The epidemiology of varicella zoster virus disease in Sweden : before and after vaccination

Abstract

Primary infection with the varicella zoster virus (VZV) presents as chickenpox, a highly contagious infection. Thereafter the virus establishes latency in nerve ganglia of the host. The virus may reactivate later in life and cause shingles, neurological and/or visceral complications. The overall aim of this thesis was to provide a baseline for the burden of chickenpox disease in Sweden and to assess the impact of vaccination on the epidemiology of VZV disease in order to contribute knowledge to an appraisal of general vaccination against the diseases in Sweden carried out by the Public Health Agency.

In Study I, we obtained data from healthcare registers and databases and found a chickenpox-related hospitalisation rate of 3.56/100,000 person-years, a consultation rate of 20.1/100,000 person-years in specialist care and 109/100,000 person-years in primary care in Sweden in 2007-2013. In Study II, we included patients hospitalised with chickenpox in Stockholm and Gothenburg in 2012-2014. Their median age was 3.6 years. 43.1% of children and 67.4% of adults had an underlying condition. Overall 87.2% and 63.0% developed complications, respectively. There was no increased risk of complications among those with underlying conditions. In addition, in a nation-wide serology study using residual samples from 2011-2013, we found a VZV seroprevalence of 66.7% in 5-year-olds and 91.5% in 12-year-olds. In Study III, we compared demographic and socio-economic factors for children hospitalised with chickenpox, influenza and respiratory syncytial virus with patients with rotavirus, in a paediatric hospital in Stockholm in 2009-2014. We found that admitted chickenpox cases were older and lived in a household with more children than the cases with rotavirus. In Study IV, we explored the impact of chickenpox vaccination on shingles incidence in a mathematical model under a range of assumptions regarding VZV immunity after an encounter with a contagious case, so-called exogenous boosting (EB). We found that EB could be strong, intermediate or weak and still not cause a surge in shingles incidence after chickenpox vaccination. In addition, the same transmission model was used to investigate the impact of various strategies for vaccination against chickenpox and/or shingles in Sweden.

In conclusion, a majority of Swedes had chickenpox in early childhood. The need for hospitalisation was low. More than half of complications were seen in previously healthy patients. Chickenpox vaccination led to a dramatic decrease in chickenpox incidence in our model, whereas the impact on shingles incidence was dependent on the assumed strength and duration of exogenous boosting. Of the assessed vaccination strategies, two-dose chickenpox vaccination in early childhood combined with shingles vaccination (RZV) at 65 years prevented the most VZV cases in the model. However, a cost-effectiveness analysis is needed to evaluate which vaccination strategy has the most reasonable costs for healthcare and society in relation to its health effects.