Recurrent Herpes Zoster

TO THE EDITOR—Estimated rates of recurrent herpes zoster, as addressed in the diagnosis code-based study of Tseng et al [1] and the accompanying editorial by Harpaz [2], presume that the studied episodes actually represent recurrent herpes zoster, as opposed to zosteriform herpes simplex. Because herpes simplex is much more common than herpes zoster and has a much higher rate of recurrence, even a low rate of misidentification of recurrent herpes simplex as recurrent zoster could result in most (clinically diagnosed) supposed recurrent herpes zoster episodes actually representing herpes simplex. Indeed, to date all patients referred to my infectious disease clinic for putative recurrent herpes zoster—always of nongenital, nonoral-labial sites—were shown by polymerase chain reaction (PCR) or culture to have herpes simplex.

Clinicians typically treat presumed herpes zoster with antiviral agents, just as they treat presumed urinary tract infection (UTI) with antibiotics, regardless of the validity of these diagnoses. Thus, a diagnostic code for herpes zoster plus a prescription for an antiviral agent no more confirms that the patient actually had herpes zoster than a diagnostic code for UTI plus a prescription for an antibiotic confirms that the patient actually had symptomatic UTI. Although the dedicated study personnel in shingles vaccine trials can identify recurrent herpes zoster with high accuracy, as validated by laboratory testing, even there some misidentification occurs, as documented in the references cited by Tseng et al and Harpaz [1^, 2]. Misidentification surely is much more common in everyday practice. Thus, the true rate of recurrent herpes zoster remains unknown, estimated rates are likely exaggerated, and clinicians should not be encouraged to believe that all—or even most—episodes of recurrent painful vesicles actually represent herpes zoster, even if they involve nongenital, nonoral-labial sites.