• Innovations on the Front Lines of Health Care

    Technology: Karen Wagner Apr 27, 2016

    Artificial intelligence is among the technological breakthroughs that are having a significant impact on care delivery.

    With so much disruption in health care today, including changes that affect the practice of medicine, many physicians are feeling frustrated and dissatisfied. Not Anthony Chang, MD.

    Chang, chief intelligence and innovation officer at Children's Hospital of Orange County (CHOC), Orange, Calif., is excited about new opportunities he sees at the intersection of data science and medicine. Chang says innovations in artificial intelligence (AI)—computer programs designed with the ability to learn from data—are improving patient care and even taking the grunt work out of care delivery.

    "Intelligence is something we get from the better use of data," says Chang, an internationally recognized pediatric cardiologist who is also founder of the Medical Intelligence and Innovations Institute (MI3) at CHOC. "Data is just the raw material we use to improve decision making and to improve patient care."

    Anthony Chang, chief intelligence and innovation officer, Children’s Hospital of Orange County.

    Anthony Chang, MD, right, chief intelligence and innovation officer, Children’s Hospital of Orange County, with (from left), Spyro Mousses, PhD, executive scientific advisor, Sharon Disney Lund Medical Intelligence and Innovation Institute, and Laura Beken, a manager with the institute. (Photo: Sharon Disney Lund Medical Intelligence and Innovation Institute)

    From word-processing programs that correct misspellings to smartphones that can detect when friends are nearby, AI is becoming an increasingly significant part of professional and personal lives. In health care, this technology is being used in day-to-day decisions on the front lines of care.

    "One example is that telepresence, the collection of data from wearable technology, and artificial intelligence are all converging to basically bring the doctor's expertise to the home in a very non-invasive, continuous way," Chang says.

    Chang and a team of colleagues at MI3 are working on an outpatient-monitoring program that incorporates the use of wearable devices and data analytics to improve outpatient follow-up on pediatric patients with conditions such as heart disease. Many high-risk patients require weekly follow-up. Rather than relying exclusively on clinic visits for these routine checkups, which can be inconvenient and time consuming for patients and their families, the children will be monitored through devices they wear outside the clinical setting. The data from these devices will be analyzed by computer algorithms for abnormalities in the patient's condition, with reports sent to the patient's physician.

    Such technology will enable routine checkups to take place outside the clinic.

    "In the future, the clinic visit will be more substantive in terms of what will need to be done," Chang says. "I think it will increase patient as well as physician satisfaction."

    The outpatient-monitoring program will begin later this year at CHOC, Chang says.

    Picking Up the Pace

    In addition to improving patient access to care and medical expertise, Chang says, AI is changing the pace of innovation within the medical world. Physicians are often unfamiliar with the advances and benefits of AI. "Most physicians are not aware that MRIs can be read better and faster with computer software analysis and, better yet, tied in with elements of medical records," Chang says.

    The latter feature is particularly important because medical records often include information on a patient's medical history, including family history, that can be relevant to the imaging results, Chang says. The imaging results, coupled with anonymized data from a patient's medical history, also can be used to advance medical knowledge through AI software that sweeps the data, searching for patterns that may be relevant to an overall patient population, for instance.

    The availability of big data, the ability to gather the data, and the new ability to automatically analyze the data represent a potent combination for solving the mysteries of medicine, Chang says. "It's going to be an amazing two decades to come in terms of uncovering all the knowledge that previously has not been available."

    Championing Change

    Just how quickly innovations using AI become widespread in the clinical setting depends on the medical profession itself, Chang says. "It could be three to five years. It could be 10 to 20 years," he says. The more physicians embrace AI, the faster the improvements will come. Chang says the rate of acceptance will depend largely on physician champions, who will be called on to act as change agents, eagerly adopting technology innovations and convincing colleagues of the benefits.

    Improved functionality will foster acceptance as well. As AI becomes more embedded in medical technology such as electronic health records (EHRs), which many physicians view as a burden, Chang says the hindrance of such technology will diminish. For example, physicians may spend less time inputting data into the EHR if the input can be done automatically, and would have more face-to-face time with patients as a result. "I think in the future, AI can liberate the physician from computers and electronic medical records because everything will be embedded and more obvious and not be so obstructive in terms of taking care of patients," he says.

    Those who feel threatened by applications of AI, Chang says, are shortsighted about its potential. For example, radiologists should be first in line to learn and appreciate what AI can do.

    "If you're a physician who is afraid of being replaced by computers, then maybe you ought to be replaced because there are so many good things that can come from being in a collaborative mode with computer science," he says. "So it's not an 'us versus them' situation. It's what we can do together with computer scientists that I think is the exciting part of medicine for the future."

    The main downside to AI is that overreliance can lead to errors, Chang says, as any traveler whose use of GPS has resulted in being led to a dead end, or worse, knows all too well. "AI can make mistakes, so you have to have human oversight," he says.

    The ROI of AI

    Of course, another potential downside of new technology is the danger of expending limited resources without reaping tangible benefits—especially in terms of quality and cost. Adopting AI has the potential to control costs, Chang says, because it applies a richer, more comprehensive database of knowledge to a particular problem or task, which in turn fosters solutions more quickly and efficiently.

    A study published in 2013 by researchers from Indiana University's School of Informatics and Computing found that machine learning, a form of AI, can improve the cost and quality of health care nationally. The researchers used an AI framework to show how simulation modeling, which understands and predicts the outcomes of treatment, could reduce healthcare costs and improve outcomes for any diagnosis or disorder.

    The study involved 500 patients with diagnoses of major clinical depression, many of whom also had chronic diseases such as diabetes, hypertension, and cardiovascular disease. The researchers compared actual physician performance and patient outcomes against the use of computerized decision-making models. The cost per unit of outcome was $497 for the physician-directed treatment and $189 for the AI model. The AI approach also produced a 30 to 35 percent improvement in patient outcomes, according to a news release, and by slightly modifying parameters of the computer modeling, researchers found that a 50 percent improvement in outcomes could be obtained.

    Such results demonstrate the potential of AI to transform the industry, Chang says. "The future is really bright with approaching data science in a different way to help patients," he says.

    Related Sidebar: Studies in Innovation: Redirecting High-Cost Care

    Related Sidebar: Studies in Innovation: Treating Stroke On-Scene

    Related Sidebar: Studies in Innovation: Sensing Trouble

    Karen Wagner is a freelance healthcare writer based in Forest Lake, Ill., and a member of HFMA's First Illinois Chapter.

    Interviewed for this article: Anthony Chang, MD, chief intelligence and innovation officer, Children's Hospital of Orange County, Orange, Calif.