Premium Project

Classify Suspected Infection in Patients

Classify patients with suspected infections using data.table and electronic health records.

Start Project
  • 12 tasks
  • 710 participants
  • 1,500 XP

Project Description

Sepsis is a deadly illness that accounts for a large portion of in-hospital deaths. It occurs when a person's organs shut down in response to a severe infection. This public health problem is a major target for research, and hospital records can help us investigate the problem. In this R project, you will identify hospital patients with severe infection using medical record data.

To successfully complete this project you should have some experience with the package data.table, including using the := operator, grouping aggregations with by, and understanding how to use the shift function.

Project Tasks

  • 1This patient may have sepsis
  • 2Which antibiotics are "new"?
  • 3Looking at the blood culture data
  • 4Combine the antibiotic data and the blood culture data
  • 5Determine whether each row is in-window
  • 6Check the I.V. requirement
  • 7Find the first day of possible sequences
  • 8Simplify the data
  • 9Extract first four rows for each blood culture
  • 10Consecutive sequence
  • 11Select the patients who meet criteria
  • 12Find the prevalence of sepsis
JoAnn Alvarez

Data Scientist at Tenet Healthcare

JoAnn is a data scientist at Tenet Healthcare and previously worked in biostatistics at Vanderbilt Medical School. She studied mathematical statistics at University of Maryland and is passionate about machine learning.

See More


  • R LogoR
  • Topics

    Data ManipulationImporting & Cleaning Data