Care Gaps (Quality Measures)

Overview

Care gaps, synonymous with quality measures, are standardized healthcare metrics defined by various entities—primarily government-affiliated organizations like the National Committee for Quality Assurance (NCQA)—to assess the performance of healthcare providers and payers. These measures evaluate whether patients receive evidence-based preventive care and chronic disease management services.

Why Care Gaps Matter

Quality measures serve several critical purposes:

1. Performance Assessment - Evaluate how well providers and payers deliver evidence-based care to their patient populations
2. Financial Incentives - Many health plans and government programs tie reimbursement to quality measure performance
3. Regulatory Compliance - Failure to meet national standards for these metrics can result in financial penalties
4. Patient Outcomes - Closing care gaps improves population health by ensuring patients receive necessary preventive and chronic care services

Types of Quality Measures

HEDIS Measures

HEDIS (Healthcare Effectiveness Data and Information Set) measures are standardized quality metrics maintained by NCQA. These measures:

• Follow strict technical specifications for eligibility criteria, numerator, and denominator definitions
• Are used nationally for health plan comparisons and reporting
• Are required for many government and commercial health plan contracts
• Change annually with updated specifications

Operational Measures

Operational measures are internally defined quality metrics that:

• May have less restrictive criteria than HEDIS counterparts
• Are tailored to Pair Team's care model and operational needs
• Help care teams identify patients who need outreach
• Complement HEDIS measures with actionable insights

Measures We Track

Our care gaps models calculate 27 distinct quality measures across several clinical domains:

Chronic Disease Management

Measure Code	Measure Name	Type	Description
cbp_hedis	Controlling Blood Pressure	HEDIS	Blood pressure control for patients with hypertension
cbp_operational	Controlling Blood Pressure	Operational	Internal BP control measure
gsd_hedis_lt8	Glycemic Status - A1C < 8%	HEDIS	Diabetes patients with A1C under control
gsd_hedis_gt9_or_missing	Glycemic Status - A1C > 9% or Missing	HEDIS	Diabetes patients with poor control or no recent test
gsd_operational_lt8	Glycemic Status - A1C < 8%	Operational	Internal diabetes control measure
gsd_operational_gt9_or_missing	Glycemic Status - A1C > 9% or Missing	Operational	Internal diabetes control measure

Cancer Screening

Measure Code	Measure Name	Type	Description
bcs_hedis	Breast Cancer Screening	HEDIS	Mammography screening for women 50-74 years
ccs_hedis	Cervical Cancer Screening	HEDIS	Pap/HPV screening for women 21-64 years
col_e	Colorectal Cancer Screening	HEDIS	Screening for adults 45-75 years

Preventive Care - Adults

Measure Code	Measure Name	Type	Description
pcp_operational	Primary Care Visit	Operational	Annual primary care visit completion
chl	Chlamydia Screening	HEDIS	Screening for sexually active women 16-24 years
dsf_e	Depression Screening and Follow-Up	HEDIS	Depression screening with appropriate follow-up
drr_hedis_1	Depression Remission at 12 Months	HEDIS	Depression improvement at 12 months (PHQ-9 based)
drr_hedis_2	Depression Remission at 6 Months	HEDIS	Depression improvement at 6 months (PHQ-9 based)
drr_hedis_3	Depression Response at 12 Months	HEDIS	Depression response at 12 months (PHQ-9 based)

Preventive Care - Children

Measure Code	Measure Name	Type	Description
cis_10	Childhood Immunization Status	HEDIS	Completion of childhood immunization schedule by age 2
ima_2	Immunizations for Adolescents	HEDIS	Completion of adolescent immunizations
lead_screening_hedis	Lead Screening in Children	HEDIS	Lead screening for children at appropriate ages
dev	Developmental Screening	HEDIS	Developmental screening in first 3 years of life
w30_6	Well-Child Visits (30+ months)	HEDIS	Well-child visits ages 30 months to 6 years
w30_2	Well-Child Visits (15-30 months)	HEDIS	Well-child visits ages 15-30 months
wcv	Well-Child Visits	HEDIS	Well-child visits in first 30 months

Maternal Health

Measure Code	Measure Name	Type	Description
ppc_operational	Prenatal and Postpartum Care	Operational	Comprehensive prenatal and postpartum care
ppc_pre	Prenatal Care	HEDIS	Timely prenatal care initiation
ppc_pst	Postpartum Care	HEDIS	Postpartum visit within 21-56 days after delivery
pds_e	Postpartum Depression Screening	HEDIS	Depression screening in postpartum period
pnd_e	Prenatal Depression Screening	HEDIS	Depression screening during pregnancy
prs_e	Prenatal Immunization Status	HEDIS	Immunization status for pregnant patients

Data Sources

Care gaps calculations draw from multiple data sources to identify qualifying events:

• Zus (Health Information Exchange data) - Encounters, procedures, conditions, medications, lab results from across the healthcare ecosystem
• Elation (EHR data) - Internal clinical encounters, vitals, lab results, procedures documented by care team
• Arc (Care management data) - Forms, assessments, care coordination activities

These sources are combined to create a comprehensive view of each patient's care history for accurate gap identification.

Model Architecture

The care gaps models follow a structured pipeline:

staging/
├── stage__care_gaps_hedis_codes_*          # HEDIS code sets (CPT, ICD-10, LOINC, etc.)
├── stage__care_gaps_periods_*              # Enrollment and calendar periods
└── stage__care_gaps_hedis_medication_*     # NDC and RxNorm medication codes

intermediate/
├── int__care_gaps_event_*                  # Event extraction by type (encounters, procedures, etc.)
├── int__care_gaps_event_union              # Combined events from all sources
├── int__care_gaps_event_matrix             # Patient-measure-period-event mappings
├── int__care_gaps_event_aggregates         # Aggregated events by measure requirements
├── int__care_gaps_aggregate_*              # Measure-specific aggregation logic
├── int__care_gaps_relative_events          # Latest readings (BP, A1C) within periods
└── int__care_gaps_measure_classification   # Final measure classification logic

marts/
├── care_gaps_patient_summary               # One row per patient per measure per period
├── care_gaps_events                        # All qualifying events for care gaps
└── care_gaps_drr_measure_detail            # Additional detail for DRR measures

Output Models

care_gaps_patient_summary

The primary output model for care gaps reporting. One row per patient per measure per time period.

Key Fields:

• person_measure_period_id - Unique identifier for patient-measure-period combination
• person_id - Patient identifier
• measure_name - The quality measure code (e.g., 'cbp_hedis', 'bcs_hedis')
• time_dimension_name - Time period identifier
• start_date / end_date - Measurement period dates
• denominator_eligible_for_measure - Boolean indicating if patient is in the eligible population
• numerator_meets_measure - Boolean indicating if patient met the measure requirements (gap is closed)

Usage:

• Calculate measure compliance rates by dividing numerator by denominator
• Identify patients with open gaps (denominator = TRUE, numerator = FALSE)
• Track gap closure over time
• Generate care team work lists for outreach

care_gaps_events

All qualifying events that contribute to care gap calculations.

Key Fields:

• event_id - Unique event identifier
• person_id - Patient identifier
• event_type - Type of event (encounter, procedure, condition, etc.)
• event_date - When the event occurred
• Additional fields describing the clinical event details

Usage:

• Audit care gap calculations
• Understand which events closed specific gaps
• Troubleshoot measure classification logic

care_gaps_drr_measure_detail

Additional detail for the three DRR (Depression Remission or Response) measures, which use PHQ-9 scores to assess depression improvement over time.

Usage:

• Track PHQ-9 score changes over time
• Audit DRR measure calculations
• Understand depression treatment effectiveness

Usage Notes

1. Data Refresh: Care gaps models are refreshed nightly in production (DBT_PROD schema)
2. Development: Use DBT_DEV_{username} schema for testing changes
3. Testing: Run dbt build --select +care_gaps_patient_summary+ to test the full pipeline
4. Lookback Windows: Most preventive care measures have multi-year lookback windows (e.g., cervical cancer screening every 3 years, mammography every 2 years)
5. Enrollment Requirements: HEDIS measures typically require continuous enrollment during the measurement period

warning

HEDIS measure specifications are updated annually and must be reviewed each year. Code sets (CPT, ICD-10, LOINC, etc.) must be updated to reflect specification changes.

Key Considerations

• Code Set Maintenance: HEDIS code sets (CPT, ICD-10, LOINC, etc.) must be updated annually to reflect specification changes
• Measure Specifications: HEDIS specifications are complex with detailed inclusion/exclusion criteria, stratifications, and allowable gaps in enrollment
• Time Periods: Different measures use different measurement periods (calendar year, rolling 12 months, etc.)

warning

Care gaps depend heavily on data from Zus (HIE) since many qualifying events occur outside Pair Team. Some gaps may appear open in our data but be closed in other systems (e.g., screenings done at non-participating providers).