doctor and nurse insert the line and was very impressed with their understanding of, and rigorous application of, the protocol. (And no, my friends did not mention to their providers that they had read all about this in my blog.)

Month ----- Infection Rate
Oct 05 ----- 1.67
Nov 05 ----- 1.28
Dec 05 ----- 2.43
Jan 06 ----- 3.07
Feb 06 ----- 1.40
Mar 06 ----- 1.07
Apr 06 ----- 0.00
May 06 ----- 0.59
Jun 06 ----- 1.15
Jul 06 ----- 0.57
Aug 06 ----- 3.03
Sep 06 ----- 2.50
Oct 06 ----- 0.00
Nov 06 ----- 2.38
Dec 06 ----- 1.87
Jan 07 ----- 0.00
Feb 07 ----- 1.15


Jon said...

Not to be a pain-- those results look wonderful-- but are they statistically significant? I assume they are, from the number of months this has been going, but I was just wondering if they passed the all-important statistical tests.

Anonymous said...

It's fantastic that you are publicizing this. This transparency will go along way to influencing other health care systems to do the same. Indeed, your efforts should be heralded. Would the BIDMC ever consider creating a site where all of it's health outcome statistics could be viewed by the public?

Anonymous said...


Statistical significance is not relevant. These are actual numbers for the entire ICU population, not a random sample. So, this is not a question of probabilities.

(It would be like asking if an election result for President is statistically significant. You can ask that question about a random sample poll taken before the election, but not about the election itself.)


Which outcomes do you think would be most valuable for the public to see?

Anonymous said...

Paul, what do you think about the outcomes listed on this website?

Or this one?

Anonymous said...

For the most part, these seem to be similar to the out-of-date data that is on most of the federal gov't and insurance company sites. For example, the DHMC surgery data is from 2005, although other things are a bit more current.

Emily DeVoto, Ph.D., said...

Paul, does the Joint Commission require reporting of infection rates such as central line infections as part of the accreditation process?