meta3-demo 0.2.0

Demo project for the META3 edc

META3 Demo Project

This is a demo project of the EDC used by the META3 clinical trial.

This demo sets up the META3 EDC to run on a Django test server in DEBUG mode on your local machine.

These are NOT the installation steps for a production system.

See also:

https://www.lstmed.ac.uk/research/departments/international-public-health/respond-africa/meta

https://github.com/meta-trial/meta-edc

https://github.com/clinicedc

https://www.djangoproject.com

https://www.fda.gov/regulatory-information/search-fda-guidance-documents/clinical-trials-guidance-documents

Installation

You’ll need setup mysql and miniconda on your local machine.

Get miniconda here

https://docs.conda.io/en/latest/miniconda.html

Now that mysql and miniconda are install on your local machine, let’s get started.

First, create the demo database

mysql -Bse 'create database meta3_demo character set utf8;'

Create a project folder and clone the repo into it

mkdir ~/clinicedc && \
cd ~/clinicedc && \
git clone https://github.com/meta-trial/meta3-demo.git

Create a conda environment named meta3_demo and activate it

conda create -n meta3_demo python=3.10 && \
conda activate meta3_demo

With the conda environment activated, install the meta-edc python package

pip install meta-edc

Now that the application is installed, we need to make some changes to the configuration.

Copy the sample environment file to a working copy

cd ~/clinicedc/meta3-demo && cp .env-sample .env

Note

Sensitive config values are stored in the environment by using an .env file and environ. (see also see https://12factor.net)

Next, edit your working copy of the environment file (.env). Look for DATABASE_URL at the top of the file. Change the value for DATABASE_URL to include your mysql user and password. The mysql account will need root or root-like permissions. Since this is a test server running locally, just use root.

# find this line
DATABASE_URL=mysql://<username>:<password>@127.0.0.1:3306/meta3_demo

# and change with your details, for example
DATABASE_URL=mysql://root:dumela@127.0.0.1:3306/meta3_demo

Next we need to create the keys used for data encryption.

Run manage.py check for the first time. This will ask django-crypto-fields to create encryption keys.

python manage.py check

Note

The system encrypts sensitive data (personally identifiable information or PII) using django-crypto-fields. The first time you run manage.py, django-crypto-fields looks for the keys. If they do not exist, it creates them.

Now go back and edit the environment file (.env). Change DJANGO_AUTO_CREATE_KEYS to False

DJANGO_AUTO_CREATE_KEYS=False

Run manage.py check again. You should see a final message System check identified 3 issues (1 silenced). Since this is a test server, you may ignore these warnings.

python manage.py check

Now you are ready to prepare the database that you created earlier.

We have a set of migrations included. Migrations are python scripts that create all the tables, relations, contraints, etc needed to run the system. To save time, we will just restore the demo data provided in this repo. The demo data is an empty mysql database archive that is exactly what you would get if you ran the migrate command.

Restore the demo data

cd ~/clinicedc/meta3-demo/demo_data && \
tar xzf meta3_demo.sql.tar.gz && \
mysql meta3_demo < meta3_demo.sql && \
cd ~/clinicedc/meta3-demo/

Note

There are a few caveats to migrating your own database instead of using the demo data. Running migrate on an empty database takes more than 30 min. Also, you will run into a few simple problems with the data migrations. See the note on`running migrations on an empty database` in Troubleshooting below.

Now that our database has the required data schema, we need to run the post-migrate signals to populate some static data. But this does not take long.

To do this we just run the migrate command.

# run migrate to trigger the post-migrate signals
python manage.py migrate

Next, import the list of holidays that will be used when scheduling appointments.

python manage.py import_holidays

Next, create a user. Do this from the command line using the createsuperuser command.

python manage.py createsuperuser

Important

The new user you just created is a superuser. Once logged in you need to remove the superuser status for this account.

Now start up the test server using the runserver command

python manage.py runserver

Open your browser and point it to

localhost:8000

You should see the login screen for the META3 trial running at Temeke Hospital in Tanzania.

Type in the credentials of the superuser account you just created.

Once logged in, go to your user account and edit the permissions on your account. You can use the link at the top right corner.

• Under the section Personal Details, fill in your name and email.

• Under the section Permissions, uncheck Superuser status.

• At the bottom of section User Profile you will see Roles. Add yourself to the following roles:

  • Account Manager

  • Staff

  • Clinician Super

• Click SAVE at the bottom of the form

Now you are ready to screen your first participant!

Click Home on breadcrumbs to the left on the top bar to get started.

Troubleshooting

Running migrate on an empty database

If you run migrate on an empty database, a few of the data migrations might fail. Since these failed migrations are data migrations and not schema migrations, it is safe to run migrate until it fails, fake the failed data migration, and continue.

python manage.py migrate

# fake the data migration
python manage.py migrate meta_prn 0035 --fake

# restart migrate
python manage.py migrate

# fake the data migration
python manage.py migrate meta_subject 0107 --fake

# restart migrate
python manage.py migrate

# fake the data migration
python manage.py migrate meta_subject 0132 --fake

# restart migrate
python manage.py migrate

Removing the demo when you are done

drop the database:

mysql -Bse "drop database meta3_demo;"

deactivate the conda environment:

conda deactivate

remove the conda environment:

conda env remove -n meta3_demo

Finally, delete the clinicedc folder.