Perspectives on validation of clinical predictive algorithms

Anne A H de Hond; Vaibhavi B Shah; Ilse M J Kant; Ben Van Calster; Ewout W Steyerberg; Tina Hernandez-Boussard

doi:10.1038/s41746-023-00832-9

Perspectives on validation of clinical predictive algorithms

NPJ Digit Med. 2023 May 6;6(1):86. doi: 10.1038/s41746-023-00832-9.

Authors

Anne A H de Hond^{1

2

3}, Vaibhavi B Shah⁴, Ilse M J Kant⁵, Ben Van Calster^{6

7}, Ewout W Steyerberg^{8

6}, Tina Hernandez-Boussard^{4

9

10}

Affiliations

¹ Clinical AI Implementation and Research Lab, Leiden University Medical Centre, Leiden, the Netherlands. a.a.h.de_hond@lumc.nl.
² Department of Medicine (Biomedical Informatics), Stanford University, Stanford, CA, USA. a.a.h.de_hond@lumc.nl.
³ Department of Biomedical Data Sciences, Leiden University Medical Centre, Leiden, the Netherlands. a.a.h.de_hond@lumc.nl.
⁴ Department of Medicine (Biomedical Informatics), Stanford University, Stanford, CA, USA.
⁵ Department of Digital Health, University Medical Center Utrecht, Utrecht, the Netherlands.
⁶ Department of Biomedical Data Sciences, Leiden University Medical Centre, Leiden, the Netherlands.
⁷ Department of Development & Regeneration, KU Leuven, Leuven, Belgium.
⁸ Clinical AI Implementation and Research Lab, Leiden University Medical Centre, Leiden, the Netherlands.
⁹ Department of Biomedical Data Science, Stanford University, Stanford, CA, USA.
¹⁰ Department of Epidemiology & Population Health (by courtesy), Stanford University, Stanford, CA, USA.

Abstract

The generalizability of predictive algorithms is of key relevance to application in clinical practice. We provide an overview of three types of generalizability, based on existing literature: temporal, geographical, and domain generalizability. These generalizability types are linked to their associated goals, methodology, and stakeholders.