Prediction Models for Post-Stroke Hospital Readmission: A Systematic Review.
Yijun Mao, Qiang Liu, Hui Fan and 6 others
PMID 39402856WHAT IT FOUND
Sixteen stroke readmission models scored from 0.520 to 0.955 on how well they separated high and low risk, but high risk of bias and little external validation limit clinical use.
Length of stay, hypertension, age, and function were common predictors.
Key findings
01Across 11 studies reporting 16 stroke readmission models, discrimination ranged from 0.520 to 0.955, with six models above 0.75.
02Length of stay, hypertension, age, and functional status were the most common predictors in the models.
03Overall risk of bias was high, and external validation was reported in only one study.
STILL TO COME
How it was doneWhat they foundWhat it means for RNs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Only English and Chinese studies were searched, and gray literature was excluded. 54.5% of studies had high risk of bias in participant selection, mostly due to retrospective design. Six studies had too few outcome events for the number of predictors. Four studies filtered predictors using univariate analysis. Six studies did not report calibration. External validation was reported in only one study.
The easy way to misread this
Do not read the highest C statistic of 0.955 as proof that a model is ready for clinical use. The review found high overall risk of bias and external validation in only one study.