Background: Following an influx of an estimated 742,000 Rohingya refugees in Bangladesh, Médecins sans Frontières (MSF) established an active indicator-based Community Based Surveillance (CBS) in 13 sub-camps in Cox's Bazar in August 2017. Its objective was to detect epidemic prone diseases early for rapid response. We describe the surveillance, alert and response in place from epidemiological week 20 (12 May 2019) until 44 (2 November 2019).
Methods: Suspected cases were identified through passive health facility surveillance and active indicator-based CBS. CBS-teams conducted active case finding for suspected cases of acute watery diarrhea (AWD), acute jaundice syndrome (AJS), acute flaccid paralysis (AFP), dengue, diphtheria, measles and meningitis. We evaluate the following surveillance system attributes: usefulness, Positive Predictive Value (PPV), timeliness, simplicity, flexibility, acceptability, representativeness and stability.
Results: Between epidemiological weeks 20 and 44, an average of 97,340 households were included in the CBS per surveillance cycle. Household coverage reached over 85%. Twenty-one RDT positive cholera cases and two clusters of AWD were identified by the CBS and health facility surveillance that triggered the response mechanism within 12 hours. The PPV of the CBS varied per disease between 41.7%-100%. The CBS required 354 full-time staff in 10 different roles. The CBS was sufficiently flexible to integrate dengue surveillance. The CBS was representative of the population in the catchment area due to its exhaustive character and high household coverage. All households consented to CBS participation, showing acceptability.
Discussion: The CBS allowed for timely response but was resource intensive. Disease trends identified by the health facility surveillance and suspected diseases trends identified by CBS were similar, which might indicate limited additional value of the CBS in a dense and stable setting such as Cox's Bazar. Instead, a passive community-event-based surveillance mechanism combined with health facility-based surveillance could be more appropriate.