Record linkage and capture-recapture estimates for underreporting of human leptospirosis in a Brazilian health district

AUTOR(ES)

Brum, Liane, Kupek, Emil

FONTE

Brazilian Journal of Infectious Diseases

DATA DE PUBLICAÇÃO

2005-12

RESUMO

Record linkage and capture-recapture models were used to estimate the number of cases of human leptospirosis in the health district of Santa Maria, RS in southern Brazil. Twelve months of laboratory, hospital and epidemiological surveillance data were matched by name, age, residence and the month of diagnosis. Only laboratory-confirmed cases were considered. The record linkage revealed more than 20 times more cases than the official estimate for the health district, indicating a leptospirosis epidemic, with an annual incidence of more than 3 per 1,000 inhabitants and a case fatality of 0.37%. Severe cases were predominantly found through hospital records, overlapping to some extent with the epidemiological surveillance data, whereas less severe cases were found almost exclusively through laboratory logs. Different combinations of data sources influenced the detection rate for low versus high severity cases. Based on log-linear capture-recapture models, stratified by case severity and taking into account possible dependencies between the data sources, an insignificant number of cases were missed by all sources.

Documentos Relacionados

• Capture-recapture methods for precise measurement of the incidence and prevalence of stroke.
• Estimating the point accuracy of population registers using capture-recapture methods in Scotland.
• Estimating the prevalence of drug misuse in Dundee, Scotland: an application of capture-recapture methods.
• Estimation of underreporting of live births by the capture-recapture
• Estimating the World Trade Center Tower Population on September 11, 2001: A Capture–Recapture Approach