Abstract

BACKGROUND: The Canadian Immunization Monitoring Program, ACTive (IMPACT) surveillance for rotavirus relies on monitoring hospital admissions. Because a diagnosis of rotavirus is not necessary for treatment purposes, and rotavirus is not a reportable disease, wide variation may exist in the admitting and testing practices for this disease. From 2005 to 2007, the number of rotavirus admissions differed significantly among IMPACT centres, and this variation could not be explained by population differences alone. Understanding this variation is important when interpreting surveillance data and estimating the cost-effectiveness of rotavirus vaccination programs.METHODS: Key informant interviews were conducted with pediatric infectious disease physicians and IMPACT nurse monitors involved with rotavirus surveillance to obtain in-depth information about rotavirus testing and admitting practices at each of the 12 IMPACT centres.RESULTS: A total of 18 of 24 interviews were completed, with at least one interview conducted with physicians and/or nurses at each centre. Four major differences were identified among the centres: case-identification methods, admission definitions, admission criteria and testing criteria. The criteria for admitting and testing patients as well as which patients were defined as admissions had the greatest influence on case totals.DISCUSSION: The present study found that differences in admitting and testing practices may contribute to significant differences in rotavirus admission totals. Given these differences, caution should be used when using local case estimates for cost-effectiveness analyses and immunization program decisions. The present study illustrates that understanding the factors that influence the identification of a disease is important when interpreting and applying surveillance data.