Outbreak of varicella-zoster virus infection among Thai healthcare workers.

OBJECTIVE: To evaluate the correlation between self-report of a prior history of chickenpox and results of varicella-zoster virus (VZV) immunoglobulin (Ig) G serologic test results in an outbreak of VZV infection among Thai healthcare workers (HCWs) and to conduct a cost-benefit analysis of establishing routine VZV immunization as part of an occupational health program on the basis of the outbreak data.

METHODS: All exposed patients received prophylaxis and the HCWs in our 3 intensive care units (ICUs) were prospectively evaluated. HCWs were assessed for disease history and serologic evidence of VZV IgG. A cost-benefit analysis was performed.

RESULTS: After 140 HCWs and 18 ICU patients were exposed to VZV, 10 HCWs (7%) with active VZV infection were relieved from work until skin lesions were crusted. Acyclovir (ACV) was prescribed to all 10 HCWs with active disease, and all 18 exposed patients received prophylaxis with ACV. Of 140 HCWs, 100 consented to longitudinal follow-up. Twenty-three (100%) of the HCWs who reported a history of chickenpox also had serologic test results that were positive for VZV IgG, compared with 30 (39%) of 77 HCWs who reported no prior history of chickenpox, yet had test results that were positive for VZV IgG. Reported history of chickenpox had a sensitivity of 43%, a specificity of 100%, a positive predictive value of 100%, and a negative predictive value of 61% with respect to VZV infection immunity. The total cost estimate for this outbreak investigation was $23,087.

CONCLUSIONS: An HCW's reported history of chickenpox was a reliable predictor of immunity; a report of no prior history of chickenpox was unreliable. Our cost-benefit analysis suggests that the costs of an occupational health program that included VZV surveillance and immunization for the next 323 HCWs would be approximately equal to the excess costs of $17,227 for the ACV therapy, HCW furloughs, and staff overtime associated with this outbreak.