Commitments and Best Practices for Healthcare Providers to Improve Data Quality Across All Levels of the Framework

The need to improve healthcare data quality comes from ongoing problems that affect patient care, public health, insurance claims, and technology use. These problems happen at three levels of healthcare data:

• Level 1: Atomic Data
• Level 2: Data Models and Schemas
• Level 3: Data Analysis Algorithms

Level 1: Atomic Data — The Foundation of Quality Healthcare Data

Atomic data is the smallest pieces of clinical information stored in Electronic Health Records (EHRs). Examples include vital signs, lab results, diagnoses, medications, and procedures. These are the basic parts that all other healthcare systems build on.

Common problems at Level 1 are:

• Synonym and Terminology Variations: For example, “systolic blood pressure” might be recorded in over 20 different ways. This causes confusion and mistakes.
• Format Errors and Missing Data: Some fields may be left blank or entered in the wrong format.
• Duplicate or Unvalidated Data: This can cause wrong medical decisions or billing problems.

To fix these issues, healthcare providers should use standard data models like the United States Core Data for Interoperability (USCDI) versions 1 and 3 along with HL7 FHIR (Fast Healthcare Interoperability Resources) standards. These give clear definitions and standard formats. This helps different organizations share data accurately.

Contracts and vendor deals now often require at least 80% of data to follow USCDI and HL7 FHIR standards by set deadlines. For example, USCDI v1 and HL7 FHIR US Core 4.0.0 are expected to be widely used within 12 months. More advanced models like USCDI v3 and HL7 FHIR US Core 6.1.0 should be adopted by January 1, 2026.

Level 2: Data Models and Schemas — Structuring Healthcare Data for Exchange

Level 2 deals with how atomic data is arranged, stored, and shared between systems. Healthcare groups use data models like OMOP (Observational Medical Outcomes Partnership) and FHIR-based schemas to organize data for sharing and analysis.

Main problems here include:

• Mapping Errors: When different groups use data models that don’t match, translating data can cause mistakes.
• Inconsistent Conformance: Providers and payers may not be sure if their systems fully follow data exchange rules.

To improve Level 2 data quality, providers and payers should:

• Use Validation Tools: Tools like the ONC’s Inferno test kit help check if systems follow the rules.
• Adopt Open Testing Methods and Scorecards: These help measure and share how well everyone is doing.
• Include Data Quality Terms in Agreements: Contracts should require following data exchange standards with over 80% compliance.

Improving data sharing reduces errors that can affect medical decisions or insurance claims. It helps build trust across the healthcare system.

Level 3: Data Analysis Algorithms — Ensuring Accurate and Useful AI Insights

This level focuses on the computer programs that analyze healthcare data. These programs help with research, population health, tracking diseases, and creating clinical measures. The results depend on data from Levels 1 and 2 being correct.

Common problems with AI include:

• Error Propagation: Bad atomic data or wrong data mapping can cause AI to make wrong or biased conclusions.
• Lack of Validation Frameworks: Without proper testing, AI tools may not work well in different healthcare situations.

Experts suggest:

• Fitness Validation: Make sure the data input matches what the AI is meant to analyze.
• Statistical and Clinical Testing: Confirm that AI models detect diseases correctly and meet rules.
• Keep Using Standards: Use USCDI and HL7 FHIR standards in data management plans.

Healthcare groups are encouraged to join data quality testing efforts with at least 80% of their partners. This helps create reliable AI systems for clinical care and policy work.

Committing to Data Quality: Industry Trends and Timelines

In August 2024, leaders from public and private sectors came together to work on improving healthcare data quality across the U.S. This group includes organizations like Leavitt Partners and the National Committee for Quality Assurance (NCQA). They want to make health data more consistent and trustworthy.

Healthcare providers and payers should:

• Use USCDI v1 and v3 data models and HL7 FHIR US Core standards in sharing clinical data.
• Include these standards in 80% of contracts and vendor agreements within 12 to 24 months.
• Use tools like ONC’s Inferno test kit to check their systems.
• Take part in pilot programs to test data quality methods.

NCQA’s Health Plan Accreditation program supports these steps by using trusted measures like HEDIS® and CAHPS® to evaluate care quality and patient experience. These data help find problems and match quality efforts with state and employer goals, like network availability and consumer protection.

NCQA has also added standards for health equity. This helps providers collect data needed to spot and reduce gaps in care among different groups.

Addressing Health Equity Through Data Quality Improvements

Health equity is an important issue for healthcare groups, especially those serving diverse patients. Studies and policies show that better data quality is key to fixing equity gaps.

Healthcare systems should:

• Learn About Health Inequities: Good data collection reveals differences by race, ethnicity, language, and income.
• Change Care Delivery: Accurate data helps clinics adjust how they work to better help patients.
• Change Organizational Culture: Leaders, especially nurse leaders, play a big role in raising equity awareness and making it part of daily work.
• Use Shared Language: Consistent terms help everyone communicate better, from care teams to policymakers.

Technology helps by enabling screening at the point of care and collecting real-time data. This lets clinicians find disparities early and change care plans when needed.

Healthcare leaders who improve data quality not only follow rules but also can create targeted programs that improve health for groups that need the most help.

AI Integration and Workflow Automation: Enhancing Data Use and Administrative Efficiency

Using AI and automation can also improve data quality and help healthcare office work run smoother. Providers who add automated phone systems see usable benefits.

AI-Powered Phone Automation: Improving Patient Engagement and Data Accuracy

Front-office phone work involves a lot of tasks like scheduling, triage, and answering questions. Using AI to handle calls can:

• Reduce Human Error: Automated systems can collect patient info, appointment times, and screening answers without mistakes.
• Improve Data Collection: AI can collect consistent information that meets Level 1 data quality.
• Enhance Patient Access: Systems that work 24/7 reduce wait times and give quick responses.
• Streamline Workflow: Staff can focus on harder tasks while AI handles routine calls.

AI in Data Management and Analysis

AI and machine learning also help clean and check data, find errors, and spot duplicates or missing items. This supports atomic data (Level 1) and better data models (Level 2).

Automated tools help keep up with standards like USCDI and HL7 by:

• Watching data inputs constantly.
• Helping data entry with standard templates.
• Supporting fast checks using built-in validation features.

Supporting Health Equity Initiatives with Automation

Automation can also help with health equity by adding standard questions about social factors during patient intake. AI can review answers to spot possible issues and help care teams respond.

Operationalizing Data Quality Through Education and Quality Improvement

Besides technology, education is important for better healthcare data. NCQA offers training programs for staff on data management, accreditation, and improving quality.

These courses focus on:

• Understanding national data standards and systems.
• Finding gaps in data quality.
• Aligning improvements with goals, including health equity.

With over 95% satisfaction, these programs help healthcare teams stay up to date on rules and good practices.

Healthcare leaders and IT managers should make training and ongoing learning a priority to keep high data standards.

Final Thoughts for Healthcare Providers in the United States

Improving healthcare data quality requires work in clinical, technical, and administrative areas. Practice managers, clinic owners, and IT staff should follow the three-level data quality framework for better patient care, meeting rules, and smoother operations.

• Using USCDI and HL7 FHIR standards is key to good atomic and structured data.
• Validation tools like ONC’s Inferno test kit and open scorecards help check data quality and sharing.
• Joining quality frameworks and accreditation programs such as NCQA helps focus on clinical and equity goals.
• AI automation, including phone answering services, lowers errors and improves data collection.
• Training staff regularly boosts the ability to maintain these improvements.

By following these steps, healthcare providers in the U.S. can make their data more reliable and useful. This helps in daily patient care and supports the broader use of AI and improved workflows for better healthcare and fairer outcomes.