Laura Institute

The COVID-19 pandemic shows that the present healthcare system is not well prepared to deal with highly infectious diseases. The simple act of seeking medical help increases the spread of the disease and exposes healthcare workers to potential infection. Besides that, the COVID-19 patient has a potential critical phase among the 6th-10th day of disease. Our solution mitigate those problems by offering an AI-powered platform for self-triage and continuous monitoring of symptoms through chatbot, detecting early signs of clinical deterioration and solely referring cases that need medical intervention to the emergency room. In addition, provides aggregated data to both City Health Departments and Hospitals to local insights. If our solution were to be scale globally, it could help local governments controlling the pandemic and possible future outbreaks of infectious diseases, optimizing the resources usage.

With 8 million cases worldwide, the COVID-19 pandemic challenged the entire world and showed that most healthcare systems are not well equipped to fight it. The current model of getting medical attention is highly inadequate when dealing with an infectious disease like COVID-19, it increases the spread of the virus and exposes healthcare professionals to potential infection. Besides that, the COVID-19 patient has a potential critical phase among the 6th-10th day of disease that usually is diagnosed too late. Examples around the world show that the most effective way of fighting the pandemic is to decrease the spread of the disease and protect the healthcare system from overflowing. Also, research points that approximately 20% of the cases need medical intervention, which means that  the majority of patients could stay at home and take simple measures to heal, decreasing the spread of the virus and the demand on the healthcare system. Although, it is still important to monitor these cases for signs of clinical deterioration, in order to refer only the patients that need medical intervention. Another huge problem is lack of real time data of patients with suspicion of COVID-19.

Laura Digital ER is an AI-powered platform that provides hospitals and health departments a solution to triage patients and monitor the progression of their symptoms. The first front is composed by a chatbot that,  through the use of Natural Language Processing (NLU), answers more than 100 questions related to COVID-19. It can be implemented on the websites of city halls and hospitals to answer questions from general population. The second front is a self-triage page where the user can inform his symptoms and, through the use of Machine Learning algorithms, classifies the case as mild, moderate or severe. Severe cases are oriented to seek an ER immediately, moderate cases can be referred to a virtual consultation depending on the availability from the healthcare system, and, lastly, mild cases are monitored through the use of cloud computing and AI ― which sends text messages daily, asking about the progression of symptoms and detecting early signs of clinical deterioration, referring these cases to an ER. From this information, the system can provide hospitals a forecast of demand, improving the management of the resources.

Our main target population are the entities from the healthcare system, hospitals, cities health departments and everyone focused on providing better healthcare to the population. Helping them to contain the pandemic, all the population benefits indirectly from our solution. One of the main challenges for hospitals is managing the limited resources available, our purpose is to relieve that burden by offering a virtual self-triage tool and monitoring of symptoms for as long as necessary to ensure that the patient is cured, decreasing the demand for the limited resources of the hospital and providing a estimate of the cases that will need them. Also, many local governments which have a precarious healthcare system,  will be benefited  with a cost effective way that helps monitoring patients and planning measures that  protect the population. In conclusion, by helping local authorities and hospitals controlling the outbreak, our solution can benefit millions of lives directly and indirectly.

Most countries proved their inefficiency when dealing with a pandemic. the  paradigm of medical consults is completely inadequate when facing a highly infectious disease.  An adaptation is needed. In a world where physical distances can be shortened by technology, why can't healthcare follow this trend? Providing a platform where cases can be quickly identified and monitored without physical examination, the current and future pandemics can be managed more effectively and millions of lives can be saved. Therefore, we propose a new way to provide healthcare in the outpatient settings: a more efficient, accessible and cost-effective one.

There are many self-triage tools and chatbots about COVID-19 in the internet:

• Isabel Symptom Checker provides a self-triage tool, but no chatbot and monitoring

• Telecovid offer virtual consults and monitoring only of severe cases

• Lifen Covid offers similar text message monitoring, but lacks the dashboard with data analysis and chatbot to answer questions

Our platform offers a complete package, consisting of a chatbot that answers questions, a self-triage tool connected to hospitals and local governments and an AI-powered monitoring solution that can classify COVID-19 cases and refer only those that need medical treatment to the ERs. Since all of those solutions are connected to a database, we can collect data from the patient's entire journey, from start of the symptoms, until the cure. All of that information is displayed in a dashboard, providing epidemiological data, estimate of demand and current status of all monitored patients. Enriching the decision makers from local governments and hospitals administration.

The core technologies that drives out platform are Artificial Intelligence and Cloud Computing. We use Rasa's Natural Language Understanding (NLU) solution to power our chatbot, enabling it to understand and answer more than 100 questions about COVID-19. To analyze the messages that are not understood we use different Machine Learning (ML) algorithms, that allow us to identify popular demand for new questions. Through the use of Amazon's Web Services (AWS), we can store all the data collected from the self-triage tool and feed the dashboard for our clients. AWS also powers our monitoring AI, that sends text messages to the patients, understand their answers and detect signs of clinical deterioration

https://vimeo.com/430391702

Getting to know our data.

Currently, our solution is present in 10 hospitals and 4 city halls, monitoring 8,000 patients.

In one year our target is 30 hospitals and 20 cities, monitoring 50,000 patients.

In five years our target is 100 hospitals and 100 cities, monitoring 1 million patients.