The burden and cost of hospitalised varicella and zoster in Australian children

Abstract

Background:

Economic analyses of varicella-zoster virus (VZV) immunisation are sensitive to the costs of hospitalised cases, so there is a need to validate VZV hospitalisation data.

Aims:

To assess the accuracy of hospital VZV coding data and to apply these parameters to a population-based sample to estimate incidence and costs.

Methods:

A 3-year retrospective chart review from one hospital to document clinical features and validate coding data. A separate 9-year analysis of discharge data from two hospitals draining a defined region of suburban Melbourne, with adjustment for miscoding and estimates of direct hospital costs.

Results:

After correction for miscoding, 224 patients were admitted to one hospital over 3 years, 79% with varicella and 21% with zoster. Miscoding resulted in a 15% underestimate of zoster cases and a 4% overestimate of varicella cases. Thirty-six percent of varicella admissions compared to 80% of zoster admissions were immunocompromised and/or had chronic disease. Compared to otherwise-healthy patients, immunocompromised patients were admitted earlier in their illness and had lower complication rates. Forty-two percent of immunocompromised/chronic disease patients with varicella had a known exposure, usually from a family member. The incidence of hospitalised varicella and zoster in under 15-year olds was 15.7 and 1.8 per 100,000 per year, respectively. This suggests that there are 615 varicella hospitalisations and 72 zoster hospitalisations in this age group each year in Australia, at a total direct cost of over AU $2.2 million.

Conclusion:

These results highlight the considerable burden of hospitalised zoster and the importance of immunising non-immune contacts of immunocompromised individuals. They also support previous estimates of the incidence of hospitalised varicella in Australian children and adolescents, although direct medical costs may be higher than those previously estimated.

Introduction

Vaccines to prevent varicella-zoster virus (VZV) infection have been licensed in many industrialized countries over the past 10 years. These vaccines are highly efficacious, and most vaccine failures are associated with only mild disease [1]. VZV vaccines were first licensed in Australia during 1999, and in 2003, the Australian National Health and Medical Research Council recommended that VZV vaccine be offered to all Australian children at an age of 18 months (with catch-up vaccine offered to adolescents aged 10–13 years with no history of varicella) [2]. However, the Australian government does not provide funding for this vaccine, so parents must pay the cost if they wish their child to be immunized according to the Australian Standard Vaccination Schedule.

Decisions on vaccine funding are based on many considerations including the likely cost-effectiveness of different immunisation strategies. Cost-effectiveness models for VZV immunisation have been developed in many countries [3], [4], [5], [6], [7], [8], [9], [10], and rely on accurate epidemiological data relating to VZV infections. An Australian VZV vaccine cost-effectiveness analysis estimated that 71% (AU $1.97 million) of the total AU $2.77 million annual direct costs of varicella-related infections were due to hospitalisations, even though only 0.4% of varicella cases were hospitalised [9]. Therefore, any economic analysis of VZV infections or immunisation will be sensitive to the number and cost of hospitalisations associated with the disease.

In order to gain accurate estimates of hospital costing parameters related to severe VZV infections, we set out to review 3 years of VZV-related hospitalisations in children and adolescents at our hospital, with the intention of applying these parameters to a population-based sample in metropolitan Melbourne. This would allow us to estimate the incidence and costs of hospitalised VZV infections in Australian children.