The program, which took effect in 2012, seemed to work as intended. Within a few years, studies appeared in prestigious medical journals showing dramatic drops in readmissions.
“They declined the most in the hospitals doing the worst — just what you’d hope for,” said Dr. Robert Yeh, who studies cardiology outcomes at Beth Israel Deaconess Medical Center in Boston. “We thought, ‘Oh, looks like it’s been successful.’”
Now, it’s not so clear. Are readmissions for those conditions really dropping as substantially as it first appeared? Or has the program’s impact been overstated?
Are Medicare patients getting better care, or are they being kept out of hospitals to avoid readmission penalties? Are people getting hurt in the process?
There’s no consensus on the answers, as research has produced conflicting results. But the questions intensified recently as two new studies helped stoke skepticism.
One study, published in JAMA, reported that deaths from heart failure and pneumonia within 30 days of discharge have risen since the program began. “There’s a cloud over this,” said Dr. Yeh, the study’s senior author.
The readmissions program had no pilot testing before it began affecting decisions in thousands of hospitals, he noted: “Why couldn’t we have rolled this out in a way that let us evaluate it better?”
A spokeswoman for the Center Medicare and Medicaid Services, citing “deep concern” about findings of increased mortality, said in an email that the agency would thoroughly review both studies and their methodologies and conclusions “to inform any future actions.”
Reservations about the program’s claims to success, focusing on how hospitals report their statistics, had already surfaced before these latest studies.
To control for the fact that some patients are much sicker than others to begin with, Medicare uses “risk-adjusted” readmission statistics to prevent unfair penalties when higher-risk patients return to the hospital.
Those designing the new program expected hospitals to reduce readmissions by improving transitional care: giving discharged patients better instructions, following up with advice and referrals, perhaps providing home visits.
But a study of seven million Medicare hospitalizations suggested a more disappointing explanation. Looking at readmissions before the program, then after its announcement and implementation, the researchers noticed a distinct drop in a single month: January 2011.
“The idea that one hospital would figure this out in a month is a stretch,” said Christopher Ody, a health care researcher at Northwestern University who led the study. “That 3,000 hospitals would figure it out in the same month is incredibly unlikely.”
What happened? Dr. Ody suspects a change in administrative practices, not an improvement in care.
In the study, published in Health Affairs, he and his colleagues point out that Medicare had just changed the electronic form used for billing. Instead of allowing hospitals to enter up to nine diagnoses for each patient, the new form permitted as many as 25.
Accordingly, the proportion of claims with fewer diagnoses plunged and those with 11 or more climbed, making patients’ conditions look more severe.
“That explains roughly half the decline in the risk-adjusted readmission rate,” Dr. Ody said. Instead of a 2.6 percent decline from 2010 to 2012, the study found, readmissions actually fell a more modest 1.3 percent.
That decline, he and his co-authors argue, is no greater than the drop in readmissions for other conditions at smaller, rural hospitals that were exempt from readmissions penalties.
That might represent good news, if it meant the program had changed hospital cultures enough to reduce readmissions even when penalties didn’t apply.
But Dr. Yeh and his group have reported, more disturbingly, that the program may contribute to unnecessary deaths. Using the Medicare database to compare mortality rates before and after the penalties were imposed, they found rising mortality within 30 days of discharge for heart failure and pneumonia (but not for heart attacks).
The rates of increase were small but growing, and may indicate that thousands of additional deaths from heart failure and pneumonia followed the program’s announcement and implementation.
“Some of those patients previously would have been readmitted, but because of the financial incentives, they were not,” said Dr. Gregg Fonarow, co-chairman of cardiology at the David Geffen School of Medicine at the University of California, Los Angeles, and a critic of the readmissions program.
In an editorial accompanying the Harvard study, he called on Congress and Medicare to revise the program.
Because both these new studies are observational, they can show readmissions falling or deaths rising, but can’t explain why. The findings could reflect other factors, like better outpatient care and higher thresholds for hospitalization.
The readmissions program still has staunch defenders, too. One is the Medicare Payment Advisory Commission, a nonpartisan panel that advises Congress.
It reported in June that the program had reduced readmissions without increasing mortality — and was saving Medicare more than $2 billion a year.
Another defender is Dr. Harlan Krumholz, a cardiologist and researcher at Yale University whose team helped develop the readmission measures Medicare uses. He’s skeptical of the recent findings.
“This suggests that doctors are putting their hospitals’ financial interests ahead of their patients’ welfare, which I haven’t seen and which the evidence doesn’t support,” he said, citing MedPAC’s study and his own.
By contrast, Dr. Fonarow said he has indeed heard that very complaint from cardiology colleagues. “They’re getting tremendous pressure from their administrations,” he said.
Now that additional conditions can result in readmission penalties — including joint replacements, pulmonary disease treatment and cardiac bypass surgery — he wants an independent investigation of the program.
“You want to have faith that the physicians making decisions with you don’t have, in the back of their minds, ‘Are we going to get dinged for this?’” he said.
While nobody wants to spend more time in a hospital, some patients need to be readmitted. But it’s hard for consumers to judge whether another stay will help them recover or needlessly expose them to additional risks and expense, so a consensus on the Hospital Readmissions Reduction Program’s safety and effectiveness would be welcome.
“Perhaps the lesson is, it’s harder to reduce readmissions than we would have thought or hoped,” Dr. Ody said.”