Big Data in an Emergency Service

• Grant number: 19/22926-9
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: January 01, 2020
• End date: December 31, 2020
• Field of knowledge: Health Sciences - Medicine - Medical Clinics
• Principal Investigator: Heraldo Possolo de Souza
• Grantee: Cauê Gasparotto Bueno
• Host Institution: Faculdade de Medicina (FM). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated research grant: 16/14566-4 - Diagnostic and prognostic markers in patients admitted to an emergency department, AP.TEM

Abstract

In recent years we have witnessed an exponential increase in data processing and storage capacity in digital form, as well as statistical analysis capable of making the analysis of this data useful information.From a generic point of view, the concept of Big Data refers to the daily generation of a huge and very diverse amount of data that can no longer be analyzed by humans alone. One of the most commonly used definitions of Big Data in health is "Big Data includes heterogeneous, multispectral, incomplete and inaccurate observations (eg, diagnosis, demographics, treatment, prevention of disease, illness, injury and physical and mental disability) derived from different sources using incongruous sampling ".In HC-FMUSP a huge amount of data is generated daily, stored in electronic medical records and laboratory databases. The analysis of these data may contribute decisively to better understand the diseases and patients treated. Unfortunately, this data is found in different databases, making integrated analysis difficult.Our project is divided into two subprojects. In the first, it is our initial objective to create a database that can receive data from the patient and laboratory electronic medical records and that is accessible to analysis tools. In the second, we use this database to determine how the clinical characteristics of acute diseases differ between young and old adults and older patients at higher or lower risk for better resource management, because, despite efforts made in the last two decades, There is still a reliable tool for tracking vulnerability in the elderly in emergency servicesThe first part will be accomplished by transforming current databases through the OMOP Common Data Model into mutually compatible databases and then joining them into a single database. This process will be performed by computer scientists, programmers, and doctors from the HCFMUSP Emergency Service;Next, we will retrospectively analyze all elderly patients (over 65 years) treated at the Emergency Service of the FMUSP Clinical Hospital and determine the clinical and laboratory characteristics that differentiate them from young patients (under 65 years) with similar diseases. The main diagnosis to be used for this analysis will be sepsis, as it is common in both young and old and its high prevalence in our service.We believe that the data obtained in this study will be extremely important both from a clinical point of view and for better management of resources spent on emergency services with elderly patients.