Severe Intraventricular Hemorrhage Prediction Models in Preterm Infants

Sixteen prediction models for severe intraventricular hemorrhage in preterm infants were identified and evaluated from a systematic review of 13,311 studies. The meta-analysis of 7 models produced a pooled AUC of 0.805. Zhiheng Zhan authored this review, which was last updated in April 2026.

Use Cases

• Compare the pooled AUC of 0.805 and its 95% confidence interval across the 7 included prediction models.
• Analyze the reported high risk of bias percentages (e.g., 93.75% for missing data handling) to assess model methodological quality.
• Examine the 16 included models for their specific predictor variables and validation methods described in the review tables.

Strengths

• Systematic review methodology screened 13,311 initial studies for comprehensive model identification.
• Meta-analysis provides a quantitative performance summary with a pooled AUC of 0.805 (95% CI: 0.756–0.853) from 7 models.
• Detailed methodological critique using PROBAST framework quantifies specific bias rates (e.g., 93.75% for missing data).

Limitations

• Dataset is a 10.7 KB summary of a review, containing aggregated statistics rather than primary patient-level data.
• All 16 included studies were assessed as having a high risk of bias, limiting the reliability of the underlying models.
• The small file size indicates limited scope, likely containing only summary tables and figures from the publication.

Provenance

Source

figshare, authored by Zhiheng Zhan.

Collection Method

Systematic review and meta-analysis of studies from PubMed, Web of Science, Embase, CINAHL, MEDLINE, SinoMed, CNKI, and Wan-Fang databases.

Time Range

Literature search covered studies up to September 30, 2025.

Freshness

Last updated April 2026, based on a literature search up to September 2025.