Friday, August 10, 2007

Quality can be hazardous to your health

"No man's life,liberty or property are safe when the legislature is in session".

This thought has been attributed to various sources but according to this website, it originated in a court decision in the 19 th century. Having read about the harmful and probably unintended consequences of some quality rules ( the word guideline has become a euphemism) I believe you could as well say : " When the quality rule makers are at work, no patient's well being is safe".

The four hour pneumonia antibiotic treatment rule has lead to patients without pneumonia getting antibiotics within four hours of arrival in the E.R. A recent posting by Dr. RW pointed out the weak evidence supporting rigid blood sugar control in sick hospitalized patients is poorly suited to mandate such an approach to all sick hospitalized patients yet guidelines mandating just that seem to be being prepared if not already proposed.Peri-operative use of beta blockers became a guideline but now there is more than a little doubt about the utility universal application.

The evidentiary best of the quality rules are supported to varying degrees by randomized clinical trials but even here one should proceed slowly as the reality of the well known efficacy-effectiveness gap tolerates simplistic approaches poorly. RCTs are often small and fail to capture the rich and perplexing diversity of relevant pathophysiological variables (known and unknown) that coexist in complex, hospitalized patients. It is in the sickest patients that rules formulated by committees based on what-ever that the most harm can be done and in whom the pathophysiological diversity is likely to overwhelm premature generalizations and an overly eager rush to mandate treatment rules.

The hubris,arrogance and naivete that I believe characterize those who make rules for how to treat individuals based solely on group data ignoring the input of the physicians on the scene is even worse that the folks who simplistically sprout out coarse grain outcome data quality indicators such as percentages of patients with HbA1c less than 7 %. In the later they are just counting beans in the former they are making rules about treatment decisions that might be determinative of the outcome. It all seems to be part of the notion that medicine is too important to be left to the individual physicians and patients.

The mandatory and quasi mandatory nature of guidelines or quality indicators as wielded by the CMS mandarins become even more dangerous with the spreading use of the "disruptive physician" doctrine. Not only must you go by the rules you can't complain about them without incurring the wrath of the hospital's disruptive physician committee.

In a recent post Dr. Robert Centor , of DM's Medical Rants quotes from a thoughtful article that correctly states that black and white answers are rare in clinical research and those that may seem clear cut today may not be as more data are accumulated. Committee derived quality indicators mandate that the grey be magically transformed into black and white and CMS and other third party payers give powerful financial valence to the process .

1 comment:

Zagreus Ammon said...

Sorry, for cross-posting, but I put this comment up on Kevin, MD and realized it more properly belonged here!

It is possible to apply the basic principles of industrial engineering to health care... BUT EXTREME CAUTION MUST BE TAKEN.

A notion exists in the management literature of mass customization. The process must be standardized and there is not reason the way an appointment is taken, or a sign-in is done or a blood test is drawn should differ from visit to visit. Yet the experience from the other side of the service line should always be unique.

This is the craft, this is the art.

But the devil is in the details and, in this case, in the execution of a good plan...

One of the major bones I have to pick with the damn "committee" is, that by misinterpreting data and the meaning of the word "guideline", they have created protocols from hell. Some people are so important to "know" something with certainty, they forgot that all knowledge is conditional and all data has limitations.

Please don't let the idiots convince you that the idea is bad. It truly is another good idea in the hands of fools.