Browsing by Subject "Spatial epidemiology"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Febrile Disease Epidemiology and Geospatial Modeling in Southern Sri Lanka
    (2011) Gong, Wenfeng

    As a subproject under the collaboration between Ruhuna University in Sri Lanka and Duke University, this study focused on the identification of socioeconomic and ecological determinants of febrile illness in Galle district. We integrated socioeconomic data from local government, ecological data from national geographic information system (GIS) database, and febrile patients' epidemiologic data from clinical investigation. The integrated database was prepared using GIS techniques and validated via field visits. Missing population data were simulated through Bayesian imputation. While the febrile disease risk is not measurable in the current study, social and ecological predictors of disease distribution (proportion of specific disease in all cases) were identified for the enrolled Teaching Hospital Karapitiya (THK) patients. These predictors are potentially the determinants of febrile disease in Galle. Due to the limitation of single-center clinical sampling, patient travel distance was highly associated with patient visits, thus, it became a strong confounder in analyses. After adjusted for the confounders, a set of patients' social/ecological exposures were found to be associated with dengue, leptospirosis, URTI, LRTI, gastroenteric infection, and/or undifferentiated febrile illness.

  • Thumbnail Image
    ItemOpen Access
    Neighborhood Disadvantage is Associated with High Cytomegalovirus Seroprevalence in Pregnancy.
    (J Racial Ethn Health Disparities, 2017-08-24) Lantos, Paul M; Hoffman, Kate; Permar, Sallie R; Jackson, Pearce; Hughes, Brenna L; Kind, Amy; Swamy, Geeta
    BACKGROUND: Cytomegalovirus (CMV) is the most common infectious cause of fetal malformations and childhood hearing loss. CMV is more common among socially disadvantaged groups, and geographically clusters in poor communities. The Area Deprivation Index (ADI) is a neighborhood-level index derived from census data that reflects material disadvantage. METHODS: We performed a geospatial analysis to determine if ADI predicts the local odds of CMV seropositivity. We analyzed a dataset of 3527 women who had been tested for CMV antibodies during pregnancy. We used generalized additive models to analyze the spatial distribution of CMV seropositivity. Adjusted models included individual-level age and race and neighborhood-level ADI. RESULTS: Our dataset included 1955 CMV seropositive women, 1549 who were seronegative, and 23 with recent CMV infection based on low avidity CMV antibodies. High ADI percentiles, representing greater neighborhood poverty, were significantly associated with the nonwhite race (48 vs. 22, p < 0.001) and CMV seropositivity (39 vs. 28, p < 0.001). Our unadjusted spatial models identified clustering of high CMV odds in poor, urban neighborhoods and clustering of low CMV odds in more affluent suburbs (local odds ratio 0.41 to 1.90). Adjustment for both individual race and neighborhood ADI largely eliminated this spatial variability. ADI remained a significant predictor of local CMV seroprevalence even after adjusting for individual race. CONCLUSIONS: Neighborhood-level poverty as measured by the ADI is a race-independent predictor of local CMV seroprevalence among pregnant women.