Deaths from chickenpox in adults are decreasing

EDITOR—On the basis of death certificates from the Office for National Statistics from 1995 to 1997, Rawson et al conclude that deaths as a result of chickenpox are increasing in adults in England and Wales.1 More up to date figures from the Office for National Statistics, however, show that chickenpox mortality is decreasing in adults (from 32 deaths in 1996 to 18 in 2000—see figure (a)). Furthermore, the number of deaths from chickenpox and case fatality rates were significantly higher in 1995-7 (period of the analysis) than at any other period. The claim by Rawson et al that deaths in adults are rising is therefore misleading.

Deaths from chickenpox (a) and annual consultation rate for chickenpox (b) in England and Wales, 1981-2000

• Download figure
• Open in new tab
• Download powerpoint

The change in age related varicella mortality is the result of a shift in the age distribution of infection. Over the past two decades there has been an increase in cases in the youngest age group (possibly due to greater attendance of day-care).2^–4 Over the same time period there has been a gradual increase in reported incidence in adults, which peaked in the late 1980s and has been falling since (figure (b)). This is broadly reflected in the gradual decrease in deaths in adults during the past decade. The exception to this trend are 1996 and 1997—exactly the time period when Rawson et al performed their study. What has caused these large shifts in the incidence of varicella in adults is still largely unexplained.

Epidemiology of chickenpox in United Kingdom needs further investigation

EDITOR—Rawson et al highlight the potential severity of chickenpox.1 They say that the age distribution of chickenpox is changeable. But recent data from Scotland, England and Wales, and the United States show that the previous shift towards increased infection in older age groups has not been sustained.2^–4 In recent years the trend has been towards decreased age at infection, with most cases now occurring among the group aged 1-4 years, rather than among children of school age.

Varicella vaccine is recommended for routine administration in the United States and Canada, among other countries, but its suitability for inclusion in the United Kingdom's childhood immunisation programme is still being considered. Further work on the epidemiology of chickenpox in the United Kingdom is therefore now particularly important.

We have proposed a one year period of enhanced active surveillance for severe complications of varicella in children admitted to hospital throughout the United Kingdom and the Republic of Ireland, using the British Paediatric Surveillance Unit's orange card scheme.5 The information gained, together with that of Rawson et al, and others, would help to determine the advisability of a universal programme for the United Kingdom, and provide a baseline against which to evaluate its impact should it be adopted.

Healthcare workers should not be forgotten

EDITOR—We agree with the conclusion of Rawson et al that, although deaths in adults from chickenpox have increased in number and proportion, this does not justify mass immunisation with varicella vaccine.1 One population, however, that would clearly benefit from vaccination is susceptible healthcare workers.

At St George's Hospital in London we identified a total of 25 cases of chickenpox in staff and students from data prospectively collected over the past three years. We were able to determine the country of birth in 22 of these and found that most cases (13/22 (59%)) occurred in people born outside the United Kingdom. This figure was higher than expected since only 39% of the St George's workforce who have contact with patients are black or from an ethnic minority. Since Rawson et al found that there was a disproportionately higher mortality among such people compared with those born here, it would be interesting to know if occupations, such as those in health care with a higher likelihood of exposure, were over-represented among the cases of fatal varicella.

Live attenuated varicella vaccine has been in use now for over two decades.2 Moreover, it has had a licence for use in susceptible individuals in the United States since 1995 and has an excellent safety and efficacy record.3 We believe that the increased mortality from chickenpox in adults of working age of between 1:1000 and 1: 5000 shown by Rawson et al may make it indefensible for NHS trusts not to offer varicella vaccine to their susceptible staff for two reasons: personal safety at work and nosocomial chickenpox. If only medical and nursing staff and students had been vaccinated in the last three years at St George's, 85% of chickenpox cases in hospital staff would have been prevented.

Chickenpox associated morbidity may be long term

EDITOR— Rawson et al analysed deaths from chickenpox during 1995-7.1 It has previously been recorded that chickenpox in healthy adults has a 25-fold greater risk of complications than in children.2 Rawson et al show a significant mortality of chickenpox in England and Wales but do not address the question about associated morbidity, an important issue when addressing the value of immunisation on a population. We recently performed a prospective study on respiratory function in adult patients with chickenpox admitted to a subregional infectious diseases unit in a United Kingdom hospital over a period of 29 months.3

Sixty six adult patients with chickenpox were admitted to hospital during the period, four of whom were immunocompromised. Thirty eight patients fulfilled the study protocol and of these, 50% had radiological evidence of pneumonia (all immunocompetent).3 Three female patients required admission to intensive care unit, two of whom were pregnant. One patient presented with chickenpox encephalitis, and five had superimposed bacterial skin infections.

Severe respiratory disease was associated with the presence of new respiratory symptoms, close contact with the index case, and a history of current smoking. On follow up at a year post-infection, 37% of patients with radiological pneumonia and 10.6% of those without pneumonia continued to have reduced single breath carbon monoxide transfer factor.

This effect was independent from the effect of smoking and may indicate permanent lung damage. It may therefore be that the morbidity relates not only to the acute infection and admission but also to longer term effects on the lung function, but the exact clinical relevance of our findings is uncertain at present. The study does, however, indicate that chickenpox causes significant morbidity in adults, which may be seen increasingly in the future. Accurate data on morbidity as well as mortality are required to inform the debate on the value of mass vaccination for chickenpox in the United Kingdom.

Extracorporeal membrane oxygenation has important role

EDITOR—Rawson et al highlighted the potentially devastating effects of varicella infection, particularly the fact that adults in the United Kingdom are dying from it and these deaths are increasing in number.1 We know the pneumonitis caused by varicella infection can lead to respiratory failure that is often the cause of death in these patients. Antiviral treatment may help in such patients, but only if their severely compromised physiology can be adequately supported until they recover.

Extracorporeal membrane oxygenation has been reported to be used successfully in cases of adult respiratory failure resulting from varicella pneumonia and we would like to bring the results of such intervention to the attention of Rawson et al.2^–5 We have treated 15 adults with this procedure for confirmed varicella pneumonitis in Leicester between August 1992 and December 1999. These 15 patients had a mean age of 36 years (range 24-61), and were significantly hypoxic on referral with a ratio of arterial oxygen tension to fraction of inspired oxygen (PaO₂/FiO₂) of 8.09 kPa. The overall survival rate in these patients was 60%. Of the 11 patients, however, we treated with venovenous extracorporeal membrane oxygenation the survival rate was 75% (compared with zero for the four patients treated with venoarterial extracorporeal membrane oxygenation).

It seems likely, therefore, that this is a treatment that should be considered for fulminant varicella pneumonitis, but the numbers treated so far are too small to be sure of the effectiveness of this invasive treatment. To resolve this uncertainty, currently all such cases in the United Kingdom can be referred for entry into the CESAR (conventional ventilation or extracorporeal membrane oxygenation for severe adult respiratory failure) trial. Suitable patients will be randomised to receive either extracorporeal membrane oxygenation or continued conventional ventilation. Further details about the trial are available from http://www.cesar-trial.org.