Intended for healthcare professionals


Varicella vaccine in pregnancy

BMJ 1996; 313 doi: (Published 21 September 1996) Cite this as: BMJ 1996;313:701
  1. Daniel S Seidman, Research fellow,
  2. David K Stevenson, Professor of neonatology,
  3. Ann M Arvin, Professor of infectious diseases
  1. Department of Obstetrics and Gynecology, Sheba Medical Center, Tel-Hashomer, Israel
  2. Department of Pediatrics, Stanford University Medical School, Stanford, CA, 94305-5119, USA

    Routine screening and vaccination should be considered

    A live attenuated varicella vaccine was licensed in the United States by the Food and Drug Administration in March 1995. The manufacturer is currently preparing to license the vaccine in Europe. It is now possible for the first time to prevent varicella infection in pregnancy by vaccination before pregnancy. This raises the question of whether the vaccine should be offered to all women of childbearing age who are not known to be immune to varicella.

    Varicella is a highly contagious infection, caused by a DNA herpes virus. Infection occurs in 0.05-0.07% of all pregnancies.1 Non-immune pregnant women are at risk, as are all susceptible adults, for associated complications and mortality when they contract varicella.2 3 Furthermore, intrauterine infection may result in stillbirth or congenital varicella syndrome, characterised by eye defects, limb hypoplasia, skin lesions, and neurological abnormalities.3 4

    Universal immunisation and serological screening of women intending to become pregnant is already implemented for the rubella virus. The major objective of this programme is to prevent maternal rubella infection and its subsequent teratogenic effects. The overall risk of fetal damage associated with rubella is about 85%,5 while the risk related to varicella is only about 2%.2 4 6 In view of the relatively low teratogenicity of varicella (table 1), the need for a similar programme of pre-pregnancy screening and maternal immunisation has been questioned.

    Table 1

    Comparison between varicella and rubella infection in women of childbearing age

    View this table:

    However, in contrast to rubella, varicella causes severe maternal morbidity, and 10-20% of infected women develop varicella pneumonia with mortality as high as 40%.2 Furthermore, if the mother develops the rash in the five days before or the two days after delivery this may result in neonatal varicella, associated with mortality as high as 30%.7 In contrast to rubella,5 reliable prenatal diagnosis of varicella is not available.8 Thus, women who develop chickenpox during pregnancy may face a difficult decision of whether to continue their pregnancy in view of the risk, though small, of serious fetal morbidity. The need to prevent varicella in pregnancy is further emphasised by the fact that no treatment is available to prevent intrauterine transmission. Neither passive immunisation within 96 hours of exposure nor treatment with acyclovir has been shown to prevent intrauterine transmission or to alleviate fetal infection.

    The varicella live attenuated vaccine has been shown to be safe and highly effective in preventing chickenpox in adults.9 The adverse reactions associated with the vaccine are generally limited to mild pain and redness at the injection site. Yet it remains to be determined who should be immunised. Theoretically, only women proved to be non-immune should receive the vaccine. However, the cost of the new varicella vaccine to doctors in the United States is $39 a dose, while testing for antibodies to varicella costs from $32 to $134. Therefore, in areas of low seroprevalence it may be cheaper to immunise all women intending to become pregnant without determining their varicella antibody status.

    The current recommendation for adults is to verify the serological status before immunisation. Assuming that all women are serologically tested, regardless of disease history, then in an industrialised country an estimated 2000 women would be screened and 100 vaccinated for every case of chickenpox prevented in pregnancy. This would cost about $70 000, but there would be savings in diagnosis and treatment, including immune testing, administration of varicella-zoster immunoglobulin (about $400 per treatment), hospitalisation, prenatal diagnosis, and, in the rare cases of congenital varicella, the costs related to fetal morbidity, including the severe sequelae such as blindness and neurological defects.10 The cost effectiveness of a screening and vaccination programme for women of reproductive age, as demonstrated above, is likely to vary considerably between populations. For developing countries vaccination does not seem cost effective at present, despite the higher incidence of varicella among young women.

    An additional consideration is whether a history of chickenpox or zoster should suffice as evidence of prior immunity. In contrast to rubella, virtually all varicella infections produce typical cutaneous signs, and the diagnosis can usually be established by clinical presentation.7 A history of previous varicella infection is therefore generally accepted as proof of immunity. However, even among subjects who could recall suffering a varicella infection 0.9-3.2% were found to benon-immune.11 None the less, screening only women with a negative history will significantly decrease the number of women who receive the vaccine. The proportion of women who would be vaccinated, based on a negative history without serological confirmation, would vary widely between different populations as the number of adults who could not recall a history of chickenpox ranged from 19% in Ohio12 to 49% in Australia.10 However, even among subjects with no, or uncertain, prior exposure to varicella, only 10-36% were non-immune.11 12

    The availability of an active varicella vaccine opens new options for preventing chickenpox in pregnancy. Moreover, doctors may possibly face medicolegal consequences if they fail to offer the vaccine during preconception consultation. The attack rate of varicella is extremely high with household contact, and a substantial number of women with a negative history are susceptible. Furthermore, women of childbearing age may be more exposed to chickenpox from non-immunised children in their household. A routine screening and vaccination programme, similar to the one currently implemented for rubella, must be considered. A history of chickenpox should be obtained from all women at preconception counselling. Serological testing or vaccination should be offered. However, women should be informed that if they have a history of varicella there is at least a 97% chance that they are immune and that even among those who can not recall varicella infection about 80% are immune. The cost effectiveness of vaccinating women of childbearing age should be determined for each population according to its rates of seronegativity. Future vaccination policies would need to be revised in areas where all children are now routinely immunised as recommended.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.