Improving the performance of virtual crossmatch results by correlating with nationally-performed physical crossmatches: Obtaining additional value from proficiency testing activities

Abstract Purpose When donor specific HLA antibodies (DSA) are identified, the predictive value of whether a certain strength of reactivity (mean fluorescence intensity, MFI) leads to a positive crossmatch is uncertain. To determine this, we compared the DSA MFI results we generated locally for nationally distributed proficiency samples against the percentage of other laboratories reporting a positive crossmatch. Method DSA MFI from single antigen beads reported by our laboratory for nationally-distributed proficiency testing survey samples was compared against the aggregate percentage of participating laboratories reporting the crossmatch positive using direct, antiglobulin-enhanced microcytotoxic (CDC-AHG), or flow cytometric methods from 2011 to 2015. Results 180 surveys were analyzed. Positive CDC-AHG and flow cytometric crossmatches were associated with MFI greater than 8554 and 2748 respectively for HLA class I, and 6919 and 3707 respectively for class II. Institutional MFI less than 3000 had high positive predictive values (0.98, 0.85, 0.81) for negative direct, AHG, and flow crossmatches, while MFI greater than 8000 had high negative predictive values for a positive direct, AHG, and flow crossmatches (1.00, 1.00, 0.97). Conclusion Review of locally-generated MFI results as part of participating in proficiency testing allow for predictability of crossmatch results against other laboratories, providing a replicable model for other participating centers.