Work

Operationalizing the CIBMTR Virtual Consortium Program to Advance Collaborative Quality Improvement in Transplant and Cellular Therapy

Year: 2026

Type: article

Abstract: Introduction The Center for International Blood and Marrow Transplant Research (CIBMTR) has envisioned a “virtual consortium” to enhance the value of its registry by securely consolidating Data Back to Center (DBtC) files across institutions. This initiative enables multi-center collaboration for quality improvement (QI) and research, improving data quality while reducing reporting burden—advancing the principle of “enter once, use many times.” The Engraft Learning Network is among the first to implement this vision, creating a shared analytic environment for collaborative learning in transplant and cellular therapy (TCT). Objectives To demonstrate how integrating CIBMTR DBtC data enhances registry accuracy and enables collaborative quality improvement in transplant and cellular therapy. Methods With data use agreements, IRB approvals, and site authorization, CIBMTR consolidated DBtC files from Engraft centers. Files were securely transmitted to the Engraft Data Coordinating Center under an honest-broker model and integrated into the Engraft Registry to supplement manually entered data and enable systematic quality checks. Discrepancies between registry and DBtC data were assessed for infusion date, birth date, sex, race, and date of death, with mismatches adjudicated at Cincinnati Children’s (CCHMC). Chronic GVHD (cGVHD) variables—diagnosis date, maximum grade, and contact date—were summarized by mean and median time to diagnosis and maximum grade by severity. Race data from DBtC were also incorporated to enhance analyses and support equity-focused evaluation. Results Across eight centers, 515 Engraft transplants matched pre- and post-TED DBtC data. Infusion date discrepancies occurred in 35 cases (7%), all from CCHMC, with DBtC correct in each. Birth date and sex discrepancies (n=9 each) and date-of-death discrepancies (n=3) were mostly corrected in DBtC. Race discrepancies were rare, though missing or unknown race was common (20 Engraft records missing values in DBtC; 49 missing in DBtC). Median time to cGVHD diagnosis ranged from 168–228 days post-transplant, with time to maximum grade and contact summarized in Figures 1–2. Conclusions Integrating CIBMTR DBtC data within Engraft improved registry accuracy, completeness, and interoperability while reducing burden. This approach demonstrates the feasibility of securely leveraging DBtC data for multi-center QI and research, advancing CIBMTR’s vision of maximizing registry impact across the TCT community.

Source: Transplantation and Cellular Therapy

DOI: 10.1016/j.jtct.2025.12.845