T4.3: Monitors Data Quality

Overview

According to sample exam question Q17, when management asks for a process to monitor incoming data quality over the lifetime of the system, the key tool is the Data Quality tool (not the data integrity check utility, configuration check utility, usage dashboards, or definition designer). The Data Quality tool provides ongoing monitoring capabilities through business rules, trending pivots, and dashboards that track data quality metrics as data flows into the EMPI from source facilities.

Effective data quality monitoring is essential for maintaining the accuracy and completeness of the EMPI. Poor data quality leads to incorrect linkages, missing composite view data, and degraded patient matching performance. The Data Quality tool enables proactive detection of quality issues before they impact clinical operations.

The Data Quality Manager

The Data Quality Manager is accessed through the EMPI Management Portal and provides:

• Data Quality Properties: Business rules that classify field values as valid, invalid, or missing
• Data Quality Trending Pivots: Definitions of pivot tables for tracking quality trends over time
• View Data Quality Dashboards: Link to analytics dashboards displaying quality metrics
• Manage Synchronization Frequency: Settings controlling how often quality cubes are updated

Navigate to Person Index > Data Quality Manager to access the Data Quality configuration and monitoring tools.

Enabling Data Quality

Before using the Data Quality tool, it must be enabled:

1. Navigate to Person Index > Settings in the Management Portal 2. Select Enable Data Quality 3. This generates the Data Quality cubes and configures default synchronization times 4. Select Rebuild Cube Data to populate the cubes with data

Important: A linkage data build must be completed before rebuilding Data Quality cube data. The cubes depend on the normalized linkage data.

Once enabled, the Data Quality tool continuously monitors incoming data and updates quality metrics based on the configured schedule.

Data Quality Properties (Business Rules)

Data Quality Properties are business rules that determine what counts as valid, invalid, and missing values for each demographic field:

Classifier Types:

• Strong: Uses the "strong" classifier rule to categorize values
• Custom: Uses a custom classifier rule for specialized validation

Classifier Rules: Each demographic property has an associated classifier rule that categorizes field values:

• HSDQ.NameRule: Classifies name fields (first, middle, last) as valid or invalid based on character content and length
• HSDQ.SSNRule: Validates Social Security numbers for format and known invalid patterns
• HSDQ.PhoneRule: Validates phone numbers for format and completeness
• HSDQ.ZipCodeRule: Validates postal codes
• HSDQ.StateRule: Validates state/province codes
• HSDQ.NameSuffixRule: Validates name suffixes
• Custom Rules: Organizations can create custom classifier rules by subclassing HSDQ.AbstractRule

Classifier Outputs:

• Valid: Field value meets quality criteria
• Invalid: Field value fails quality criteria (e.g., name contains numbers)
• Missing: Field value is null or empty

Configuring Properties: 1. Navigate to Data Quality Manager > Data Quality Properties 2. Select a property row to view and edit rules 3. Choose Classifier Type (Strong or custom rule name) 4. Save changes 5. Rebuild cube data to apply new rules

Data Quality Trending Pivots

Trending pivots track data quality metrics over time, enabling detection of quality degradation:

Default Pivots:

• HSPI Data Quality Trending/BirthDateTime Validity Reason: Tracks validity of birth date values over time
• HSPI Data Quality Trending/City Validity Reason: Tracks city field quality
• HSPI Data Quality Trending/FamilyName Validity Reason: Tracks last name quality
• HSPI Data Quality Trending/GivenName Validity Reason: Tracks first name quality
• HSPI Data Quality Trending/PhoneNumber Validity Reason: Tracks phone number quality
• HSPI Data Quality Trending/SSN Validity Reason: Tracks SSN quality
• HSPI Data Quality Trending/State Validity Reason: Tracks state/province quality

Each pivot captures the count of valid, invalid, and missing values over time, broken down by facility, allowing identification of facilities with declining data quality.

Customizing Pivots: Organizations can define additional trending pivots to track specific quality metrics relevant to their operations.

---

Overview

Ongoing data quality management requires establishing regular review processes, defining quality baselines and thresholds, and implementing corrective action workflows. Effective monitoring transforms data quality metrics into actionable insights that drive continuous improvement in source system data quality.

The Data Quality tool provides the metrics, but successful quality management depends on organizational processes that review metrics, identify trends, engage with facilities, and track improvements.

Regular Review Schedule

Establish a regular schedule for data quality reviews:

• Weekly Reviews: For new implementations or facilities with known quality issues
• Monthly Reviews: For mature implementations with stable quality
• Quarterly Reviews: For executive reporting and long-term trend analysis

Data quality reviews should involve:

• EMPI Operators: Front-line users who interact with data quality issues daily
• HIM Leadership: Representatives from Health Information Management who can engage with facilities
• IT Support: Technical staff who can investigate system-level quality issues
• Facility Representatives: When specific facilities show quality degradation

Dashboard Review Process

During each review, examine the Data Quality dashboards:

1. Navigate to Dashboards: Person Index > Data Quality Manager > View Data Quality Dashboards 2. Review Overall Trends: Look at facility-level quality metrics across all properties 3. Identify Anomalies: Flag facilities or time periods with sudden quality changes 4. Drill Down: Use pivot table drill-down capabilities to investigate specific issues 5. Compare Facilities: Identify facilities with consistently lower quality than peers

Key Metrics to Track:

• Percentage of Valid Records: Overall data quality score per facility
• Missing Data Rates: Percentage of null or empty values for critical fields
• Invalid Data Rates: Percentage of values that fail validation rules
• Trend Direction: Whether quality is improving, stable, or degrading over time

Facility Scorecards

Create facility scorecards that summarize quality metrics for each data source:

Scorecard Components:

• Overall Quality Score: Weighted average of valid percentages across all properties
• Critical Field Quality: SSN, name, date of birth quality metrics
• Trend Indicator: Arrow or color showing improving/stable/declining quality
• Last Review Date: When the facility's data quality was last reviewed
• Action Items: Open corrective actions for the facility

Scorecards provide an at-a-glance view of facility performance and facilitate prioritization of quality improvement efforts.

Baselines and Thresholds

Establish quality baselines and alert thresholds:

Baseline Establishment:

• During initial go-live, establish baseline quality levels for each facility
• Document baseline metrics as the expected "normal" quality level
• Recognize that baseline quality varies by facility based on source system capabilities

Alert Thresholds:

• Yellow Alert: Quality drops 5-10% below baseline
• Red Alert: Quality drops >10% below baseline or critical field quality <80%
• Missing Data Alert: >20% missing values for critical fields (name, DOB, SSN)

Thresholds trigger proactive engagement with facilities before quality issues impact operations.

---

Overview

Monitoring incoming data feeds for quality degradation enables early detection of source system problems, registration workflow changes, or interface issues. Proactive monitoring prevents quality issues from accumulating over time and ensures that corrective actions are taken promptly.

The Data Quality tool provides the metrics, but effective feed monitoring requires alerting mechanisms, investigation workflows, and facility engagement processes.

Feed-Level Quality Tracking

Track quality metrics at the data feed level:

• Feed Identification: Each facility typically has one or more data feeds (ADT feed, lab feed, etc.)
• Per-Feed Metrics: Track quality separately for each feed to isolate problems
• Temporal Analysis: Compare current quality to historical averages for the feed
• Volume Correlation: Correlate quality changes with message volume changes

Use the Data Quality dashboards to filter by facility and time period to focus on specific feed quality.

Real-time Quality Alerts

Implement alerting for quality threshold violations:

Alert Mechanisms:

• Dashboard Annotations: Highlight facilities exceeding thresholds on dashboards
• Email Notifications: Send automated emails when thresholds are violated
• Event Log Entries: Log quality alerts to the production Event Log
• Escalation Workflows: Define escalation paths for unresolved quality issues

Alert Content:

• Facility name and feed identifier
• Specific quality metric that violated threshold (e.g., "SSN invalid rate: 25%")
• Threshold value and current value
• Time period of the violation
• Recommended actions

Root Cause Investigation

When quality degradation is detected, perform root cause analysis:

Investigation Steps: 1. Verify Trend: Confirm that quality change is real, not a reporting anomaly 2. Identify Timing: Determine when quality degradation began 3. Review Facility Changes: Contact facility to ask about recent system changes, workflow updates, or staff changes 4. Sample Data Review: Examine sample records from the feed to identify specific quality issues 5. Interface Review: Check interface logs for errors, format changes, or missing fields

Common Root Causes:

• Source System Upgrade: EMR or registration system upgrade changed data format or content
• Workflow Change: New registration workflow omits previously required fields
• Interface Modification: HL7 interface mapping changed, dropping or altering fields
• Staff Training: New staff not trained on proper data entry procedures
• System Error: Source system bug generating invalid or missing data

Facility Engagement Process

Engage with facilities to address quality issues:

1. Initial Contact: Notify facility HIM or IT contact of quality issue with specific examples 2. Data Review: Provide sample records showing quality problems 3. Root Cause Discussion: Work with facility to identify source of quality degradation 4. Corrective Action Plan: Define actions facility will take to improve quality 5. Follow-up Monitoring: Track quality metrics over subsequent weeks to verify improvement 6. Escalation: If quality does not improve, escalate to facility leadership or organizational quality committee

Documentation:

• Document all facility contacts, issues identified, and corrective actions in a quality tracking system
• Maintain a log of quality issues by facility to identify chronic problem sources
• Report quality trends to EMPI governance committee

---

Overview

The Data Quality tool provides rich analytics capabilities through dashboards, pivot tables, and reports. These visualizations transform raw quality metrics into actionable insights. Effective use of these tools enables data quality teams to identify trends, prioritize actions, and communicate quality status to stakeholders.

Data Quality Dashboards

Access Data Quality dashboards via Person Index > Data Quality Manager > View Data Quality Dashboards.

Available Dashboards:

• Data Quality Overview: High-level summary of quality metrics across all facilities
• Facility Quality Scorecard: Detailed quality metrics for individual facilities
• Trending Analysis: Quality trends over time, showing improvement or degradation
• Field-Specific Quality: Quality metrics for specific demographic fields (name, SSN, DOB)

Dashboard Features:

• Interactive Filtering: Filter by facility, time period, or field
• Drill-Down: Click on metrics to drill down to underlying data
• Export: Export dashboard data to Excel or PDF for reporting
• Refresh: Update dashboards with latest cube data

Pivot Table Analysis

Pivot tables provide detailed drill-down analysis:

Using Pivots: 1. Select a Data Quality Trending Pivot from the list 2. Choose rows, columns, and measures to display 3. Filter by time period, facility, or other dimensions 4. Drill down to individual records if needed 5. Export pivot results for further analysis

Example Analysis:

• Row: Facility
• Column: Month
• Measure: Percentage of Valid SSN values
• Result: Monthly SSN quality trend by facility, enabling identification of facilities with declining SSN quality

Custom Reports

Create custom quality reports for specific stakeholder needs:

Report Types:

• Executive Summary: High-level quality scorecard for leadership
• Facility Detail: Comprehensive quality report for facility HIM departments
• Corrective Action Tracker: Status of open quality improvement initiatives
• Trend Report: Long-term quality trends for governance committees

Report Creation:

• Use InterSystems Analytics reporting tools to create custom reports
• Define report schedules for automated generation and distribution
• Configure report recipients and delivery methods (email, portal)

Establishing Baselines and Thresholds

Define quality baselines and alert thresholds to guide monitoring:

Baseline Definition:

• During initial go-live, collect 30-60 days of quality metrics
• Calculate average quality levels for each facility and field
• Document these as baseline quality expectations
• Recognize that baselines vary by facility based on source system maturity

Threshold Configuration:

• Acceptable Range: Baseline ± 5%
• Warning Threshold: >5% below baseline
• Critical Threshold: >10% below baseline
• Missing Data Threshold: >20% missing for critical fields

Threshold Application:

• Configure automated alerts when thresholds are violated
• Use color coding on dashboards (green/yellow/red) to show threshold status
• Prioritize facilities in red or yellow for corrective action

Automated Alerting

Implement automated alerts to enable proactive quality management:

Alert Configuration:

• Define alert rules based on quality thresholds
• Configure alert recipients (EMPI operators, HIM leadership, facility contacts)
• Set alert frequency (real-time, daily digest, weekly summary)
• Specify alert delivery methods (email, Event Log, dashboard notifications)

Alert Content:

• Facility name and affected feed
• Quality metric that violated threshold
• Current value vs. threshold value
• Time period of violation
• Recommended actions or contacts

Alert Response Workflow: 1. Alert is triggered when threshold is violated 2. Designated recipient receives alert notification 3. Recipient reviews Data Quality dashboard for details 4. Root cause investigation is initiated 5. Facility is contacted with specific examples and corrective action request 6. Follow-up monitoring tracks improvement 7. Alert is closed when quality returns to acceptable range

---