CORDIS - EU research results

AI-based infectious diseases diagnosis in seconds

Periodic Reporting for period 1 - EDAS HEALTHCARE (AI-based infectious diseases diagnosis in seconds)

Reporting period: 2019-11-01 to 2020-02-29

Due to the lack of available and reliable means to diagnose respiratory infectious diseases at the primary point of care, all having similar symptoms, physicians commonly prescribe antibiotics based on these symptoms, even for viruses. This results in mistreatment and in approximately 30% to 50% of inappropriate or unnecessary antibiotic prescriptions - both lead to a huge financial and operational burden over healthcare systems due to revisits to GPs, unnecessary admission days, unnecessary lab tests, etc. Anti-Microbial Resistance (AMR) caused by such overuse of antibiotics, is considered to be one of the main causes of death within the coming 20 years. From a clinical perspective, providing the right treatment within the first 24-48 hours of is key to better clinical results, while keeping infectious patients away from the point-of-care has become a necessity in today’s Healthcare delivery environment. EDAS Healthcare devised a Machine Learning based infectious disease diagnostic aid, which serves as a decision support system for physicians and enables accurate, instant and remote diagnosis without a visit to the clinic or hospital. The EDAS solution improves treatment, while reducing costs and overuse of antibiotics. Amid the COVID-19 epidemic, a tool that can allow physicians to remotely diagnose patients, without compromising possible isolation thus preventing additional spread of the virus is a unique and important ability made possible with the EDAS system.
During the project period we were focusing on achieving the main goal of preparing for the EIC (previous called phase 2) application.
1. Implement our solution at the Hadassah Medical Center infectious disease department as a real-time diagnosis service for physicians to be valid for TRL 6 level {TRL 6 – Technology demonstrated in relevant environment (industrially relevant environment in the case of key enabling technologies)}.
2. After securing TRL 6, we analyze the results with a Hadassah Medical center professor and defined the product area needed for improvements to be made for allowing further use at the Hospital.
3. Made all preparations needed for retrospective study at Sheffield Teaching Hospital trust (UK) and based on requirements, plan the clinical test with the community GP by defining a protocol and registered our research in the IRAS system.
4. Hired HaMaDa service to define the right CE classification and additional NHS regulations needed for our solution. The assessment yielded the following classification: “The software provides information used to take decisions for diagnostic or therapeutic purposes. Since the decision is taken by the physician or medical professional the information provided may not cause death, irreversible or serious deterioration in person’s state of health thus the classification is IIa”, according to “MDR 2017/745 annex 8, chapter 3, section 6.3 Rule 11(page 144)”.
5. Patent strategy check and decision to abandon the second patent application, due to patent definitions changes in the US.
6. Based on market research and interviews with leaders and stakeholders from the UK Healthcare market, EDAS will focus on primary care with special emphasis on Telemedicine services, being the only technology allowing accurate infectious disease diagnosis support from remote.
The company follow a direct sales approach to its HDO customers in the first couple of years. This will allow the company to be close to its customers, have a first-hand feedback to improve its solution and better understand its customers’ needs. Nevertheless, the company sees great value in partnerships that can be leveraged to grow its business. Such partnerships and channels to market can be:
• Associations, trusts (UK) and national healthcare entities – Facilitating the reach to public health organizations.
• Tele-Medicine technology providers – Facilitating the use of remote diagnosis by HDOs as part of their technological solutions
• Payers and insurance companies – Endorsing the EDAS solution as a mean to cut its expenditure. These partners may well become our “second wave” customers as soon as the company reaches a significant market share within a specific country.
• Other ad-hoc channels – Facilitating the reach to private healthcare organizations and lobbies.
The EDAS system provides a reliable prediction of infectious disease's causative agent, upon first patient encounter with a physician, with no physical examination. The system will have multiple effects on healthcare systems:
• Allowing remote treatment of respiratory infectious diseases, thus keeping infecting patients away from medical centers and optimizing treatment delivery and costs.
• Reducing overuse of antibiotics by 30% thus contributing to global antimicrobial stewardship targets to reduce antimicrobial resistance.
• Reducing unnecessary lab tests by 35%, thus allowing better economics and better use of resources.
• Reducing unnecessary hospitalization days by 15%, thus contributing to the efficiency of healthcare delivery organizations.

With these effects, the EDAS system will impact the global healthcare system:
• Patients – With a rapid correct diagnosis, patients will be subjected to less unneeded antibiotics leading to significantly better treatments. Patients will have a shorter hospital stay or shorter physician visit time, which will also lower their chance to be infected by additional pathogens such as MRSA (very common in hospitals). Furthermore, patients will also be able to use a telemedicine service for respiratory diseases, eliminating the need to go to a hospital or clinic.
• Physicians – The EDAS system will assist physicians in optimizing their diagnosis, rapidly, which will lower the rate of treatment changes and results with fewer return visits at the primary care and fewer hospitalization days.
• Health Delivery Organizations and Insurers – By lowering the above-mentioned factors (e.g. overuse of antibiotics, unnecessary lab tests and unnecessary hospitalization days) HDOs and insurers will be able to optimize their treatment delivery costs and quality, leading to better healthcare to all. The resulting effect will allow Healthcare systems to better cope with treating the growing epidemiologic events, as seen lately.
• National and local infection disease and control agencies – Will be able to better control and contain outbreaks using accurate, timely and online Live Heatmaps – showing such outbreaks on of a single neighborhood level. This will allow local prevention activities, better track of the spread dynamics of the virus and prediction of the next possible infected area.

Based on our studies results at Hadassah Medical Center, our system can save:
• €380 on the overall patient treatment cost in the Hospital ER by 30% reduction of diagnostic mistakes leading for extra hospitalization time, extra lab tests and extra medical treatment (€340M/annually for the EU).
• €63 per GP visit by reducing 30% of unnecessary antibiotics and preventing return visits (€471M/annually for the EU).
• Telemedicine service is an evolving technology and specific numbers are not yet available, however, researchers estimate that in the US the number of virtual visits will surpass physical visits to GPs by 2024, and is estimated to be over 600M annual visits. Such estimations may even grow with recent Coronavirus outbreak and realization of the value of remote treatment.