Primary Care

Identification of potential candidates for varicella vaccination by history: questionnaire and seroprevalence study

BMJ 2004; 329 doi: (Published 02 September 2004) Cite this as: BMJ 2004;329:551
  1. Eithne MacMahon, consultant (eithne.macmahon{at},
  2. Lisa J Brown, research scientist1,
  3. Sarah Bexley, specialist nurse adviser2,
  4. David C Snashall, senior lecturer3,
  5. Dipti Patel, lecturer3
  1. 1 Department of Infection, Guy's and St Thomas' Hospital NHS Trust, London SE1 7EH
  2. 2 Occupational Health Department, Guy's and St Thomas' Hospital NHS Trust
  3. 3 Occupational Health Department, Guy's, King's, and St Thomas' School of Medicine, London SE1 9RT
  1. Correspondence to: E MacMahon


    Guidance from the Joint Committee on Vaccination and Immunisation regarding varicella vaccination of workers in healthcare settings is imminent.1 The new recommendations will advise that all those with a negative or uncertain history of chickenpox or shingles at pre-employment assessment should be tested for varicella zoster virus IgG. Vaccination with the newly licensed varicella vaccine will be advised for those with a seronegative result.

    A history of chickenpox has a high positive predictive value for immunity among healthcare workers in Europe, where the seroprevalence is as high as 98.5%.2 The validity of a history of chickenpox is unknown in those from tropical countries, however, where the mean age of infection is in early adulthood.3 Guy's and St Thomas' Hospital NHS Trust in inner London has some 8000 staff and 4700 healthcare students of increasingly heterogeneous origins.4 5 We conducted a questionnaire and seroprevalence study at the hospital to ascertain the relations between history of chickenpox, countries of birth and residence, and serological status.

    Participants, methods, and results

    From September 2001 to July 2002 a nurse administered questionnaire was completed for 747 staff and students consecutively attending pre-employment screening. Country of birth or of first residence was noted, together with age range during subsequent residence in other countries. Recruits were asked about prior chickenpox or shingles, with details recorded. Serology was requested in line with usual practice. Immune status was determined by using the Diamedix enzyme immunoassay kit for varicella zoster virus IgG (Miami, FL, USA).

    Of 629 (84%) recruits tested, six yielded equivocal results and were excluded from further analysis. Those who denied or were unsure of past infection were considered as one group. We assigned participants to “temperate” or “tropical” subgroups according to country of birth or first residence. We compared differences in seroprevalence and proportions of groups with a positive history by using the χ2 test; we also analysed other variables (unadjusted and adjusted) using logistic regression.

    The table shows the results. Additional information about prior chickenpox (nature and site of rash, residual scar) did not increase the positive predictive value. Shingles had a positive predictive value of 100%, however, with all 22 participants testing IgG positive. Twenty nine participants in the temperate group had not lived solely in temperate zones during their first 12 years and had a reduced risk of testing seropositive (odds ratio 0.35 compared with those born and brought up in a temperate climate)—similar to that of those born and brought up in a tropical climate (odds ratio 0.36).

    Results and statistical analysis of differences between temperate and tropical subgroups

    View this table:

    Participants from Sub-Saharan Africa and the Caribbean were disproportionately represented among the 52 seronegative participants (29% (15) and 17% (9) respectively, compared with just 18% (113) and 5% (31) respectively among all new recruits). Among participants with a history of prior infection, the false positive rate was 9.3% in the tropical group versus 3.7% in the temperate group (P = 0.019) (table).


    Candidates for varicella vaccination (seronegative staff) will inevitably be missed if a blanket policy of screening by history alone is implemented. As shown here, their number will increase considerably if groups with a significantly higher false positive rate account for a substantial proportion of the workforce.

    The overall seroprevalence and positive predictive value are lower than figures usually quoted for healthcare workers in Europe, but the distribution of geographical origins largely explains the discrepancy. It is a weakness of this study that some of the countries defined as temperate in the questionnaire—for example, India—also contain regions with tropical climates. Serological testing of recruits from these areas should also be considered. Individuals born or raised in tropical climates should have serological screening regardless of a history of chickenpox.

    What is already known on this topic

    A history of chickenpox is a reliable indicator of past infection in temperate climates

    What this study adds

    Seronegative candidates for varicella vaccination will be missed if a blanket policy of screening by history alone is applied to people born or raised in tropical climates


    • This article was posted on on 23 July 2004:

    • We thank the Guy's and St Thomas' Hospital NHS Trust's occupational health nurses for conducting the questionnaire, its department of infection staff for laboratory work and secretarial help, and Sue Chinn for help with statistical analysis.

    • Contributors EMacM and DP designed the study, wrote the paper and did the statistical analysis. LJB did data entry, manipulation, and analysis. SB coordinated data collection and entered the data. DCS contributed throughout the study. EMacM is the guarantor.

    • Funding None.

    • Competing interests In the past five years, EMacM has received sponsorship from Launch Diagnostics Biokit, GlaxoWellcome, SmithKlineBeecham Pharmaceuticals, and Aventis Pasteur MSD towards the cost of attending conferences. DP has been funded by GlaxoSmithKline to give a talk on a subject unrelated to this study.

    • Ethical approval St Thomas' Research Ethics Committee.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    View Abstract