Adverse Event Frequency in Paediatric Critical Care Transports, London 2016–2021

1,519 interhospital transports of critically ill children conducted by a specialist paediatric critical care team in London between 2016 and 2021. The dataset contains frequency data for adverse physiological events, likely used to train and evaluate machine learning models for predicting respiratory and cardiovascular deterioration up to 15 minutes in advance. It was authored by Milan Kapur and shared under a CC-BY-4.0 license.

Use Cases

• Benchmarking lightweight predictive models based on vital sign time-series and clinical data.
• Analyzing the temporal frequency of adverse events during paediatric interhospital transport.
• Developing interpretable risk prediction tools for deployment on edge devices in clinical settings.
• Studying the relationship between patient diagnoses and physiological deterioration during transport.

Strengths

• Data is based on 1,519 real-world paediatric critical care transports.
• Covers a 5-year time period from 2016 to 2021.
• Models trained on this data achieved reported AUROC scores of 0.851 and 0.792.
• Dataset is small (5.5 KB) and focused, facilitating quick inspection.

Limitations

• Row count and specific column definitions are unknown, requiring manual inspection after download.
• Geographic scope is limited to a single specialist team's operations in London, which may limit generalizability.
• Description metadata is limited; actual data quality and structure require verification.

Provenance

Source

Milan Kapur via figshare.

Collection Method

Likely collected from continuously streamed vital signs and clinical data during specialist paediatric critical care transports.

Time Range

2016–2021

Freshness

Last updated 2026-05-19 17:28:25; freshness should be verified.

Geography

London, United Kingdom