Skip to content

Latest commit

 

History

History
38 lines (24 loc) · 2.15 KB

assignments.md

File metadata and controls

38 lines (24 loc) · 2.15 KB
Raw

Assignments

Sepsis

The goal here is to understand what types of patients get admitted with Sepsis, what happens to them in the ICU, and explore if we can do early detection of adverse events.

Questions

  1. Identify patients who are admitted with a diagnosis of sepsis or have a diagnosis of sepsis during their stay. How many such patients are there? Hint: This is not a trivial question to answer. it may be worth thinking deeply about how to identify these patients.

  2. What are some characteristics (demographics, other diagnosis, procedures, medications, any other history you can get in the data) of patients with sepsis?

    • list at least 10 things that this group has in common

  3. How many of these patients die in the near future? within a day? a week? a month?

  4. For the patients who die within a month of admission, what characteristics do they have in common?

    • list at least 10 things that they have in common

  5. How are these characteristics different from people who are admitted with sepsis but do not die within a month?

    • list the most important 10 things that they don't have in common

Diabetes

The goal here is to understand what types of patients get diagnosed with Type 2 Diabetes and explore if we can do early detection. You should start from the assignment from Day 7 where you counted patients with Diabetes, prescribed Metformin, and getting insulin.

Questions

  1. What are some characteristics (demographics, other diagnosis, procedures, medications, any other history you can get in the data) of patients who have type 2 diabetes?
  • list at least 10 things that this group has in common
  1. How are these charactersitcs different for patients who don't have type 2 diabetes?
  • list at least 10 things that they don't have in common
  1. Now can you identify patients who have many/most of the characteristics you identified in question 1 but do not have type 2 diabetes in the data?
  • how many such patients can you identify?
  • how many of the characteristics do they have from the list you created in question 1?
  1. Of the patients you identify in question 3, could some of them have type 2 diabetes? why or why not?