What is data linkage in healthcare and what methods are used?

Data linkage combines records from multiple sources (primary care, secondary care, pharmacy, social care, mortality registers) relating to the same individual. Deterministic linkage uses exact agreement on unique identifiers like the NHS Number — fast and specific but fails when identifiers are missing. Probabilistic linkage (Fellegi–Sunter model) assigns match weights to partial identifiers and is more flexible. Privacy-preserving approaches include Trusted Third Parties, federated linkage, and cryptographic hashing.