Changing Technology Paradigms - Machine Learning in the Hospital Setting
As consumer technology continues to progress at an incredible rate, the healthcare field is experimenting with how these advances can shape patient care. Investment in healthcare technology is at an all-time high, with private equity and corporate venture capital in the digital health sector growing 74 percent in the first three-quarters of 2016 compared to the same time period in 2015. Consumer technologies, applied in health care, can help tackle some of the industry’s toughest challenges and work in tandem with human resources to provide high-quality care.
The most exciting part about how quickly technology develops is that we no longer have to wait for the hardware and software to catch up to our ideas–all of the equipment exists now. For example, the mechanics of your in-home security camera, applied in the healthcare space, could prevent a nurse from accidentally providing a patient with incorrect medication. The code that helped IBM’s Watson beat Ken Jennings and Brad Rutter at Jeopardy! could help care teams identify the best course of treatment for individual patients, based on their medical history and outcomes from similar patient populations. All of these programs utilize machine learning, a technology that has the power to revolutionize health care. Machine learning’s ability to continually absorb information and improve itself as it gains access to more data and experiences is being used across industries to tackle some of today’s biggest challenges. In healthcare, companies are beginning to utilize machine learning to help clinicians detect breast cancer metastases in lymph nodes, develop a tool to help prevent blindness in patients with diabetes and identify people at high-risk for cardiac arrest.
To capitalize on machine learning technology, the healthcare field needs to continue to evolve with these innovations. One of the first places where evolution can occur is in the health system. Hospitals are excellent proving grounds for new healthcare technologies, as they are a microcosm of the greater health care system: patients, providers, payers, caregivers and manufacturers all have a presence in the hospital setting.
Right now, the current technology in use by hospitals does not adapt on its own; it must first be programmed to respond to a specific trigger. For example, if a patient presses the call button for a nurse, that technology is programmed to respond to the action by triggering a light or sound at the nurses’ station. Much, if not all, of the technology used in a hospital, operates in this type of closed system where tools’ functions do not evolve without manual program updates. In a machine learning system, programs learn on their own and become smarter and more responsive over time. No one needs to program the computer to alert the nurses’ station when a patient presses the call button–it has learned to identify certain characteristics of a patient in need without prompting and takes appropriate action. This type of system, because it can identify other triggers, can be applied in a variety of places around the hospital to solve an array of challenges, from ensuring that medication is stored at the proper temperature to predicting whether or not a patient will be readmitted, and for how long. Getting stakeholders within the health system to how else this technology can be applied and wherein both the short and long-term.
Machine learning’s ability to continually absorb information and improve itself as it gains access to more data and experiences is being used across industries to tackle some of today’s biggest challenges
In order to capitalize on machine learning, health systems must also analyze their organizational workflow and how human resources currently interact with technology. When machine learning technology is implemented, it will constantly gather and aggregate data, and make predictions and decisions based on that information. Some of that information will need to be transmitted to the people working in the hospital, but there will need to be ongoing discussions about how that notification happens. Is it through a vibrating bracelet? Does there need to be a wearable device mandated with a display to explicitly state what is occurring? These questions will need to be addressed head-on and collaboratively with a variety of stakeholders, from providers to technology companies. While traditionally, partnerships between technology innovators and health systems are still in their infancy, creating and deepening those relationships will be vital for the future of patient care.
The healthcare field has always been a proving ground for new technologies–from FDA-cleared wearable trackers that monitor and assist in the treatment of Parkinson’s disease, to therapies that use a person’s living tissue to create a medicine individualized to them, technology and health care are inextricably linked. But, in order to truly realize the promise of machine learning, hospitals and health systems need to evolve to capitalize on this new technology. It starts by understanding the current technology systems in place within the hospital and continues by evaluating how the hospital’s systems and workflow support the connection between human resources and technology. Machine learning will revolutionize the way we provide care for patients–we need to be ready for it.