Source: Masters Abstracts International, Volume: 57-05.

Adviser: Anne Marie Z. Jukic.

Introduction: Herpes simplex virus type 1 (HSV-1) and type 2 (HSV-2) cause serious disease in neonates (neonatal herpes (nHSV)), with substantial mortality and morbidity. Population-based surveillance and administrative hospital discharge data are two proposed methods for evaluating the burden of disease in a population. We compare the incidence of nHSV disease using these two data sources and examine the rate of post-diagnosis hospitalization for nHSV survivors in New York City (NYC) from April 2006 through June 2015.

Objectives: First, to determine the incidence rate of nHSV in NYC using population-based surveillance and to compare this rate with the rate obtained from administrative hospital inpatient discharge data. Second, to measure the sensitivity of select ICD-9 codes in identifying laboratory-confirmed nHSV cases in administrative hospital discharge data. Third, to compare the post-diagnosis hospitalization rate between a cohort of lab-confirmed nHSV cases after their initial herpes diagnosis with a comparison cohort of hospital-matched controls.

Methods: From April 2006 through June 2015, routine population-based surveillance was conducted using laboratory reports of HSV infection in infants aged 60 days or younger, hospitalized in NYC, and followed by case investigation. We identified cases of nHSV in administrative hospital discharge data using a case definition of infants aged 60 days or younger, having an ICD-9 code for herpes simplex infection, and hospitalized in NYC during the same time frame. We linked lab-confirmed surveillance cases to their administrative hospital discharge data to examine the sensitivity of using select ICD-9 codes to detect lab-confirmed nHSV cases in administrative data. We examined the hospitalization rate and subsequent cost of hospitalization between lab-confirmed survivors of the initial nHSV infection and a comparison cohort of hospital-based infants matched with cases on birth hospital, sex, date of birth, race/ethnicity, and birth weight.

Results: A total of 115 laboratory-confirmed cases of nHSV were detected though population-based disease surveillance (overall incidence, 10.99 per 100,000 live births). The case-fatality rate was 17.86% among lab-confirmed surveillance cases. Using administrative hospital discharge data, 139 nHSV cases were identified (13.29 per 100,000 live births) during the same time period. Our administrative case definition identified 68 of the lab-confirmed cases detected through surveillance (sensitivity, 60.71%). Compared to a comparison cohort of hospital-matched controls, cases in the nHSV cohort were 8.65 (95% CI: 4.41--16.98) times more likely to be hospitalized within two months. Kaplan-Meier survival curves indicated significant differences in the time without hospitalization between cohorts across two, twelve, and twenty-four months of follow-up (p < 0.001). In addition, the mean cost of the first hospitalization during each follow-up period was higher for the nHSV cohort than the comparison cohort (p < 0.001).

Conclusions: Administrative hospital discharge data may be adequate to approximate the overall incidence of nHSV as compared to a population-based surveillance system, but performs poorly in identifying cases of lab-confirmed nHSV. In addition, infants that survive the initial nHSV infection are at higher risk for additional hospitalization and exhibit a faster time-to-hospitalization event, with higher mean costs, compared to a comparison cohort of hospital-matched controls.