ACO REACH data quality requirements and trust infrastructure needs

ACO REACH is the largest risk-bearing accountable care model CMS has ever run. More than 200 ACO REACH entities manage care for over 2 million Medicare beneficiaries. Every one of those entities is financially accountable for outcomes it can only manage if the underlying data is accurate, complete, and timely. CMS publishes data quality requirements. But what CMS requires and what trust actually demands are two different standards.

What CMS actually requires from ACO REACH participants

CMS structures ACO REACH around total cost of care benchmarks, quality performance, and health equity reporting. Participants receive prospective beneficiary alignment files, claims data feeds, and are expected to submit clinical quality measures and demographic data.

The model requires participants to collect and report social determinants of health (SDOH) data, including screening for housing instability, food insecurity, and transportation barriers. CMS also mandates the use of Certified Electronic Health Record Technology (CEHRT) for a minimum percentage of participating providers.

But here is the gap: CMS specifies what data to submit, not what quality standard it must meet before submission. There is no trust score. There is no provenance chain. There is no automated check for whether the SDOH screening data was collected via a validated instrument or free-text entry by a medical assistant who was told to "just put something in."

What are the quality measures used in the ACO REACH model?

ACO REACH uses three primary quality measures to evaluate performance:

ACO-quality measures based on CAHPS (Consumer Assessment of Healthcare Providers and Systems): Patient experience survey scores measuring access, communication, and care coordination.

Claims-based quality measures: Including all-cause unplanned admissions and timely follow-up after acute exacerbation of chronic conditions.

Health equity measures: CMS added requirements for demographic data collection and stratified quality reporting to surface disparities.

These measures are tied directly to the quality withhold. CMS holds back a percentage of shared savings and only releases it based on quality performance. For Performance Year 2024, the quality withhold is 5% of gross savings.

The problem is that every one of these measures depends on underlying data quality that CMS does not independently verify at the record level. CAHPS surveys depend on accurate beneficiary contact information. Claims-based measures depend on correct ICD-10 and CPT coding. Health equity measures depend on self-reported race, ethnicity, and SDOH data that is frequently incomplete or inferred.

What is the definition of data quality in healthcare?

Data quality in healthcare refers to the degree to which data is accurate, complete, timely, consistent, and fit for its intended use. The most commonly cited dimensions are:

Accuracy: Does the data correctly represent the real-world clinical event?

Completeness: Are all required fields populated with meaningful values?

Timeliness: Was the data captured and made available close enough to the event to be actionable?

Consistency: Does the same patient show the same diagnosis across systems?

Validity: Does the data conform to expected formats, code sets, and value ranges?

But traditional data quality frameworks stop at the record level. They do not ask: where did this record come from? Was the patient aware their data was being used for risk adjustment? Has this record been modified since it was created? These are questions of data trust, not just data quality. The distinction matters.

Key statistics

Data Trust Index (DTI) dimension weights for ACO REACH data scoring

ACO REACH and accountable care data represent some of the highest-stakes data environments in U.S. healthcare. The numbers make the trust gap concrete.

200+ ACO REACH entities manage care for over 2 million Medicare beneficiaries under total cost of care risk arrangements.

5% quality withhold on gross shared savings is tied directly to quality measure performance, making data accuracy a direct financial variable.

CMS reports average savings of approximately $930 per beneficiary in ACO REACH relative to fee-for-service benchmarks, but these calculations depend entirely on accurate claims and clinical data.

SuperTruth's imaware case study showed that standardizing 105,000 diagnostic records reduced processing time from 3 weeks to 2 hours, a 95% reduction, and saved 200+ hours per month of manual reconciliation.

Studies estimate that 30% of SDOH screening fields in EHRs are either blank or populated with default values, undermining the health equity measures CMS requires.

The trust infrastructure gap in accountable care

ACO REACH entities receive beneficiary-level data from CMS, including claims history, demographic files, and prospective alignment lists. They combine this with clinical data from their provider networks, pharmacy data from PBMs, and increasingly, SDOH data from community-based organizations.

Each of these data sources has a different provenance chain. Claims data arrives with a 30 to 90 day lag. Clinical data quality varies by EHR vendor and site configuration. SDOH data from CBOs often lacks standardized formats and may arrive as PDFs, spreadsheets, or unstructured notes.

No single system scores all of these inputs on a common trust framework before they are used for risk stratification, care management, or quality reporting. The result is that ACO REACH entities are making million-dollar risk decisions on data they have aggregated but not verified.

This is not a hypothetical problem. When an ACO REACH entity uses inaccurate HCC (Hierarchical Condition Category) codes for risk adjustment, it either leaves money on the table or triggers a CMS audit. When SDOH screening data is incomplete, the health equity adjustment fails. When clinical quality data is inconsistent across sites, the quality withhold is not released.

What is the downside of an ACO?

ACOs carry financial risk that scales with data quality failures. The most significant downsides include:

Financial exposure from inaccurate risk adjustment. ACO REACH entities in the Global track bear full downside risk. If their attributed population appears healthier than it actually is due to coding inaccuracies, the benchmark is set too low and the ACO loses money on every beneficiary.

Operational cost of data reconciliation. Most ACO REACH entities spend hundreds of hours per month reconciling claims, clinical, and demographic data across systems. This is the same problem SuperTruth solved for imaware, where 200+ hours of monthly manual work disappeared after DTI-based standardization.

Audit liability from unverifiable data. CMS conducts retrospective audits of risk adjustment and quality reporting. If an ACO cannot demonstrate the provenance of the data it used for HCC coding or quality measure numerator/denominator calculations, it faces recoupment.

Health equity penalties from incomplete demographic data. CMS is increasing the weight of health equity in ACO REACH scoring. Entities that cannot demonstrate complete, validated demographic and SDOH data will face scoring penalties that directly reduce their shared savings.

What are the four key principles of effective quality improvement programs?

Effective quality improvement in healthcare, including in ACO REACH, follows four principles:

Measurement before intervention. You cannot improve what you cannot measure. In ACO REACH, this means having accurate baseline data on utilization, cost, and quality before designing care management programs.

Standardization of processes. Quality improvement requires that care processes and data collection methods are consistent across sites. An ACO REACH entity with 50 participating practices needs all 50 to screen for SDOH using the same validated instrument, not 50 different workflows.

Continuous feedback loops. Data must flow back to clinicians and care teams in near-real time. A quality measure result that arrives 6 months after the measurement period is useless for improvement. This is where recency scoring becomes critical.

Accountability at every level. Quality improvement fails when accountability is diffuse. In ACO REACH, this means that data quality accountability cannot sit only with a central analytics team. Every provider site that generates data is responsible for the quality of that data.

These four principles all depend on a trust infrastructure that can score incoming data at the point of ingestion and flag records that fail to meet minimum quality thresholds before they contaminate downstream analytics.

Why existing health data networks do not solve the ACO REACH trust problem

ACO REACH entities often rely on health information exchanges (HIEs), CommonWell, Carequality, and direct EHR integrations to aggregate clinical data. These networks solve a connectivity problem. They do not solve a trust problem.

CommonWell and Carequality move data between systems, but they do not score it. A CCD document that arrives via Carequality may contain duplicate problem lists, stale medication data, or missing allergy entries. The receiving ACO REACH entity has no automated way to know.

Similarly, EHR vendors like Epic and Oracle Health provide data exchange capabilities but do not provide trust scoring. An Epic installation at one ACO REACH site may capture SDOH data in structured FHIR resources while another captures it in free-text clinical notes. Both are "in the EHR." Only one is computationally usable for quality reporting.

What accountable care data trust actually requires

Trust infrastructure for ACO REACH must operate at the record level, not the system level. It must score every incoming data element across dimensions that matter for accountable care:

Provenance (25% of DTI weight). Where did this record originate? Was it generated by a CEHRT-certified system? Has it been modified since creation? For risk adjustment, provenance determines whether an HCC code will survive a CMS audit.

Consent (20%). Was the beneficiary's data collected and shared in compliance with HIPAA and any state-level restrictions? For SDOH data especially, consent governance is complex because screening data may flow from CBOs that are not covered entities.

Recency (15%). Is this record current enough to inform care decisions? A blood pressure reading from 18 months ago should not be used for a current quality measure numerator. Claims data that has not been adjudicated should be flagged.

Quality (10%). Are required fields populated? Are values within expected ranges? Is the ICD-10 code valid for the date of service?

Concordance (10%). Does this record agree with other records for the same patient? If two systems show different primary diagnoses for the same beneficiary, which one is correct?

These are the dimensions that the Data Trust Index scores on a 0-100 scale for every record. ACO REACH entities that deploy DTI scoring at the point of data ingestion can set minimum trust thresholds. Any record below the threshold gets flagged for manual review before it enters the analytics pipeline.

The financial math of data trust in ACO REACH

Data reconciliation time: before vs after DTI scoring (imaware case study)

Consider an ACO REACH entity managing 10,000 beneficiaries with average savings of $930 per beneficiary. That is $9.3 million in potential shared savings. The 5% quality withhold means $465,000 is at risk based on quality measure performance.

If 10% of the clinical quality data used for measure calculation is inaccurate, stale, or incomplete, the entity's quality scores drop. A single quality measure failure can reduce the quality score enough to forfeit the entire withhold.

Now consider the cost of data reconciliation. If the entity spends 200 hours per month on manual data matching and validation (the pre-DTI baseline from the imaware case study), that is 2,400 hours per year. At $75 per hour for analyst time, that is $180,000 annually in labor costs that automated trust scoring eliminates.

The ROI is not abstract. It is $465,000 in protected quality withhold plus $180,000 in reduced reconciliation costs, against the cost of deploying a trust scoring infrastructure.

SDOH data: the most fragile layer in ACO REACH reporting

CMS requires ACO REACH entities to screen beneficiaries for social risk factors and report demographic data stratified by race, ethnicity, and social needs. This requirement exposes the most significant data quality gap in the model.

SDOH data comes from multiple sources: AHC-HRSN screening instruments, community health worker assessments, CBO intake forms, and patient self-report. Each source has different provenance characteristics, different consent models, and different recency profiles.

Address-level SDOH data used for geographic risk stratification depends on geocoding accuracy. Housing instability data from public records may be stale. Food insecurity screening data may be captured in non-standard formats.

Without a trust scoring layer that evaluates each SDOH data element for completeness, recency, and provenance, ACO REACH entities are reporting health equity metrics on data they cannot stand behind.

What changes with trust-scored accountable care data

When every record entering an ACO REACH data pipeline receives a DTI score, three things change:

First, risk adjustment accuracy improves. Records with low provenance or concordance scores get flagged before they are used for HCC coding. This reduces both under-coding (lost revenue) and over-coding (audit risk).

Second, quality measure integrity improves. Only records meeting minimum trust thresholds are included in measure numerator and denominator calculations. This means the quality scores reported to CMS reflect actual clinical performance, not data artifacts.

Third, health equity reporting becomes defensible. When CMS asks how an ACO REACH entity collected and validated its demographic and SDOH data, the entity can point to a trust score for every record, with provenance documentation and consent verification built in.

This is not a compliance exercise. It is the infrastructure that makes accountable care financially viable at scale.

The DTI Engine scores every health data record 0-100 across 8 trust dimensions before your AI model or analytics pipeline sees it. If your ACO REACH entity, health system, or managed care organization is building risk-bearing contracts on data it has not verified, the trust gap is a financial liability. Talk to the SuperTruth commercial team. Schedule a conversation or call (215) 918-4140.

Further reading:

DTI™ Engine

Health systems solution

CMS Innovation Center value-based model data requirements

Health data integrity for value-based care programs

Claims data lag: what 30-90 day reporting delays cost AI models