Using health and demographic surveillance for the early detection of cholera outbreaks: analysis of community- and hospital-based data from Matlab, Bangladesh

Abstract

Background: Cholera outbreaks are a continuing problem in Bangladesh, and the timely detection of an outbreak is important for reducing morbidity and mortality. In Matlab, the ongoing Health and Demographic Surveillance System (HDSS) data records symptoms of diarrhea in children under the age of 5 years at the community level. Cholera surveillance in Matlab currently uses hospital-based data. Objective: The objective of this study is to determine whether increases in cholera in Matlab can be detected earlier by using HDSS diarrhea symptom data in a syndromic surveillance analysis, when compared to hospital admissions for cholera. Methods: HDSS diarrhea symptom data and hospital admissions for cholera in children under 5 years of age over a 2-year period were analyzed with the syndromic surveillance statistical program EARS (Early Aberration Reporting System). Dates when significant increases in either symptoms or cholera cases occurred were compared to one another. Results: The analysis revealed that there were 43 days over 16 months when the cholera cases or diarrhea symptoms increased significantly. There were 8 months when both data sets detected days with significant increases. In 5 of the 8 months, increases in diarrheal symptoms occurred before increases of cholera cases. The increases in symptoms occurred between 1 and 15 days before the increases in cholera cases. Conclusions: The results suggest that the HDSS survey data may be able to detect an increase in cholera before an increase in hospital admissions is seen. However, there was no direct link between diarrheal symptom increases and cholera cases, and this, as well as other methodological weaknesses, should be taken into consideration.