Patient Corrections Should Be Preserved as Clinical Data, Not Overwritten
Victor Kuznetsov argues that when patients correct information in their health records, the correction should be treated as new clinical data with provenance, not as a simple edit. He emphasizes that overwriting the original answer can erase critical context about what changed and when, which is essential for clinical decision-making. The article suggests using standards like HL7 FHIR to preserve the history of corrections.
The article distinguishes between consumer-app behavior and healthcare data handling, noting that a patient's correction can arrive at different workflow stages—before clinical review, during note drafting, or after a treatment decision is recorded—each requiring a different response. The author proposes storing immutable answer events with links to superseded versions, preserving question version, original and corrected values, timestamps, actor, and source.
The piece points to HL7 FHIR standards as the technical foundation, specifically QuestionnaireResponse statuses (amended, entered-in-error) and the Provenance resource for describing how a record reached its current state. It also emphasizes dependency mapping so systems know which summaries, rules, or documents consumed the old answer and who owns the next action when changes are material.
This approach could reduce medical errors stemming from silently overwritten patient data, particularly in cases where clinicians rely on outdated summaries after a correction arrives. Patients may benefit from greater transparency about how their corrections propagate through care workflows, while healthcare organizations could face implementation costs and integration challenges. The emphasis on provenance could also strengthen audit trails for compliance and liability purposes, though smaller practices without robust IT infrastructure may struggle to adopt these practices.