Rethinking hospital readmissions

There’s long been a belief that hospital readmissions are largely preventable, and that if a patient is readmitted after an earlier hospital stay, it’s a sign of lower-quality care. In recent years, hospitals in the U.S. and elsewhere have poured untold amounts of time and resources into lowering their readmission rates.

A new study from Canada has come along to challenge this thinking. The study, which appeared this week in the Canadian Medical Association Journal, analyzes emergency readmissions among 4,800 patients at 11 hospitals in Ontario and concludes that only about one in five could have been prevented.

Moreover, the researchers learned that the number of readmissions may be lower than previously thought. In this particular study, 13.5 percent of patients in the sample group had an urgent, unplanned readmission within six months after a hospital stay.

One study doesn’t make a trend. For one thing, the analysis didn’t include patients who were readmitted after being discharged to a nursing home and who may have been older, more frail and more vulnerable. For another, the determination of whether an admission was avoidable was to some extent subjective.

But it certainly calls into question the widespread belief, especially among health policy folks, that readmissions are a reliable indicator of whether a hospital is “good.” It also challenges the assumption that readmissions can be avoided if only hospitals work diligently and hard enough.

Among the many nuggets of information in the Canadian study: The majority of patients who had an urgent readmission were more likely to have chronic or serious health conditions and a history of previous hospital admissions than the sample group as a whole.

There appeared to be multiple reasons why these patients had to be readmitted. Case reviewers found that some readmissions were obviously avoidable but others were less clear-cut. Errors in patient management were among the most common factors. So were surgery-related complications and medication-related issues. Some patients were readmitted because they developed a hospital-associated infection, others because there was a diagnostic error during their initial hospital stay.

One point worth noting: Readmissions that were deemed preventable mostly occurred within a few days after the patient left the hospital. This could suggest many things – for instance, that this subset of patients was perhaps simply more sick to begin with, or not quite ready yet to be sent home.

Another crucial point: The case reviewers couldn’t always clearly determine whether a readmission was truly avoidable and often needed more information before classifying it as preventable or not.

Dr. Carl van Walraven, a clinical epidemiologist at the Ottawa Hospital Research Institute and lead author of the study, told the Ottawa Citizen, “Not all urgent readmissions are avoidable, despite the care that is provided. This means that a lot of them are caused by a patient’s condition, or other factors that are not treatable or modifiable.”

The use of metrics, or statistically measurable indicators of hospital care, is widespread in the industry. Increasingly, the federal government, payors and quality assurance organizations are tieing metrics to how much hospitals are paid – more money for hospitals who meet the standard, less for those that don’t. But what if the metrics are based on faulty assumptions, i.e. that most unplanned readmissions are avoidable?

If anything, this study seems to underscore how difficult it is to define and measure quality in health care. Can every pressure ulcer be prevented? Perhaps not, although the most serious pressure ulcers are probably avoidable and the goal should be to make them a very rare occurrence. Much more can be done to lower the rate of hospital-acquired infections but it may not be possible to get down to zero.

This isn’t to say that high-quality hospital care doesn’t matter, because it does. But hospitals deal with sick human beings in all their infinite variety and the results aren’t always standard or predictable. The risk with metrics is that they can reduce the definition of quality to “that which can be measured” instead of the complex, nuanced, many-faceted creature that seems to be emerging the more we study it.

Photo: Wikimedia Commons

Leave a Reply

Your email address will not be published. Required fields are marked *