Determine the availability and implementability of vertical statistical analytics methods for health data

Determine the extent of available statistical analytics methods for vertically partitioned data and ascertain the degree to which these methods can be implemented with real-world health databases to enable statistical inference.

Background

Health research often requires integrating complementary variables that are split across institutions, creating vertically partitioned datasets where individual-level data cannot be shared due to regulatory constraints. Vertical methods aim to enable statistical inference without pooling individual-level data.

Before conducting their scoping review, the authors note that the overall availability of vertical statistical analytics methods and their practical implementability with real-world health databases had not been established, motivating a systematic assessment of existing methods and their properties.

References

At this stage, the extent of the availability of vertical statistical analytics methods has not yet been determined and the degree to which these existing methods can be implemented with real-world health databases is unclear.

— Multi-Site Health Research Integrating Complementary Data Sources: A Scoping Review of Statistical Inference Methods for Vertically Partitioned Data  (2604.02595 - Domingue et al., 3 Apr 2026) in Section 1. INTRODUCTION