Purchase this article with an account.
David Poulsen, Grace Honik, David Gritz; Case Control Study of Herpes Zoster Ophthalmicus in the Bronx Using Population-based and Clinic-based Controls. Invest. Ophthalmol. Vis. Sci. 2013;54(15):2905.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Herpes zoster ophthalmicus (HZO) has an estimated incidence of 1.5 to 6.4 cases per 10,000 persons per year in the general population. As far as the authors are aware, no previous study has reported the epidemiology of HZO in the Bronx, or estimated the risk of developing HZO among persons with HIV. HIV is 3.6 times more prevalent in the Bronx than in the U.S. overall, allowing examination of the association between HIV and HZO. Montefiore Medical Center is the largest healthcare provider in the Bronx, caring for about one-third of the Bronx population each year.
This is a retrospective case-control study of the risk of developing herpes zoster ophthalmicus among patients that are HIV-positive and living in the Bronx. Cases were Bronx residents diagnosed with new-onset HZO during the ten year study period (May 1, 2002 to April 30, 2012) at Montefiore Medical Center. The study utilizes two different case control approaches: 1. A population-based control group taken from the 2010 U.S. Census and 2010 New York City HIV/AIDS Annual Surveillance Statistics, matched to the Zip codes of cases; 2. Clinic-based controls chosen randomly on the same appointment date as cases in a 20:1 ratio. The data were used to calculate odds ratios of HIV infection and HZO. The data was analyzed using STATA 12 (College Station, TX) and statistically significant relationships were assessed at p=0.05.
During the study period, 106 incident cases of herpes zoster ophthalmicus met inclusion criteria. Compared to the population-based controls (n=1,405,127), infection with HIV showed an increased association of having HZO relative to those not known to be HIV-positive by an odds ratio of 11.027 (95% CI: 6.324-19.229; p<0.001), using multivariate logistic regression to control for age group, zip code, and gender. Compared to clinic-based controls (n=2,120), HIV infection was shown to increase the odds of HZO by 10.281 (95% CI 5.422-19.493) using multivariate regression to control for age, gender, and race.
Infection with HIV is a significant risk factor for having an incident case of herpes zoster ophthalmicus. This is particularly significant in the Bronx where the prevalence of HIV infection is relatively high. Patients, particularly young patients, who develop HZO should be tested for HIV, especially if other risk factors for HIV are present.
This PDF is available to Subscribers Only