The MARS pilot project: implementing real-time measles and rubella surveillance during elimination phase in Canada
OBJECTIVES: Measles and rubella are nationally notifiable, vaccine-preventable diseases targeted for elimination by the Pan American Health Organization (PAHO). To support national and international elimination efforts, surveillance optimization is important to ensure rapid case detection, document endemic transmission interruption, identify susceptible populations and inform immunization strategies. While current national surveillance captures confirmed-case data, its performance cannot be assessed using PAHO-recommended surveillance indicators as suspect-case investigation data are required for their estimation. In Canada, the investigation of clinically-suspect measles-like illness (MLI) is highly dependent on laboratory evidence, providing an opportunity to use laboratory data to estimate MLI investigation rates. The Measles and Rubella Surveillance (MARS) pilot project was developed to address existing surveillance challenges with the central hypothesis that (I) ‘it is feasible to develop and implement a real‐time, web‐based measles and rubella surveillance system in the Canadian setting’, and the following sub‐hypotheses: (II) ‘implementation of real‐time surveillance in MARS pilot provinces will result in increased timeliness of national measles and rubella surveillance when compared with established confirmed-case surveillance’, and (III) ‘it is possible to use augmented laboratory data to estimate the performance of national measles and rubella surveillance using adapted PAHO indicators’. METHODS: A MARS application was designed to support centralized real-time measles/ rubella investigation reporting and alerting with integration of non-nominal laboratory and epidemiological data, then developed and piloted using the web-based Canadian Network for Public Health Intelligence platform in British Columbia, Alberta and Newfoundland from June/2011-May/2012. Pre- and post-pilot laboratory surveys were conducted to retrospectively assess national surveillance performance in ‘outbreak’ and ‘non-outbreak’ settings during the 2005‐2011 and pilot years using various surveillance indicators and attributes. Measles IgM serology testing was used as a laboratory-based proxy for MLI investigation to support indicator estimation. RESULTS: Real-time, integrated surveillance was successfully implemented in MARS pilot provinces as modeled within the context of established reporting roles, and surveillance indicators and attributes were estimated using augmented laboratory data. MARS surveillance was more timely than confirmed-case surveillance, and real-time MARS reports exceeded all laboratory-related PAHO targets evaluated: 100% met ‘sample collection’ and ‘receipt’ timelines, and 91.7% met ‘result' timelines (Targets:≥80%); 99.8% of all MLI investigations were discarded (Target:≥95%). A national ‘non-outbreak’ baseline rate of 14 MLI investigations/100 000 population was estimated, whereas MARS pilot sites averaged 22 MLI investigations/100 000 population during the pilot year. While ‘non-outbreak’ investigation rates varied between provinces, all annual provincial and national rates estimated for the 2005‐2011 and MARS pilot years exceeded the PAHO investigation target of ≥2 suspected cases/100 000 population in settings attempting elimination. CONCLUSIONS: The MARS model supported more timely and integrated national measles and rubella surveillance, and enabled indicator‐based performance assessment. Results underscore the importance of laboratory data when evaluating and documenting surveillance performance to support elimination efforts. Consideration should be given to national MARS implementation and its use as a model adaptable to the case-based surveillance of other nationally notifiable diseases.
measles, rubella, microbiology, epidemiology, virus, surveillance, infectious disease, real-time, Congenital Rubella Syndrome, IgM, indicators, incidence