METHODS: Electronic laboratory reporting (ELR), automated transmission of reports into a surveillance database, real-time geocoding of case addresses, daily matching of case residence building identification numbers (BINs) to previously reported cases and congregate living facilities, and routine cluster detection analyses have been implemented by the New York City (NYC) Department of Health and Mental Hygiene (DOHMH) for legionellosis and other reportable communicable diseases. Cluster detection analyses include weekly application of the historical limits method and daily application of the space-time permutation scan statistic. When feasible, DOHMH staff reviewed electronic health records using health information exchange (HIE) patient lookup portals. Molecular tests on clinical Legionella isolates are reported electronically from the public health laboratory to the surveillance database. Walking routes between patient homes and destinations were estimated using ArcGIS Network Analyst to identify potential locations of common exposure. Multi-focused cluster tests and descriptive spatial statistics were used to assess the degree of clustering of cases around potential environmental sources. Methods for detecting and managing recent outbreaks of legionellosis in NYC were described with attention to detection methods, their timeliness, and whether an environmental source was identified for remediation.
RESULTS: Three community legionellosis outbreaks occurred in NYC during November 2014–November 2015. These outbreaks were first detected by the historical limits method, the space-time permutation scan statistic, and by an astute epidemiologist, respectively; the days from first outbreak-associated case report to earliest cluster detection were 22, 9 and 4, respectively. Each outbreak was also detected by other methods, including hospital calls and BIN matches. Initial clinical and radiologic data were found readily in an HIE, but not for the majority of cases. In each outbreak, a cooling tower was immediately remediated and subsequently implicated as the source by laboratory evidence.
CONCLUSIONS: New electronic surveillance tools are valuable additions to the public health armamentarium, supplementing traditional epidemiologic methods, enabling rapid detection of legionellosis outbreaks, and providing the opportunity for timely and effective public health intervention. HIE patient portals provide timely access to clinical data, but facility participation and completeness of data must improve to maximize gains in efficiency. These tools will likely help public health agencies confront other threats, but require mature and well-supported informatics infrastructure and expertise.