BackgroundVaricella and herpes zoster are both caused by varicella zoster virus (VZV) infection or reactivation and may lead to complications associated with a (severe) societal burden. Because the epidemiology of VZV-related diseases in the Netherlands remains largely unknown or incomplete, the main objective of this study was to study the primary care incidence, associated complications and health care resource use.MethodsWe investigated the incidence of VZV complications in the Dutch general practitioner (GP) practices and pharmacies in a retrospective population-based cohort study (2004–2008) based on longitudinal GP data including free text fields, hospital referral and discharge letters from approximately 165,000 patients.ResultsThe average annual incidence of varicella GP-consultations was 51.5 per 10,000 (95% CI 44.4-58.7) overall; 465.5 per 10,000 for 0–1 year-olds; 610.8 per 10,000 for 1–4 year-olds; 153.5 per 10,000 for 5–9 year-olds; 8,3 per 10,000 for >10 year olds. When only ICPC coded diagnoses were analyzed the incidence was 27% lower. The proportion of complications among varicella patients was 34.9%. Most frequently complications were upper respiratory tract infections. Almost half of the varicella patients received medication. The referral rate based on GP consultations was 1.7%. The average annual incidence of herpes zoster GP-consultations was 47.5 per 10,000 (95% CI 40.6-54.4). The incidence increased with age; 32.8 per 10,000 for <60 year-olds; 93.1 per 10,000 for 60–64 year-olds and 113.2 per 10,000 for >65 year olds. When estimating herpes zoster incidence only on ICPC coded information, the incidence was 28% lower. The complication rate of herpes zoster was 32.9%. Post herpetic neuralgia was seen most often. Of patients diagnosed with herpes zoster 67.8% received medication. The referral rate based on GP consultations was 3.5%.ConclusionsFor varicella the highest incidence of GP-consultations was found in 1–4 year-olds, for herpes zoster in the >65 years olds. The occurrence of complications was not age-dependent but varies per complication. When estimating incidence of VZV-related diseases in primary care, based on diagnostic codes only, one should be aware of a gross underestimation of the incidence. Our analysis may have important implications for the outcomes of upcoming cost-effectiveness analyses on VZV vaccination.