Development and clinical implementation of a local automated kidney allocation platform integrating virtual crossmatch

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Background Highly sensitized kidney transplant candidates possess preformed anti-human leukocyte antigen (HLA) antibodies that substantially reduce their likelihood of receiving a compatible deceased-donor organ. Manual virtual crossmatch (VXM) workflows are sequential and labor-intensive, often limiting histocompatibility laboratories to screening only a subset of eligible candidates during time-sensitive donor evaluations and potentially overlooking compatible recipients. Methods We designed and developed an automated kidney allocation system (KAS) that links the hospital information system with HLA Fusion antibody data to perform candidate scoring and VXM prediction across the entire eligible active waitlist for every deceased donor workup. We conducted a single-center implementation study consisting of a cross-sectional analysis of 219 active kidney transplant candidates and a retrospective comparison of 67 deceased donor match runs processed by the automated KAS and the historical manual VXM workflow. The primary outcome was identification of potentially compatible highly sensitized patient (HSP) events. Analyse. Results Of 219 candidates, 98 (44.7%) were highly sensitized (cPRA ≥80%). Elevated prevalence of HSP was associated with prior transplantation, pregnancy, and transfusion. HSP candidates waited far longer than non-sensitized peers (median 7.10 vs. 3.81 years; p < 0.001). Across 67 retrospective match runs, the automated KAS identified 13 potentially compatible HSP events compared with 4 under the manual workflow, more than threefold higher (rate ratio 3.25; 95% CI 1.06–9.97; p = 0.039). Automated runs completed in 17 ± 0.3 minutes, compared with 152 ± 1.6 minutes for manual workups (p<0.0001). Conclusions A locally developed automated KAS substantially improved compatible donor identification for highly sensitized kidney transplant candidates while reducing VXM turnaround from hours to minutes. By enabling comprehensive evaluation of every eligible candidate during each deceased donor workup, automated allocation represents a practical and potentially scalable strategy that could improve equity and operational efficiency.