Objectives: To introduce statistical methods to combine the different prediction models published for the same outcome or target population into a single prediction model based on aggregate data only, and augmented with minimal individual participant-level data (IPD).

Commonly, numerous prediction models have been developed for the same outcome or target population. For example, there are over 100 models for predicting outcome after traumatic brain injury, over 60 for breast cancer and over 40 for diabetes type 2. Rather than developing the next prediction model for a particular outcome or target population, systematic reviews of risk prediction models have become timely. These raise questions about whether, and how, previously published prediction models should and can be combined in a meta-analytical manner. Recently, innovative methods have been developed to meta-analyse (combine) previously published prediction models, given that particular aggregate data from these studies is available. Also the combination of such aggregate data plus minimal individual level participant data from one’s own single study, is illustrated.

We describe whether and how different published prediction models for the same outcome or target population can be combined into a single prediction model. We provide strategies for improving upon the generalizability and applicability of the so-derived meta-analytical prediction model, accounting for heterogeneity across the studies.