Making patients safer through user-friendly ICUs

Today's intensive care units bear little resemblance to their late-20th-century counterparts. They house advanced, sci-fi modish medical technology: intricate devices and systems intended to improve patient care and save lives.

But newer doesn't necessarily mean better. Is too much information and sophisticated machinery an unnecessary hardship for clinicians? Can refined and simplified systems better serve doctors and nurses and make patients safer? A multidisciplinary effort at Johns Hopkins will endeavor to find out.

Johns Hopkins Medicine's Armstrong Institute for Patient Safety and Quality will lead an effort with the university's Applied Physics Laboratory, Whiting School of Engineering's Systems Institute, and Berman Institute of Bioethics to design, implement, and deploy an integrated set of interventions to reduce medical errors in intensive care units. The project is funded by an $8.9 million grant from the San Francisco-based Gordon and Betty Moore Foundation.

The grant—the first award in a new $500 million 10-year Patient Care Program designed to eliminate preventable harms that patients experience in the hospital—will fund a two-year demonstration project at three sites: the Surgical Intensive Care Unit at The Johns Hopkins Hospital; Johns Hopkins Bayview Medical Center; and the University of California, San Francisco Medical Center.

Work will begin this fall.

The Johns Hopkins group's aim is to improve outcomes by using a systems engineering approach to health care, leveraging technologies and creating better processes to ensure that patients always receive the therapies and treatments they need and that clinicians work as effectively and efficiently as possible, says Peter Pronovost, senior vice president for patient safety and quality at Johns Hopkins Medicine and director of the Armstrong Institute. The grant-funded work also seeks to better engage patients—and their families—in their own care, making them an integral part of the health care team.

"Despite heroic efforts by clinicians, patients continue to suffer preventable harm, in large part because health care is grossly under-engineered. Devices don't talk to each other, treatments are not specified and ensured, and outcomes are largely assumed rather than measured," Pronovost says. "This project will seek to change that by enlisting systems engineers to ensure patients always get the treatments they should, by engaging patients in every aspect of their care, and by creating a health care system that continually improves." While sophisticated health care technology has improved mortality and morbidity rates in hospitals, the increasing use of these innovations has also resulted in patient safety and care quality challenges. Studies show that at least one in every five of the estimated 4 million patients treated in intensive care units every year is harmed during his hospital stay. The ramifications range from ICU-acquired weakness and delirium to ventilator-associated infections such as pneumonia, adult respiratory distress syndrome, bloodstream infections, deep vein thrombosis, and pulmonary embolus.

A significant problem is that the medical devices commonly used in ICUs—infusion pumps, ventilator systems, defibrillators, electrocardiogram analyzers—are not integrated or interoperable, resulting in systems that do not share data or functionality across the health care enterprise. The devices can also be hard to decipher, even by experienced human users, says Adam Sapirstein, an associate professor of anesthesiology and critical care medicine at the School of Medicine and director of the Learning Laboratory at the Armstrong Institute.

"Medical devices don't do a good job of communicating with clinicians," he says.

Sapirstein says that many current devices require a wealth of experience to interpret the data and put them into clinical use. The data and readouts, he says, are often complex and not meaningful enough.

He likens the operation of these systems to driving a car with a dashboard that presents information in binary code, with key displays such as miles per hour or the check-engine light either not in view or nonexistent.

"When you design an automobile, you don't want to make it more complicated and harder to drive. You want to easily interpret data on the dashboard and have all the information you require at your fingertips," he says. "In the ICU, we sometimes come across information overload. We need to find ways to present information in a more useful and perhaps simpler way, like a car with a backup camera to improve situation awareness. But at the same time, you can't lose sight of that big tree on the side of the car."

APL, which will receive roughly $3.2 million of the grant, will lead the systems engineering component to ensure that the technologies developed support improving patient outcomes, and enhance the effectiveness and efficiency of clinicians. "We will demonstrate that a systems approach to the management of patient care can improve both the processes of care and the outcomes of patients, and facilitate the meaningful engagement of patients and families," says Alan Ravitz, of APL's Research and Exploratory Development Department.

The team will apply systems engineering principles and best practices to improve care and reduce a variety of negative effects, not only those that harm the patient physically but also those that can damage the dignity and respect of patients and their families. Whiting School faculty will work on the math and computer components of these systems, looking, for example, for improved algorithms and ways to better collect and process the data.

Berman Institute of Bioethics faculty will ensure that the patients' rights and sensitivities are being addressed.

APL and Johns Hopkins Medicine have collaborated on other health care improvement projects that employed a systems engineering approach, such as an effort to improve the design and operations of medical treatment facilities and a safety initiative aimed at improving the safety of infusion pumps. The Johns Hopkins Systems Institute—based at the Whiting School—provided the research and development funding, as well as the technical oversight, for the development phase of the latter project.

"Through these efforts, and the initial funding and oversight support that was provided by the WSE's Systems Institute, we've learned key lessons that we intend to apply to the patient safety ICU demonstration project," Ravitz says.

Established in 2000, the Gordon and Betty Moore Foundation seeks to advance environmental conservation, scientific research, and patient care around the world. Gordon Moore is the co-founder of Intel Corp. and chairman emeritus of its board of directors.

The foundation chose to work with Pronovost, whose team has already shown it can dramatically reduce central line-associated bloodstream infections in ICUs by using a simple checklist coupled with culture change that empowers staff members to speak up when proper procedures aren't being used. The checklist program has been instituted in hospitals across the United States and in several countries abroad.

Many more harms need to be tackled, Pronovost says.

"A patient in the ICU or with multiple chronic conditions may need to receive scores or hundreds of therapies a day, yet there is no list of what needs to be done, no feedback about whether they have been performed, and they largely happen by memory rather than automatically," Pronovost says. "Contrast that with all of the automatic safety checklists in cars. No wonder health care continues to harm patients while driving deaths continue to decline."