Quality gains, with a catch
Quality campaign makes headway, but lack of standardized data blurs picture
Fewer early elective deliveries, lower 30-day readmission rates, greater levels of engagement by patients and families—those are among the victories being touted by the leaders of the Partnership for Patients, a sweeping, $1 billion patient-safety initiative led by the federal government.
But CMS officials and participants acknowledge the program’s lack of standardized measures or validation protocols could make it hard to reach conclusions about across-the-board performance improvement, both nationally and within networks set up to coordinate hospitals’ efforts.
Launched in 2011, the Partnership for Patients aims to reduce hospital-acquired conditions—including pressure ulcers, adverse drug events, falls and ventilator-associated pneumonia—by 40% and preventable all-cause 30-day readmissions by 20%, all by the end of 2013. To those ends, HHS in December 2011 enlisted the help of 26 hospital engagement networks, which include large health systems, state hospital associations and other groups charged with leading improvement efforts among their member hospitals.
At last count, more than 3,700 hospitals were participating in the campaign, led by HENs such as the Charlotte, N.C.-based Premier healthcare alliance; 37-hospital Dig-
“In the long run, we need common metrics, but in the short term, we are trying to provide some flexibility.”
—Sean Cavanaugh, CMS Innovation Center
nity Health, San Francisco; the Health Research & Educational Trust; and Irving, Texas-based VHA.
Encouraging outcomes data have begun to trickle in from the HENs, Dennis Wagner, codirector of the Partnership for Patients and the associate director for campaign leadership at the Center for Medicare and Medicaid Innovation, said during a Feb. 26 webcast held by VHA. He cited dramatic reductions in early elective deliveries before 39 weeks’ gestation, a commonly targeted obstetrical harm.
Specifically, rates of early elective deliveries plummeted 48% among 681 hospitals in 20 HENs, based on data from January 2010 to June 2012, Dr. Paul McGann, co-director of the Partnership for Patients and chief medical officer for campaign leadership at the CMS Innovation Center, said during the webcast.
Thirty-day readmission rates have also improved across participating hospitals, officials said. They cited federal data released Feb. 28 that showed the national 30-day, all-cause readmission rate fell to 17.8% in the last half of 2012, after averaging 19% for five years.
Progress has been seen across other types of patient harm, too, though not as dramatic, Wagner said. “The trends on other harm areas are moving in the right direction but not fast enough,” he said, urging more hospitals to join the initiative.
Nagging concerns persist, however, about
whether the federal government can accurately judge the improvement work of hospitals engaged in the initiative without the benefit of commonly used quality metrics. HHS granted significant latitude to the HENs regarding which measures their member hospitals use and how they evaluate data.
The Texas Center for Quality and Patient Safety, for instance, has noted big safety improvements among many of the 68 hospitals in its HEN, said Rachel Cicerchi, the center’s program manager for the Partnership for Patients.
“We’ve been pleasantly surprised by the progress we’ve seen so far, especially because a lot of our hospitals are small and rural,” she said. “For a big portion of our network, this is one of their first large quality initiatives.”
But Cicerchi conceded that the Texas Center for Quality and Patient Safety has struggled to gather HEN-wide data because participating hospitals are using vastly different measurement strategies. The center’s solution has been to evaluate each hospital independently, she said.
“We did allow them to choose their process and outcomes metrics,” Cicerchi said of the hospitals. “In an ideal world, it probably would have been much easier for CMS to have one measure that they could use, but the initiative was not set up that way. They told us we could use whatever metrics we wanted.”
Sean Cavanaugh, the CMS Innovation Center’s deputy director, called the Partnership for Patients’ end-of-2013 goals achievable, as evidenced, he said, by high-performing organizations across the country. But he acknowledged that the program’s relaxed measurement requirements have made gath- ering and analyzing data difficult.
“There is a tension between the need to collect data uniformly and the burden it places on the HENs,” Cavanaugh said in an interview. “It’s a sensitive balance, and we struggle with it every day. In the long run, we need common metrics but in the short term, we are trying to provide some flexibility.”
One obstacle the Partnership for Patients has faced is that it was deployed very quickly, reducing the amount of time planners could spend defining measures and performing due diligence, said Dr. Donald Goldmann, chief medical and science officer at the Cambridge, Mass.-based Institute for Healthcare Improvement, and a clinical professor of pediatrics at Harvard Medical School, Boston.
“It’s left to the individual HENs to figure out how they will measure, and individual hospitals can be measuring differently with different levels of reliability,” said Goldmann, who co-authored a May 2012 perspective piece in the Journal of the American Medical Association about the need for a more rigorous measurement strategy within the Partnership for Patients. “That puts us on very shaky ground when we’re trying to assess what is actually happening.”
Some HENs are in a better position than others, he said. Premier, for instance, whose HEN includes more than 450 hospitals, has a robust data system and fairly strong measurement capabilities, he added.
The Partnership for Patients is still a worthwhile endeavor, said Goldmann, who praised the initiative’s focus on safety and the use of evidence-based guidelines. But he cautioned that without a more accurate gauge of harm— he advised a retrospective analysis of hospitals’ performance using the IHI’s Global Trigger Tool—it would be very difficult to reliably track improvement.
“That is going to be a real challenge,” he said.