Avey’s research team, consisting of experts in medical and computer sciences, conducted the study using the standard clinical vignette approach established by the Harvard Medical School for this kind of studies. The approach requires that external medical doctors with certain characteristics test the symptom checkers using gold-standard vignettes (i.e., anonymized patient cases) and record all their findings. All the vignettes were made publicly available to facilitate the reproducibility of the study and support future related studies. Furthermore, all the recorded findings were posted online to establish a standard of full transparency and allow researchers across the globe to cross-validate the results, a step much needed in health informatics.

The study compared six symptom checkers: Avey, Ada, WebMD, K Health, Buoy, and Babylon. The results showed that Avey, an AI based model, was the most accurate among the six symptom checkers. To this end, Avey’s performance was compared against 3 primary care physicians with an average experience of 16.6 years. Avey slightly outperformed this committee of experienced physicians in some accuracy metrics, including the ability to rank diseases correctly within the generated differential lists.

The increasing popularity of symptom checkers stems from the earlier trend of investigating the symptoms of search engines such as google, which is a great educational source for most topics. However, search engines cannot mimic clinical reasoning thus lack credibility when it comes to medical expertise.

The symptom checkers, on the other hand, can emulate clinical reasoning, especially when they use AI. The AI-based ones are trained to make medical experts–like judgments on behalf of patients. In particular, a patient can start an assessment session with a symptom checker via inputting a chief complaint.

Akin to human doctors, the symptom checker will ask the patient a few questions before generating a differential diagnosis that explains the causes of the patient’s symptoms.Avey stood out as the most accurate symptom checker tool in the study due to its advanced AI diagnostic model that uses a proprietary probabilistic graphical model, namely a Bayesian network. The model was fully researched and built from scratch at Avey for more than 3.5 years.

“Symptom checkers are a revolutionary tool that makes the healthcare system more accessible and reliable in an era in which information is abundant. It empowers the healthcare system to make better-informed decisions, improves patient-provider communication, and puts the patient in control of their health. Nonetheless, while they offer a helpful initial step in symptom assessment, they shall not substitute for professional medical evaluation.” ~ Dr. Mohamad Darmach, Doctor at Avey shared.

The study conducted by Avey is now considered the most comprehensive in the field. It does not only position Avey as the most accurate AI-based symptom checker worldwide but also highlights the potential of AI in transforming healthcare. The future of healthcare is leaning towards patient-centricity, allowing patients to take control of their health journey. Avey is serving in this transformation and promising to empower health.

About Avey

Avey is a deep AI company dedicated to transforming health via unlocking the full potential of clinical care and making every health journey delightful. It empowers physicians with proprietary, clinical-grade AI that provides instant diagnostic support, recommends personalized care plans, automates medical documentation, and accurately codes and adjudicates medical concepts, right at the point of care.