Aim: To increase the rigor of our quality control (QC) processes within the constraints of our current vendor software and laboratory information system. In response to presumed gradual changes in vendor kit performance over time, we implemented dynamic ranges that are recalculated with each run. Additionally, to treat each bead, of which there are greater than 250, as a unique analyte, we applied bead-specific positive control serum thresholds alongside the standard positive control bead metric.
Methods: We developed a software tool that parses XML reports exported from Fusion and evaluates each Luminex antibody screen run against a set of configurable QC criteria: (1) Negative control beads must be <1500 MFI in both negative and positive control sera; (2) Positive control beads must exceed 4500 MFI; (3) All non-control beads in the negative control serum must remain <2000 MFI, our current threshold for a positive result; and (4) Each allele-specific bead in the positive control serum must fall within four standard deviations of its MFI values from the previous 50 runs. The system maintains historical data and dynamically recalculates thresholds during each QC session. Runs that fail QC must be repeated; runs that pass are forwarded to the LIS for patient-level review. The system was developed entirely in-house and is accessible through any web browser without requiring local installation. It runs on a secure hospital-based server and is maintained by our team. Built on a modular, API-based architecture, it allows for centralized updates and makes it straightforward to add new tools as needs evolve. We refer to the platform as Houston Methodist HLA Quality Utilities (HQU), and it serves as a foundation for future laboratory informatics development.
Results: The system enforces consistent QC for every antibody screen and adapts to variability introduced by reagent performance over time. Compared to Fusion’s static metrics, our tool provides greater specificity, flexibility, and responsiveness to observed data.
Conclusion: A dynamic, software-enforced QC process improves oversight of antibody screening, reflects real-world assay behavior, and reduces the likelihood of accepting suboptimal data. This approach enhances the reliability of clinical interpretation and establishes a scalable framework for adaptive reagent monitoring.
