Research Studies

Peer-reviewed publications

Peer-reviewed research is essential for improving the quality, accuracy, and safety of healthcare technology. Here we present research papers that showcase the value and impact of Infermedica’s AI-powered solutions for healthcare professionals and decision-makers.


Using AI-based virtual triage to improve acuity-level alignment of patient care seeking in an ambulatory care setting

  • Gellert et al. (2024) / International Journal of Healthcare

A study evaluating the impact of AI-based virtual triage (VT) on patient care-seeking behavior in an outpatient care setting. Researchers analyzed 8,088 online encounters to understand how VT influenced patient behavior regarding care levels.


  • 35% of patients changed their care-seeking behavior

  • 12.5% decrease in video and in-person consultations

  • 19.1% increase in patient engagement through virtual care

The results show how VT helped reduce unnecessary in-person visits, streamline care delivery, and ensure patients received suitable care, potentially improving overall healthcare efficiency in ambulatory settings.

The potential of virtual triage AI to improve early detection, care acuity alignment, and emergent care referral of life-threatening conditions

  • Gellert et al. (2024) / Frontiers in Public Health

A study evaluating the extent to which users intended to seek emergency care when reporting symptoms of five severe/acute conditions to a virtual triage (VT) engine. The five conditions in question have high potential mortality; heart attack, stroke, asthma, pneumonia, and pulmonary embolism (PE). The study highlights misalignments between perceived seriousness and the actual risk of life-threatening symptoms and how VT could be used to address this.


  • 76% of users with symptoms of a heart attack didn't plan urgent care help

  • 33.5% of users with symptoms of investigated acute conditions hadn’t planned medical consultation before completing virtual triage

  • 2 in every 1000 virtual triages performed could lead to life-saving decisions

Early detection of emergency conditions can significantly reduce the risk of severe health outcomes and death. This study shows the strong influence of VT on modifying care-seeking behavior, which could result in reduced healthcare expenses due to timely intervention.

A comparative performance analysis of live clinical triage using rules-based triage protocols versus artificial intelligence-based automated virtual triage

  • Gellert et al. (2023) / Journal of Hospital Administration

A comparison of the triage care referral accuracy of AI-based virtual triage to rules-based triage protocols (RBTP). The study concludes that Infermedica’s AI tools perform comparably to the Schmitt-Thompson protocols — the industry gold standard for patient triage.


  • Infermedica and Schmitt-Thompson produced zero critical mis-triage situations — cases where an “emergency” is identified as “self-care”

  • Infermedica provides triage advice as safe as Schmitt-Thompson

  • Infermedica collects 4x more patient initial symptoms than Schmitt-Thompson for more data-driven assistance

  • Infermedica provides reliable and accurate results and increases triage levels in case of risk factors (especially for children)

How Virtual Triage Can Improve Patient Experience and Satisfaction: A Narrative Review and Look Forward

  • Gellert et al. (2023) / PubMed Central (PMC), National Institutes of Health (NIH)

Review and synthesis of the literature on patient experience and satisfaction as informed by emerging evidence, indicating potential for virtual triage (VT) to favorably impact these clinical care objectives and outcomes.


  • Delivery of earlier and faster, more acuity-level-appropriate care

  • Digital triage as a front door to reduce a patient’s need to visit a medical facility for low-acuity conditions

  • The use of VT to enhance clinical effectiveness through early detection and referral

  • VT as an opportunity for providers to make patient health care experiences more personalized

The Role of Virtual Triage in Improving Clinician Experience and Satisfaction: A Narrative Review

  • Gellert et al. (2023) / PubMed, National Institutes of Health (NIH)

A review and synthesis of evidence on clinician satisfaction indicating a potential for virtual triage (VT) to favorably impact clinician experience, sense of effectiveness, efficiency, and reduction of administrative task burden.


  • Importance of clinician experience and satisfaction to organizational performance

  • Contributors to poor clinician satisfaction

  • Current and potential role of virtual triage in delivering positive clinician experiences and streamlining communication with patients 

  • Benefits of using virtual triage technology for healthcare delivery organizations in general and with respect specifically to improving clinician experience and satisfaction

Cardiovascular Disease Prevention Recommendations From an Online Chat-Based AI Model

  • Gellert, G. A., & Jaszczak, J. (2023) / Journal of the American Medical Association (JAMA)

A response to a recent report discussing how a dialogue-based AI model can assist clinical workflows. The comment and subsequent response from the authors address the safe use of AI and large language models (LLMs) in healthcare.


  • Generative AI vs. Symbolic AI in healthcare applications

  • Potential and limitations of large language models (LLMs)

  • Investment and research into healthcare AI

A Multinational Survey of Patient Utilization of and Value Conveyed through Virtual Symptom Triage and Healthcare Referral

  • Gellert et al. (2023) / Frontiers in Public Health

A study based on an online survey of 2,113 web-based patient-users of Symptomate to describe the use patterns, impact, and derived patient-user value of a virtual triage/symptom checker.


  • 80.1% of patients indicated substantial overall satisfaction with the experience and value delivered

  • 80% of users indicated that they will use triage recurrently in the future

  • 25.4% decrease in post-triage patients who remained uncertain of their care path

The quality of diagnosis and triage advice provided by free online symptom checkers and apps in Australia

  • Australian Department of Education, Skills and Employment

This Systematic Performance Case Study presents the investigation of 36 symptom checkers providing a medical diagnosis or triage advice, where each was tested with 48 medical condition vignettes (1,170 diagnosis vignette tests and 688 triage vignette tests).


Symptomate has the highest accuracy results:

  • 77% correct possible conditions listed in the top 3 results

  • 61% correct possible conditions listed as first

  • 81% correct possible conditions listed in the top 10 results

Diagnostic Accuracy of Web-Based COVID-19 Symptom Checkers: Comparison Study

  • Data Science Department, Symptoma, Vienna, Austria

  • Medical Department, Symptoma, Attersee, Austria

  • Department of Internal Medicine, Paracelsus Medical University, Salzburg, Austria

The study compares 10 web-based COVID-19 symptom checkers by assessing 50 COVID-19 case reports alongside 410 non–COVID-19 control cases. The aim of this study is to evaluate and the diagnostic accuracies of web-based COVID-19 symptom checkers.


  • A good balance between sensitivity and specificity achieved only by two symptom checkers – COVID-RA by Infermedica and Symptoma

  • Infermedica and Symptoma were ranked at the top in terms of the accuracy for COVID-19 evaluations

Virtual Care to Increase Military Medical Centre Capacity in the Primary Health Care Setting: A Prospective Self-Controlled Pilot Study of Symptoms Collection and Telemedicine

  • Singapore Armed Forces Medical Corps, Singapore

The study focuses on virtual care (VC) at a military medical center. 28 patients underwent on-premises VC, comprising of digital symptoms collection and telemedicine, in addition to the usual in-person physician consultation. While waiting for the telemedicine consultation, patients used the Symptom Checker app. This bespoke web application utilized Infermedica API.


  • 85.8% of patients were satisfied with the virtual care experience

  • 85.7% of patients agreed that the symptom checker was easy to use

  • Both patients and physicians felt that the symptom checker allowed for clearer and faster consultations

Heuristic Evaluation of COVID-19 Chatbots

  • University of Luxembourg

The results of a heuristic review of 24 COVID-19 chatbots on different channels (webchat vs messengers) for diverse topics (symptom-checker vs FAQ) and with varying interaction styles (visual-centric vs content-centric vs conversation-centric).


  • Infermedica ranked in the top 3 best scoring chatbots

  • Infermedica as one of the web chat bots with the most attractive graphical user interface

Application of a web-based self-assessment triage tool during the COVID-19 pandemic: descriptive study

  • Provincial Specialist Hospital them. J. Gromkowski, Wrocław

  • Infermedica, Wrocław

  • Department of Infectious Diseases and Hepatology, Wroclaw Medical University, Wrocław

This paper demonstrates experiences with the COVID-19 risk assessment (CRA) tool. The study tries to determine who the user of the web-based COVID-19 triage app is and compares this group with patients in the infectious diseases ward’s admission room to evaluate who could benefit from implementing the COVID-19 online symptom checker as a remote triage solution.


  • The CRA tool screened a vast audience—248,862 people with potentially worrisome symptoms indicating COVID-19

  • Getting results and next-step recommendations provided reassurance and evidence-based information

  • The CRA tool served as a preclinical triage for individuals with worrisome symptoms

How we develop, maintain, and test our technologies

A carefully curated medical information database lies at the very heart of the Infermedica platform. Get a better understanding of the processes used to build, maintain, and test our technologies.

Pipette ilustration
Content Development Process

The creation, validation, and maintenance of the medical knowledge base follow rigorous and well-established procedures.

Acceptance Test Cases

Learn how we test the quality of our inference engine by carrying out assessments on thousands of well-documented clinical cases.


Selected modules marked as class I medical device

Compliant with the General Data Protection Regulation

Compliant with standards for patient data protection

Quality Management System compliant with PN-EN ISO 13485:2016

Processes and Development Procedures certified with ISO 27001:2017

The proof of trust

The excellence of Infermedica’s technology is confirmed by real-life implementations in ecosystems of leading healthcare service providers.

Let's discuss your challenges

Complete the form below to connect with our team of experts on implementing digital primary care solutions. Find the right path to reduce costs and improve patients’ experiences using our AI-powered platform.