Planning Your Enterprise AI Investments? Discover How to Choose the Right AI Use Case →

Solution Accelerator

Solution Accelerator

Master Patient Index · Databricks

Master Patient Index (MPI) Framework

Reference workflows and reusable PySpark code to match patient records across EHR, LIS, RIS, pharmacy, and billing systems, create Master Patient IDs and golden records, and manage reviews with audit history.

Language

Python / PySpark

Deployment

Databricks Notebooks

Type

Identity Resolution Framework + Reference Code

Trusted By :

How it works

Resolve patient identities with controlled matching

Patient information often differs across EHR, laboratory, radiology, pharmacy, and billing systems. Variations in names, addresses, identifiers, and date formats can create duplicate records and prevent teams from seeing one complete patient record. Linking records without confidence and safety controls can also create incorrect matches.

DataTheta’s Master Patient Index Framework cleans and standardizes each source, applies deterministic and fuzzy matching, and calculates weighted scores from 0 to 1. HIGH matches that pass every safety rule link automatically. MEDIUM matches move to human review, while LOW matches remain separate until new evidence becomes available.

  • Configure sources and policies: Define source layouts, ID columns, field mappings, date formats, matching weights, safety rules, review policies, golden-record rules, privacy settings, and pipeline destinations in the JSON files under configs/.
  • Standardize and compare records: Collect data from EHR, LIS, RIS, pharmacy, and billing systems. Clean names, addresses, dates, and coded values before applying deterministic and fuzzy matching.
  • Classify and link matches: Assign HIGH confidence to scores of at least 0.80 when every safety rule passes. Send scores from 0.55 to below 0.80, or any pair that fails a safety rule, to MEDIUM review. Keep scores below 0.55 separate as LOW confidence.
  • Build the master record: Create a unique Master Patient ID, apply configured golden-record rules, preserve mappings back to every source record, and expose the final patient records through one unified table.
  • Review and correct safely: Present pending pairs in one bulk-review table. Apply Approve, Reject, Split, or Unsure decisions, support controlled merge and unmerge actions, and preserve audit history.
AI systems in production
0 +
Avg. time to first outcome
0 Weeks
Forecast accuracy
0 %
Faster decision cycles
0 X
Revenue influenced by AI
$ 0 M+
Manual processing eliminated
0 %

40+

AI systems in production

8 Weeks

Avg. time to first outcome

34%

Forecast accuracy improvement

3×

Faster decision cycles

$180M+

Revenue influenced by AI

68%

Manual processing eliminated

Key Capabilities

What this accelerator helps you do

Multi-System Data Standardization

Collect patient records from EHR, LIS, RIS, pharmacy, and billing systems. Standardize names, addresses, dates, identifiers, and source-specific codes before matching.

Deterministic and Fuzzy Matching

Use exact matching and weighted fuzzy comparisons to identify records that may belong to the same patient. Configure blocking rules and field weights for each implementation.

Confidence and Safety Controls

Classify candidate pairs as HIGH, MEDIUM, or LOW confidence. Apply safety rules and cluster guards before records can be linked automatically.

Master ID and Golden Record

Create a unique Master Patient ID and one golden record using configurable field-selection rules and source priorities. Maintain mappings to the original source records.

Human Review and Audit History

Review pending matches in bulk and record Approve, Reject, Split, or Unsure decisions. Safely merge or unmerge records while preserving review activity and audit history.

Solution Accelerators

Strengthen healthcare data foundations with reusable accelerators.

Explore reusable accelerators for patient identity resolution, PHI and PII protection, data quality, reconciliation, synchronization, governance, and Databricks data engineering.

Ready to build a trusted master patient index?

Use DataTheta’s Master Patient Index Framework to resolve duplicate patient records, create governed golden records, manage uncertain matches, and maintain a reviewable identity history in Databricks.

©2026 Copyright DataTheta – Lance Labs Inc.

Let's Build Your Next Data & AI Solution

Tell us about your business goals, data challenges, or technology requirements. Our team will review your enquiry and connect with you to discuss the right solution.

No spam. Your information stays private.