Projecto Laura

COMPOSTA+
21.02.2020
VEEVER
21.02.2020

Area of work:  Medical care

Website: https://www.laura-br.com/about.html

Challenge overview

According to data from the World Health Organization, 31 million cases of sepsis per year worldwide are registered, of which 6 million resulting in the patient’s death. In Brazil, the rates indicate 670 thousand cases of sepsis annually, 50% of which involve the death of individuals. Sepsis is a disease that can be contracted in the hospital (between 30% and 40% of cases) and in other environments through viruses, fungi or bacteria. The disease involves “widespread inflammation of the body itself against an infection that can be localized to any organ. This inflammation can lead to the failure of one or more organs, with the risk of death when not discovered and treated promptly” (Setor Saúde, 2018).

Baby Laura was one of the victims who did not resist sepsis in May 2010 after a premature birth and the struggle for survival for 18 days while remaining in the Neonatal Intensive Care Unit. Jackson, driven by the pain of losing his daughter Laura, and aiming to help lower these rates, created the world’s first risk-management cognitive robot named LAURA.

The care provided in hospital environments reflects factors of structural and technological nature, as well as human resources, hospitals are faced with. Such factors can directly affect hospital mortality. The demand for medical and hospital services encompasses a good amount of patients, a huge amount of data, and unmonitored processes that can generate human or technical system failures. Just as operational inefficiency, the delay in defining which patients require priority care, diagnosis, and treatment lead to increased mortality and costs to patients, their families, and the institutions. In this respect, the LAURA project arises from an experience lived in a hospital context, and from the observation of the aforementioned demands.

Enterprise`s Establishment

Shortly after Laura’s death and before the creation of the robot, between 2010 and 2012, Jackson studied sepsis, deepening in the issue, in hospital problems, and in the dynamics surrounding the system, based on the perceptions of a “layman”. In 2011, with the theme artificial intelligence emerging in the market and from his background as a professional in the area of ​​information technology, Jackson began to analyse the opportunity that technology could bring to the problems observed. He identified the possibility of autonomously isolating a possible risk trigger without requiring a validation curator. Thus, between 2012 and 2013, based on statistical information, evidences in his favour, and the observation of cases that demonstrated whether autonomy in curating was possible, Jackson structured a technology-based solution to promote a risk rating, subsidizing information for decision-making by health professionals. After structuring the solution, it was necessary to invest. The first operational prototype of Robot LAURA came in 2015.

The first test took place in a hospital in the interior of Parana, not bringing the expected results, meanwhile showing improvement points for the prototype. In mid-July 2016, the Beta version of robot LAURA went live at the Nossa Senhora das Gracias Hospital (Curitiba/Parana/Brazil), and in September 2016 version 1.0, with several updates already implemented. (Fressatto, 2019).

Robot LAURA is now operating in 13 Brazilian hospitals, including Santa Casa da Misericórdia (Porto Alegre/ Rio Grande do Sul), Márcio Cunha Hospital (Ipatinga/ Minas Gerais), AC Camargo Hospital (São Paulo), Erasto Gaertner Hospital (Curitiba/ Parana), São Francisco Xavier Foundation (Ipatinga/ Minas Gerais). The project counts today with such important partners as: Nvidia and Intel.

Currently, robot LAURA helps save 12 lives a day. The software reads the patient’s information and issues alerts that are sent every 3.8 seconds to medical staff to flag patients at risk of widespread infection, and warns in advance other cases of clinical deterioration.


Objectives and Activities 

LAURA is considered a cognitive robot that manages risk by collecting and organizing data, followed by performing complex calculations, comparing results with probabilistic ranges and accurate inferences about the favourable or unfavourable conditions of succeeding a risky event. Its operation is based on artificial intelligence, highlighting two technologies, cognitive computing and machine learning, so LAURA is considered a “learning robot”.

In this sense, in the case of a cognitive robot, the process involves making cognitive connections, as with individuals and in the case of computer technology this activity involves the ability to identify data and categorize it, storing it as cognitive artefacts. The analysis of this data starts from a large number of occurrences, in a matter of microseconds, the machine learning algorithm detects patterns and determines with high reliability the chances of a favourable or unfavourable result to the analysed event, very similar to the deduction process of humans.

Robot LAURA is considered as artificial intelligence that empowers people with a powerful cognitive robot capable of collecting and analysing a huge volume of data. The project provides an efficient way to identify process errors and to notify in real time what is happening and where it is occurring.

With the objective of promoting subsidies for assertive and efficient decision-making, the LAURA project already points out benefits promoted by the robot to the following four health challenges already observed:

  • Patient deterioration: robot LAURA provides more than 12 hours of advanced instability warning and provides early intervention when combined with continuous monitoring of vital signs
  • Alert Fatigue: Compared to other protocols, robot LAURA improves conditions for early identification of clinical deterioration. Compared to other protocols, every 100 alerts, 50 are false positives, with LAURA only 25%
  • Huge amount of information: Artificial intelligence “reads” and interprets data very quickly beyond human capacity, assisting in data-based decision-making
  • Accreditations: robot LAURA contributes to organizations that adopt it to become paperless (reducing paper consumption in the institution), which facilitates the acquisition of hospital accreditations.

In this sense, robot LAURA’s actions reduce negative outcomes, unnecessary costs, overloaded staff, repeated processes, miscommunication, and legal risk. It also increases certifications, awards and accreditations, productivity, scientific research, more patients served, protocol adherence, and risk and process management.

Robot LAURA’s activities begin with inpatients and data collection through the patient’s electronic medical record, observing vital sign data and examinations. This information is deposited in LAURA’s system and sent to devices used at the hospital by the nursing staff, such as smartphones/tablets, computers, and panels allocated in the corridors. The devices inform and “alert” nurses, followed by doctors about patients at risk. The information robot LAURA collects and manages enables medical decision-making reports.

The work carried out by the LAURA project team in the hospital context includes, besides the collection of information and treatment issues for decision-making, a framework of practices that involves hospital administration, medical team, nursing team, and the information technology team, as well as the patient’s experience. In each of these spheres, robot LAURA supports the activities and performance of these professionals, hospital infrastructure, and the direct impact on the beneficiary:

  • For hospital administration, robot LAURA’s contributions cover issues such as:
    • reduces overall hospitalization costs
    • efficiency in bed turnover (more patients attended)
    • generates reports and shows trends in real time
    • performs a digital transformation.
  • In the medical team this support involves:
    • predicting patient deterioration with the use of artificial intelligence
    • perform more effective interventions with data-based decisions
    • generate reports and show trends in real time
    • analyze grouped clinical information (patient timeline)
    • alerting and activating the Quick Response Team
  • The nursing team receives the following information:
    • early warnings so that the care team can act on patients at higher risk
    • reduces work overload
    • reduces alertness and information overload fatigue
    • increases the operational efficiency of the team
    • empowers the care team to prioritize their tasks (information for action)
    • generates reports and shows trends in real time
    • gets fast training (in minutes) and online
  • In terms of information technology, the connection to the electronic medical record occurs in real time and the basic infrastructure used is the one provided by the hospital, without costly equipment costs
  • Patients, in turn, realize that they are being taken care in full time in a hospital environment with modern and advanced technologies

Challenges

Jackson’s biggest challenge is getting the project purpose to go along with the business plan, without allowing the business plan to destroy the purpose. A significant factor in the process is to believe in people, otherwise, there is no motivation to do something for them. Given this, it is necessary to promote life, for only in this way will things happen. Moreover, promoting life occurs in moments of peace, mainly because individuals are sensitive and emotional beings (Fressatto, 2019).

Impact

The impact promoted by robot LAURA is observed in average service time, which indicates reduced hospital time by 7 hours per patient, and 25% reduced mortality overall. The system connects 2.5 million patients and records 12,289 lives saved (12 lives per day) with the help of robot LAURA, since it started being used. Another impact is identified in the care provided by the hospital care team. Robot LAURA increases staff efficiency by up to 85%, thus optimizing time and resources. According to records from one of the hospitals that use the robot LAURA, it indicates that the decrease in patients’ internment time saves the hospital 5.5 million dollars a year.

The impact of the LAURA project is also proven and recognized through awards already received, such as:

  • Pfizer Challenge 2016
  • Innovation Winner 2017 (Awards – The Institution of Engineering and Technology)
  • Challenge Cup – 1776 (largest competition for social impact start-ups in the world)
  • HIMSS – Elsevier Digital Healthcare Award Brazil and Latin America 2019
  • HIMSS – Transforming Healthcare Through IT
  • HIS – Hospital Innovation Show 2016
  • III Empreenda Saúde (Undertake Health) Award
  • Accelerate 2030 – United Nations

All the work carried out by robot LAURA founder and project team is also recognized through media outlets in various local media. To increase its impact and establish bridges for research and knowledge sharing, in 2019 the project established a partnership with the Brazilian Ministry of Science, Technology, Innovation and Communications to access the network of national and international research centers, in favor of knowledge exchange in this context (Fressatto, 2019).

Lessons learnt

The main lesson learnt by its founder is that the LAURA Project doesn’t misappropriate the solutions it generates. The product and algorithm belong to Jackson and his team, but the result should not be owned by the project. The results and knowledge produced must be shared with all intellectual agents that are part of the medical field (doctors, professionals, academia, among others), since the exchange brings new perceptions to the process. It was observed that sharing knowledge brings forward new opportunities and other work fronts. An example of this involves research, as in certain hospitals where the technology was implemented. There was an improvement in the quality of care professionals and doctors, who became more comfortable conducting research in these environments, risking “a bit more” in certain procedures.

Another good practice involves promoting learning environments, involving system actors such as doctors, nurses, lay people, and academia to think together and test project tools and applications (Fressatto, 2019). Finally, it is noteworthy that despite all the impacts generated, the founder of robot LAURA still has bold goals for the project, one of which being positively impacting 1 billion lives.

Notes and remarks. Bibliography

1. The website of the Laura project: https://www.laura-br.com/about.html.

2. Fressatto, J. Interview granted to Flavia Roberta Fernandes, 02 Dec 2019.

3. Setor Saúde (2018). Sepse: Brasil registra 670 mil casos por ano, sendo 50% fatais. Available at: https://setorsaude.com.br/sepse-brasil-registra-670-mil-casos-por-ano-sendo-50-fatais/. Accessed 08 Dec. 2019.

4. Credit to Carlos Olavo Quandt, Flavia Roberta Fernandes, Mari Regina Anastacio, Sara Regina Hokai and Ubiratã Tortato