Geospatial Analysis of Environmental Risk Factors for Missing Dementia Patients.

Abstract

BACKGROUND\nDementia-related missing incidents are highly prevalent but still poorly understood. This is particularly true for environmental/geospatial risk factors, which might contribute to these missing incidents.\n\n\nOBJECTIVE\nThe study aimed to conduct a retrospective, observational analysis on a large sample of missing dementia patient case records provided by the police (n\u200a=\u200a210), covering dates from January 2014 to December 2017. In particular, we wanted to explore 1) whether there were any hotspot regions of missing incidents and 2) the relationship between outdoor landmark density and missing incidents.\n\n\nMETHODS\nGlobal spatial autocorrelation (Moran s I) was used to identify the potential hotspot regions for missing incidents. Meanwhile, spatial buffer and regression modelling were used to determine the relationship between outdoor landmark density and missing incidents.\n\n\nRESULTS\nOur demographics measures replicated and extended previous studies of dementia-related missing incidents. Meanwhile, no hotspot regions for missing incidents were identified, while higher outdoor landmark density leads to increased missing incidents.\n\n\nCONCLUSION\nOur results highlight that missing incidents do not occur in isolated hotspots of regions but instead are endemic in patients regardless of location. Higher outdoor landmark density emerges as a significant geospatial factor for missing incidents in dementia, which crucially informs future safeguarding/intervention studies.