Review of Prognostic Models in Traumatic Brain Injury
Author Information
Author(s): Pablo Perel, Phil Edwards, Reinhard Wentz, Ian Roberts
Primary Institution: London School of Hygiene & Tropical Medicine
Hypothesis
The systematic review aims to assess the quality and characteristics of existing prognostic models for traumatic brain injury.
Conclusion
Prognostic models for traumatic brain injury are often poorly developed, rarely validated externally, and not presented in a user-friendly manner.
Supporting Evidence
- 53 reports describing 102 prognostic models were identified.
- 93% of the models were derived from high-income country populations.
- Only 15% of models reported a loss to follow-up of less than 10%.
Takeaway
This study looked at different models that try to predict how well people will do after a brain injury, but found that many of them are not very good or easy to use.
Methodology
The review included studies that combined at least two variables to predict outcomes in patients with traumatic brain injury, assessed through independent review of titles and abstracts.
Potential Biases
Potential bias due to high loss to follow-up and lack of discussion on predictor rationale.
Limitations
Many models had small sample sizes, poor methodological quality, and were not validated on external populations.
Participant Demographics
Almost half of the models were derived from adult patients, with a majority from high-income countries.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website