kandimallab / 410.712_finalproject_emscompare

JHU - 410.712 Advanced Practical Computer Concepts for Bioinformatics Final Project

PROGRAM

JHU - 410.712 Advanced Practical Computer Concepts for Bioinformatics Final Project

AUTHOR

Bhavya Kandimalla

ABOUT

This program is an online quality control tool that asses EMS performance for training and process improvement purposes by comparing EMS impressions to eventual hospital diagnosis.

The program's algorithm calculates a EMS quality care score for each event using a crosswalk to quantify EMS performance based on a 5-3-1-0 point system assigned to impression-diagnosis matches. 5 points are awarded if the EMS impression matches the first hospital diagnosis, 3 points are awarded if the impression matches the second diagnosis, 1 point is awarded if the impression matches the third diagnosis, and 0 points are awarded if the impression does not match any of the three hospital diagnoses.

The results, which can be filtered by score and impression, are viewed at: http://bfx.eng.jhu.edu/bkandim1/410.712_Advanced_Practical_Computer_Concepts_for_Bioinformatics/final/

REQUIREMENTS

The recommended Python version is Python 2.6.6.

The recommended MySQL version is 5.1.73.

The program requires a stable internet connection. Any web browser can be used.

The EMS patient raw data file must be in the following format:

"Name",DateOfBirth,DispatchDate,"EMS Impression"

The hospital patient raw data file must be in the following format:

"Name",DateOfBirth,AdmitDate,"Reason For Visit","Diagnosis1 Code", "Diagnosis1", "Diagnosis2 Code","Diagnosis2","Diagnosis3 Code","Diagnosis3"

USAGE

SETUP AND BACKEND USE:

1. Raw data files of the EMS Patients and Hospital must be uploaded to the server as source files

2. To de-identify EMS patients and create patient-event table, un-comment lines 19, 23-25, and 30-31 of the insert.py script and, on the server, run command:

   python insert.py

3. To de-identify hospital patients and create patient-event table, un-comment lines 20, 23-25, and 33-35 of the insert.py script and, on the server, run command:

   python insert.py

4. To create crosswalk table, un-comment lines 18 and 27-28 of the insert.py script and, on the server, run command:

   python insert.py

5. To run the data through the program and upload to database

   1. Log into the MySQL server:

      mysql -u USERNAME -h localhost -p

   2. Use the bkandim1 database:

      use bkandim1;

   3. Direct the source to the views.sql script:

      source /FULLPATH/views.sql;

USER USE:

6. View and query the results at: http://bfx.eng.jhu.edu/bkandim1/410.712_Advanced_Practical_Computer_Concepts_for_Bioinformatics/final/

FILE NOTES:

EMS Patients.txt - raw EMS patient file
Hospital Patients.txt - raw hospital patient file
insert.py - de-identifies patients and inserts data into input tables
tables.sql - defines structure of input tables
views.sql - input table queries to create 8 views and 1 table for the program
querydb.cgi - runs queries and returns results to the webpage
query.sql - script to test queries
query.py - test script

DEMO DATA

The project source files can be found here: https://github.com/kandimallab/410.712_FinalProject

A sample dataset can be viewed at: http://bfx.eng.jhu.edu/bkandim1/410.712_Advanced_Practical_Computer_Concepts_for_Bioinformatics/final/