Case Study  ·  ACO LEAD Readiness  ·  Illustrative Engagement

From Attribution Gaps to
LEAD-Ready Infrastructure in 8 Weeks

4,832
Attributed Beneficiaries
6.2% → 1.8%
Attribution Gap Reduced
38 → 74
LEAD Readiness Score
8 wks
Engagement Duration

This case study is available as a PDF for sharing with your leadership team or board.

↓ Download PDF
01 — Situation

A mid-sized ACO REACH entity in the mid-Atlantic region — 4,832 attributed Medicare beneficiaries across 14 primary care practices — engaged Integris Health Data Systems in early 2026 ahead of the ACO LEAD model launch. The organization had been successfully operating under REACH since 2023, but leadership recognized that the infrastructure built for a 4-year model cycle was not designed for a 10-year accountability commitment with monthly attribution reconciliation.

Their data team consisted of two part-time analysts managing CCLF files manually in Excel, a data warehouse that had not been updated since 2024, and a vendor-provided dashboard they could not customize or extend. They had no BCDA API integration, no dbt transformation layer, and no equity reporting capability — despite the LEAD model’s explicit equity reporting mandate.

“We knew we were going to be in LEAD. We didn’t know how unprepared our data infrastructure was until Integris ran the assessment. Some of the gaps were things we had just never thought about.”

— ACO Executive Director (representative, illustrative engagement)
02 — Assessment Findings

Integris conducted a structured LEAD Readiness Assessment across six domains using the bcda-client library against the CMS sandbox and a CCLF file sample. The initial score was 38/100 — classified as Critical Gaps.

DomainInitial ScoreSeverityKey Finding
Data Infrastructure20/100CriticalNo BCDA API connection. CCLF files processed manually in Excel. No automated pipeline.
Attribution Accuracy35/100Critical6.2% of roster MBIs absent from BCDA Patient responses — 301 beneficiaries missing from care management programs.
Equity Reporting10/100CriticalNo race/ethnicity stratification. No SDOH data. LEAD equity reporting mandate unaddressed.
Care Management Data30/100HighNo care gap registry. No ADT feed. TCM workflow not connected to claims data.
FHIR Compliance60/100ModerateUsing BCDA v1 (FHIR STU3). LEAD requires FHIR R4. Migration path straightforward.
Analytics Maturity25/100HighVendor dashboard not extensible. No gold-layer analytics. No reproducible transformation layer.
Attribution Gap Deep Dive

The most operationally significant finding was a persistent attribution gap between the ACO’s CMS-issued roster and the BCDA Patient NDJSON data. Across six months of retrospective analysis, an average of 6.2% of attributed beneficiaries present in CCLF8 were absent from BCDA Patient responses — a documented CMS Beneficiary FHIR Data Server limitation many ACOs do not actively monitor.

141
Nov
125
Dec
153
Jan
101
Feb
92
Mar
87
Apr

Roster-only MBIs declining after reconciliation pipeline deployment

03 — Work Performed
Weeks 1–2
BCDA API Integration & First Production Pull
Deployed bcda-client to ACO environment. Configured credentials, ran first BCDA production pull (Patient, EOB, Coverage NDJSON). Validated MBI extraction against CCLF8. Identified and documented 301 roster-only MBIs.
Weeks 2–3
CCLF Parser & Reconciliation Engine
Built automated CCLF0–CCLF8 ingestion pipeline replacing Excel workflows. Deployed three-way reconciliation engine: roster CSV ↔ BCDA ↔ CCLF8. Generated ReconciliationReport with gap classification for each missing beneficiary.
Weeks 3–5
dbt Bronze / Silver / Gold Transformation Layer
Deployed dbt project with Bronze (raw ingestion), Silver (normalized patient + claims master), and Gold (population summary, care gap registry, equity dashboard, utilization trends) layers. Replaced vendor dashboard dependency with owned analytics.
Weeks 5–6
Equity Reporting Infrastructure
Built equity dashboard stratifying population by race, dual-eligible status, and geography. Mapped CCLF8 race codes to LEAD-required reporting categories. Documented baseline disparity metrics for LEAD Year 1 comparison.
Weeks 6–8
Care Gap Registry & Automated Scheduling
Deployed care gap registry — 1,847 open gaps identified, prioritized by HCC relevance and risk tier. Built Airflow DAG automating weekly BCDA pull, CCLF reconciliation, dbt transformation, and Slack notification to care management team.
04 — Results
38
Before · Critical Gaps
74
After · LEAD Ready
Domain Score Improvement
Data Infrastructure
85/100
Attribution Accuracy
78/100
Equity Reporting
65/100
Care Management Data
60/100
FHIR Compliance
90/100
Analytics Maturity
70/100
Key Operational Outcomes
MetricBeforeAfterStatus
Attribution match rate (BCDA vs. roster)93.8%98.2%Resolved
Beneficiaries missing from care management30187Resolved
CCLF processing time2–3 days (manual)<4 hours (automated)Resolved
Open care gaps identifiedUnknown1,847 (prioritized)New Capability
Equity reporting baselineNone6 stratification categoriesLEAD Compliant
ADT feed integrationNoneScoped, in progressIn Progress

“The attribution gap alone — 301 beneficiaries we weren’t tracking — justified the entire engagement. Those are real patients who were attributed to us but invisible in our care management system.”

05 — Ongoing & Next Phase

Following the 8-week engagement, the ACO retained Integris on a monthly analytics retainer covering:

A methods paper documenting the BCDA field-mapping gap discovered during this engagement — with longitudinal attribution concordance data — is in preparation for submission to JAMIA or Applied Clinical Informatics in Q3 2026.

06 — About Integris Health Data Systems

Integris Health Data Systems

Virginia-based health AI research and data infrastructure firm founded by an MD-trained biomedical informatics scientist (PhD, FAMIA) with a decade of national health data program leadership. We build the data infrastructure that ACOs, FQHCs, and health systems need to operate in value-based care — with the domain knowledge to understand what the data means and the engineering execution to make it run reliably.

Our work is open-source first, equity-by-design, and grounded in peer-reviewed research. Every pipeline we build is also a contribution to the published literature on health data infrastructure.

All data in this case study is synthetic and representative of the types of outcomes achievable through Integris’s ACO data infrastructure engagements. No PHI was used. The BCDA field-mapping gap described reflects a documented limitation of the CMS Beneficiary FHIR Data Server acknowledged by CMS. This document illustrates Integris’s service delivery approach and methodology.