Full definition
Patient matching is the operational discipline of correctly identifying the same patient across systems and records. The challenge: same patient frequently exists with slightly different demographic data across systems — different name spelling, date format, phone number, address. Without reliable matching, the same patient appears as multiple records, fragmenting clinical history and creating safety risks.
Production patient-matching uses combinations of: deterministic matching (exact match on identifier — national ID, MRN), probabilistic matching (Fellegi-Sunter algorithms scoring multi-attribute similarity), AI/ML matching (machine-learning models trained on labeled match decisions), and human review for ambiguous cases. Reference matching against authoritative sources (national patient identifier where available — NHS Number in UK, BSN in Netherlands) substantially improves accuracy.
For clinic technology: patient matching is foundational. Modern platforms implement multi-strategy matching with configurable thresholds, manual-review queues for ambiguous cases, and ongoing identity-resolution discipline.
MOVO-X patient matching uses MyKad / NRIC / Aadhaar / national ID where available for exact match, plus name + DOB + phone + address composite for fallback. Manual review queue for ambiguous cases. Cross-system matching for FHIR-based integration.
Where patient matching is used
- Patient registration across visits
- Cross-system EHR integration
- Health information exchange
- Population-health analytics
- Identity-resolution programmes
- Master Patient Index (MPI)
Types of patient matching
Deterministic matching
Exact match on identifier.
Probabilistic matching
Multi-attribute similarity scoring.
AI/ML matching
Machine-learning models on labeled match decisions.
Reference matching
Match against authoritative source (national patient ID).
Master Patient Index (MPI)
Centralised system maintaining unique patient identity.
Enterprise MPI (eMPI)
MPI spanning multiple systems within an organisation.
Quantified benefits
- ▸Unified clinical history per patient
- ▸Avoided duplicate records
- ▸Cross-system continuity of care
- ▸Foundation for population-health analytics
Frequently asked
How does MOVO-X match patients?+
National ID where available (MyKad, NRIC, Aadhaar) for deterministic match; name + DOB + phone + address composite fallback; manual review queue for ambiguous cases.
What if same patient registers with different IDs (e.g. before and after MyKad renewal)?+
Manual review queue identifies likely duplicates. Identity resolution merges records with audit trail.
Patient matching for cross-border medical tourism?+
Composite of passport + name + DOB + home-country phone. Lower confidence than national-ID match but workable.
How does this relate to MPI?+
MPI is the centralised identity-resolution system; patient matching is the algorithmic discipline that powers MPI.