The VEDx dashboard is a flask/d3 application developed at the University of Washington for the visualization and interactive exploration of the ICD-9-CM diagnosis codes in a medical record or claims dataset. ICD-9-CM is a hierarchical coding standard used for diagnosis and billing at hospitals around the world. While ICD-9-CM was discontinued in 2015, and ICD-10-CM replaced it, the vast majority of retrospective Electronic Health Record data still currently has ICD-9-CM as the diagnosis code. For more information on ICD-9-CM, see ICD9Data.com.
Clone this repository and enter the project
git clone https://github.com/UWMooneyLab/VEDx.git
cd VEDx
Use pip install on the requirements file to download dependances
pip install -r requirements.txt
To start the application, simply run
python application.py
The dashboard should fire up. If your browser doesn't automatically open, go to you localhost to see the dashboard.
The dashboard contains a built-in tutorial that users can go through to learn how to use the tool and interact with the data.
The key area that will break the dashboard is bad data formats. It's important that the csv files contain the ICD9 codes and their associated scores or counts in the correct format.
For a collection of ICD 9 counts, format the csv file as such:
| ICD 9 Code | counts |
|---|---|
| 830.0 | 127 |
| 830.1 | 98 |
| 941.01 | 28 |
| 941.02 | 89 |
| 941.03 | 36 |
The first column may be named "ICD 9 Code", "code", "codes", "icd 9 code", or "icd9 code". Everything else will break the dashboard (error messages coming soon). The second column must be called "counts".
For a collection of ICD 9 scores (e.g. generated via statistical enrichment), format the csv file as such:
| ICD 9 Code | scores |
|---|---|
| 830.0 | 127 |
| 830.1 | 98 |
| 941.01 | 28 |
| 941.02 | 89 |
| 941.03 | 36 |
As in the case above, the first column may be named "ICD 9 Code", "code", "codes", "icd 9 code", or "icd9 code". Everything else will break the dashboard (error messages coming soon). The second column can be named "scores" or "score".
Another important note: The uploaded csv files do not need to be just two columns. All that is required is that the file have two columns that fit the naming schema above. The dashboard will ignore all columns not in the above examples.
Authored by Tim Bergquist.
Contact: trberg@uw.edu