Digital decision support using a phospholipid-dependent biomarker for early exclusion of negative cases in lupus anticoagulant diagnostics

Lupus anticoagulant (LAC) testing is central to antiphospholipid syndrome diagnostics and thrombotic risk assessment, but guideline-based workflows can be operationally complex, resource-intensive, and highly imbalanced. In our real-world cohort, completed LAC evaluations required an average of approximately 7 downstream clot-based assays, while nearly 90% of final classifications were negative. We developed and evaluated MINI-LAC™ (Minimal Lupus Anticoagulant), a biologically grounded digital decision-support framework designed for early exclusion of cases highly likely to yield negative final LAC classifications using routinely available pre-LAC coagulation parameters. Across 7454 LAC evaluations from a tertiary medical center, the phospholipid-poor versus phospholipid-rich activated partial thromboplastin time ratio (PTT-FSL/PTT-FS) demonstrated strong discrimination across LAC outcome categories ( p < 10 −53 ) and consistently emerged as the dominant predictive feature. In the independent prospective cohort and complete-case analyses, MINI-LAC achieved approximately 98% negative predictive value, with preserved performance in analyses not relying on imputation of the primary biomarker, while maintaining full guideline-based evaluation for all non-negative or uncertain cases. These findings support integration of biologically meaningful functional coagulation signals into digital laboratory decision support to improve efficiency in LAC diagnostics.