Mediktor COVID-19 Online Assistant

What did you do the last time you felt symptoms? You rushed into the hospital, you scheduled an appointment with an specialist, you googled them, you called your mom,... We are certain that 70% of the time people make wrong decisions that may be unnecessarily expensive, inconvenient and clinically unsafe. 

Unlike past highly infectious virus outbreaks such SARS and MERS, nowadays people can count on Artificial Intelligence (AI) and Machine Learning (ML) to better manage COVID-19.

Mediktor aims to screen patients massively, identify those with risk of COVID-19 infection and inform them what are next steps. Always according to the federal and healthcare organizations protocols and rules.

WHO declared COVID-19 a pandemic as of 03/11/20. Our AI/ML-based solution for self-triage is the only way to screen, identify and advice population on a large scale without collapsing health systems' facilities of any given region.

The scale of the problem is enormous and affects individuals globally. Latest data includes over 2,000,000 cases in USA, 235,000 cases in Italy, 178,000 cases in Iran and 15,000 cases in South Korea. In some countries, cases are doubling every 24 hours and contention measures don't seem to be as successful as expected.

The deployment of our solution addresses the lack of real-time data on origin and the lack of personalized recommendations about next steps. Such insights should be used by health systems and government agencies to better plan the fight against and provide patients with on-demand healthcare guidance at the first onset of symptoms on a massive scale and automatically.

Mediktor is the most advanced and accurate AI based technology for triage, pre-diagnose and decision-making support. By combining multiple technologies and techniques such as AI/ML, Natural Language Recognition (NLR), Mediktor replicates physician’s logic and acumen to assess symptoms and come up with preliminary conclusions on potential conditions. Once potential conditions are identified, clients’ protocols and rules are applied to connect patients with the most appropriate level care. Under any circumstance Mediktor replaces physicians/clinicians but gathers and analyzes information that enables smart patient routing and enhances patient/physician healthcare experience.

Mediktor provides a highly accurate differential diagnosis including the likelihood of each listed condition. We differentiate between primary conditions (most likely) and secondary based on a probability scoring model. We also identify and select relevant symptoms that may have an impact on the level of urgency and the recommendation of next steps. In the unlikely event that a specific condition may not be identified (vague, unspecific symptoms with little to no additional symptoms and signs) a level of urgency is always indicated based on the symptoms reported. The list of potential conditions and the associated symptoms that led to that conclusion is always shared with clinicians providing the remote consultation services.

Our target populations is humankind. We strongly believe that consumers have the right to access healthcare on their mother tongue and using their preferred interface.

This is the reason why Mediktor is available in 12 languages and dialects: English (UK), English (USA), Spanish (ES), Spanish (MX), Protuguese (BR), French, Italian, Greek, German, Chinese (simplified) and Catalan.

Also, Mediktor is available in multiple interfaces including web, app (iOS and Android), chatbot and voice.

COVID19 is here to stay for a long time until a vaccine is commercialized ( 2nd semester 2021)

For the time being, healthcare systems must take measures to prevent COVID19 from spreading worldwide again. The only way to control COVID19 is through AI-based technology solutions like Mediktor for early detection of people at risk for COVID19 that will play a key role in this first line of defense.

Testing people for COVID19 is expensive and not as scalable as an AI-based solution. Clinical study shows that 8/10 COVID19 cases can be predicted without test by using AI/ML: https://www.ictandhealth.com/n...

The list of our competitors includes Infermedica, WebMD and Gyant. There are a number of features that allow Mediktor stand out of its competitors:

Clinical study - Mediktor is the only AI-based technology for triage, pre-diagnose and decision-making support scientifically validated to be 91.3% accurate.

Differential diagnosis - Mediktor provides a highly accurate differential
diagnosis including the likelihood of each listed condition. We
differentiate between primary conditions (most likely) and secondary
based on a probability scoring model. We also identify and select
relevant symptoms that may have an impact on the level of urgency and
the recommendation of next steps. The choice to share with the users the
list of potential conditions is up to our clients. In any case, the
list of potential conditions and the associated symptoms that led to
that conclusion is always shared with clinicians providing the remote
consultation services.

Care Navigation - Once potential conditions are identified, Mediktor enables the health system to direct the patient to the most appropriate level care always according to the health system’s protocols and rules - or CDC guidelines by default.

Virtual visits - Synchronous -  Mediktor enables health systems to
engage with patients through its chat platform.

White label - Mediktor white labels its solution. Indeed, white
labeling is at the core of our philosophy and clearly defines our
business model. This is one of the reasons why Mediktor has a SDK
available so health systems have the option to use their
look&feel/brands.

Once the user has entered his/her symptoms the system will begin a brief questioning to assess the following important points to discover about the user:

- possible conditions the user may suffer
- medical specialties that best could serve the user
- urgency of their case

 Mediktor lays in the knowledge of a self-made database relating 40.000 symptoms with up to 4000 questions and 660 conditions. This database is a huge random tree that relates symptoms with conditions as are related in the medical books. We expand this database with AI algorithms. The approach is based in finding the most suitable question from the 4000 questions we have in the system.

Mediktor uses the intuition generated by more than 4 million consultations in Mediktor database. The intuition takes into account all the profile fields (more than 40), the location of the user, and the main symptoms detected. Mediktor combines hundreds of parameters to identify the meaningful patterns and finally proposes a related symptom or question to the user. This feature helps a lot at the beginning of the interview when we do not have medical evidence to go to the next question.  Mediktor uses machine learning techniques to build up this intuition. 

Mediktor uses Natural Language Processing (NLP) to identify the symptoms from the reason of consultation. That NLP is based on a self-made database of symptoms synonyms. 

Mediktor uses neural networks techniques to build the intuition and to incrementally adapt it to new evidences.

Mediktor is the first and so far, only symptom checker in the market clinically validated. Its level of accuracy for any symptom/condition is 91.3% With the 15 most prevalent conditions in western countries accuracy increases to a 95%. Using Harvard Medical Schools criteria for their benchmark published in 2015, Mediktor’s accuracy is of 97% (independently certified by Munich RE). Our level of accuracy is scientifically validated by a first clinical trial at Hospital Clinic (Barcelona, Spain) in 2017 with a sample of 1,000 patients that visited Clinic’s ED. The results of the clinical trial were published in the scientific magazine EMERGENCIAS.
 
It is also worth pointing out that Mediktor is one of the few symptom checkers in the market that has validated its value proposition for insurance companies and hospitals. Mediktor and DKV (MUNICH RE Affiliate) have jointly announced at HiMSS Europe 2018 the results of the deployment of DIGITAL DOCTOR, the jointly developed tool offered to the 800,000 members of DKV in Spain. Those results include the fact that 60% of users claim that the joint utilization of AI and a remote consultation to a Dr., helped them prevent a face to face consultation with a Doctor.

Change is circumstantial and invariably accelerated by external factors. AI-based solution for clinical triage and telemedicine technologies have been in the market for over 15 years without achieving a significant market share. COVID19 has changed this overnight.

Innovation happens when there is a specific use case that achieves a superior outcome by leveraging the use of existing/new technology and, as a result, change becomes real. AI-based solutions like Mediktor solves the very specific need to control COVID19 at large. This use case will allow AI-solutions gain traction and initial market share in the short term.

Nevertheless, true change becomes the new reality in the long-term. Why assess the risk for COVID19 only? when Mediktor can assess up to 780 conditions that account for 90% of most prevalent conditions you may find in the ER of an urban area in the western world. This is precisely our long-term vision: become the digital front door where patients will (virtually) knock when they feel symptoms and want to access to healthcare.

Mediktor was used by 1,303,693 unique users worldwide for the last 12-month period ( Source Google Analytics)

Mediktor access to population depends on our partners who pay for the solution (health systems, health plans,...). Mediktor has partnered with 25 organizations. We plan to collaborate with a total of 75 organization by the end of this year and, as a result, we expect to serve a total population of over 4,000,000 people.

In 5 years, Mediktor expects to serve to up to 50,000,000 people on a daily basis.

Our main goal for next year is to help more health systems make the most of virtual care. Post-COVID19 world will see how virtual care plays a key role in the care continuum and our solution is going to become the digital front door that will enable health systems direct patients to the appropriate level of care.

While large health systems - those with an annual turnover over $1Billion USD- were already interested in virtual care before COVID19, mid-sized health systems that represent around 70% of all health systems operating in the US Market are willing to launch virtual care projects. Our goal is to start working with at least 15 of them within the first year.

Our goals are not only on the revenue side but also on the product side. Our product roadmap for the next 12 months include:

1. • Additional conditions added (mainly pediatrics & mental health)

   • Additional languages: Hindi, French, Dutch, Arabic,

   • Verticals per pathology (chronic mainly) for the pharma industry (in collaboration with IBM)

   • Integrate with professional virtual assistants (IPSoft’s Amelia)

Our goals for the next five years will be to bringing on board more
mid-sized health systems worldwide once we have developed best practices
with American health systems. While we expect that the business model will be easy to replicate in other Western countries, developing countries in Africa and Asia may require some adjustments to be fully aligned with people needs and health systems capabilities.

The main barrier for adoption is cultural. People is used to go right to the hospital and visit a doctor in person whenever they feel symptoms. While changing this habit is complicated, the truth is that it has certainly started to change due to COVID19. Patients noticed that their only option to access healthcare during COVID19 outbreak for non-urgent conditions was virtual care.

As a result, people became used to virtual care overnight and they are starting to see its benefits but they are not quite convinced yet. There are still certain cultural barriers for virtual care.

As per Mediktor, the main barrier is that we are a foreign company trying to expand our services in the US Market, which is unarguably the most competitive market for healthcare.

We anticipate that both barriers will disappear relentlessly in the coming years and we hope that in five year we won't face challenging barriers anymore.

The main way to overcome cultural barriers is evangelizing early adopters. This requires human resources and access to media to amplify the message. This is the reason why we expect to hire more people and invest heavily in Media and run Webinars where we will share best practices learned with existing partners.

As per Mediktor, we decided to open our NYC branch to become less of a foreign company. We also hired to American professionals. We are already seeing progress and Mediktor has recently closed a partnership with Yale New Haven Health System and another one with Austin Regional Clinic.

We believe that Solve support would help us greatly to speed up the process to overcome both barriers. A partner like Solve would certainly open doors that we are not capable to open on our own. Additionally, board members from Solve ecosystem will certainly help us adjust the message and make introductions to their networks.

Mediktor employs 23 full-time staff and 3 part-time staff.

Mediktor team is ready to make a positive impact in the healthcare industry. It's worth pointing out the following reasons:

- Mediktor was co-founded by a physician and currently employs a few physicians to make sure that our solution complies with clinical regulation and highest standards.

- Mediktor team combines the perfect balance of tech and commercial skills. This has allowed us to navigate the initial stages of development and become a rising star in the ecosystem (https://medium.com/idinvestpar...)

- While Mediktor team is quite young (age avg. 34), c-level positions are coped by professionals with extensive professional experience in healthcare - both as physicians and in corporations.

Mediktor has partnered with top-notch organizations worldwide, our partners include:

- Yale New Haven Health System (CT, USA)

- Austin Regional Clinic (TX, USA): https://www.austinregionalclinic.com/services/covid-19-care/covid-19-assessment

- Munich Re (Munich, Germany)

- Scap (Mexico): https://www.medicaldoctor.ai/

- Hadiel (Nigeria): https://hadiel.mediktor.com/

- Hospital Clinic (Barcelona, Spain)

- Hospital King Faisal (Kigali, Rwanda)

1. Mediktor has in place a 3-tier business model:

   1. License fee: One-time payment. License fee includes basic integration fees.

   2. Maintenance: Yearly payment. 20% of the one-time license fee.

   3. Variable payment: variable payment with a flexible approach (from Per use, to PMPM, PUPM, Revenue share,...) based on the business model and incentives of our clients. Long term, win-win models are set up in place base on business case jointly developed with our partners.

   Occasionally, additional cost may apply: translations, SDKs, specific complex integrations,... Those additional services are driven by specific client’s request.

Mediktor is close to reach break-even and cover expenses. Annual turnover was over $800K USD in 2019 and we expect to close 2020 with revenues over $1MM USD.

In the meanwhile, Mediktor raises funds from private investors (VCs, family offices,...) and from public grants. To date, Mediktor has raised over $6.1MM USD and it's currently raising a series B of $10MM USD (We already have $3MM USD in hard commitments)

Introducing innovation in the market is a tough, complex process and we believe that Solve would help us in a number of ways. Mainly, there is nothing like the support from a well-known organization to open doors that now remain closed for us (partners, investors,...)

Generate a breakthrough like the one we aim at requires the alignment of lots of stakeholders and we believe that Solve can be a powerful catalyser and help us bring together some of the key partners we are looking for. It would be great to have a board member from Solve ecosystem!

Communication is a key element when it comes to innovate. Your solution does not exist unless is present in the media. We believe that a partnership with Solve would provide us with brand awareness and get the attention from the media.

Mediktor is interested in partnering with:

- American academic health care institutions such as Yale New Haven Health System

- American academic technology partners such as MIT

- American healthcare organizations such as American America's Physicians Group (APG)

- Amercian health plans such as Cigna

- Solve members with powerful networks in the aforementioned segments.

Mediktor is built upon AI and ML since its inception. Many years ago when AI was not a buzz word, we saw its potential and created a solution based on AI/ML exclusively.

Ever since we have perfected Mediktor's AI/ML skills but there is still room to improve. Innovation in AI/ML is expensive and we wish we had more funds to employ a small division focused on streamlining and executing an ambitious product roadmap.