Spatial measurement errors in the field of spatial epidemiology.

Area of Science:

• Epidemiology
• Geographic Information Science
• Spatial Analysis

Background:

• Spatial epidemiology utilizes GIS, remote sensing, and GPS, alongside specialized statistical methods.
• The increasing use of spatial data in epidemiology necessitates understanding common spatial measurement errors.

Purpose of the Study:

• To review and analyze common spatial measurement errors in spatial epidemiological analysis.
• To propose a systematic framework for understanding and addressing these errors.

Main Methods:

• Comprehensive literature search of Google Scholar, Medline, and Scopus databases up to December 20, 2014.
• Keywords included terms for location (space, geography, location, position) and measurement error (measurement error, inaccuracy, misclassification, uncertainty).
• Review of identified papers and their citations to ensure relevance.

Main Results:

• Spatial measurement errors were classified into four main groups: pure spatial location errors, location-based outcome errors, location-based covariate errors, and covariate-outcome spatial misalignment errors.
• Specific examples within these categories include geocoding errors, address proxies, and missing outcome measurements.
• A proposal for unifying these error types within an integrated theoretical model was discussed.

Conclusions:

• Spatial measurement errors pose a significant threat to the validity of spatial epidemiological research.
• A systematic framework is proposed to understand the mechanisms generating these errors.
• Practical examples are provided to illustrate these ubiquitous threats.