# A Novel Alternating Joint Longitudinal Model for Post-ICU Hemoglobin Prediction

**Authors:** Gabriel Demuth, Curtis Storlie, Matthew A. Warner, Daryl J. Kor, Phillip J. Shulte, Andrew C. Hanson  
**arXiv:** [2302.09110](https://arxiv.org/abs/2302.09110)  
**Date:** 2023-02-17  
**Summary page:** https://www.summarizepaper.com/en/paper/2302.09110v1/

## Key points

- Anemia is common in patients after being discharged from the ICU
- Researchers introduced a novel Bayesian joint longitudinal model for hemoglobin over time that includes specific parametric effects for hospital admission and discharge
- The model accurately estimated hemoglobin levels and predicted anemic status and hospital readmission within 30 days post-discharge with AUCs of .82 and .72, respectively
- This innovation allows for a much richer and more dynamic representation of the interplay between longitudinal and event data
- The study presents a comprehensive approach to predicting anemia in post-ICU patients using Bayesian joint longitudinal modeling techniques
- Future research could explore the potential of using this model to inform clinical decision-making and improve patient care.

## Summary

Anemia is a common condition in patients after being discharged from the intensive care unit (ICU). To address this issue, researchers have introduced a novel Bayesian joint longitudinal model for hemoglobin over time that includes specific parametric effects for hospital admission and discharge. The study found that the model accurately estimated hemoglobin levels and predicted anemic status and hospital readmission within 30 days post-discharge with AUCs of .82 and .72, respectively. This innovation allows for a much richer and more dynamic representation of the interplay between longitudinal and event data. Overall, this study presents a comprehensive approach to predicting anemia in post-ICU patients using Bayesian joint longitudinal modeling techniques. Future research in this area could explore the potential of using this model to inform clinical decision-making and improve patient care.