Managing Clinical Data Lineage in Distributed Healthcare Integration Environments: A Metadata Instrumentation Framework for End-to-End Provenance Tracking
DOI:
https://doi.org/10.63282/3050-9262.IJAIDSML-V7I1P159Keywords:
Data Lineage, Provenance Tracking, Healthcare Interoperability, Integration Engine, HL7, FHIR, Metadata Governance, W3C PROV, Distributed SystemsAbstract
Healthcare integration engines transform clinical messages through multiple processing stages before delivery to downstream systems, yet no standardized framework exists for tracking field-level transformation provenance within middleware architectures. This gap impedes data quality root-cause analysis, complicates regulatory audit response, and undermines trust in derived clinical datasets. This paper proposes the Healthcare Integration Lineage Instrumentation Framework (HILIF) a metadata-driven architecture for capturing field-level transformation provenance across distributed integration engine topologies. HILIF introduces a lineage event model grounded in W3C PROV ontology, extended for HL7-specific transformation semantics. A sidecar instrumentation layer captures transformation events asynchronously through a lock-free ring buffer, persisting them to a dual-store architecture (relational + graph) that supports three query patterns: forward-trace, backward-trace, and impact analysis. Evaluation across a simulated enterprise environment 120 channels, 14 transformation stages, 450 messages/second sustained over 72 hours demonstrates a mean latency overhead of 3.2 ms per message, sustained throughput reduction of 1.5%, and lineage event capture completeness of 99.9999%. Forward-trace queries resolve in under 200 ms median, and FHIR Provenance resource generation achieves 97.8% structural validation. HILIF provides a replicable framework for audit-ready provenance tracking in enterprise healthcare integration.
References
[1] R. Haux, "Health information systems past, present, future," Int. J. Med. Inform., vol. 75, pp. 268–281, 2006.
[2] D. Bender and K. Sartipi, "HL7 FHIR: An agile and RESTful approach," in Proc. IEEE CBMS, 2013, pp. 326–331.
[3] C. Dolin et al., "HL7 Clinical Document Architecture, Release 2," JAMIA, vol. 13, no. 1, pp. 30–39, 2006.
[4] HL7 International, "US Core IG v6.1.0 Basic Provenance," 2023.
[5] P. Buneman et al., "Why and where: A characterization of data provenance," in Proc. ICDT, 2001, pp. 316–330.
[6] L. Moreau and P. Missier, "PROV-DM: The PROV Data Model," W3C Rec., 2013.
[7] W3C, "PROV-O: The PROV Ontology," W3C Recommendation, 2013.
[8] Z. Ives et al., "Dataset management and versioning for ML," in Proc. NeurIPS Workshop, 2017.
[9] ONC, "USCDI," U.S. DHHS, 2023.
[10] ONC, "TEFCA," U.S. DHHS, 2024.
[11] HHS, "Information Blocking Penalties Final Rule," Federal Register, 2024.
[12] D. An, M. Lim, and S. Lee, "Challenges for data quality in the clinical data life cycle," J. Med. Internet Res., vol. 27, e60709, 2025.
[13] S. T. Rosenbloom et al., "Data from clinical notes," JAMIA, vol. 18, no. 2, pp. 181–186, 2011.
[14] A. R. Hevner et al., "Design science in IS research," MIS Q., vol. 28, no. 1, pp. 75–105, 2004.
[15] J. Walonoski et al., "Synthea," JAMIA, vol. 25, no. 3, pp. 230–238, 2018.
[16] M. Thompson et al., "Disruptor: High performance bounded queues," LMAX, 2011.
[17] HL7 International, "FHIR R4 Provenance Resource," 2019.
[18] D. W. Bates et al., "Reducing errors in medicine using IT," JAMIA, vol. 8, no. 4, pp. 299–308, 2001.
