Christopher G Hudson1, Varda Soskolne. 1. School of Social Work, Salem State University, 352 Lafayette Street, Salem, MA 01945, United States. chudson@salemstate.edu
Abstract
OBJECTIVE: The aim of this research is to test and apply a model of the disparities and variations in serious mental illness (SMI) to estimating prevalence in local areas throughout Israel. METHODS: This study employs a secondary analysis of data from the 2003/2004 Israel National Health Survey of 4859 adults aged 21 and over from the household population of legal residents and citizens. It uses small area estimation methods (SAE), specifically to: (i) estimate and test a multivariate logistic model of disparities in the risk of serious mental illness; (ii) use the foregoing model for computing estimates, using census data, for local areas; (iii) validate these estimates against the rate of psychiatric hospitalizations. RESULTS: The model uses standard demographic and socioeconomic variables to successfully predict 92.5% of respondents' statuses as SMI, with a sensitivity of 26.9%, specificity of 95.9%, and an AUC index of .797. The resulting estimates of the percentage of adults with an SMI in the 16 subdistricts ranged between 3.7% and 7.7%, with a national mean of 5.0%. The estimates have a partial correlation of .63 with rates of psychiatric hospitalization in Jewish localities, but elevated rates have not been validated in Arab localities. CONCLUSION: The use of small area estimation methods demonstrated the capacity for deriving local prevalence rates of serious mental illness, ones that can be validated against psychiatric hospitalization for the majority population group in Israel.
OBJECTIVE: The aim of this research is to test and apply a model of the disparities and variations in serious mental illness (SMI) to estimating prevalence in local areas throughout Israel. METHODS: This study employs a secondary analysis of data from the 2003/2004 Israel National Health Survey of 4859 adults aged 21 and over from the household population of legal residents and citizens. It uses small area estimation methods (SAE), specifically to: (i) estimate and test a multivariate logistic model of disparities in the risk of serious mental illness; (ii) use the foregoing model for computing estimates, using census data, for local areas; (iii) validate these estimates against the rate of psychiatric hospitalizations. RESULTS: The model uses standard demographic and socioeconomic variables to successfully predict 92.5% of respondents' statuses as SMI, with a sensitivity of 26.9%, specificity of 95.9%, and an AUC index of .797. The resulting estimates of the percentage of adults with an SMI in the 16 subdistricts ranged between 3.7% and 7.7%, with a national mean of 5.0%. The estimates have a partial correlation of .63 with rates of psychiatric hospitalization in Jewish localities, but elevated rates have not been validated in Arab localities. CONCLUSION: The use of small area estimation methods demonstrated the capacity for deriving local prevalence rates of serious mental illness, ones that can be validated against psychiatric hospitalization for the majority population group in Israel.