Efficiency in the United States healthcare industry has been experiencing a slow but certain decline over the last seventy five years, and patients seem to be paying the price in more ways than one. According to the Organization for Economic Cooperation and Development (OECD), healthcare-related costs in the U.S. now account for roughly 16% of the nation’s gross domestic product, and healthcare spending per capita is nearly twice that of other industrialized countries. As the cost of patient treatment has continued to increase, the average life expectancy of Americans has subsequently begun to fall. Despite the outdated nature of our healthcare system, however, there may still be hope for reform in the near future thanks to recent developments in artificial intelligence (AI). Tech companies are beginning to create machine learning algorithms that can reverse the downward trajectory of productivity in the healthcare industry by offering solutions to drastically improve how we diagnose, treat and prevent disease. The potential for accurate diagnoses and precise predictions with a near-zero margin of error has understandably drawn attention towards the application of artificial intelligence in every facet of the healthcare industry. In addition, an estimated 14% of unnecessary health care spending — nearly $91 billion — is the result of inefficient bureaucracy. AI can be utilized to automate a large percentage of the industry’s current administrative tasks which often have little to do with the treatment of patients, thus saving money for healthcare organizations. But perhaps the single most promising feature of artificial intelligence in regards to its application in the medical field is its ability to help patients on a day to day basis with notifications and reminders about things like timely prescription refills, taking the right medications at the correct times, and getting enough exercise. By constantly monitoring vital signs for sudden changes, machine learning algorithms can provide patients with recommendations and interventions in real-time.

Remote home health monitoring isn’t a particularly new concept, but rather, it’s a subdivision of telehealth that includes the collection, evaluation, transmission, and communication of patient medical data from electronic devices. These remote devices include wearable sensors, implanted equipment, and handheld appliances. A number of distinguished American healthcare institutions including United Healthcare, Partners Healthcare, and the Johns Hopkins School of Medicine have conducted pilot studies on individual health monitoring in recent years which have largely yielded positive results. However, these American studies have yet to utilize the full potential of AI with regards to making split-second judgements on behalf of patients based on perpetual data analysis.

Outside of the United States, however, the integration of AI into the realm of healthcare and patient data analysis has already begun. Under a pilot program conducted by the National Health Service (NHS) in southeast England, a group of hospitals that served roughly 500,000 people collectively began sending patients home with Wi-Fi enabled, fitted armbands following their initial treatment. The armbands remotely monitored the patient’s vital signs, including their oxygen levels, blood pressure, respiratory rate, internal body temperature and pulse. According to the NHS, this led to a significant decrease in hospital readmission rates and emergency room visits. In addition, the number of costly home visits dropped by around 22%, and long term adherence to prescribed treatment plans increased to 96% – nearly twice the industry average. 

Prior to the implementation of this pilot program, staffers were dispatched to drive up to an hour and a half round-trip for home monitoring visits and patient check ups several times a month. However, with machine learning algorithms actively monitoring and analyzing patient data, and instantly notifying doctors and patients about sudden changes, patients were able to receive the best care possible at a lower cost to their healthcare provider. 

Through the application of artificial intelligence, this National Health Service pilot program was able to expand out of England and make its way into international markets. The company that created the AI-powered remote monitoring devices used in the program – Current Health – received clearance from the FDA last year to run a pilot of the system in the United States. In a statement from Current Health following the U.S. pilot approval, the company wrote: “The post-acute FDA clearance comes as Current Health is experiencing overwhelming customer demand for its platform, combining its all-in-one wireless wearable, which provides ICU-level accuracy and tracks more vital signs than any other all-in-one wearable available today, with analytics to derive actionable insights.” Trials are currently underway at Mount Sinai Hospital in New York City as part of an effort to mitigate patient readmissions, which cost hospitals in the U.S. an average of $40 billion per year.

The overt success of the NHS program illustrates a key concept regarding the employment of AI into the emerging patient-centered medical marketplace. By focusing on potential solutions to individual issues within the current healthcare sphere  – for instance, developing a machine learning algorithm to reduce expensive hospital readmission rates – companies can actually solve multiple problems simultaneously. In the case of Grady Hospital, the largest public hospital in the state of Georgia, the application of an artificial intelligence software which identified at-risk patients and alerted designated clinical teams saved the hospital $4 million over the course of two years by reducing readmission rates by more than 30%. Hospitals are initially financially incentivized to adopt these AI tools, but the enhanced patient treatment and experience that follow as a result are arguably even more important.

Improving efficiency within the healthcare industry using artificial intelligence will come with its own set of unique challenges, but the advancements that have already been made possible with AI are promising enough to warrant overcoming those obstacles. Healthcare organizations must work together towards simplifying and standardizing data and processes in order to ensure compatibility with AI algorithms. Additionally, organizations within the industry would be wise to reevaluate the effectiveness of certain positions that don’t improve patient outcomes and add little value to the business. 

Digital transformation and innovation in healthcare has come about much more incrementally and reluctantly than in most other massive industries in the United States. However, by allowing AI to assist in administrative and analytical capacities within the medical field, wellness providers can devote more time and resources to patient care and treatment, potentially saving lives at a substantially lower cost to healthcare institutions.