Quality Measures

Overview

Code

The Quality Measures data mart is where we are building publicly available quality measures. You can see the roadmap in this section. If there is a publicly available measure you would like to see added you can submit an issue on GitHub.

Check out the Knowledge Base article for an overview of the data mart and a walkthrough example for calculating a quality measure.

Measure Name	Measure ID	Specification	Status
Breast Cancer Screening	CMS Star C01, MIPS CQM 112, NQF/CBE 2372	Link	Released
Colorectal Cancer Screening	CMS Star C02, MIPS CQM 113, NQF/CBE 0034	Link	Released
Controlling High Blood Pressure	CMS Star C11, MIPS CQM 236	Link	Released
Diabetes: Eye Exam	CMS Star C09, MIPS CQM 117, NQF/CBE 0055	Link	Released
Diabetes: Hemoglobin A1c (HbA1c) Poor Control (> 9%)	CMS Star C10, MIPS CQM 001, NQF/CBE 0059	Link	Released
Documentation of Current Medications in the Medical Record	CMS Star C06, MIPS CQM 130	Link	Planned 2024 Q3
Falls: Plan of Care	CMS Star C12, NQF/CBE 0101	Link	Released
Follow-up after ED Visit for People with Multiple High-Risk Chronic Conditions	CMS Star C18	Link	Planned 2024 Q4
Hospital-Wide All-Cause Readmission (HWR)	CMS Star C15, MIPS CQM 479	Link	Released (Readmissions mart)
Influenza Immunization	CMS Star C03, MIPS CQM 110, NQF 0041	Link	Released
Medication Adherence for Cholesterol (Statins)	CMS Star D10, NQF 0541	Link	Planned 2024 Q3
Medication Adherence for Diabetes Medications	CMS Star D08, NQF 0541	Link	Planned 2024 Q3
Medication Adherence for Hypertension (RAS antagonists)	CMS Star D09, NQF 0541	Link	Planned 2024 Q3
Medication Reconciliation Post-Discharge	CMS Star C15, NQF 0097	Link	Released
Osteoporosis Management in Women Who Had a Fracture	CMS Star C08, MIPS CQM 418, NQF/CBE 0053	Link	Released
Pain Assessment and Follow-Up	CMS Star C07, MIPS CQM 131	Link	Planned 2024 Q3
Statin Therapy for the Prevention and Treatment of Cardiovascular Disease	CMS Star C16, MIPS CQM 438	Link	Released
Statin Use in Persons with Diabetes (SUPD)	CMS Star D12	Link	Planned 2024 Q4
Transitions of Care	CMS Star C17	Link	Planned 2024 Q4
Urinary Incontinence	CMS Star C13, MIPS CQM 48	Link	Released

Instructions

Data Requirements

This data mart uses the following tables from the Tuva Core Data Model:

• condition
• encounter
• lab_result
• medication
• observation
• patient
• procedure
• medical_claim
• pharmacy_claim

Note: The Tuva Project will generate these Core tables. You just need to map your data to the input layer and run the project.

Variables

The data mart includes logic that allows you to choose a measurement period end date.

• quality_measures_period_end defaults to the current year-end
• snapshots_enabled is an optional variable that can be enabled to allow running the mart for multiple years

To run the data mart without the default, simply add the quality_measures_period_end variable to your dbt_project.yml file or use the --vars dbt command. See examples below.

dbt_project.yml:

vars:
    quality_measures_period_end: "2020-12-31"
    snapshots_enabled: true

dbt command:

# Uses defaults or vars from project yml, runs all marts
dbt build

# Runs only the Quality Mesures mart using defaults or vars from project yml
dbt build --select tag:quality_measures

# Overrides vars from project yml, executes snapshots
dbt build --select tag:quality_measures --vars '{quality_measures_period_end: "2020-12-31", snapshots_enabled: true}'

Data Mart Structure

Staging

The staging tables show what tables and fields are used from the Core data model.

Intermediate

The intermediate tables contain the logic for calculating each quality measure. The subfolder for each quality measure contains that measure's specific logic for calculating the denominator, numerator, and exclusions. Many measures use the same logic for calculating exclusions, such as dementia or hospice. This shared logic can be found in the shared exclusions subfolder.

Final

The final tables are an aggregated view of all quality measures and your population.

• Summary Counts: Reporting measure counts with performance rates.
• Summary Long: Long view of the results for the reporting version of all measures. Each row represents the results a measure per patient. A null for the denominator indicates that the patient was not eligible for that measure.
• Summary Wide: Wide view of the results for the reporting version of all measures. This model pivots measures on the patient level (i.e. one row per patient with flags for each measure. The false flags can be treated as care gaps as exclusions have been included in the pivot logic.

Data Dictionary

summary_counts

Column	Data Type	Description	Terminology

summary_long

Column	Data Type	Description	Terminology

summary_wide

Column	Data Type	Description	Terminology

Analytics

Quality Measure Performance

select
      measure_id
    , measure_name
    , performance_period_end
    , performance_rate
from quality_measures.summary_counts
order by performance_rate desc

Exclusion Reason Breakdown

select
      measure_id
    , exclusion_reason
    , count(patient_id) as patient_count
from quality_measures.summary_long
where exclusion_flag = 1
group by
      measure_id
    , exclusion_reason
order by
      measure_id
    , exclusion_reason

Patient Pivot

select * from quality_measures.summary_wide