If we want to rein in the costs of the U.S. health-care system — now equal to nearly 18 percent of the nation's gross domestic product — we cannot ignore the fragmented technologies used to help heal and save lives.
At first glance, the devices, monitors, electronic health records and machines found in today's hospitals might inspire awe. Look beyond the slick displays with blinking lights, however, and the picture is less reassuring. Rather than working as an integrated whole, these technologies rarely "speak" to one another, reducing productivity and increasing costs. As a result, time that clinicians might spend at the bedside or discussing patient cases with colleagues is used to fill in the gaps between uncoordinated technologies.
For example, nurses scribble down a physician's instructions for a drug infusion from one computer screen, do math to find the right dose, and then walk to the medication pump and enter the order. Every high-risk medication requires a second nurse to double-check that the pump is programmed accurately — a task that drains staff time. On a 12-bed intensive care unit, these double-checks add up to two full-time nursing positions, we found at Johns Hopkins. Yet if the electronic orders communicated with the pump, this extra work would not be needed, and the risk of entering the wrong dose would diminish.
Clinicians who want to gauge a patient's progress and risk of complications need to make hundreds of clicks on computer screens, check devices and eyeball the settings on the patient’s bed. In one of our ICUs, the work of ensuring that patients received steps to prevent seven common harms — such as blood clots and infections — takes nearly 20 minutes. Integrate systems and that time can be cut by two-thirds, we found in a pilot project.
Hospital beds typically come with alarms that alert staff when a patient starts to get out of the bed on his or her own. Yet sometimes staff forget to turn the alarm back on — for instance, after the patient returns to the bed following an X-ray. An integrated system would quickly flag patients who are deemed a high fall risk but whose bed alarms aren't activated. Instead, in many hospitals, staff still must walk from room to room to visually check that bed alarms are on for patients who need them.
Other fields have used systems integration approaches to become high-performing, less costly and ultra-safe. In aviation, for example, controls, instruments and mechanical systems communicate with each other, enhancing pilots' situational awareness. Pilots don't spend time poring over topographical maps to make sure they don't fly the plane into a mountain; the plane gets that data and warns them if they are approaching a cliff. Yet health care has not embraced similar approaches.
Such a change is overdue. There are certainly other solutions to improve health-care productivity and decrease costs. Complications and substandard care make hospital stays longer or lead to readmissions. Staff spend too many hours on required documentation. Hospitals devote enormous administrative resources to working with hundreds of different insurers. Drug prices are constantly rising. Yet we cannot ignore the dysfunctional systems at the heart of how we care.
If we improve productivity by integrating technologies and extracting their data, we can do more than reduce costs. We can transform the patient experience: fewer complications, more "quality time" with their care team, and more opportunities to engage in their care.
This post was first published in The Wall Street Journal's Experts blog.