How AI Can Reduce Costs in Population Health

By Infinx
June 20, 2018

Artificial intelligence (AI) can reduce costs in population health. That’s not a statement to be taken lightly. In fact, it is a statement of towering proportions. Why? Because the health of the population drives healthcare costs. Reducing the costs of population health reduces costs of the U.S. healthcare system at large — the goal of nearly every national initiative being implemented today.

First, let’s define population health. It is generally regarded as the health of a group of individuals, grouped by demographics, geographic area, community or nation. It also includes the distribution of health and health outcomes within those groups. So, for example, artificial intelligence holds the potential to reduce costs for the U.S. population and to improve outcomes within that group. That’s powerful potential.

Improving U.S. Population Health Means Addressing Chronic Disease

We know that chronic disease causes the majority of deaths in the United States. Nearly 900,000 deaths annually are due to five things:

  1. Heart disease
  2. Cancer
  3. Lung disease such as emphysema and chronic bronchitis
  4. Stroke
  5. Unintentional injuries such as those on roads or caused by medication overdoses.

According to the CDC report on potentially preventable death, more than a third of those deaths are preventable.

Here’s the problem. Until recently, the United States has largely ignored population health as a factor in improving the U.S. healthcare system. It’s one reason why we have the highest healthcare costs and worst healthcare outcomes of 11 developed nations. We spend $3 trillion on healthcare (16.3 percent of our GDP), but only nine percent of our GDP on things that directly impact population health: social services, employment programs, and support housing.

Artificial Intelligence Pinpoints Targets

Artificial intelligence can reduce costs by identifying precise demographics and geographic locations where population health issues exist. Within the United States, there are neighborhoods, communities, and rural areas where specific health issues occur frequently. When AI reviews large sets of data, healthcare experts can easily identify populations that are at risk for chronic disease caused by environmental, socio-economic, or accessibility issues. When AI can pinpoint those problems, treatment programs and education can be targeted to address them head on. That can improve outcomes while reducing costs.

The challenge with population health is that it can be very expensive to influence and improve. For example, one smoking cessation campaign launched by the CDC cost $48 million. Artificial intelligence can put a finer point on the issue and discover what populations are smoking the most. What ages and demographics are still taking up smoking? Rather than launching a national campaign, educators would be able to focus education or treatments in the neighborhoods, counties, states, or age groups with the highest prevalence rates.

Take that example and apply it to heart disease, diabetes, obesity, and other chronic conditions and you can see how targeted treatments could begin to improve outcomes and reduce costs. It is a vast improvement over the current method of one doctor, hospital, or health system connecting one patient at a time to community support services.

Artificial Intelligence Saves Time

One of the biggest and most expensive problems in healthcare today is the amount of time that physicians are required to spend on administrative tasks — a minimum of 8.7 hours a week. It’s one reason why the United States spends so much money on healthcare. It creates undue pressure on physicians and takes them away from patient-facing time. The more that AI can conduct administrative tasks, the more time and money physicians save. The more that AI can analyze test results, the more efficient physicians can be in ordering expensive imaging, and that reduces costs.

At this year’s World Medical Innovation Forum, talk centered around the use of AI to improve physician’s decision making through data. Associate Director of the Healthcare Transformation Lab at Massachusetts General Hospital Maulik Majmudar said, “Despite all the progress we’ve made over the past century, there is still a tremendous amount of scientific uncertainty. Only about 20% of all medical decisions are actually based on high-quality evidence.”

Ziad Obermeyer, MD, assistant professor of emergency medicine at Brigham and Women’s Hospital, illustrated how AI could make imaging more efficient and cost effective. “Low-risk patients are getting over-tested, and high-risk patients aren’t getting tested enough. The fact that nobody is getting this balance right is the key to seeing how an algorithm can do so much better. If an algorithm was making the decisions, we could cut tests by about 40% and still find about as many patients who will go on to have cardiovascular interventions.”

These are just a few of the ways in which AI can improve population health and reduce costs. It’s also the reason why in a recent survey, more than one-third (39%) of respondents rated AI as the most significant health IT topic of 2018.

About the Author



Infinx provides innovative and scalable payment lifecycle solutions for healthcare practices. Combining an intelligent, cloud-based platform driven by AI with our trained and certified coding and billing specialists, we help clients realize revenue, enabling them to shift focus from administrative details to billable patient care.

Leave A Comment