- NIH Using Artificial Intelligence to Evaluate Stem Cell Quality
- What Are Precision Medicine and Personalized Medicine?
The press release also indicates that tools like these have significant potential due to the complexity of the hematopoietic stem cell (HSC) transplant process. HSC transplants, also known as bone marrow or stem cell transplants, are used to treat many types of diseases in adults and children. Part of the process involves using chemotherapy to wipe out a patient’s existing HSCs, which are then replaced by new HSCs from a donor.
The new cells need time to establish and proliferate throughout the bone marrow in a process known as engraftment. The time from HSC infusion to engraftment is key to preventing and addressing potential complications, such as infection, bleeding issues, and emergency re-transplant, according to the press release.
The serious complications that could follow HSC transplant led to federal regulations requiring clinicians to report time to engraftment for each patient to the Center for International Blood and Marrow Transplant Research (CIBMTR). Data collected by CIBMTR’s registry is then used to inform research and guidelines for transplant patient care.
To calculate engraftment, clinicians must manually extract relevant patient data and lab results from potentially disparate sources. Time to engraftment can also be impacted by whether patients had platelet transfusions, whether their neutrophil counts dipped after infusion, and other variables, making calculations difficult, time-consuming, and error-prone.
CHOP researchers built an application to combat this issue. The application, which is capable of complex data acquisition and manipulation, was embedded into the tool used within the health system’s Cell Therapy Program. The application works by extracting data from multiple sources and calculating engraftment dates based on CIBMTR rules.
Researchers integrated the tool into routine care in April 2021. The researchers found that between April 2021 and April 2022, neutrophil engraftment reporting was incorrect in two of 53, or 3.8 percent, of cases. However, both of those inaccuracies were due to typographical errors, not the application or its analysis. Comparatively, for engraftments that were calculated manually prior to April 2021, 15 percent of cases had incorrect neutrophil engraftment reports.
Similar improvements were observed for platelet engraftment reporting. From April 2021 to April 2022, platelet engraftment reporting was found to be incorrect in one of 53, or 1.9 percent, of cases, which was also the result of a typographical error. The platelet engraftment reporting error rate was also much lower than before the tool’s implementation when it was 28 percent.
"The success of this tool and the speed with which we were able to build it underscores the value of having an informatics team embedded within the Cell Therapy program, rather than under organization-wide software or analytics groups," Kadauke said. "Our consistent cellular therapy focus has allowed us to move more nimbly to complete projects and pivot to new initiatives. This is a model we believe could be emulated by other academic medical centers."