Rural hospital leaders are faced with a conundrum when it comes to measuring quality.
They want to jump onboard the quality-reporting train that's charging through the industry, but many believe the current measures by both government and private sectors can't accurately or fairly quantify the level of care that's provided at rural facilities. Executives also have expressed concern because their hospitals haven't been a focus of many of the early quality projects -- which focused on the services urban acute-care hospitals offer -- and this worries rural providers, especially with pay-for-performance initiatives looming a few stops down the line.
"We do not want to leave out rural hospitals and pay-for-performance," says Mary Wakefield, director of the Center for Rural Health at the University of North Dakota.
Specialty hospitals such as psychiatric hospitals also say the acute-care hospital surveys can't be used to fairly judge the quality of care they provide. For that reason, the Joint Commission on Accreditation of Healthcare Organizations is working with the National Association of Psychiatric Health Systems and the National Association of State Mental Health Program Directors to develop quality indictors for psychiatric hospitals.
Wakefield acknowledges that many of the government and private-sector measures now used for reporting on quality of care can be applied to rural hospitals. One example of a measure that is valid at all acute-care hospitals examines emergency care for heart attack patients by rating how many patients showing symptoms of a heart attack receive aspirin upon arrival. Another is the percentage of patients given counseling on the health effects of smoking.
Such indicators are good for measuring the care provided at all hospitals, but other measures that rate aspects of care unique to rural hospitals also should be developed as quality indicators, she says.
"Much of the (research) work is derived from urban tertiary-care units with the expectation that you can generalize," Wakefield says. "But when you look at rural hospitals, the providers are different; the patient mix is different; the severity level is different; the technology is different. The patients themselves are different."
Wakefield served as chairwoman on an Institute of Medicine panel that published a November 2004 report on improving quality at rural health facilities. The report concluded that more research was needed to accurately gauge the level of care being provided.
"What's out there already?" she asks. "Not a whole lot."
Wakefield says comprehensive rural-specific quality indicators would measure both the unique and common aspects of services that rural hospitals provide. The nation's nearly 2,200 rural hospitals annually serve about 55 million patients, or about 20% of the population.
The trouble with transfers
Many rural hospitals don't provide the services that their urban counterparts provide, so a large role that rural facilities play in the continuum of care is stabilizing and transferring patients to facilities that offer more services. These transfers open the possibility for mistakes.
"Whenever there's a handoff of a patient to an emergency room or to the medical floor, we know that the handoff is susceptible to medical errors," Wakefield says. "In rural hospitals, those handoffs are common."
The CMS contracted with Minnesota's quality-improvement organization, Stratis Health, which partnered with the University of Minnesota Rural Health Research Center and HealthInsight, the QIO for Nevada and Utah, to study rural hospital indicators. The center and the two QIOs identified 20 measures from existing quality indicators and performance-measurement systems that are relevant for rural hospitals with fewer than 50 beds. The measures were field-tested at 22 hospitals in Minnesota, Nevada and Utah. In addition, indicators were developed for common rural hospital functions -- such as emergency department timeliness of care and transfer communication -- that were not considered in existing measurement sets.
The field tests found that transfer quality measures can collect relevant data on the transfer practices of small rural hospitals if the hospitals received appropriate training and support from QIOs, says Ira Moscovice, director of the Minnesota center. Some of the measures have now been proposed by the CMS to be part of a QIO project that would measure quality at critical-access hospitals starting this summer.
Moscovice says the work takes one of the broadest looks at developing quality indicators for rural hospitals. He adds that much of the previous research has examined specific hospital units or departments, and he hopes this report leads to the CMS funding more research and developing rural quality measures.
The transfer measures in the field test are basically a checklist of information the medical staff needs to provide when transferring patients, but Moscovice said transfer measures could go beyond that.
"There are no measures that look at decisionmaking," he says. "That's a big hole."
For example, a survey that the federal Agency for Healthcare Research and Quality helped develop asks about transfers, but the question doesn't ask about specific procedures. The question asks respondents to rate how strongly they agree or disagree with the statement, "Things `fall between the cracks' when transferring patients from one unit to another."
Barry Kitch, a physician and senior scientist for the Institute for Health Policy, which was founded by Partners HealthCare and Massachusetts General Hospital, both in Boston, presented a research project on hospital quality surveys at a patient-safety conference in May. That study examined 13 surveys and found five asked transfer questions, but they referred to transfers from shift-to-shift or within a hospital, Kitch says, and nothing specific to rural hospitals.
An article on the CMS-contracted study by the QIOs and the Minnesota center ran in the fall 2004 issue of the Journal of Rural Health and said it's common to transfer rural-facility patients to urban hospitals. Also common is transferring patients from urban to rural acute-care or post-acute hospitals because rural residents often undergo a procedure in an urban hospital for specialized care and then are transferred to a hospital closer to their home for recovery.
Whenever patients are transferred there's a potential for data to be lost. Also, since rural hospitals tend to have fewer resources and a staff with less expertise, there's a greater chance diagnosis could be incomplete when the patient is transferred.
The paper states that rural hospitals in general have fewer specialists and they rely on generalists to make more of the decisions. Therefore, measures should try to gauge the quality of the decisionmaking regarding where patients are treated. Measures that show how efficiently hospitals communicate with each other would also be important, Moscovice says. Communication at larger hospitals is often more difficult, but the paper noted that rural hospitals tend to have smaller staffs, which can make it easier for them to improve safety practices.
"The smaller size should result in less isolation and more interaction among staff, which may make it easier to develop a shared culture," according to the paper.
Darlene Bainbridge, founder of hospital consultant group D.D. Bainbridge & Associates, says she has seen examples of such staffing efficiencies firsthand. "Sometimes, I'll suggest a rural hospital make a change and by the end of that day, they'll say, `OK, we did that. What's next?' " she says.
But with the smaller size also comes less money to invest in adding staff and services. That's one reason why the Leapfrog Group, a patient-safety coalition, makes optional for rural hospitals -- but not for urban facilities -- a question in its hospital quality survey about computerized physician-order entry, says Suzanne Delbanco, the group's executive director. The measure is designed to show how many hospitals are using CPOE, and it's optional for rural hospitals because many can't afford to install the often-pricey systems.
Leapfrog, which started surveying hospitals in 2001, invited rural hospitals to respond in 2004 after it added 27 questions to its survey, which now has 30 questions. Rural hospitals weren't invited initially because their low volume and lack of resources might have skewed survey results, Delbanco says. Although some rural hospitals were OK with not being included, others wanted to participate, saying they wanted to be part of the quality movement, she says.
Many rural hospital officials worry that if they don't report all quality data measures it would appear they aren't being transparent; at the same time, when reporting data on all measures, rural hospitals also can be at a disadvantage because one bad outcome or decision can unfairly change their overall quality scores.
"The challenge is we have to be better than urban hospitals," says Michael Gillen, chief executive officer of 36-bed Sterling (Colo.) Regional MedCenter.
When Colorado launched a Web site in April reporting the mortality rates of the state's hospitals, it tried to address the volume problem by bundling years for the lower-volume hospitals. If hospitals have a low number of cases for a measure, then the most recent three years of data are combined.
"It's better than not bundling," Moscovice says.
However, bundling has limited value for comparative purposes. Most hospitals like to use the quality report cards to address problem areas. If the data were 3 years old, the staff members who could have been the cause of the low score may no longer work at the hospital, Moscovice says. Also, it takes longer for hospitals to improve a negative score when years are bundled, he says.
The issue of volume is a common concern rural hospitals cited, and many in the industry weren't too happy with the way the CMS decided to present that information on its quality reporting Web site, Hospital Compare, which was unveiled April 1 (hospitalcompare.hhs.gov).
The site labels hospitals -- both urban and rural -- that have low volumes for a given measure with a yield sign and an ominous message that says, "Be careful when drawing conclusions for these hospitals because of the small number of patients treated." That warning, according to Tim Size, executive director of the Rural Wisconsin Health Cooperative, says, "You don't want to go there."
Size also says he doesn't like how the site divides hospitals into two categories: "acute care (general hospitals)" and "critical access (small, remote hospitals)." He says the site should be one searchable database. Although the CMS hails the Web site as a great tool for consumers, Size says it does a disservice to rural residents because they can't compare urban facilities to their own.
"Critical access is a payment system, not a way to measure quality," Size says.
Size also noted that many of the hospitals on the site lack a phone number and some are listed as being in the wrong county.
The CMS says a computer glitch led to some phone numbers not appearing on the site, according to a statement from spokesman Peter Ashkenaz, and that the site will be updated by late May.
The CMS didn't provide interviews on this topic but Ashkenaz, in e-mailed comments, said the decision to break out the critical-access hospitals was made because the agency was following requests consumers made in focus groups. "Consumers mentioned that maintaining this distinction was important to them, because of their perceived differences in the range of services that might be available in the two groups of hospitals," according to the statement.
Ashkenaz added that the hospitals were divided into groups because the noncritical-access hospitals were entitled to payment incentives as part of the prospective payment system. Those that don't report performance data will receive a payment increase for fiscal 2006 that is four-tenths of a percentage point lower than those that do report data.
Critical-access hospitals, which are reimbursed through a cost-based system, were given the option to participate in the performance survey, but their payments are not affected.
Slightly more than half of the critical-access hospitals reported on at least 10 of the measures and more than 20% reported on all 17 of the voluntary measures, according to Ashkenaz.
Wakefield and representatives of other rural organizations such as the National Rural Health Association say there are problems with the CMS Web site, but they are confident -- as in the case of developing rural-specific quality measures -- the problems can be worked out.
"There are technical glitches that we can overcome with time," she says. "The labeling is problematic."
What do you think?
Write us with your comments. Via e-mail, it's [email protected]; by fax, 312-280-3183.