A machine learning-based workflow for predicting transplant outcomes in patients with sickle cell disease

Area of Science:

• Hematology
• Medical Informatics
• Nephrology

Background:

• Allogeneic haematopoietic cell transplantation (HCT) is a curative option for sickle cell disease (SCD).
• Current risk stratification for HCT in SCD is limited.
• Identifying novel risk factors is crucial for improving patient outcomes.

Purpose of the Study:

• To develop a risk stratification system for patients undergoing HCT for SCD.
• To identify novel clinical and imaging risk factors using a machine learning approach.

Main Methods:

• A machine learning (ML) model was developed.
• The model integrated clinical and imaging variables.
• Risk factors for HCT in SCD patients were identified.

Main Results:

• Red cell distribution width was identified as a significant risk factor.
• Renal organ damage emerged as another critical risk factor.
• The ML algorithm demonstrated potential for risk factor discovery.

Conclusions:

• Red cell distribution width and renal organ damage are important risk factors in HCT for SCD.
• The developed ML-based algorithm can aid in risk stratification.
• This approach may be applicable to similar studies for risk factor discovery.