Causal blind spots when using prediction models for treatment decisions
Abstract
Prediction models are increasingly proposed for guiding treatment decisions, but most fail to address the special role of treatments, leading to inappropriate use. This paper highlights the limitations of using standard prediction models for treatment decision support. We identify 'causal blind spots' in three common approaches to handling treatments in prediction modelling and illustrate potential harmful consequences in several medical applications. We advocate for an extension of guidelines for development, reporting, clinical evaluation and monitoring of prediction models to ensure that the intended use of the model is matched to an appropriate risk estimand. For decision support this requires a shift towards developing predictions under the specific treatment options under consideration ('predictions under interventions'). We argue that this will improve the efficacy of prediction models in guiding treatment decisions and prevent potential negative effects on patient outcomes.
Authors 15
-
London School of Hygiene & Tropical Medicine
Affiliation as printed
) London School of Hygiene and Tropical Medicine , London , United Kingdom (
-
Affiliation as printed
APH - Methodology
Affiliatie UvA
Medical Informatics
-
Delft University of Technology
Affiliation as printed
) Delft University of Technology , Delft , The Netherlands (
-
Affiliation as printed
Affiliatie UvA
Medical Informatics
-
University Medical Center Utrecht
Affiliation as printed
) University Medical Center Utrecht , Utrecht , The Netherlands (
-
Affiliation as printed
) University of Amsterdam , Amsterdam , The Netherlands. (
-
University of Washington · Ghent University
Affiliation as printed
) Ghent University , Ghent , Belgium (
) University of Washington , Seattle , United States (
-
Hein Putter Aachen
Leiden University Medical Center
Affiliation as printed
( ) Leiden University Medical Center , Leiden , The Netherlands (
-
Affiliation as printed
) University of Manchester , Manchester , United Kingdom (
-
Affiliation as printed
Affiliatie UvA
Doctoral School
-
Affiliation as printed
) Utrecht University , Utrecht , The Netherlands (
-
Leibniz Institute for Prevention Research and Epidemiology - BIPS
Affiliation as printed
) Leibniz Institute for Prevention Research and Epidemiology -BIPS , Bremen , Germany