tuva-health / docs Goto Github PK

Currently these links point to the docusaurus Github repo.

Is your feature request related to a problem? Please describe.
When you try to query the example SQL for readmission data quality issues, it fails because it cannot find a column multiple_primary_diagnoses_flag in readmissions.encounter_augmented. I double checked that table and didn't see any similarly named column.

https://github.com/tuva-health/knowledge_base/blob/main/docs/tuva-data/example-sql.md#readmissions

Describe the solution you'd like
Remove the column or rename to the appropriate column

Describe alternatives you've considered

Leave the SQL with a syntax error.

Additional context

• check on netlify pricing for open source
• get DNS/registrar credentials
• configure DNS to point to netlify hosted main builds

Is your feature request related to a problem? Please describe.
Feature request is getting more into detail for risk adjustment, further expansion of issue #163

• Date ranges and dates of service
• MMR initial enrollment and 26 sweep
• How age is exactly determined
• Rounding / sig figs
• Nitty gritty in regards to eligibility

Describe the solution you'd like
Addition to documentation

Describe alternatives you've considered
NA

Additional context
NA

The quality measures data dictionaries are not populating in knowledge base

Currently our data marts documentation is extremely out of date - we don't even list all of the data marts on our docs site. Every sub-folder under the models folder in our dbt package is a data mart. We should documentation for all of these in thetuvaproject.com.

For every data mart we should have documentation on the following:

• About: Describes what analytics you can do with the data mart and the methodology of the data mart
• Data Dictionary: Describes the final output tables of the data mart
• Example SQL: Describes how you can query the final output tables to do analytics

The Tuva Project is not using SSL - need to require all access to use HTTPS.

This allows for easy previews especially for non-technical contributors.

• Set up netlify account
• Get on Netlify Open Source plan
• Invite netlify team members
• configure the right balance of PR comments (we want builds but not much more)

When you flip to dark mode in the site the current logo looks dumb.

Add section describing how to group acute inpatient claims into acute inpatient encounters, including:

• A conceptual explanation of how to do this
• Data quality issues that can occur
• How to do acute inpatient analytics

Review all data models to make sure they reflect reality.

Encounter table needs admission and discharge dates.

Is your feature request related to a problem? Please describe.
The project has no license and is therefore subject to default copyright laws.

Describe the solution you'd like
Choose an open source license.

Change icon that appears on chrome tab from docusaurus dinosaur to tuva project logo.

• Finalize the style guide in notion so that it's ready to publish in our docs
• Hand-off to Aaron to publish in docs

Create the ability to download all terminology, reference data, and synthetic datasets from S3 via CSV and have a button on our Docs site that will trigger download.

Remove terminology_sql folder from Tuva project and remove from Docs (e.g. quickstart)

The data dictionary for the claims data model can be found in the following yml file: https://github.com/tuva-health/the_tuva_project/blob/main/integration_tests/models/_claims__models.yml

We should make this available to users in a more friendly format on thetuvaproject.com (i.e. our docs).

Let's add a main section (perhaps below Claims Data Fundamentals and above Data Marts) called Claims Data Model. Under this section add its data dictionary by pulling the information from the above YML file.

Note: You can follow the pattern of the data dictionaries from the other data marts to see how to do this, but it will need to work a bit differently because it references a different folder.

Is your feature request related to a problem? Please describe.
It's difficult to find the guide on using the mart on the CMS HCC page under Knowledge Base.

Describe the solution you'd like
It would be nice to have a separate guide for how to use this data mart. This would include:

1. Overview of main use cases for running this mart (e.g. you want to prospectively estimate HCC risk scores for your patient population)
2. Describe how to run it (i.e. dbt + tuva instructions)
3. Describe how to analyze the tables

Main sections of the website are:

• Getting Started
• Connectors
• Terminology
• Preprocessing
• Analytics
• Appendix

Need to update all of these sections so that they have meaningful and accurate content.

Getting data tables working is a separate issue.

The links in the terminology catalog are broken - if you click one they go to "page not found".

https://thetuvaproject.com/terminology/about

Is your feature request related to a problem? Please describe.
It's difficult to find the guide on using the mart on the Quality Measures page under Knowledge Base.

Describe the solution you'd like
It would be nice to have a separate guide for how to use this data mart.

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

This section currently exists under claims data fundamentals.

Need to add examples for the different scenarios (see bottom) and fix the formatting of the tables.

The ED Classification data dictionary is missing column descriptions and doesn't fully match the model in the data mart.

https://thetuvaproject.com/data-dictionaries/ed-classification

Document the methodology for breast cancer screening and colorectal cancer quality measures.

I don't know if the issue should go here or in the tuva project, but the ED classification data dictionary is missing most of its info. There is an entry for the model in the correct place in the yaml, but the columns in the yaml are almost entirely different than the columns in the actual model, so we end up with capital column names and no descriptions.

• set up GCP account
• get correct credentials for GCP account
• get access to new or existing GA4 account
• provision infrastructure (IAM etc) required for BigQuery
• set up BigQuery data transfer from GA4 account

static/img/readmissions/planned_admission_algorithm.png:
terminolgoy dataset #4 -> terminology dataset #4
terminolgoy dataset #5 -> terminology dataset #5

(Thanks for putting together this resource! I found the Claims Data Fundamentals very helpful.)

Shoutout to the Tuva Project + Knowledge Base - you can see a lot of great work going into this project and I can say it has already been super helpful!

Is your feature request related to a problem? Please describe.
The claims section has been a great read. I would love to see a section on claims linking - essentially what are some methods for linking a claim not only to a record on CMS return files ie. MAO-004 or MOR etc but also how can we (at least attempt) to associate a claim with encounters and then the encounter's entities (progress notes, conditions, providers etc).

Describe the solution you'd like
Some industry best practices for linking a claim to an encounter and to records on the CMS return files.

Describe alternatives you've considered
Most attempts we have made are tied pretty heavily to the dates of service between a claim + encounter to try and "link" this to an office visit as well as some record or records from various CMS return files. If this is the best way, it would be great to see some documentation around this. If there are alternatives, it would be great to see!

Additional context
N/A

Add Demo section to docs. The purpose is to show users how to quickly run the_tuva_project_demo and start exploring our synthetic data, terminology sets, and data marts. This is primarily a video showing how someone can get started on a free version of Snowflake.

Explain the methodology in detail using examples, so that someone can read this and completely understand how it works. Use flow chart diagrams to illustrate clearly.

Both data dictionaries for the 2 tables in the PMPM data mart are not fully documented.

https://thetuvaproject.com/data-dictionaries/financial-pmpm

• get access to new or existing GA4 account
• Add GA4 tracking code to Docusaurus

the current images for knowledge, code, and community are from the docusaurus default

The terminology links from the data dictionary pages are broken.

Description of issue:
On thetuvaproject.com under Example SQL, there is a query called Distribution of Patient by Age Group. The case statement in this query buckets a patient's age into an age group. The logic to do this is incorrect. All comparison operators are less than signs which does not properly categorize ages.

An enhancement to this query would be to change all the age group names too. There is overlap e.g. 2-18 and 18-30. To match the logic, the age group should be 2-18 and 19-30.

Current Behavior:
Patients are categorized in the wrong age group

Expected Behavior
Patients are categorized in the correct age group.