Although the validity of diagnostic codes for shingles was slightly lower for females than for males in an American study, shingles was still more common in females than in males [16]. The higher rates of medically attended shingles in females than males might

be related to gender differences in immunosuppressive disease or therapies [17]; we were not able to examine PKC inhibitor this. One may also speculate that there might be gender differences in immune responses to latent viral infections. Gender differences in health seeking behaviour could also contribute to the observed higher rate of shingles in females than males; for persons aged less than 65 years, rates of health service utilization are higher for females than for males in Alberta (Alberta Health, unpublished). Among the youngest age-group (i.e., less than 10 years of age), medically attended shingles rates have declined in the post-vaccine era for both females and males. This is not surprising as this is the age-group that would have received chickenpox vaccine, and the rate of shingles among those immunized is lower than among persons who have had wild disease [18]. The data used for the analysis were assembled and analyzed at the individual level prior to aggregations being created. Although we used individual level data to estimate shingles rates, we did not have individual level data to assess chickenpox vaccination. Therefore,

it is possible that some factor other than the introduction of the publicly funded chickenpox vaccination

program might be responsible for part of the observed changes in shingles rates over the periods of examined. Adriamycin in vitro Thus our findings may be prone to the ‘ecologic fallacy’ where the results from aggregate data may not fully apply at the individual level [19]. We unless did not attempt to generalize overall Modulators trends within any age/sex group to the individual level. Other possible explanations for the increasing rates of shingles among older persons over time include possible secular trends (increases) in the occurrence of immunosuppressive diseases or therapies [17] and [20]. Having a co-morbid health condition was strongly associated with medically attended shingles rates for both sexes among persons aged less than 65 years. Although the proportion of medically attended shingles cases with a co-morbid condition in the 12 months prior to medically attended shingles episode is less than 2%, this proportion may be increasing among females compared to males in the public availability period for shingles vaccine. Although we found that only 4% of medically attended shingles cases were hospitalized, this is an over-estimation of the proportion of cases where the hospitalization is attributable to shingles. It has been observed elsewhere that two-thirds of hospitalizations that included zoster codes in any position of a permitted15 diagnostic codes for hospitalization were incidental to the hospitalization[21].