Abstract. The Unified Requirements Modeling Language (URML) was created to support early systems engineering; permitting the capture of process, potential variation points, hazards, threats and mitigations during the elicitation process. It has been used on Siemens projects, but to date they have been proprietary, precluding public exposure and limiting the critical review needed to refine and improve the language. To that end, a standard medical process, phlebotomy, was modeled in detail to determine the efficacy of the language. One of the authors of this paper is an expert in the area of medical diagnostics and provided the domain expertise. The modeling experience exceeded our expectations, and provided some unanticipated benefits. The results of our study are presented in this paper.

We evaluated a new protocol for measurement of cyclosporine A (CsA) 2 H after dose (C2) on the V-Twin® analyzer. Imprecision, recovery, and linearity were determined using CsA-spiked blood pools. Accuracy was evaluated using specimens from renal, cardiac, and liver transplant patients, and results were compared with those from liquid chromatography-tandem mass spectrometry (LC-MS/MS) and the Abbott TDx®/TDxFLx® assay. Cross-reactivity and interferences were assessed in the presence of 800 ng/mL CsA. Imprecision coefficients of variation were 3.3%-4.8% (within run) and 5.9%-8.7% (total). Recovery was within 10% of the expected values. Linearity was 350-2,000 ng/mL. Calibration was stable for ≥ 2 weeks. Method comparison showed regression statistics: V-Twin® = 1.01 × LC tandem MS + 36.1, r = 0.971; V-Twin® = 1.13 × Abbott - 92.4, r = 0.969. Metabolite cross-reactivity and interference (endogenous substances and drugs) were within ±10%. The C2 protocol on the V-Twin® analyzer provides acceptable assay performance and accurate determination of whole blood CsA drawn at 2 H after dose.