Risk factor analysis and prediction model construction for delayed erythroid hematopoietic reconstitution after allogeneic hematopoietic stem cell transplantation: a single-center retrospective study.

BACKGROUND

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) represents a potential therapeutic option for various hematological disorders. With continuous advancement in transplantation techniques and the expansion of indications, clinical issues associated with delayed erythroid hematopoietic reconstitution (DEHR) have become increasingly prominent.

METHODS

This retrospective study identified independent risk factors through univariate and multivariate logistic regression analyses, subsequently constructing a prediction model. Model discrimination, calibration, and clinical utility were assessed using the area under the curve (AUC), calibration curves, Hosmer-Lemeshow tests, and decision curve analysis. Overall model performance was evaluated via Nagelkerke R², with internal validation conducted using the bootstrap method.

RESULTS

Logistic multivariate analysis results showed that the ferritin level before transplantation (OR = 1.001, 95% CI 1.000-1.001), the severity of anemia before transplantation (OR = 3.740, 95% CI 1.115-12.541), and whether the ABO blood type of the donor and recipient were compatible (OR = 3.283, 95% CI 1.262-8.540) were independent risk factors (P < 0.05). And the model's AUC was 0.835 (95% CI: 0.759-0.910), with the slope of the calibration curve approaching 1, the H-L test P = 0.346, and Nagelkerke R² = 0.349.

CONCLUSION

The nomogram constructed based on pre-transplant ferritin levels, pre-transplant anemia severity, and donor-recipient ABO blood group compatibility demonstrates high predictive value for the occurrence of DEHR after allo-HSCT. This provides a reference for the early identification of high-risk individuals.