Technology meant to help patients can also put them at risk. According to Dr. David Classen, faculty for IHI's Improving Patient Safety with Health Technology Expedition, the rapid implementation of health IT over the last 10 years has unintentionally exacerbated risks to patient safety. In this post, Dr. Classen describes the need for a continuous learning system to help address workarounds, glitches, and the unintended consequences of health technology.
Q: What are the main reasons health technology poses patient safety risks?
Health care organizations in the US have broadly adopted health technology (or health IT) over the last 10 years for a variety of reasons, among them to improve patient safety and reduce costs. Organizations also received federal financial incentives through the CMS Meaningful Use program to implement electronic health record technology. Rapid implementation of health IT overlays a very complex health care system, and we're beginning to see a range of unintended problems associated with this process. For example, usability issues that make life more difficult for doctors, nurses, and other health care workers using the technology; or unintentional harm to or death of a patient resulting from incorrect use of the technology.
These are just some examples of issues we need to address to improve health care IT safety going forward. There’s an almost magical way of thinking about health IT: If I just put it in and turn it on, it will improve safety. That's not the way it works. You have to put it in, turn it on, and then endlessly improve it, refine it, and enhance it to make a measureable impact on safety.
Of course, we want to make sure that technology doesn’t harm patients; yet, we’re seeing ways in which technology is causing patient harm. For example, copying and pasting physician orders into the wrong [patient’s electronic health record] and the patient dies as a result. Inserting information from one patient’s record into [another] patient’s record, leading to a misdiagnosis and serious injury. These examples push us to further evaluate, significantly redesign, and continually optimize these complex adaptive health IT systems.
Q: What are some of the keys to making health technology safer for patients?
I was a member of the Institute of Medicine committee that wrote a report called Health IT and Patient Safety: Building Safer Systems for Better Care, which outlines a roadmap for making health care IT safer for patients. A key message in that report is that health care IT is part of a complex sociotechnical environment. Technology exists within a complicated system of people, processes, and the organizations in which we work, but unfortunately health IT is often designed and implemented as if it’s in its own standalone silo. As a result, health IT systems can be a major source of job dissatisfaction for the health care professionals who are using them. For example, staff spend a lot of unproductive time using these systems, which detracts from time they could be spending with patients, or they create workarounds because the health IT systems are not designed for their actual workflow.
Now, after we’ve already broadly implemented health IT, we’re trying to go back and optimize it, fix it, and deal with the sociotechnical issues. As we redesign and optimize these systems, we need to include patients in the process. Health care is increasingly moving to the ambulatory arena and to the medical home. It’s mainly the patient who takes care of themselves in these settings. Right now, patients are on the outside of these health IT systems looking in. We give them access to some things – such as lab values — without much training or understanding, but technology is growing rapidly with new sensors and patient monitoring systems, especially in the ambulatory and medical home settings. We have devices that can record a patient’s heart rate, blood pressure, and weight automatically at home. But we’re still a long way from successfully engaging patients in this rapidly growing use of health IT. We’ll need the involvement of patients to make effective use of these systems and the volumes of data they create.
We’re also beginning to realize that this is such a complex and rapidly evolving health care environment that it’s possible that – no matter how hard we try — we can’t design health IT right the first, or even the fifth, time. Perhaps the best approach to implementing health IT systems is to assume that something will go wrong, so we should build learning systems that help us identify problems so that we can learn and continuously improve.
Q: What does it mean to have a health IT learning system?
I’ll give you a couple of examples. Let’s say a patient is on the intensive care unit and the electronic medical record (EMR) system shuts down. This happens at a lot of hospitals because of the complexity of these systems and often upgrades can inadvertently shut down the whole system. What most hospitals do is just turn the system back on again. They reboot it and turn it back on, but they don’t learn why the system goes down in the first place. Also, when these complex systems shut down and they reboot, they sometimes shut down the critical safety checks, leading to health care workers thinking the system is performing critical safety checks when, in fact, it is not.
When this happens in the airline industry, for example, they not only try to determine the cause of the shutdown, but before they put it into operation again, they perform critical safety checks of the rebooted system. In health care, we never do. We don’t learn what happens when these systems get turned off, and having a learning system would mean that we would.
A learning system would also be one that studies itself. For example, an innovative group of researchers in New York City started programming their health IT system to tell them how often a doctor placed an order on the wrong patient. In a year, physicians placed over 6,000 orders in the health IT systems on the wrong patients. Without this type of continuous systems testing, you wouldn’t have any idea that was happening. Now this health care organization can use this type of data to both continuously monitor and learn how to make these complex adaptive IT systems operate much more safely.
Q: Technology is now so prevalent throughout health care. Where should an organization start to address potential safety issues?
One place to start is with the free SAFER Guides issued by the Office of the National Coordinator for Health Information Technology. Organizations can use these guides to conduct a sociotechnical assessment of their current health IT. Even the most sophisticated health care organizations can learn a lot from this type of assessment.