Friday, February 11, 2011

Limited "coarse grain" data suggest following certain pneumonia guidelines can be bad for survival

While that headline sounds like a typo that is what a recent article in Lancet seemed to show.Further we are not talking about all pneumonia guidelines,as , for example the guidelines for the treatment of community acquired pneumonia (CAP) actually work out rather well. The focus in the above headlined article was on the ATS and IDSA guidelines for the treatment of hospital acquired pneumonia (HAP).

When I read about that finding my first thought was to look more closely at the guidelines and importantly what was the evidence underlying the recommendation. As has happened more than once, Dr.RW saved me the trouble. See here.

Dr RW's analysis suggests that the evidentiary basis of the recommendation of the IDSA and does not belong on the top of the classical,mythical evidence based medicine (EBM) evidence hierarchy in which randomized clinical trials and meta-analyses perch at the top.

My take on this article is that we might be cautious in accepting the findings on face value. After all this was a retrospective observational study replete with all the potential biases this type study might possess. This is what I call coarse grain data without the fine grain detail that might be provided by detailed patient level analysis. For example, the authors speculated that perhaps the side effect of the double gram negative antibiotic combination may have contributed to the increased mortality in the group treated in accord with the guidelines. Maybe so, but more detailed analysis might provide support or refute that speculation.

No comments: