AUTOLOGOUS stem-cell transplantation remains central to the management of multiple myeloma, and new data suggest that outpatient stem-cell mobilisation may safely reduce hospital bed use while maintaining clinical oversight.
Background and Clinical Need
Stem-cell mobilisation is a prerequisite for autologous stem-cell transplantation and in Germany is commonly delivered through inpatient chemotherapy-based protocols. While effective, this approach places sustained pressure on hospital capacity. Investigators at the University Medical Center Göttingen assessed whether a data-driven strategy could support a shift towards outpatient stem-cell mobilisation without compromising patient safety.
The study analysed 109 patients with multiple myeloma undergoing stem-cell mobilisation and collection. Successful collection was achieved in 97% of cases, confirming the effectiveness of the overall approach. However, 69% of patients experienced severe adverse events that required hospitalisation, highlighting the importance of careful monitoring and timely intervention when considering outpatient pathways.
Predicting Risk During Stem-Cell Mobilisation
To address this challenge, the research team trained machine learning models using clinical data from the cohort. The aim was to predict which patients were likely to develop adverse events requiring admission and to estimate when these events might occur. Classification models demonstrated strong performance for several adverse event types. Prediction of elevated creatinine reached an area under the receiver operating characteristic curve of 1.0, indicating excellent discrimination. By contrast, neutropenic fever remained difficult to predict, with an area under the curve of 0.67.
In parallel, regression models were developed to forecast the timing of adverse event onset. These models achieved a mean prediction error of just over one day, suggesting potential value for proactive ward planning and early intervention strategies during stem-cell mobilisation.
Implications for Outpatient Care
Simulation analyses indicated that a risk-stratified outpatient stem-cell mobilisation protocol could reduce inpatient bed usage by at least one third without compromising safety. Such an approach could help ease healthcare constraints while maintaining high standards of care for patients with multiple myeloma.
Together, these data outline a pragmatic roadmap for integrating predictive analytics into clinical decision-making. With careful patient selection and monitoring, outpatient stem-cell mobilisation supported by machine learning may represent a feasible evolution in resource-efficient transplant care.
Reference
Schwarz F et al. Predicting adverse events for risk stratification of chemotherapy based stem cell mobilization in multiple myeloma. Npj Digital Medicine. 2026; https://doi.org/10.1038/s41746-026-02394-y.
