Christopher Cheney, for HealthLeaders Media , February 5, 2015

Individual hospital performance accounts for less than half of the variation in pooled readmission rates across the United States, researchers find.

County-based data collected from across the country show hospitals are far from solely responsible for readmission rates.

An analysis of the data, which is slated for publication this month in the journal Health Services Research, features information collected from 4,000 hospitals for patients with three conditions: acute myocardial infarction, heart failure, and pneumonia. The key finding of the study, “Community Factors and Hospital Readmission Rates,” is that 58% of the variation in readmission rates was related to community characteristics outside a hospital’s control.

Jeph Herrin, PhD, lead author, says several elements of a community’s health capabilities that are not under hospital control help drive readmissions. “The health system outside the hospital, independent of any socioeconomic status characteristics, is important to understanding geographic differences in readmission rates,” Herrin said in an interview.

“Our results indicate that at least some of the accountability should be shifted away from hospitals,” he says.

The study comes as hospitals are facing growing financial penalties over readmissions. The Centers for Medicare & Medicaid Services’ the Hospital Readmissions Reduction Program cuts a hospital’s aggregate Medicare reimbursement if a facility reports higher-than-expected 30-day, risk-adjusted readmission rates for patients 65 years and older. The penalties were phased in, starting with up to a 1% Medicare reimbursement cut starting Oct. 1, 2012, and rising to up to 3%, effective Oct. 1, 2014.

Herrin and his co-authors examined county data for three types of community characteristics:

  • “Sociodemographic” factors such as living alone and educational levels
  • Access-to-care measures including general practitioners per capita
  • The number and quality of nursing homes in a county

More than half, “58% of the total variation in publicly reported hospital 30-day readmission rates was attributable to the county where the hospital was located. Expressed differently, the results suggest that individual hospital performance accounts for only 42% of the variation in pooled readmission rates across the United States,” the study says.

While socioeconomic status (SES) factors such as educational level were associated with hospital readmission rates, nursing home density and quality were found to be more significant factors.

CMS is planning to roll out a readmissions reduction program that targets skilled nursing facilities. The Protecting Access to Medical Care Act of 2014, last year’s congressional patch of Medicare’s reviled Sustainable Growth Rate formula for physician reimbursement, includes a value-based purchasing (VBP) program for skilled nursing facilities.

Beginning in October 2018, under the VBP program, CMS is expected to hold skilled nursing facilities accountable for hospital readmissions through financial incentives such as linking Medicare payment rates to performance standards.

Spreading Responsibility for Hospital Readmissions
The Los Angeles-based physician who wrote an editorial to accompany Herrin’s readmissions study says the analysis breaks new ground. “It’s the first study that I’ve seen that really has done a rigorous and in-depth look at the factors happening outside the hospital,” Teryl Nuckols, MD, a hospitalist and director of the Cedar-Sinai Medical Center Division of General Medicine, said in an interview.

She says the research Herrin and his team conducted shows the necessity to hold more parties accountable for hospital readmission rates. “There’s definitely a need for greater coordination of care. There’s a need for increased collaboration between the in-patient and out-patient settings. What the [HRRP] policy does is it makes the hospitals accountable for all of it,” Nuckols says. “The readmissions penalties for hospitals are meaningful. They have created an incentive [to reduce readmissions].”

Future research on the impact of SES on readmission rates should focus on rural and inner-city areas. “It really warrants additional study,” Nuckols said.

A CMS spokesman says the agency is gauging its hospital readmission reduction efforts carefully: “We are establishing a detailed plan to comprehensively analyze the impact of SES factors for Medicare payment systems and programs, and investigating data sources that would enable accurate measurement of SES.”

No ‘Magic Answer’ to Hospital Readmission Puzzle
The lead author of another recent study on readmissions that cast doubt on the effectiveness of readmission reduction programs says the research Herrin and his team conducted accurately reflects the challenge.

“If you have a well-greased community, readmissions are a manageable problem,” says Ariel Linden, DPH. “Providers are humming along; there’s outpatient coordination. You don’t need to be as concerned about readmissions. People are going to fall through the cracks, but not as much [as in communities with fragmented healthcare services or low socioeconomic status.]”

He says future research is likely to show a direct correlation between hospital readmission rates and the level of economic distress in a community. “In a disadvantaged community, you have Medicaid patients and the uninsured. Doctors don’t want to see them because reimbursement rates are low,” Linden said. “These patients are the high-fliers in the emergency room; and when they get out of the ER, there’s nothing for them out in the community to keep them on track.”

As Herrin’s research implies, a broad and flexible approach is needed to reduce hospital readmission rates, Linden said.

“I don’t think there’s a magic answer here,” said Linden, an adjunct associate professor at the University of Michigan’s School of Public Health in Ann Arbor and president of Linden Consulting Group. “Reaching individual patients through nurses or other readmission reduction programs takes a tremendous amount of resources to make a dent at the population-health level.”

To achieve significant reductions in hospital readmissions, he says the healthcare industry has to find a way to deploy a broad set of solutions.
“We need more doctors. We need to pay doctors to see patients who don’t have insurance. We need to reimburse higher for Medicaid patients. We need more coordinated care. The key is doing all of it. I don’t think doing any one thing is going to solve much.”

Health Services Research is a publication of AcademyHealth. A grant from The Commonwealth Fund, a private foundation based in Washington, DC, financed Herrin’s research.

Christopher Cheney is health plans editor at HealthLeaders Media.