The consistency of shingles and its significance for health monitoring.

Accurate estimation of monitored populations is essential for epidemiological study. Many countries do not have systems of patient registration and routine disease surveillance is thereby hindered. We studied the incidence of shingles over time and investigated the hypothesis that the incidence is consistent and could be used as a proxy for estimating the monitored population. Annual incidence rates of shingles reported in the Weekly Returns Service (WRS) since 1970 and in the Dutch Sentinel Network (DSN) over the period 1998--2001 were studied. Gender specific annual rates (1998--2001) were compared after standardising for age. The population in the DSN was estimated by applying the WRS incidence rates to the numbers of DSN incident cases. The incidence of shingles was annually and seasonally consistent. Incidence in males was similar in both networks and in females approximately 18% greater in the WRS: in age groups 15-64 years, incidence was similar in both networks, but in children 0-14 years and in persons 65 years and over, it was higher in the WRS. The total populations in the DSN estimated from average age/gender specific rates in the WRS were within 12% of the observed in each of the 4 years surveyed. The incidence of shingles in the two countries was sufficiently close to estimate the surveyed population aged 15-64 years from knowledge of incident cases in the community. Routine monitoring of shingles in sentinel practice networks is commended as a method of assuring recording quality and as a means of estimating the survey population where the registered population is not known.