Al Ozonoff1, L Forsberg, M Bonetti, M Pagano. 1. Department of Biostatistics, Harvard School of Public Health, 655 Huntington Avenue, Boston, MA 02115, USA. pagano@hsph.harvard.edu
Abstract
INTRODUCTION: Statistical analysis of syndromic data has typically focused on univariate test statistics for spatial, temporal, or spatio-temporal surveillance. However, this approach does not take full advantage of the information available in the data. OBJECTIVES: A bivariate method is proposed that uses both temporal and spatial data information. METHODS: Using upper respiratory syndromic data from an eastern Massachusetts health-care provider, this paper illustrates a bivariate method and examines the power of this method to detect simulated clusters. RESULTS: Use of the bivariate method increases detection power. CONCLUSIONS: Syndromic surveillance systems should use all available information, including both spatial and temporal information.
INTRODUCTION: Statistical analysis of syndromic data has typically focused on univariate test statistics for spatial, temporal, or spatio-temporal surveillance. However, this approach does not take full advantage of the information available in the data. OBJECTIVES: A bivariate method is proposed that uses both temporal and spatial data information. METHODS: Using upper respiratory syndromic data from an eastern Massachusetts health-care provider, this paper illustrates a bivariate method and examines the power of this method to detect simulated clusters. RESULTS: Use of the bivariate method increases detection power. CONCLUSIONS: Syndromic surveillance systems should use all available information, including both spatial and temporal information.
Authors: Al Ozonoff; Thomas Webster; Veronica Vieira; Janice Weinberg; David Ozonoff; Ann Aschengrau Journal: Environ Health Date: 2005-09-15 Impact factor: 5.984