Intended for healthcare professionals

Education And Debate

Controversies in Management: Acyclovir for childhood chickenpox

BMJ 1995; 310 doi: (Published 14 January 1995) Cite this as: BMJ 1995;310:108
  1. M W McKendrick, consultant physiciana
  1. a Department of Infectious Diseases and Medicine, Royal Hallamshire Hospital, Sheffield S10 2JF

    Cost is unjustified

    Chickenpox is usually a disease of childhood, but it can occur at any age. In Sheffield (population 500000) about one immunocompetent patient a year requires admission to the intensive care unit for chickenpox pneumonia, and there has been one death in the past 12 years (unpublished observations). Pneumonia is a rare complication in children, in whom the disease is usually mild and uncomplicated. But in immunocompromised patients, including infants acquiring infectionperinatally and patients taking steroids, the illness can be severe, and antiviral drugs may be life saving.

    Mortality figures from the United States suggest an overall death rate from chickenpox of about one in 40000, though for patients aged over 19 years the rate is one in 1460.1 Severe infection with chickenpox in childhood is rare. One of the commonest causes of death was Reye's syndrome, but this should now be less common as its association with aspirin has been recognised and aspirin is not given to children. Death due to chickenpox or its complications is very rare in immunocompetent children.

    In 1935 Bullowa and Wishik observed that chickenpox is the disease of childhood which is taken least seriously by the general public and by the medical profession.2 Children with secondary infection from someone in their household develop more skin lesions and have a higher temperature, probably because of the greater inoculum of the infecting virus compared with that in primary infection acquired “in the playground.”3 The most common complication of childhood chickenpox requiring medical treatment is secondary bacterial sepsis, but there is no confirmed evidence of a relation between this complication and the number of lesions.

    Clinical benefit

    What then are the arguments for using antiviral drugs routinely in chickenpox? Acyclovir, the only drug currently licenced for treatment of primary varicella zoster virus infection, is successful and safe. Though poorly absorbed, with only 15 to 30% oral bioavailability, therapeutically effective concentrations can be reached by oral administration. This has been shown by the significant effects on viral shedding in acute herpes zoster.4

    Two studies of childhood chickenpox in which acyclovir has been started within 24 hours of the onset of the rash have been reported (815 and 105 children respectively).5 6 These showed a significant reduction in the number of lesions, reduced duration of new lesion formation, accelerated progression towards healing with less itching, but nodifference in systemic complications such as bacterial sepsis. Similar findings have been found in adolescents with chickenpox.7 Acyclovir was well tolerated.

    Are these statistical differences clinically relevant? In Dunkle's study of 815 children a scoring system of 0 to 3 was used for anorexia, lethargy, and fever to indicate none, mild, moderate, or severe.5 The peak mean score for the “constitutional” illness (maximum 9) was reached on the second day after enrolment and was 2.18 for the placebo group and 1.29 for the acyclovir group. In Balfour's study the times to return to normal temperature and to reach maximum numbers of lesions were each reduced by one day.6 Compliance could be a problem in children. In Balfour's study all patients took the prescribed doses but in Dunkle's paper non-compliance was given as one of the reasons for exclusion from the analysis. The use of acyclovir is unlikely to prevent spread in schools as the greatest contagiousness seems to be during the prodrome.8

    Economic arguments

    There are two economic arguments for antiviral treatment. Firstly, if an infected child recovers more quickly, the parent may be able to return to work more quickly. But it is usually the appearance of the lesions and not the infectivity that dictates when a child returns to school, and a “spotty child” is unlikely to be welcomed back into a school environment earlier even if certified free from infection. Secondly, if complications could be reduced this would reduce costs, but the above trials did not show any benefits of using acyclovir on complications.

    The published studies have included only children treated within 24 hours of onset. This would be difficult to achieve for primary infection, though secondary and tertiary cases within a family could realistically be treated within this time frame. The cost must also be considered. A five day course of acyclovir at a dose of 400 mg (recommended dose aged 2–5 years) four times daily in the United Kingdomcurrently costs about £32. Over the age of 6 years the recommended dose is 800 mg four times daily for five days, and over the age of 12 it is 800 mg five times daily for seven days. There is also the cost of medical consultation. The direct costs of treating all children with varicella would be enormous, and the cost of drugs alone has been estimated to be about $250m (£167m) in the United States.9 This compares with a total care cost (including hospitalisation, doctor's visits, and drugs) of about $15m in 1988. The potential economic benefits of earlier return to work of parents seem unlikely to offset the cost of the drug and medical time.

    Disadvantages of treatment

    Widespread use of an antiviral drug has theoretical disadvantages. The immune response may be impaired in some, and this could alter the natural pattern of the infection and result in earlier reactivation in the form of zoster. Studies of antibody titres after acyclovir treatment in childhood chickenpox indicate that the humoral immunological response is unmodified at 28 days and one year,10 but it is unclear whether the cell mediated immune response is altered.

    The question of antimicrobial resistance is repeatedly raised in relation to antibiotics and bacteria. There is no evidence at present to suggest that resistant strains of varicella zoster virus will occur in the immunocompetent community with antiviral use, but this will need to be monitored. Resistance has occurred rarely in immunocompromised patients with varicella zoster virus infection; prolonged viral replication in these patients provides an optimum environment for selection of resistant strains.

    The mildness and self limiting nature of chickenpox is confirmed by the placebo groups in the American studies. Though an argument can be made for treating secondary and tertiary cases and adults with acyclovir as well as any immunocompromised patient, the modest clinical benefit from what would be a very costly programme do not support the routinetreatment of chickenpox in normal healthy children.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    View Abstract