AI Symptom Checkers Market Map: Guiding Patients to the Right Care
This is part of Elion’s weekly market map series where we break down critical vendor categories and the key players in them. For more, become a member and sign up for our email here.
Navigating the healthcare system can be overwhelming, particularly for the 1 in 3 Americans who don’t regularly see a PCP. Many who experience symptoms end up in emergency rooms for non-urgent conditions or delay care for more serious issues, not knowing where to go. Considering only commercially insured patients, an estimated 2 out of every 3 emergency department visits are avoidable, and could be handled in an office setting at roughly 10% the cost.
For healthcare organizations, this presents an opportunity: AI symptom checkers integrated into provider organizations’ websites can triage patients effectively, reducing the burden on emergency departments while improving patient satisfaction and decreasing revenue loss on unpaid bills.
What Are AI Symptom Checkers?
AI symptom checkers are digital tools that are embedded into provider organizations’ websites and use artificial intelligence to assess patient-reported symptoms and guide patients to the right care setting—whether it’s self-care, telehealth, or an in-person appointment. Unlike B2C products offering a similar user experience (which monetize by charging consumers for telehealth appointments), our category refers to B2B solutions designed to act as a digital front door for health systems and providers.
How AI Symptom Checkers Work
Traditional symptom checkers often used basic decision trees, which can lead to generic or inaccurate advice. The newer generation of models might use LLMs to process user input into structured clinical formats like UMLS/SNOMED, then traverse a knowledge graph of symptoms and potential diseases, with a separate agent asking follow-up questions to help narrow down the diagnosis, ultimately arriving at a list of likely diseases and their associated triage levels. These new architectures offer several advantages over traditional symptom checkers:
Improved accuracy based on intent identification and similarity search
Personalization, using conversation history and context summarization
Efficiency, reducing response time and enhancing user experience
It’s important to acknowledge, though, that existing solutions are not yet perfect. A 2022 study that looked at a variety of tools found that diagnostic accuracy was only 19-37.9%, however triage accuracy was 48.8-90.1%. In other words, the tools are not very effective at diagnosing specific issues, but are better at recognizing severity and what type of care is needed. We’ve seen similar results across other studies, as well.
As such, it’s important to consider these tools part of a broader digital front door strategy, that does a tolerably good job of triage, without functioning as a true diagnosis or clinical decision support product. Instead, good solutions—for now—must focus more on connecting symptom assessment with other systems such as EHRs, scheduling workflows, and telehealth platforms. Their other benefit is patient access outside of business hours and reduction in call volume.
Vendor Landscape: Key Players and Differentiation
There are two main groups of products within this category. The first are independent platforms that providers and health systems can partner with for referrals. For example, Buoy Health combines symptom checking with patient education, delivering personalized care recommendations and guidance in addition to provider referrals. Ubie is primarily a mobile app provider organizations can partner with to be referred new patients. Ubie can also be embedded on clinics’ websites, to serve a similar function to the platform solutions described below.
The second group of products are bought by providers to be embedded within their own websites and potentially EHRs, and many are part of a broader contact center or digital front door solution. These include: Clearstep, Ada, Fabric, Loyal, Infermedica, Mediktor, Talkdesk, and Xund.
Where AI Symptom Checkers Are Headed
It’s no secret that technology’s ever-increasing capacity to collect and process data is rapidly changing every aspect of the patient experience. The “digital front door” subset of AI symptom checkers are just one component of the growing shift toward integrated care ecosystems. Here are a few of the directions we see this category going:
As AI models improve and the market matures, we expect these tools will offer increasingly precise recommendations with the potential to integrate data from wearable devices, personal and SDoH information, and even EHRs if there’s an appropriate authentication mechanism.
Solutions could also analyze patient inputs over time, aiming to predict health trends and alert providers to outbreaks of particular illnesses, bolstering data coming from population health data and analytics and scheduling optimization platforms.
Given the disparate number of uninsured patients seen in the ED, we wouldn’t be surprised to see symptom checkers work alongside cost estimate tools to help uninsured and underinsured patients select the most cost-effective care context.