Ruwan Ratnayake1, Robert Allard. 1. 1. Canadian Field Epidemiology Program, Public Health Agency of Canada, Ottawa, Ontario, Canada 2. Public Health Department, Montréal Health and Social Services Agency, Montréal, Québec, Canada. ruwan.ratnayake@rescue.org.
Abstract
OBJECTIVES: Though rare in Montréal, meningococcal disease continues to cause serious morbidity and mortality. In an era of declining incidence, our objective was to evaluate the sensitivity and the timeliness of case reporting and the capacity to statistically detect disease clusters. METHODS: We used the public health department's reportable disease database (RDD) to calculate the timeliness of reporting by physicians and laboratories for the period 1995 to 2008. The sensitivity of case reporting was evaluated through capture-recapture estimation using the RDD and the hospitalization discharge database (MED-ECHO). To evaluate the detection of cases clustered by time and proximity, we applied scan statistics to the RDD with cases coded by time and geographic location for the period 1992 to 2008. RESULTS: While the system sensitivity was judged to be high at 94%, physicians reported only 54% of cases. A total of 92.3% of cases were notified by physicians or laboratories within seven days, meaning that in theory, 13 cases were not notified in time to conduct thorough contact tracing and offer chemoprophylaxis to close contacts. In high-incidence years, scan statistics detected two statistically-significant clusters one to two weeks earlier than traditional detection through the manual monitoring of cumulative cases. CONCLUSIONS: To improve system performance, we recommended increasing the emphasis of laboratory reporting, reinforcing early reporting by physicians and if incidence increases, using scan statistics to identify clusters that can add to a public health practitioner's initial "hunch" of an emerging outbreak.
OBJECTIVES: Though rare in Montréal, meningococcal disease continues to cause serious morbidity and mortality. In an era of declining incidence, our objective was to evaluate the sensitivity and the timeliness of case reporting and the capacity to statistically detect disease clusters. METHODS: We used the public health department's reportable disease database (RDD) to calculate the timeliness of reporting by physicians and laboratories for the period 1995 to 2008. The sensitivity of case reporting was evaluated through capture-recapture estimation using the RDD and the hospitalization discharge database (MED-ECHO). To evaluate the detection of cases clustered by time and proximity, we applied scan statistics to the RDD with cases coded by time and geographic location for the period 1992 to 2008. RESULTS: While the system sensitivity was judged to be high at 94%, physicians reported only 54% of cases. A total of 92.3% of cases were notified by physicians or laboratories within seven days, meaning that in theory, 13 cases were not notified in time to conduct thorough contact tracing and offer chemoprophylaxis to close contacts. In high-incidence years, scan statistics detected two statistically-significant clusters one to two weeks earlier than traditional detection through the manual monitoring of cumulative cases. CONCLUSIONS: To improve system performance, we recommended increasing the emphasis of laboratory reporting, reinforcing early reporting by physicians and if incidence increases, using scan statistics to identify clusters that can add to a public health practitioner's initial "hunch" of an emerging outbreak.
Entities:
Keywords:
Disease clustering; Neisseria meningitidis; evaluation; public health; surveillance