All Content

Know the Score


With hospitals facing increasing pressure to improve safety based on measurements, hospitalists need to build a key role in improving quality by developing safety scorecards, say leading hospital medicine experts.

A framework for designing scorecards was recently suggested by researchers at Johns Hopkins University in Baltimore in an article published by the Journal of the American Medical Association.1 The commentary suggests a framework to help healthcare organizations develop safety scorecards, evaluate their validity, and understand measures appropriate to present as rates.

Their framework is intended to build scorecards that monitor progress in improving patient safety over time or relative to a benchmark. The authors urged organizations to think of safety on a continuum and look for improvements, rather than regard practices as either safe or unsafe. They also stated that their term “safety scorecard” acknowledges an overlap between quality and safety.

To build their framework, the researchers adapted elements of the “Users’ Guide to the Medical Literature: A Manual for Evidence-Based Clinical Practice,” to address three key questions:

  • Is the measure important?
  • Is the measure valid? and
  • Can the measure be used to improve safety in healthcare organizations?

The resulting worksheet to evaluate a scorecard guides hospitals through questions aimed at determining whether their institution meets the three criteria.

Initial Reaction

“This worksheet would be very important to follow, a good step forward in efforts to improve quality because its questions make clear where a group might be falling down in developing a scorecard,” says hospitalist Brian Bossard, MD, director of Inpatient Physician Associates at BryanLGH Medical Center in Lincoln, Neb. Dr. Bossard, who is also the medical staff quality designee, says he believes large national hospitalist groups should be involved in planning scorecards as part of a multidisciplinary team.

Other experts also tout hospitalists’ importance in the vanguard of creating these vital instruments.

“Hospitalists should be on or chairing safety committees, and there should be investments in training them in these areas,” says Eric Kupersmith, MD, division head of hospital medicine and assistant professor of medicine for the Cooper Health System, University of Medicine and Dentistry of New Jersey Robert Wood Johnson Medical School in Newark. “Because we are frontline physicians experiencing what is happening with patients and orchestrating as well as delivering care, we should provide feedback.”

Dr. Kupersmith, who is on his facility’s patient safety committee and has participated in a root-cause analysis of hand-offs as well as worked on medical reconciliation and pneumonia core measure performance improvement, says SHM “should help take the lead in bringing together specialists, administrators, and nurses with hospitalists who should have a major impact in designing a scorecard as a society.”

As hospitalists mature as clinicians, they become system- and process-oriented, says Dr. Kupersmith. As a result, “We should be part of re-engineering efforts because of our experience bringing people together,” he says. “We’re in a good position to analyze the process.”

Use with Care

There is also value in hospitalists’ anecdotal experiences, Dr. Kupersmith says, but “there needs to be a filter between anecdotes and a facility’s leadership to ensure that the information provided is broadly important. Decisions shouldn’t be made on anecdotes, but creative ideas can come from them.”

Randy Ferrance, DC, MD, and chief of the medical staff at Riverside Tappahannock Hospital in Tappahan-nock, Va., regards anecdotal information in much the same light. “It should be seen as guidelines, not rules,” he cautions. “There is still an art to medicine even though it is clearly science. Sometimes the best available evidence may be anecdotal. It’s not hard data, but it can be valuable.”

Dr. Ferrance at one time chaired his hospital’s quality improvement committee; the panel now reports to him. He believes hospitalists are fortunate that “we became a specialty after evidence-based medicine really came to the forefront. We are fortunate to have the backing of much hard data.”

Still, he acknowledges the difficulty of establishing proof that an action affects patient outcome. “It’s hard, but what we can do is look at what might help result in things like decreased morbidity, length of stays and complications, and a faster return to a patient’s normal functions. Then we might see influence on patient outcomes.”

Beyond Core Measures

Richard Rohr, MD, vice president for medical affairs at Cortland Regional Medical Center in New York, works extensively on quality improvements. He urges hospitalists and institutions to go beyond required measures.

“A lot of what we’re doing now is imposed by government, insurers or the Joint Commission,” says Dr. Rohr. “In day-to-day work, we should look for areas to go beyond what is required.”

He believes hospitalists should look at specific issues underlying the Johns Hopkins framework’s three core questions.

For example, within the first core question “Is the measure important?” he suggests hospitalists consider what their facility’s priorities are. “There are thousands of things that could address safety,” he says. “Hospitalists should look at how a measure fits in to their organization’s priorities. At an institution known for cardiac care, look at safety measures in cardiology.”

He also suggests looking at a facility’s potential problem areas to help determine if a measure is important. “I worked at a facility that had a rule that Coumadin had to be ordered one day at a time,” says Dr. Rohr. “Since patient use of the medication was closely monitored, there were very few patients with serious bleeding.”

The point, he says, is that hospitalists should “make sure the safety process they’re interested in truly addresses a true problem and is not already in place. Try to add something of value.”

Hospitalists are in a good position to do this, he says, because of their day-to-day perspective on patients. “Hospitalists should start by looking at what their organization has addressed and what’s causing patient problems day to day and then set priorities,” he advises.

Institutional Support

There are thousands of things that could address safety. Hospitalists should look at how a measure fits in to their organization’s priorities. At an institution known for cardiac care, look at safety measures in cardiology.

—Richard Rohr, MD, vice president for medical affairs at Cortland Regional Medical Center, Cortland, N.Y.

Of critical importance also, Dr. Rohr says, is to see what their organization can support before deciding what safety measure to explore.

“Is it feasible to collect data to use this measure?” he asks. “That’s partly dependent on where your facility is in using electronic medical records. Does the benefit of researching and implementing outweigh the cost? You may have to spend some staff time to decide what is worthwhile.”

The last task may be easier for hospitalists working at more academic hospitals, he says, which is also an important part of trying to answer the Johns Hopkins framework’s second question: “Is the measure valid?”

It’s often hard to answer that one, Dr. Rohr acknowledges. “Research has at times shown that a process may show statistically significant improvement, but it does not show up clinically,” he says. “Aspirin and beta-blockers for heart patients, for example, has a statistically significant difference—but it is small.”

Most hospitalists try to see patients and do this kind of work on the side, he concedes: “They should do some research, but value what they see when treating patients. You have a good sense of what has helped patients.”

Dr. Wright

Julia Wright, MD, associate professor of medicine at the University of Wisconsin School of Medicine and Public Health and medical director for hospital medicine at the University of Wisconsin Hospital in Madison, agrees.

“The expertise of our specialty is that we deliver care that is not just clinical, asking, ‘Did I meet the guidelines?’’’ she says. “We’re with patients. We should help determine how quality and safety models are addressing how care is delivered.”

She also believes hospitalists should work closely with hospital administrators on these issues. “Hospitalists have an intrinsic sense of value in delivering care,” she notes. “We are unique in that we can combine consideration of hospital goals with knowledge of care at patient levels. This provides great value to the institutions.”

Culture of Safety

While he agrees with the importance of involving hospitalists deeply in safety efforts, Dr. Kupersmith believes institutions should strive to create a culture that focuses on safety and looks at all its processes in that light.

“You shouldn’t just track hard outcomes,” he suggests. “Track the outcomes of your processes. This gives an overall sense of safety awareness in all personnel. If you focus on the process and culture, you might find a significant change in outcomes.” This also helps address the difficulty of finding data on outcomes, he says.

He agrees with the researchers’ view that safety is on a continuum, and he thinks acknowledging that can help establish an institutional culture around safety. “There is always going to be patient danger,” he says. “You want to get to a point where it is minimized because of an awareness of actions. That focus on safety will lead to less danger.”

As a result, he believes quality improvement strategies must address culture. “You need to provide education for all on safety and provide oversight and monitoring with expectations that can be tracked,” he says. “You need to create this mandate and speak in the quality language from the top. Then you start to have people bring in information that affects outcomes.” TH

Karla Feuer is a journalist based in New York.


  1. Pronovost PJ, Berenholtz SM, Needham DM. A framework for health care organizations to develop and evaluate a safety scorecard. JAMA. 2007;298(17):2063-2065.

Next Article:

   Comments ()