Welcome, Welcome, Welcome!
Week 1 is upon us and we are going to spend it getting everyone situated, goiong over the primary paper, and beginning to read up on the domain of sepsis care in a critical care unit. As well as using EHR data for clinical care research.
Topics
You will be touching on these topics over the first week
- What is sepsis, how it can be diagnosed/treated and why it is very deadly from an epidemiological point of view
- Severity of illness scores
- EHR data
- Reproducibility/replicability in data science
Tasks
- Gain access to the MIMIC-III and MIMIC-IV dataset, see instructions on home pagecritical
- Complete all assigned readings
- Begin to familairize yourself with the domain by thumbing through the Domain Expertise page.
- Submit question answers to the online form (linked below)
- Join the class Discord Channel: TBD
Readings
- Read the capstone primary paper by Zador et al. linked on the home pagecritical
- MIMIC-III paper by Johnson et al. link - if you already did not finish it
- MIMIC-III web description link - if you already did not finish it
Questions
Answer the following questions using this google form link
- What are the primary goals of the research paper we aim to replicate? Additionally, what do you anticipate to be the major hurdles in this process? Lastly, define what outcomes would you consider as a successful completion of Phase I in this capstone project.
- Provide a detailed explanation of what an Electronic Health Record (EHR) is. How does EHR relate to and integrate with the MIMIC datasets?
- Explain the concept and utility of Illness Scores in healthcare. Why do multiple illness scores exist? Select one illness score from the following - OASIS, SAPS II, or SOFA. Write a brief overview, including its applications and key attributes. Note: The aim is for each team member to gain expertise in at least one illness score, so ensure that all three scores are explored by different members.