West Virginia Medical Institute: Using IT to Improve Patient Safety in Small, Rural Hospitals

In rural West Virginia, a patient safety project supported by the Agency for Healthcare Research and Quality (AHRQ) is proving that, with the right kind of assistance, all hospitals - regardless of their size or resources - can usestate-of-the-art information technology (IT) to vastly improve attention to patient care.

The three-year AHRQ grant has stimulated broad participation by small, rural hospitals across the state in an online patient safety network designed to prevent medical errors. Twenty-eight hospitals currently take part in the program; prior to the AHRQ grant, only six had signed up.

"What is extraordinary is that more than 50 percent of West Virginia hospitals are participating and they're doing so voluntarily," says Patricia Ruddick, MSN, project director at the West Virginia Medical Institute (WVMI), the quality improvement organization that oversees the patient safety program. "Some states that want to adopt these systems have had to pass laws mandating participation."

John G. Brehm, MD, chief medical officer at WVMI, explains that the program was designed to pinpoint and correct common problems that can lead to tragic outcomes. That required, as he puts it, "strength in numbers." But in a state where so many hospitals have less than 25 beds, the only way to get those numbers was to create a single system for fielding reports from multiple facilitie. And that meant wiring remote, financially strapped hospitals for high-speed Internet connections and then training staff, many of whom were new to IT, on how to use special reporting software.

Sounds daunting, but WVMI, leading a public-private partnership that includes the state, the state's hospital association, an IT vendor (Quantros, Inc.), and Verizon, has made it happen. To date, participating hospitals have contributed to a database that has catalogued more than 24,000 "events," information that WVMI and hospital staff have used to identify errors and determine how to prevent future errors.

According to Ruddick, online reporting can have a powerful effect on hospital practices. Before joining the program, she says, most hospitals relied on a paper-based system to flag possibleerrors. And it wasn't uncommon for those papers to get lost in the shuffle or fail to reach the person who could accurately assess the severity of the situation.

The new reporting system makes it easy for any employee - a doctor, a nurse, even housekeeping staff - to stop at a computer terminal and report a potential problem, and for supervisors to receive an immediate email alert. For example, at one hospital, a nurse used the system to report what seemed to be a minor, isolated incident involving non-sterile equipment. But a hospital manager who read the alert saw it as an indication of a potentially far-reaching safety concern. He immediately issued a broad recall of equipment that might be contaminated.

Meanwhile, on the other end of the system, WVMI staff aggregate data from participating hospitals to analyze how rates of events such as patient falls and drug reactions compare with national trends. In some cases, the data are used to increase hospitals' attention to problems that might be rooted in medical errors.

For example, as Brehm examined patient safety data from the system, he began to wonder about the extent to which adverse drug reactions might be connected to improper dosing procedures. One hospital, Bluefield Regional Medical Center, agreed to participate in a pilot project that would record the use of "rescue drugs" in patients to reverse an adverse event and determine whether the event should have been reported as a potential error.

Bluefield's Celesta Starling worked with WVMI to conduct a six-month study, during which the hospital pharmacy every two weeks produced a list of patients who received one of eight commonly used rescue drugs. Starling identified 109 cases that should have triggered a report; only six had been entered into the system.

"The issue here is education, not neglect," she says. "Some people might not view these as incidents. From their perspective, there was patient who had a reaction to a drug, they were treated, and now they're fine. But the situation really should be reported, because when you have these reports, they make everyone stop and look at the process."

The data also show that physicians rarely generate potential error reports; most come from nurses or other hospital staff. At one hospital, where physician-reported alerts accounted for less than 3 percent of electronic safety flags, leaders are trying to change their organizational culture by ensuring that interns and residents are trained in how to use the reporting system, Ruddick notes.

Gail Bellamy, PhD, principal investigator on the AHRQ grant, says that hospitals are seeing the benefits of consistent, accessible, and real-time reporting of patient safety issues. The challenge now lies in the program's future. Bellamy says that hospitals also need to see the connection between improving patient safety and improving their financial stability.

"There needs to be a strong business case for the time and technology investments required," Bellamy says. "We want to develop indicators that help our partners understand that this kind of system offers a good return on investment and that they should make a commitment to the program for the long term."