2019
DOI: 10.1002/sim.8296
|View full text |Cite
|
Sign up to set email alerts
|

Assessment of heterogeneity in an individual participant data meta‐analysis of prediction models: An overview and illustration

Abstract: Clinical prediction models aim to provide estimates of absolute risk for a diagnostic or prognostic endpoint. Such models may be derived from data from various studies in the context of a meta‐analysis. We describe and propose approaches for assessing heterogeneity in predictor effects and predictions arising from models based on data from different sources. These methods are illustrated in a case study with patients suffering from traumatic brain injury, where we aim to predict 6‐month mortality based on indi… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

1
63
0

Year Published

2020
2020
2024
2024

Publication Types

Select...
5
2

Relationship

3
4

Authors

Journals

citations
Cited by 47 publications
(64 citation statements)
references
References 61 publications
1
63
0
Order By: Relevance
“…Preferably, extensions are sought to non-Western hospitals, with large numbers of patients, high-quality data, and long-term follow-up. Such extensions may reveal heterogeneity in performance, motivating locally adapted versions of the calculator, that is, model updating ( 6 ).…”
mentioning
confidence: 99%
“…Preferably, extensions are sought to non-Western hospitals, with large numbers of patients, high-quality data, and long-term follow-up. Such extensions may reveal heterogeneity in performance, motivating locally adapted versions of the calculator, that is, model updating ( 6 ).…”
mentioning
confidence: 99%
“…Prediction models should ideally be developed and validated in large samples from multiple populations and settings. [17,23,20,24,25] This requires research groups to join efforts by sharing their individual participant data and subsequently applying adequate statistical methods to synthesize the data across studies or research centers. To account for heterogeneity between settings and populations (randomeffects) meta-analysis can be used, which appropriately weights the evidence from each study.…”
Section: Individual Participant Data Meta-analysismentioning
confidence: 99%
“…This model may account for clustering by including random intercepts and/or predictor effects. [25,116,232,228] A disadvantage of the one-stage approach in IECV is that the data from each cluster needs to be used K-1 times to fit a model in the one-stage approach. On the other hand, in the two-stage IECV approach the data from each cluster only needs to be used for model fitting once, as the second stage comprises metaanalysis of different combinations of coefficients and their standard errors.…”
Section: Model Fittingmentioning
confidence: 99%
See 2 more Smart Citations