🏁 Quickstart

Pre-requisites

There are a couple pre-requisites for using the Tuva:

dbt
Healthcare data (i.e. claims or clinical data) in a data warehouse

Tuva supports the following data warehouses:

Snowflake
Bigquery
Redshift
Databricks
DuckDB

Running the Synthetic Data Tutorial

We created this demo project to make it easy for people to explore Tuva, even if you don't have access to healthcare data. The demo includes a 1,000 patient synthetic claims dataset based on Medicare claims data. Follow the instructions below to load the dataset into your data warehouse and run the Tuva Project.

Steps:

Clone or fork the demo project repo (link above)
Connect the demo project to your data warehouse
Run dbt deps to import Tuva and dbt build to run the project

Running Tuva on Your Data

Here we assume you've already mapped and loaded data to your data warehouse and connected your dbt project to your data warehouse.

Import Tuva: You need to import Tuva into your dbt project. To do this, add the following code to your packages.yml file:
```
packages:
  - package: tuva-health/the_tuva_project
    version: [">=0.5.0","<1.0.0"]
```
And then running the following command to import the package:
```
  dbt deps
```
Mapping: Running Tuva on your data requires mapping your raw healthcare data to the Input Layer. This involves creating custom SQL models to transform your data into the Input Layer data model.
Set Variables: Specific models in Tuva are designed to work with specific types of data. For example, the Financial PMPM data mart only runs on claims data. Set the following variables based on the type of data you have. For example:
```
  vars:
    claims_enabled: true
    clinical_enabled: false
```
Run dbt: Execute the following command to build the models that map to the Input Layer and the entire Tuva data model.
```
dbt build
```

Other Variables

Tuva uses variables to set default behavior for the data marts. These defaults can be found in the package's dbt_project.yml. You can change these values here or set them in the dbt_project.yml of your project.

tuva_last_run: The date and timestamp of the dbt run that will populate the tuva_last_run column in all models. Default timezone is UTC.
cms_hcc_payment_year: The payment year for the CMS HCC mart. Defaults to the current year.
cms_hcc_model_version: The risk model used for the CMS HCC mart.
quality_measures_period_end: The reporting date used to calculate the performance periods for Quality Measures. Defaults to the current date.
snapshots_enabled: Some data marts use the dbt snapshot feature. These are disabled by default. To enable them add this variable and set the value to true (snapshots_enabled: true).

Alternatively, you can set these in the CLI. Example:

dbt build --select the_tuva_project --vars '{cms_hcc_payment_year: 2020}'

Running a Single Data Mart

Mapping an entire healthcare dataset to the Input Layer can take several hours or more. You can also get started by running a single data mart, which is just a subset of the Tuva data model. Every data mart in the Tuva Project can be run individually. To run a single data mart complete the following steps:

Map to staging: Each data mart has its own staging layer. The staging layer is the set of models you need to create in order to run the data mart. You can find this in the staging folder under the data mart.
Set the variables: You need to enable the data mart you want to run. Do this by adding the variable for that data mart to your dbt_project.yml file and set the value to "true". For example:
```
  vars:
    cms_chronic_conditions_enabled: true
```
Run the package to build the marts you enabled.
```
dbt build
```

Running Terminology and Reference Data Only

If you just want to load the Terminology and Reference datasets and not worry about the Tuva data model, that's easy as well. You can disable the data marts and just load the terminology seeds by completing the following steps:

Add the variable tuva_marts_enabled to your dbt_project.yml file and set the value to "false".
```
   vars:
     tuva_marts_enabled: false
```
Run the package to build the seeds.
```
 dbt seed 
```

🏁 Quickstart

Pre-requisites​

Running the Synthetic Data Tutorial​

Running Tuva on Your Data​

Other Variables​

Running a Single Data Mart​