The recent flurry of press generated by last week’s Consumer Electronics Show (CES) has me thinking a lot about technology and its role in healthcare and improving health outcomes. Health-related technology was front and center at last week’s show as more than 100 companies displayed their latest health-related devices, software and services in the hope of gaining exposure and credibility in the crowded health-tech market. For example, consumer “wearables” were bigger than ever this year. Devices to monitor your ECG, hydration levels, brain activity, skin pH and even a device to measure baby diaper activity were all introduced. A new device from Black & Decker was particularly interesting…a voice-recognizing digital assistant optimized for home healthcare. Time will tell if it is effective, but I appreciated the effort to help seniors age in place rather than a nursing home or assisted-living center.
Obviously, CES had many innovations and cool products…some of which may have a positive impact on healthcare delivery. Unfortunately, however, products that will deliver on the goals of improving health and reducing costs are still mostly a dream. As stated in a recently published research article in Health Affairs, there is little evidence that leading digital health companies have yet demonstrated a substantial impact on disease burden or cost in the US healthcare system. Maybe next year?
AI Today
Moving beyond CES, another tech development of interest is the growing use of Artificial Intelligence (AI) in healthcare. Cleveland’s MetroHealth is using AI to reduce the impact of no-shows and improve access to provider appointments. Cleveland Clinic is using AI to track hospital bed use and to increase surgical suite capacity and utilization. Others are using AI to identify possible misuse of drugs and devices in health facilities, apply automatic customization of patient payment plans to ability to pay, identify high-utilizing patients, and match patients with appropriate specialists. AccuHealth collects patient data using sensors in the home and uses AI to build predictive alerts. When an alert is triggered, remote providers can provide an earlier intervention. AxisPoint Health uses claims and consumer data to identify proxies for social determinants of health. This data is then digitally scrutinized and used to provide care managers with profiles and predictions for patients and populations. These are just a sampling of the application of AI in healthcare.
But while AI is proving to be especially effective for health administration and predictive analytics, its widespread use for imaging and diagnostics is still in development. For example, IBM’s Watson supercomputer was trained to diagnose and treat real cancer patients. In spite of massive efforts to develop the supercomputer’s clinical abilities, the results have been underwhelming and sometimes wrong. Experts think better integration with medical records and more data on real patients are needed to make Watson better for healthcare. For now, it appears the human element provided by doctors, perhaps assisted by AI, is the preferred and safer route. A final note here is that AI has made inroads on very specific diagnostic scenarios. Recently, for example, IDx-DR became the first autonomous AI to be commercially authorized by the FDA for diagnosis of diabetic retinopathy.
Tying to Genetics
Another area where technology is making inroads in healthcare is in the field of genetics. We know that genetic predisposition drives a significant portion of health problems and premature death in the US. Performing genetic screening on individuals and populations accelerates the ability of researchers to access and apply data for health management. An exciting example is the Healthy Nevada Project led by Renown Institute for Health Innovation in Reno, Nevada. Renown has partnered with the Desert Research Institute to analyze and model public health risks in Nevada and aims to serve as a national model for future population health studies working to improve overall health through clinical care integration. The Healthy Nevada Project combines genetic, clinical, population and environmental data into one of the world’s fastest-moving population sequencing projects. Given Reno’s locally high incidence of death due to heart disease, cancer and chronic lower respiratory disease, the search for non-behavioral causes is considered a critical part of the disease management paradigm.
Technology vs. the Human Touch
The irony of all this technological development and the cost associated with it is that the largest drivers of impaired health and premature death in the US are individual behavioral patterns — especially obesity, inactivity and smoking…very low-tech conditions! Given the human factors associated with these conditions, can technology alone help us to eradicate their behavioral causes? The short answer, according to many studies, is no. However, that is not to say there is no role for technology in improving these personal behaviors. Experts tell us that technology can be highly effective in supporting behavioral change when integrated into a well-planned system of care. Whether caring for an individual patient or a population of thousands, the system of care must be designed to holistically address the health needs of patients. Health needs are driven by genetics, social circumstances, environmental exposures, behavioral patterns and health care delivery. As illustrated above, technology can certainly aid in some fashion in each of these domains. However, systems of care designed and operated by humans are required. Ultimately, contact between patient and caregiver is what will likely drive the changes necessary to proactively address and reduce the incidence of poor health. If only there was an app for that!