Understanding prognostic models: The example of the PHASES score for unruptured intracranial aneurysms
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Prognostic models require rigorous validation before clinical use. Unvalidated models for conditions like unruptured intracranial aneurysms (UIAs) should not guide patient care.
Area Of Science
- Medical Statistics
- Clinical Epidemiology
Background
- Prognostic studies inform disease course and clinical decisions.
- Significant challenges exist in developing reliable prognostic models.
Purpose Of The Study
- To review the construction of prognostic models.
- To highlight inductive problems in prognostic study design.
- To emphasize the critical need for model validation.
Main Methods
- Summarized a study on the natural history of unruptured intracranial aneurysms (UIAs).
- Reviewed steps in constructing prognostic models.
- Focused on inductive problems from event extension to risk class definition.
Main Results
- Model development involves selecting baseline variables based on existing knowledge and associations.
- An infinite number of models can fit the same data, necessitating external validation.
- Prognostic models for UIAs have not been externally validated.
Conclusions
- Prognostic studies are fraught with inherent difficulties.
- Unvalidated prognostic models pose risks and should be avoided in clinical practice.
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.