Commentary
Since writing a blog about the 30s ("The magic of 30 and 35"), I have thought about artificial time lines in general but also the skewing of incentives.
The penalties meted out by the Centers for Medicare and Medicaid Services for hospitals that exceed specific 30-day readmission rates for heart failure are based on two very faulty premises. First, that sociodemographics don’t matter, since patient zip code and related factors do not enter the calculation. Second, that there is something meaningfully different about a patient readmitted on day 29 vs. day 31. Does anyone really think that there is a physiological, structural, or other difference? The 30-day construct is artificial, though as I write this sentence I can hear the response from the regulators: "Well, we had to make a cut somewhere!"
Maybe so, but how should we judge a health care system in which a hospital spends significant resources to prevent a 30-day readmission but then may not pay much heed to the initiative once a financial penalty is no longer a risk. In other words, a system in which the penalty for day 31 is lower (zero) than for 29?
©kyoshino/thinkstockphotos.com
It’s also interesting that we talk about length of stay for heart failure, but not quite as much as we used to. Indeed, lengthening the length of stay might reduce readmissions, if we can take a lesson from our European colleagues.
Expounding a bit more on the topic of artificial metrics, why does the Food and Drug Administration want 180 days of safety information for a study in which in-hospital patients are randomized to active study drug versus placebo for acute decompensated heart failure? How about 145.6 days instead? Is the patient with a left-ventricular ejection fraction of 36% safe, while one at 34% is a "must do" defibrillator implant? Does the patient with dyspnea and a b-type natriuretic peptide of 101 pg/mL have heart failure but at 99 pg/mL they are safe? I won’t name names here, but such conversations have been duly noted by your blogger/columnist, which prompts me to discuss receiver operator curves. Unfortunately, I don’t seem to get very far.
Frankly, a small dose of nihilism and a bit more understanding of statistics are needed in our profession. It’s also time to reestablish some common sense. The number 31 is not lower than 29 even if regulations make us (and hospital administrators) think it is.
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. He is an associate editor for Circulation: Heart Failure and blogs while staring out his office window at the Arch.