Custom tool sharply reduces errors in stem cell transplant engraftment reporting
Click Here to Manage Email Alerts
Time to engraftment after hematopoietic stem cell transplantation is an important outcome that institutions are legally required to report to the Center for International Blood and Marrow Transplant Research.
Researchers at Children’s Hospital of Philadelphia (CHOP) have created an application that automates determination of engraftment. The custom application supplants a laborious manual process and improves the accuracy of HSCT engraftment reporting.
The performance of the new tool has been characterized in a paper published in Transplantation and Cellular Therapy.
“At our center, we found up to a 28% error rate in the manual reporting process for platelet engraftment, likely due to the complexity of patients’ clinical courses, the difficulty of the manual process and the complexity of the rules,” David S. Anderson, MD, pediatric hematology/oncology and clinical informatics fellow at CHOP and leader of the project, told Healio. “If the identification of engraftment dates was automated, it would not only save significant time and effort for coordinators, but also lead to a more accurate nationwide assessment of transplant success and risk as research continues to advance in this exciting area of medicine.”
Anderson spoke with Healio about the development of the tool, its performance in the study and the potential of this application for widespread use.
Healio: How does the tool you developed automate the determination of engraftment?
Anderson: The application automatically extracts data from our institutional data warehouse, as well as from a separate stem cell transplant data management program. It matches patients by medical record number and displays their dates of transplantation along with other clinical data, graphs of platelet and neutrophil trends, visual indication of platelet transfusions and calculated engraftment dates based on Center for International Blood and Marrow Transplant Research (CIBMTR) rules. The graphs are interactive and contain specific visual cues to help the user interpret and validate the reported engraftment date.
Healio: How did your prospective study analyze the accuracy of this tool?
Anderson: First, we retrospectively validated the application's performance using platelet and neutrophil engraftment data that our institution previously reported from 2016 to 2020. Starting in April 2021, HSCT reporting staff used the engraftment application as part of their workflow of collecting and reporting transplantation data to the CIBMTR. The [quality improvement] team continued to manually review all engraftment data, as had been done in the retrospective review. Prospective data collection was performed from April 2021 through April 2022.
Healio: How did the tool perform?
Anderson: Reporting errors significantly decreased after application implementation. We found neutrophil engraftment to be incorrect in 3.8% of cases (two of 53), a significant decrease from 15% (P = .003), and platelet engraftment to be incorrect in 1.9% of cases (one of 53), a significant decrease from 28% (P < .001). The three postimplementation errors (two in neutrophil engraftment date and one in platelet engraftment date) were typographic — including transposition of two digits and mistakenly keying in an adjacent number on a keyboard — not because of incorrect calculation of engraftment by the application. Staff were enthusiastic about using the application in lieu of the manual legacy process.
Healio: What is next for the development of this approach? Do you expect it to be used widely at other centers?
Anderson: At our institution, we hope to continue to expand the scope of the application. which is facilitated by its open-source, modular framework. As of January 2022, we began to display T-cell infusions by connecting to an existing internally hosted database. We also hope to include other reportable HSCT-related measures, such as infections, end-organ complications, and post-HSCT outcomes and relapses. Further, by publishing this work and continuing discussions with CIBMTR, we hope to show that other academic institutions could adopt a similar system to increase nationwide automation of the engraftment calculation process.
For more information :
David S. Anderson, MD, can be reached at Children’s Hospital of Philadelphia, 34th St. and Civic Center Blvd., Philadelphia, PA 19104; email: andersons3@chop.edu.
Collapse
Click Here to Manage Email Alerts