Prediction Models for Post-Stroke Hospital Readmission
Author Information
Author(s): Yijun Mao, Qiang Liu, Hui Fan, Erqing Li, Wenjing He, Xueqian Ouyang, Xiaojuan Wang, Li Qiu, Huanni Dong
Primary Institution: Xianyang Central Hospital
Hypothesis
This study aims to evaluate the predictive performance and methodological quality of post-stroke readmission prediction models.
Conclusion
The review found that existing readmission prediction models for stroke generally exhibit good predictive performance but require further validation and adaptation for practical use.
Supporting Evidence
- Eleven studies and 16 prediction models were included in the review.
- The sample sizes of the studies ranged from 108 to 803,124 cases.
- Common predictors included length of stay, hypertension, age, and functional disability.
- Twelve models reported an area under the curve (AUC) ranging from 0.520 to 0.940.
- Only one model included both internal and external validation.
- Six models had internal validation.
- High readmission rates may indicate unresolved issues at discharge.
- Reducing unnecessary readmissions is crucial for quality improvement.
Takeaway
The study looked at different models that predict if stroke patients will need to go back to the hospital after being discharged, and found some good options but also noted that they need more testing.
Methodology
A systematic review of 11 studies and 16 prediction models was conducted, assessing their predictive performance and methodological quality.
Potential Biases
High risk of bias was found in 11 studies, particularly in data analysis.
Limitations
The generalizability of the models is uncertain due to methodological limitations and high risk of bias in many studies.
Participant Demographics
The studies included stroke patients, with sample sizes ranging from 108 to 803,124.
Statistical Information
P-Value
0.13
Confidence Interval
0.08–0.18
Statistical Significance
p<0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website