User login
Each month, a hospital quality scorecard is presented to the medical staff for heart failure and myocardial infarction. In it, compliance with the Joint Commission (formerly the Joint Commission on Accreditation of Healthcare Organizations [JCAHO]) performance measures is reported. Each figure is highlighted by a color, and if it is red, it means your performance is below the mean for similar hospitals. It is obviously better to be highlighted in green because that means you are above the mean.
The remarkable observation is that the difference between red and green can be tenths of a percentage point, so at 97.3% you might be red and at 97.8% you might be green.
Does this have any clinical meaning? Of course not. It’s statistical gamesmanship, but with serious overtones – because reimbursement can be tied to this version of your "quality." Aside from the fact that adherence to performance measures is not clearly related to outcomes, at least in patients with early post discharge (JAMA 2007;297:61-70), Mark Twain would remind us in situations like this that there are three types of lies: "Lies, Damn Lies and Statistics."
Unfortunately, the bean counters don’t understand regression to the mean or the lack of value that often accompanies a P value. And they most certainly don’t read Mark Twain anymore.
Dr. Hauptman is professor of internal medicine and assistant dean of clinical-translational research at Saint Louis University and director of heart failure at Saint Louis University Hospital, and a member of the Cardiology News Editorial Advisory Board.
Each month, a hospital quality scorecard is presented to the medical staff for heart failure and myocardial infarction. In it, compliance with the Joint Commission (formerly the Joint Commission on Accreditation of Healthcare Organizations [JCAHO]) performance measures is reported. Each figure is highlighted by a color, and if it is red, it means your performance is below the mean for similar hospitals. It is obviously better to be highlighted in green because that means you are above the mean.
The remarkable observation is that the difference between red and green can be tenths of a percentage point, so at 97.3% you might be red and at 97.8% you might be green.
Does this have any clinical meaning? Of course not. It’s statistical gamesmanship, but with serious overtones – because reimbursement can be tied to this version of your "quality." Aside from the fact that adherence to performance measures is not clearly related to outcomes, at least in patients with early post discharge (JAMA 2007;297:61-70), Mark Twain would remind us in situations like this that there are three types of lies: "Lies, Damn Lies and Statistics."
Unfortunately, the bean counters don’t understand regression to the mean or the lack of value that often accompanies a P value. And they most certainly don’t read Mark Twain anymore.
Dr. Hauptman is professor of internal medicine and assistant dean of clinical-translational research at Saint Louis University and director of heart failure at Saint Louis University Hospital, and a member of the Cardiology News Editorial Advisory Board.
Each month, a hospital quality scorecard is presented to the medical staff for heart failure and myocardial infarction. In it, compliance with the Joint Commission (formerly the Joint Commission on Accreditation of Healthcare Organizations [JCAHO]) performance measures is reported. Each figure is highlighted by a color, and if it is red, it means your performance is below the mean for similar hospitals. It is obviously better to be highlighted in green because that means you are above the mean.
The remarkable observation is that the difference between red and green can be tenths of a percentage point, so at 97.3% you might be red and at 97.8% you might be green.
Does this have any clinical meaning? Of course not. It’s statistical gamesmanship, but with serious overtones – because reimbursement can be tied to this version of your "quality." Aside from the fact that adherence to performance measures is not clearly related to outcomes, at least in patients with early post discharge (JAMA 2007;297:61-70), Mark Twain would remind us in situations like this that there are three types of lies: "Lies, Damn Lies and Statistics."
Unfortunately, the bean counters don’t understand regression to the mean or the lack of value that often accompanies a P value. And they most certainly don’t read Mark Twain anymore.
Dr. Hauptman is professor of internal medicine and assistant dean of clinical-translational research at Saint Louis University and director of heart failure at Saint Louis University Hospital, and a member of the Cardiology News Editorial Advisory Board.