This review underscores the underutilization of geospatial modeling in kidney research, showcasing improved precision and applicability through spatial models, as demonstrated in a case study using U.S. nationwide chronic kidney disease data.

The following is a summary of “Geospatial modeling methods in epidemiological kidney research: An overview and practical example,” published in the January 2024 issue of Nephrology by Buchalter et al.

Geospatial modeling methods in population-level kidney research have been underutilized, with limited studies conducting comprehensive spatial analyses linking risk factors and exposures to kidney conditions and outcomes. These spatial models offer distinct advantages over traditional approaches, providing an enhanced estimation of statistical variation and a more precise estimation of coefficient effect direction/magnitudes by accounting for the spatial data structure. Given the prevalence of geographically referenced data in population-level kidney research, there is a pressing need for a deeper understanding of geospatial modeling to assess associations between individual geolocation, care processes, and clinical outcomes.  

This review elucidates common spatial models, offering insights into their execution, and presents a case study to demonstrate the impact of integrating geographic structure. Leveraging U.S. nationwide 2019 chronic kidney disease data from the CDC’s Kidney Disease Surveillance System and 2006-2010 EPA environmental quality index (EQI) data, the researchers applied a non-spatial count model alongside global spatial models (spatially lagged [SLM]/pseudo-spatial error model [PSEM]) and a local spatial model (geographically weighted quasi-Poisson regression [GWQPR]). Results indicated that the SLM, PSEM, and GWQPR models enhanced model fit compared to the non-spatial regression, with the PSEM model attenuating the positive relationship between EQI and CKD prevalence. The GWQPR uncovered spatial heterogeneity in the EQI-CKD relationship. 

In conclusion, spatial modeling emerges as a promising clinical and public health translational tool, and the case study serves as an illustrative example of how these analyses can be conducted to enhance the precision and applicability of research findings.

Source: sciencedirect.com/science/article/pii/S2468024924000184