Value of 30-day Readmissions Measure Questioned

A paper by the chairman of the Department of Medicine at Tufts Medical Center finds "very little evidence" that the 30-day readmissions measure of quality "has any effect on overall survival." In fact, he says it, "drives a lot of doctors and hospitals crazy."

Not all quality measures are created equal: Some are good, some are head-scratchers, and a handful may even be dangerous to patients.

That's the argument of "Quantity Over Quality: How the Rise in Quality Measures is Not Producing Quality Results," a perspective piece in the Journal of General Internal Medicine.

"All quality measures are not equal, although they're treated as equal… some of them are very importantly backed by science, and some of them are not very well-backed at all," says Deeb N. Salem, MD, physician-in-chief and chairman of the Department of Medicine at Tufts Medical Center in Boston, and first author of the paper. "It raises the fact that we may not be scrutinizing quality measures as well as we should."

According to the paper, quality measures take clinical guidelines a step further, into the realm of a "mandate that affects fiscal compensation and public reputational standing. QMs influence patient care, institutional compliance, and organizational financial well-being."

And that's where things can get cloudy.

For instance, Salem points to 30-day readmissions, the quality measure that he calls the "one I hate the most" and that "drives a lot of doctors and hospitals crazy."

"The 30-day readmissions one is filled with unanswered questions," Salem says. "Why are we using this one? Why has it become something that if we don't do well on, the hospital is going to lose money?"

Salem says he has "no idea" how or why those 30 days were arrived upon in the first place, and sees "very little evidence that this has any effect on overall survival."

Moreover, the paper notes there are multiple factors that lead to readmissions, not all of which are within a hospital's control. The paper cites research showing that readmissions vary by patient population, with some attributed to factors such as poverty, mental illness, and lack of social support, so hospitals serving large numbers of those populations may be unfairly judged.

"There are social factors that may play a big role in that," Salem says. "We're maybe punishing hospitals that are taking care of the kinds of people that can't take care of themselves, and we may be hurting those hospitals more than we're helping anybody."

The paper also includes discussion of two quality measures—both of which have now been eliminated—that not only were eventually shown to be ineffective, but actually ended up being harmful to patients.

The "scariest" example of this, according to Salem, was the widespread use of perioperative beta blockers, a quality measure based on research that was later shown to be fraudulent. In fact, "a meta-analysis published in 2013 with consolidated outcomes of over 10,000 randomized participants showed that initiation of [perioperative beta blockers] before surgery caused a significant increase in mortality," the paper says.

"It took years to figure this out," Salem says, but by then, that faulty quality measure had already been used for a long time.

Other quality measures are flawed in other ways, the paper says. For instance, it states that length of hospital stay isn't positively correlated with quality of care, and patient satisfaction in surgical care has "no association with hospital compliance with surgical quality measures."

The point of this discussion isn't to disparage all quality measures, Salem says. He believes there are valid quality measures that have saved many lives, such as the focus on reducing central line-associated bloodstream infections.

Infection Prevention Saves Lives and Money

"It sounds like we're trashing the whole system; we're not," Salem says. "We're just saying, 'back up a little bit.'"

Instead of focusing on or prematurely adopting quality measures that may not be useful and that may not differentiate good practices—or worse, actually harm patients in the long-term—Salem believes that there should be a peer-review-type process that thoroughly investigates and vets potential quality measures so health insurers and Medicare don't "jump into them without much proof." Instead, doctors and hospitals should be able to focus on the quality measures that are most effective for improving patient outcomes.

"We need to develop a process of being careful and making sure that there is quality in these measures," Salem says.