How reliable is the history of chickenpox? Varicella serology among children up to 14 years of age.

OBJECTIVES: The aim of this study was to assess the seroprevalence of antibodies to varicella-zoster virus (VZV) in children of northern Greece and to estimate the reliability of varicella history.
METHODS: A serosurvey of 632 children, aged 13 months to 14 years (median 5.2 years), was conducted between April 1999 and July 2001. Serum samples were tested by enzyme-linked immunosorbent assay (ELISA) for IgG antibodies to VZV (IgG Genzyme Virotech GmbH). A history of varicella in these children was obtained from the parents of all these patients. Also, a check of state health cards of the patients was done.
RESULTS: Two hundred and forty-eight (39%) of the children were seropositive for VZV. Two hundred and thirty (36%) of the 632 children claimed to have had previous varicella infection; 87.8% were seropositive, and 12.2% lacked antibodies to VZV. One hundred and seven of the 230 children with a history of varicella had the information about the disease confirmed, as it was reported on their state health card by a pediatrician; 10.2% were seronegative for VZV. Absence of history of varicella was reported in 402 (63.6%) of the 632 children; 88.6% of those were seronegative, and 11.4% were seropositive. The percentage of incorrect negative history ranged from 6% (13-60 months of age) to 48% (120-168 months of age).
CONCLUSIONS: A large proportion of the study group (61%), mainly below 7 years of age, is susceptible to varicella. The positive predictive value of a history of varicella is 87.8%, whereas the negative predictive value of a negative history is 11.4%, which means that there is an 88.6% probability of a negative history being correct. Varicella serology may be reasonable prior to vaccination in children >10 years old with a negative chickenpox history. However, if one excludes cost considerations, it is also reasonable to vaccinate all children, irrespective of serostatus.