Integrations#
VA Explorer supports integrations with a variety of services, some built-in, some external, in order to facilitate the quick and efficient movement or processing of VAs. In this section you will find information on these integrations as well as details on how to enable them in VA Explorer if needed.
ODK Central
KoboToolbox
DHIS2
Algorithm Support
pyCrossVA
InterVA5
Networking#
When integrating services with VA Explorer, operators should be aware of additional networking configuration if multiple docker-compose services are run on the same server.
docker-compose creates all containers that are part of its compose file under a default bridge network to allow intercommunication for that service. This setup does not allow one set of docker-compose managed containers to talk to another (VA Explorer -> ODK for example) by default.
To support this use case, operators can create an external bridge network and connect it to the necessary containers like so:
docker network create <external network name>
docker network connect <external network name> va_explorer_django_1
docker network connect <external network name> <other service's container name>
VA Explorer provides some additional network configuration to enable use of this
external network via .env
variables:
Variable Name |
Default Value |
Description |
---|---|---|
|
|
|
|
|
An http scheme ( |
These .env
variables should supplement the variables described below in whichever
service(s) you wish to integrate with.
ODK Central#
VA Explorer supports integration with ODK Central to provide both manual and automatic import of VA questionnaire responses. You can read more about ODK Central via its documentation.
To perform manual imports of ODK Central data, please see Data Admin in the Admin training guides.
To configure VA Explorer for automatic import of VAs, the relevant
.env
variables are detailed below. Once properly set, rundocker-compose up -d
to push the new configuration. ODK integration features within VA Explorer should now appear and function.
Variable Name |
Default Value |
Description |
---|---|---|
|
|
|
|
|
|
|
|
Value indicating the ID of the project in ODK Central that holds Verbal
Autopsy forms or data. Typically, a number found in the url when viewing
the project in ODK. Defaults to |
|
|
Value indicating the ID of the Verbal Autopsy form within the given project. Typically, found in the form list under the “ID and Version” column. Defaults to the ID of an example VA form. |
|
|
Value indicating the email of the account you wish to use to login to ODK Central. Defaults to an example email. Recommended to customize. |
|
|
Value indicating the password for the provided email’s account. Defaults to a blank string. Recommended to customize. |
KoboToolbox#
VA Explorer supports integration with KoboToolbox as an alternative to ODK that also allows for both manual and automatic import of VA questionnaire responses. You can read more about KoboToolbox via its documentation.
To perform manual imports of KoboToolbox data, please see Data Admin in the Admin training guides.
To configure VA Explorer for automatic import of VAs, the relevant
.env
variables are detailed below. Once properly set, rundocker-compose up -d
to push the new configuration. Kobo integration features within VA Explorer should now appear and function.
Variable Name |
Default Value |
Description |
---|---|---|
|
|
|
|
|
|
|
|
Value used to authenticate requests to the KoboToolbox API. A string of
~40 lowercase letters/numbers. Found for your user at |
|
|
Value identifying the project used to collect Verbal Autopsies within
KoboToolbox. Typically found in the url when viewing the project on the
instance (Ex. |
DHIS2#
VA Explorer supports integration with DHIS2 to export VAs that have been processed and assigned causes of death along for potential further use. You can read more about DHIS2 via their documentation. Details on configuring DHIS2 for use with Verbal Autopsies metadata is also available via the DHIS2 VA program open-source project.
To configure VA Explorer for automatic export of VAs, the relevant .env
variables are detailed below. Once properly set, run docker-compose up -d
to
push the new configuration. DHIS2 integration features within VA Explorer should
now appear and function.
Variable Name |
Default Value |
Description |
---|---|---|
|
|
|
|
|
|
|
|
Value indicating the root Organization Unit UID within your DHIS2
organizational hierarchy. Should be available using the DHIS2 API Query:
|
|
|
Value indicating the username of the account you wish to use to login to DHIS2 Defaults to DHIS2’s existing default username credential. Recommended to customize. |
|
|
Value indicating the password for the provided username’s account. Defaults to the default credentials for DHIS2’s default admin account. Recommended to customize. |
If you encounter any issues during integration, please reference the Troubleshooting section.
Algorithm Support#
VA Explorer provides built-in support for cause of death assignment via the InterVA5 computer coded VA (CCVA) algorithm without additional configuration. If you would like to change how these services are configured (including the HIV or Malaria prevalence variables for InterVA5) or point to a custom service instead, this section will help with that. As a future expansion, VA Explorer may provide built-in support for additional CCVA algorithms.
pyCrossVA#
pyCrossVA is provided as a built-in docker service to allow VA Explorer to transform VAs into input suitable for cause of death assignment algorithms. You can read more about it via its project page.
Variable Name |
Default Value |
Description |
---|---|---|
|
|
|
InterVA5#
InterVA5 is provided as a built-in docker service to allow VA Explorer VAs to be processed for likely cause of death. You can read more about it via its project page.
Variable Name |
Default Value |
Description |
---|---|---|
|
|
|
|
|
One of the enumerations “h” (high), “l” (low), or “v” (very low). Used to indicate the prevalence of Malaria within the population. Defaults to “l” for low. |
|
|
One of the enumerations “h” (high), “l” (low), or “v” (very low). Used to indicate the prevalence of HIV within the population. Defaults to “v” for very low. |
|
|
|