Correction: Vol. 18, No. 2

To the Editor: In our earlier assessment of the etiologic role of human bocavirus 1 (HBoV1) among 109 constitutionally healthy children, seroconversion measured by a standard IgG enzyme immunoassay (EIA) was the prime marker of acute infection, and signs and symptoms during sampling intervals were interpreted from hand-written questionnaires and study nurse notes (1). In that study, we found that acute HBoV1 infection was associated with upper respiratory tract illness (URTI) and acute otitis media. However, after discovery of 3 new human bocaviruses (HBoV2–4), we used a competitive second-generation IgG EIA to differentiate between type-specific and cross-reacting IgG responses and reassessed the etiologic associations of each virus in the same pediatric cohort. We found only a weak correlation of HBoV1 infection with acute otitis media and no correlation with URTI (2).

There are 2 reasons for this discrepancy. First, the competitive second-generation EIA generated a different set of HBoV1-infected children. A total of 27 children initially interpreted as having HBoV1 seroconversions had, on reassessment, HBoV2 or HBoV3 seroconversions (2). Second, after improved reevaluation of clinical symptoms of the entire replenished cohort by 3 of the co-authors, we found that URTI symptoms of 38 children were classified differently. Despite tight sampling, intervals of 3 and 6 months might still have been too long for clinical assessment for young children with numerous respiratory infections per year. The authors regret any inconvenience to readers.