{"id":22256,"date":"2022-10-19T09:15:50","date_gmt":"2022-10-19T13:15:50","guid":{"rendered":"https:\/\/healthit-gov.go-vip.net\/buzz-blog\/?p=22256"},"modified":"2026-01-17T21:36:09","modified_gmt":"2026-01-17T21:36:09","slug":"ai-ml-in-health-care","status":"publish","type":"post","link":"https:\/\/healthit.gov\/blog\/ai-ml\/ai-ml-in-health-care\/","title":{"rendered":"Two Sides of the AI\/ML Coin in Health Care"},"content":{"rendered":"\n

As we\u2019ve <\/span>previously discussed<\/span><\/a>, algorithms\u2014step by step instructions (rules) to perform a task or solve a problem, especially by a computer\u2014have been widely used in health care for decades.  One clear use of these algorithms is through evidence-based, <\/span>clinical decision support interventions<\/span><\/a> (DSIs). Today, we see a rapid growth in data-based, predictive DSIs, which use models created using machine learning (ML) algorithms or other statistical approaches that analyze large volumes of real-world data (called \u201ctraining data\u201d) to find patterns and make recommendations. While both evidence-based and predictive DSI types (models) could be used to address the same problem, they rely on different logic that\u2019s \u201cbaked into\u201d their software.\u202f<\/span> <\/span><\/p>\n\n\n\n

But before we explore the two approaches, we should first revisit a key challenge posed in our previous post in this series: capitalizing on the potential of artificial intelligence (AI), particularly ML and related technologies, while avoiding risks (such as potential harm to a patient) from these technologies. In this blog post, we\u2019ll dig a little deeper into what some of those risks are and where those risks can originate.  <\/span> <\/span><\/p>\n\n\n\n

Evidenced-Based vs. Predictive Decision Support Interventions\u202f\u202f<\/span><\/b> <\/span><\/h2>\n\n\n\n

DSIs that <\/span>use evidence-based guidelines or other expert consensus<\/span><\/a> generate recommendations based on how the world <\/span>should work<\/span><\/b>. Generally,<\/span> they represent the implementation of expert consensus emerging from high-quality clinical trials, observational studies, and other research. Evidence-based DSIs are usually \u201cfixed rules,\u201d essentially, a series of \u201cif-then\u201d statements that form an algorithm. For instance, \u201cif a woman is between the age of 45-54 and if she is of average risk of breast cancer, then she should get a mammogram every year.\u201d\u202f<\/span> <\/span><\/p>\n\n\n\n

Predictive DSIs, by contrast, generate recommendations (outputs) to support decision-making based on recognized patterns in the way the world <\/span>actually works<\/span><\/b>,<\/span> filling in knowledge gaps with real world data. It\u2019s up to humans then to determine the recommendation\u2019s relevance in a given context.<\/span> This makes predictive DSIs powerful tools because they can, at least in theory, be used to predict anything about which the technology collects data\u2014whether that image looks like a tumor, whether a patient is likely to develop a specific disease, or whether a patient is likely to make it to their next appointment, to name a few. In part, because expert clinical guidelines have not been established for many topics, predictive DSIs can provide important guidance on a wide range of topics that evidence-based DSIs currently do not touch. <\/span>At their best<\/span>,<\/span><\/a> predictive DSIs can identify patterns in data earlier or more precisely than health care professionals, or even uncover patterns not previously known, and recommend decisions across many facets of health care.\u202f<\/span> <\/span><\/p>\n\n\n\n

Magnifying Existing Risks Resulting from Emerging Technology\u202f<\/span><\/b> <\/span><\/h2>\n\n\n\n

While predictive DSIs have enormous potential to improve many aspects of health care, they also present several potential risks that could lead to adverse impacts or outcomes. These risks may be magnified because of their potential to \u201clearn\u201d rapidly and produce predictions across many hundreds or thousands of patients. In particular, predictive DSIs in health care can:\u202f<\/span> <\/span><\/p>\n\n\n\n