Citation

Congdon P. Eur. J. Popul. 2000; 16(3): 251-284.

Copyright

(Copyright © 2000, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1023/A:1026587810551

PMID

unavailable

Abstract

A significant fall in suicide mortality relative to England and Wales levels has occurred in London though with wide variation between its 33 constituent boroughs in the extent of mortality reduction. A Bayesian random effects approach is used is to model differential changes in suicide by borough and time over a 16 year period, 1979-94. Of particular concern in such modelling are persistent differences between boroughs in suicide risk (temporal correlation) and spatial clustering in relative risk. It is also important to represent the changing impact on suicide of socio-economic factors such as social deprivation. The data used are defined by deaths through de-jure suicide (ICD9 categories E950-E959) and those through undetermined injury, whether accidental or purposely inflicted (ICD E980-E989).

Language: en

Keywords

Bayesian analysis; mortality; suicide; United Kingdom