P-011 Laboratory medicine Poster

Implementation of an automated amino acid result entry and verification pipeline in a public sector laboratory

Audience Choice Award

To rate this poster, scan the QR code on the poster itself — voting only works on site. Every attendee has 3 stars and can give at most one per poster. A star is final and cannot be taken back.

Authors

K. Magolego presenting

Division of Chemical Pathology, Department of Pathology, University of Cape Town and National Laboratory Service, South Africa

#0
J. Croxford

Division of Chemical Pathology, Department of Pathology, University of Cape Town and National Laboratory Service, South Africa

#1
H. Vawda

Division of Chemical Pathology, Department of Pathology, University of Cape Town and National Laboratory Service, South Africa

#2
B. Southon

Division of Chemical Pathology, Department of Pathology, University of Cape Town and National Laboratory Service, South Africa

#3
J. Rusch

Division of Chemical Pathology, Department of Pathology, University of Cape Town and National Laboratory Service, South Africa

#4

Abstract

Following the discontinuation of in-house amino acid analysis, results were received from an external laboratory as PDF reports without interpretive comments and required manual transcription into the Laboratory Information System (LIS). Each patient report required the entry of 29 amino acid results together with an interpretive comment, making the process labour-intensive, time-consuming, and vulnerable to transcription errors.

To address these challenges, a Python-based automation framework was developed and implemented within routine service. The system uploads results directly into the existing LIS through a controlled user-interface workflow. The solution was designed to integrate with existing laboratory infrastructure without requiring vendor modification, middleware development, or direct database access.

Governance and patient safety considerations were central to implementation. The automated workflow was validated against the existing manual process prior to deployment, retained standard laboratory authorisation procedures, and generated timestamped audit logs for traceability. Automated verification functionality independently compared uploaded results against source data to identify discrepancies. Direct modification of laboratory databases was prohibited, ensuring alignment with existing information governance requirements and minimising implementation risk.

The system was validated using 503 amino acid reports, comprising approximately 14,500 individual analyte results and interpretive comments. Manual processing and verification required an average of 289 minutes per batch of 31 reports, compared with 54 minutes following implementation of the automated workflow, representing an 81% reduction in processing time. No transcription errors were identified during validation. Users reported reduced fatigue, improved workflow efficiency, and greater opportunity to focus on clinical interpretation and result authorisation.

Key lessons learned were that successful digital transformation depends as much on workflow understanding, governance, and user engagement as on software development. Low-cost, locally developed automation can safely deliver substantial efficiency gains in diagnostic laboratories when implemented within existing quality management frameworks. This approach provides a practical model for modernising manual reporting processes in resource-constrained healthcare settings.