Using artificial intelligence to solve public health problems

According to the most recent data from the Centers for Disease Control and Prevention (CDC), approximately 2.6 million people die annually in the United States. Data suggests that about one-third of these deaths could be prevented through efforts such as population health programs.

The main causes of death—heart disease, cancer, respiratory diseases and stroke—have all decreased this decade partly due to aggressive education and awareness programs on topics such as smoking cessation, nutrition and exercise. Yet convincing a nation to stop smoking, eat better and exercise more takes significant effort and dollars. Just one smoking cessation campaign, featuring emotional pleas from ex-smokers suffering from throat cancer and other ailments, cost $48 million, according to the CDC.

One reason for the high cost of population health programs is the broad reach of education campaigns. By pinpointing specific demographics or geographies where population health issues exist, artificial intelligence (AI) and machine learning can help to target and precisely implement education and treatment programs and reduce spending waste.

AI enables computers to mimic the cognitive function of human minds, and machine learning gives computers the ability to learn without being explicitly programmed. By using AI and machine learning to review vast sets of real-time data, health experts can identify at-risk populations for any number of diseases, from diabetes to heart disease.

Opioid addiction is one area that can benefit from AI and machine learning. Prescription and illicit opioids are the main drivers of drug overdose deaths in the United States. According to the CDC, opioids were involved in more than 42,000 deaths in 2016. That’s two-thirds of all drug overdoses, and five times higher than in 1999, which is why opioid addiction is now considered an epidemic. On average, 115 Americans die every day from an opioid overdose.

While opioid overdoses are a problem across the country, certain counties, states and regions experience significantly higher rates of opioid abuse and misuse than others. In 2016, the five states with the highest rates of death due to drug overdose per 100,000 residents were West Virginia (52.0), Ohio (39.1), New Hampshire (39.0), Pennsylvania (37.9) and (Kentucky (33.5), while states such as Nebraska (6.4), South Dakota (8.4) and North Dakota (10.6) had significantly fewer deaths per capita. Limited dollars for a public education program to combat opioid misuse and abuse are better spent in areas with higher overdose rates.

In Indiana, where more than 1,500 people died from drug overdoses in 2016, the Indiana Management Performance Hub (MPH) is using the hc1 Opioid Dashboard, a new solution that uses AI and machine learning to gain live insight into opioid usage trends across the state. The hc1 Opioid Dashboard organizes and analyzes billions of anonymous lab test results from diagnostic labs, government testing databases, employers and providers to identify the drug positivity rate down to the zip code level. This allows officials to proactively implement programs where they are needed most, before a public health crisis explodes.

Indiana has earmarked an estimated $100 million to fight opioid abuse by implementing public education programs, educating prescribers, equipping law enforcement with the overdose reversal drug Naloxone, and establishing specialized drug treatment centers quickly in the areas that are most in need. The hc1 Opioid Dashboard allows officials to identify potential hot spots of abuse early on, while measuring the impact of programs aimed at prevention and treatment. By tracking progress, public health officials, program managers, and law enforcement can pinpoint areas that need improvement and direct funds to those areas.

Indiana officials recognize that opioid misuse and addiction—like other population health issues—is an incredibly complex and expensive issue to address. Key factors contributing to the problem include abusers actively seeking access, physicians over-prescribing, illicit dealing, and increasing availability of lethal synthetic drugs including fentanyl. While prescription drug monitoring programs (PDMP) have been helpful in reducing unnecessary prescribing, a multi-faceted approach is required to tackle this epidemic.

Ultimately, to meaningfully decrease the devastating impacts of opioid misuse and abuse, along with other public health issues, a targeted, data-driven approach that is fiscally sound will be required. AI holds great promise to make population health programs more targeted to achieve this goal.

Brad Bostic is chairman and CEO of

The views, opinions and positions expressed within these guest posts are those of the author alone and do not represent those of Becker's Hospital Review/Becker's Healthcare. The accuracy, completeness and validity of any statements made within this article are not guaranteed. We accept no liability for any errors, omissions or representations. The copyright of this content belongs to the author and any liability with regards to infringement of intellectual property rights remains with them.

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Whitepapers

Featured Webinars