What question did this study set out to answer?

The aim is to evaluate the concordance of the Short-Form PROBAST with full PROBAST in assessing risk of bias in prediction models.

March 22, 2026Open Access

Concordance analysis of the Short-Form of the PROBAST as a Screening Tool: Two-Step Risk of Bias Assessment showing only Underestimation of Risk of Bias

Key Points

The aim is to evaluate the concordance of the Short-Form PROBAST with full PROBAST in assessing risk of bias in prediction models.
Applied SF-PROBAST and compared it with PROBAST scores for methodological quality assessment.
Conducted a two-step risk of bias assessment.
Recommended using full PROBAST on studies flagged with low or unclear risk of bias by SF-PROBAST.
SF-PROBAST showed good agreement with PROBAST scores.
Identified a consistent underestimation of risk of bias in certain studies with SF-PROBAST.

Abstract

SF-PROBAST showed good agreement with PROBAST scores and offers a reliable, simplified, time-efficient two-step alternative for assessing the methodological quality of prediction models. To assess potential misclassification, we recommend applying the full PROBAST to studies scored as low or unclear ROB on the SF-PROBAST.

Concordance analysis of the Short-Form of the PROBAST as a Screening Tool: Two-Step Risk of Bias Assessment showing only Underestimation of Risk of Bias

Key Points

Abstract

Cite This Study