Problem

The main point of communication between doctors in a hospital is called a handover. It is when a doctor's shift ends, and they hand over all their patients (20-30 patients) to the next doctor.

Due to the shear amount of information doctors need to synthesize, 37% of handovers are deemed unsuccessful. 1 In the typical hospital, about 1.5k handovers are failed every day. 1

Sentinel events are events of course of the reason the patient is in the hospital that can result in severe temporary harm, permanent harm, or even death to a patient. The root cause for 70% of sentinel events, is poor communication. 2

By improving communication is hospitals, we can prevent up to 800 sentinel events in the typical hospital every year. 3

What it does

This early prototype analyzes a patient's medical record and returns the 3 most crucial details about this patient, making sure that the doctor on the receiving end of the handover doesn't miss anything important.

How I built it

rake_nltp is a natural language processing library in python. The bulk of the functionality runs in a python script which was programmed in vs code.

Challenges I ran into

The main challenge for building this is that I had no prior experience with most of what I used. I had never built a project which uses Natural Language Processing, which means that I needed to QUICKLY learn and synthesize information about it if I wanted to use it in a 48 hour hackathon.

What's next for No Missing Details

Next steps:

  1. Make a web app to make public testing easier
  2. Test with real medical record (waiting for mine to arrive)

Built With

Share this project:

Updates