Estimates suggest that the United States wastes more than $1 trillion each year on health care (one-third of all health care spending). This waste includes over-treatment, seeing multiple doctors unnecessarily, care delivery failure and lack of care coordination.
Unfortunately, health care cost-cutting measures have historically made matters worse in the United States, often cutting back access to care in a misguided race to the cost floor.
A perfect example of these cost-cutting measures is the creation of narrow network plans, which limit choice and network access in an attempt to reduce costs. According to McKinsey, 70 percent of plans on the Affordable Care Act exchanges in 2014 featured a limited network, and narrow-network plan premiums were on average 17 percent cheaper.
While less choice does typically mean lower premiums, often patients are deprived of seeing the physicians and services they need in order to receive appropriate care. There should be a better way.
How then can we cut costs while designing a more responsive, efficient and effective health care system?
Data analytics: The path forward to increasing high-value care
In the United States, more than 1.2 billion clinical documents are produced each year, and we have barely mined this information treasure trove for insights. With data analytics and artificial intelligence, we can demystify health care spending, design more practical provider networks, and help providers more wisely sift through care delivery options.
Here are four ways in which data analytics and AI can help bend health care’s cost curve while improving quality.
1. Suggesting the most effective care providers
A 2014 United States Preventive Services Task Force noted that imaging tests to evaluate headaches were a common culprit contributing to millions in wasted health care dollars while offering little value to patients’ health outcomes.
Imagine if, from the moment the patient with acute headaches began interacting with the health care system, their specific experience was matched with the doctor most knowledgeable about their condition who could then better filter out unnecessary tests.
Hospitals and provider networks track the health and wellness of their patient populations and have data about their patient mix based on their medical history. Leveraging data analytics, physicians can then be matched with patients based on past success in treating similar patient profiles.
This acknowledges the unique needs of individual patients, matches them to the unique capabilities of individual doctors, and reduces medical care runaround that drives up non-essential spending.
Imagine a physician asking a voice-activated medical AI (think Amazon Alexa) about the preferred, least costly surgical treatment for her patient. The artificially intelligent assistant returns the best matched surgeons, medically clears the patient for surgery, and software conducts a real-time auction among preferred surgeons. The surgical center then has options to find the best price for the patient based upon patient profile and risk.
2. Creating more detailed patient portraits
Currently, documentation for medical encounters remains effectively locked up within each hospital, clinic, surgical center, office and urgent care facility. Efforts to share and aggregate medical records for an individual are limited, but by aggregating records and using computers to decipher content, we could assemble full individual portraits of patient care.
With data from different sources, we can paint a more complete picture and personalize their treatment. Doctors would no longer have to guess at and fill gaps in documentation.
3. Identifying the best course of treatment
Machines are able to sift through millions of health records to learn what treatments work in similar patient profiles. Rather than depending on narrowly-designed studies, health care providers can base treatments and outcomes on data relevant to individual patients, their environments and how they live their lives
4. More efficiently leveraging health care’s most valuable resource: providers
With the rise of data analytics and AI, there will be a greater need for primary care physicians and less demand for some types of specialists as machines will play a greater role in analysis and diagnosis.
Certainly, the next generation of doctors and health care providers will need to understand the underlying mechanisms of health and disease and the basis for treatment, focusing on delivering what computers cannot: guidance, counseling and advocacy to help patients in the best way possible
But in this new world of health care, computers will be a vital aid in care delivery, automating processes and making sense of data. Meanwhile, physicians will have more time to lay hands. Given the high rate of burnout among physicians today, this should come as a welcome development.
In the next 20 to 30 years, data analytics and AI have the potential to transform the practice and consumption of health care from more of an art to a science and finally reach the utopia of high-quality, low-cost care.
The assumption in U.S. health care of tradeoffs between cost and key outcomes no longer has to be accepted. The world’s most expensive health care system, with AI, can become its most successful.