T3.3: Explains System Settings Impact on Worklist

InterSystems EMPI (formerly HealthShare Patient Index) uses system-wide configuration settings to control how the matching engine operates and which record pairs appear on the Worklist for human review. These settings are stored as global variables and can be modified through the terminal prompt in the registry namespace.

The technical specialist must understand that system settings are not arbitrary values—they represent organizational policy decisions about patient identification. Each setting reflects how aggressive or conservative the organization wants to be with automatic linking, and these decisions have direct consequences on worklist volume, data quality, and operational workflow.

The primary settings that affect worklist behavior include threshold values, data source configurations, audit trail settings, and user interface limits. These settings work together to determine which record pairs require human intervention versus automatic processing.

Organizations must balance the desire to minimize manual review workload against the risk of incorrect matches. Conservative settings (higher autolink threshold) create more worklist items but reduce the risk of false matches. Aggressive settings (lower autolink threshold) reduce workload but increase the risk of linking records for different people.

---

The three threshold settings are the most critical configuration parameters affecting worklist behavior:

Review Threshold: This is the minimum link weight score required for a record pair to appear on the worklist as a potential link. Pairs scoring below this threshold are considered non-matches and are automatically excluded from linking. Pairs scoring between the review threshold and autolink threshold appear on the worklist in the "Review" category, requiring human judgment to determine if they should be linked.

Autolink Threshold: This is the decision boundary for automatic linking. Record pairs scoring at or above this threshold are automatically linked without human intervention, creating a shared MPIID for the records. The autolink threshold represents the organization's confidence level—the score at which they believe two records almost certainly represent the same person.

Validate Threshold: This threshold sits above the autolink threshold and identifies automatically-linked pairs that should be reviewed for quality assurance. Pairs that score above the autolink threshold but below the validate threshold are linked automatically but appear on the worklist in the "Validate" category, allowing staff to confirm the system made the correct decision.

These three thresholds create distinct operational zones:

• Below Review: Automatic non-links (no worklist entry)
• Review to Autolink: Manual review required (Review category)
• Autolink to Validate: Automatic links requiring validation (Validate category)
• Above Validate: High-confidence automatic links (no worklist entry unless other issues exist)

During initial implementation, organizations typically set conservative thresholds that create larger worklist volumes. As the matching configuration is tuned and confidence increases, thresholds are adjusted to reduce manual review requirements while maintaining data quality.

---

The relationship between threshold settings and worklist volume is inversely proportional to organizational risk tolerance. Organizations that are highly risk-averse regarding incorrect matches will set higher autolink thresholds, resulting in more record pairs requiring human review.

Conservative Configuration Impact: When the autolink threshold is set high (e.g., 30 out of 40 possible points), many record pairs that score between 20 and 30 will appear on the worklist for manual review. This creates significant workload for worklist staff but minimizes the risk of false positive matches. In the early stages of EMPI implementation, this approach allows the organization to manually review borderline cases and gain confidence in the matching algorithm.

Aggressive Configuration Impact: As organizations mature and gain confidence in their matching configuration, they may lower the autolink threshold (e.g., to 24 out of 40 points). This reduces the number of pairs requiring review, decreasing staff workload. However, it increases the probability that some automatically-linked pairs may represent different people, potentially creating overlay problems that are more difficult to resolve later.

Threshold Adjuster Tool: EMPI provides a Linkage Definition Threshold Adjuster tool that displays a histogram showing the distribution of link weights across all record pairs. This visualization helps technical specialists and administrators understand how threshold changes will impact worklist volume. The histogram shows:

• Number of pairs at each link weight score
• Current positions of review, autolink, and validate thresholds
• Projected impact of moving thresholds up or down

Technical specialists should work with operational staff to monitor worklist volume trends and matching accuracy. If the worklist becomes unmanageable, thresholds may need adjustment. If false matches are discovered, thresholds may need to be made more conservative.

---

System settings in EMPI encode organizational policies about patient identification. The technical specialist must understand how to translate business requirements into technical configuration:

Deterministic Identifier Policies: Organizations may decide that certain identifiers (such as corporate IDs from trusted sources) should always force a link or unlink, regardless of other matching parameters. For example, if two records from a trusted hospital system have the same Medical Record Number (MRN), they must be linked. Conversely, if they have different MRNs from the same facility, they must not be linked (this creates duplicates on the worklist). These policies are implemented through deterministic identifier configuration.

Data Source Trust Levels: Not all data sources are equally reliable. Organizations may configure EMPI to treat matches from highly-trusted sources differently than matches involving less-reliable sources. This affects how link weights are calculated and which pairs appear on the worklist.

Domain Conflict Handling: Organizations must decide how to handle duplicates—two records from the same facility (data domain) that represent the same person but have different MRNs. The standard policy is to NOT automatically link duplicates, instead placing them on the worklist with a "Duplicate" category and "Domain Conflict" link reason. The organization's procedure should be to send these pairs back to the originating facility for MRN merge resolution.

Custom Status Configuration: Organizations define custom status values (e.g., "Sent to Hospital", "Pending MRN Merge", "Under Investigation") to track worklist item workflow. These status values must be created in system configuration before they can be assigned to worklist items.

Audit Trail Settings: The skipAudit and skipDemographicAudit settings control whether EMPI maintains detailed audit logs of all changes. Organizations with strict compliance requirements typically enable full auditing, while others may disable demographic auditing for performance reasons. This decision reflects the organization's regulatory environment and governance policies.

---

Several system settings specifically control user interface behavior and performance:

auditLogMaxResultRows: This setting limits the maximum number of rows returned when displaying the audit log. Without this limit, attempting to view the complete audit history for a heavily-modified record could cause performance issues or browser timeouts. The default value balances the need to see recent audit history with the need for responsive display.

worklistMaxSortRows: This setting limits the number of worklist rows that can be sorted in the user interface. When the worklist contains thousands of items, attempting to sort the entire dataset by link weight or timestamp could cause significant delays. This setting protects system performance by restricting sorting to a manageable number of rows.

These UI limits reflect organizational decisions about the trade-off between comprehensive data display and responsive system performance. Organizations with large EMPI deployments (millions of records) must be more conservative with these settings than smaller implementations.

Row Pagination: The worklist interface allows users to specify how many rows to display per page (e.g., 20, 50, 100). This pagination setting works in conjunction with the maxSortRows setting to ensure acceptable performance. Users can navigate between pages using First, Previous, Next, and Last navigation controls.

Column Selector: Organizations can configure which columns are displayed by default in the worklist through the Worklist Column Selector. This affects initial display performance—fewer columns means faster initial rendering.

Technical specialists should monitor worklist query performance and adjust these settings as needed based on actual usage patterns and system response times.

---

Technical specialists should recommend that organizations document their problem resolution workflows for each type of worklist item. This documentation ensures consistency across different staff members and shifts, and provides training materials for new employees.

Category-Specific Procedures: Each worklist category (Review, Validate, Duplicate, Overlap, Overlay, Open-chaining, Deterministic) requires different resolution procedures:

• Review Category: Document the criteria for deciding whether to link potential matches. What constitutes sufficient evidence? What tools should be used (Comparison Detail, Whole Record Viewer)? How should staff handle cases with missing or conflicting demographic data?
• Validate Category: Define the validation workflow for automatically-linked pairs. What checks should be performed? Under what circumstances should an automatic link be reversed?
• Duplicate Category: Document the procedure for handling same-facility duplicates. Who should be notified? What information should be sent to the originating facility? How should the worklist item be tracked during MRN merge resolution?
• Overlap/Overlay Category: Provide detailed procedures for resolving MPIID conflicts. These are among the most complex worklist items and require careful analysis using the Whole Record Viewer and Linkage Graph tools.
• Open-chaining Category: Document the logical analysis process for resolving transitive linking conflicts (A linked to B, B linked to C, but A not linked to C).

Decision Criteria Documentation: Organizations should document the specific criteria staff should use when making linking decisions. This might include:

• Minimum number of matching data elements required for manual link
• Specific identifiers that should be considered definitive (e.g., SSN, MRN)
• How to handle name variations, nicknames, hyphenated names
• Date of birth tolerance (e.g., transpose errors, data entry errors)
• Address matching considerations (e.g., PO Box vs. street address)

---

Not all worklist items can be resolved by standard procedures. Organizations should define escalation processes for complex or ambiguous cases:

Escalation Criteria: Document specific situations that should trigger escalation to more experienced staff or subject matter experts:

• Record pairs with conflicting deterministic identifiers
• Large linked groups with multiple overlaps or overlays
• Cases involving VIP or sensitive patients
• Situations where organizational policy is unclear
• Suspected data quality issues affecting multiple records
• Linkage conflict categories (Overlap, Overlay, Open-chaining)

Tiered Expertise Model: Organizations should establish expertise tiers:

• Level 1 - Standard Staff: Handle routine Review and Validate category items with clear decision criteria
• Level 2 - Lead/Supervisor: Handle Duplicate category items, complex Review cases, and provide guidance to Level 1 staff
• Level 3 - Technical Specialist: Handle Overlap, Overlay, and Open-chaining categories, investigate systemic matching issues, recommend configuration changes

Escalation Documentation: When cases are escalated, the documentation should include:

• Original worklist item details (MPIID, MRN, names, scores)
• Reason for escalation
• Analysis performed at lower level
• Decision made by higher level
• Rationale for decision
• Any configuration changes recommended

Linkage Conflict Categories: The technical specialist should emphasize to customer staff that linkage conflict categories (Overlap, Overlay, Open-chaining, Deterministic conflicts) need careful research by experienced staff members. These situations often indicate:

• Previous incorrect manual decisions
• System configuration issues
• Data quality problems in source systems
• Complex family relationships or identity scenarios

These cases should not be resolved quickly without thorough investigation, as incorrect resolution can create cascading problems affecting many records.

---

Organizational policies have direct, measurable impacts on EMPI linking behavior and operational workload:

Patient Safety Policies: Organizations that prioritize preventing false matches (linking records for different people) will implement conservative policies:

• Higher autolink thresholds (more manual review)
• Strict deterministic unlink rules (domain conflicts always prevent linking)
• Comprehensive validation requirements
• These policies create larger worklist volumes but reduce patient safety risk

Operational Efficiency Policies: Organizations facing resource constraints may implement policies that favor automation:

• Lower autolink thresholds (less manual review)
• Relaxed validation requirements
• Streamlined resolution procedures
• These policies reduce workload but may increase data quality issues

Data Governance Policies: Organizations with strong data governance programs typically implement:

• Comprehensive audit trail requirements (no skipAudit)
• Documented data source trust levels and handling procedures
• Formal change control for system settings modifications
• Regular data quality monitoring and threshold tuning

Compliance and Regulatory Policies: Healthcare organizations must consider regulatory requirements:

• HIPAA audit trail requirements
• State-specific patient matching regulations
• Organizational policies on record access and modification
• Joint Commission or other accreditation standards

Technical Specialist Role: The EMPI technical specialist must be able to articulate to customer staff how different policy decisions translate into system configuration settings and operational outcomes. This includes providing data-driven recommendations:

• "If we lower the autolink threshold from 24 to 20, we estimate the Review worklist will decrease by 30%, but we may see an increase in overlap/overlay issues."
• "Our current duplicate handling policy requires manual notification to source facilities, which takes an average of 15 minutes per case. With 50 duplicates per week, this represents 12.5 hours of staff time."

---

System settings should not be "set and forget"—they require ongoing monitoring and adjustment as the EMPI implementation matures and data characteristics change.

Worklist Volume Monitoring: Track key metrics over time:

• Total worklist items by category (Review, Validate, Duplicate, etc.)
• Average time to resolve worklist items by category
• Backlog trends (growing or shrinking)
• Staff utilization and capacity

Matching Accuracy Analysis: Periodically analyze actual linking outcomes:

• Sample manually-reviewed cases to assess decision quality
• Investigate discovered false matches to identify patterns
• Review high-scoring non-links to identify potential algorithm improvements
• Monitor overlap and overlay rates as indicators of matching issues

Threshold Adjuster Reviews: Schedule regular reviews (quarterly or semi-annually) using the Threshold Adjuster tool to visualize the current distribution of link weights and assess whether threshold adjustments are warranted. The histogram display shows:

• How many pairs fall into each threshold zone
• Whether the distribution has shifted over time
• Impact projections for threshold changes

Configuration Change Management: Establish a formal process for modifying system settings:

• Document current settings and rationale
• Propose changes with justification and impact analysis
• Test changes in non-production environment when possible
• Implement changes during low-activity periods
• Monitor results and document outcomes

Staff Feedback Integration: Operational staff working the worklist daily have valuable insights:

• Which types of cases are most difficult to resolve
• Whether current tools and procedures are adequate
• Suggestions for workflow improvements
• Identification of data quality issues in source systems

The technical specialist should facilitate regular meetings between technical staff, operational staff, and management to review EMPI performance metrics and adjust settings and procedures as needed.

---

A critical KSA for the HS-EMPI Technical Specialist exam is the ability to review organizational person identification decisions reflected in system settings and discuss their impact on system behavior with customer staff.

Translation Skills: Technical specialists must be able to explain complex configuration concepts in terms operational staff and business stakeholders can understand:

• Instead of: "The autolink threshold is set to 24 with a validate threshold of 34"
• Say: "The system automatically links records that score 24 or higher out of 40 possible points. We review the lowest-scoring automatic links (those between 24 and 34) to ensure quality."

Impact Analysis: When discussing configuration changes, provide concrete operational impact estimates:

• "Lowering the review threshold from 14 to 10 would add approximately 200 pairs per month to the worklist, requiring an estimated 10 additional staff hours."
• "Enabling full demographic auditing would add approximately 50MB per month to database storage and may increase response time for audit log queries by 10-20%."

Stakeholder Facilitation: The technical specialist often needs to facilitate discussions between different stakeholder groups:

• Operations Staff: Concerned with workload, user interface usability, resolution procedures
• IT Management: Concerned with system performance, database sizing, integration issues
• Clinical Leadership: Concerned with patient safety, data quality, regulatory compliance
• Data Governance: Concerned with policies, audit trails, data stewardship

Decision Documentation: After configuration decisions are made, ensure comprehensive documentation:

• What settings were changed and why
• What organizational policies or requirements drove the change
• What operational impact is expected
• How success will be measured
• When the next review is scheduled

This documentation becomes part of the organization's EMPI governance framework and provides valuable context for future technical specialists or during system upgrades.

---