Dataset Information

Evaluation of the Gini Coefficient in Spatial Scan Statistics for Detecting Irregularly Shaped Clusters.

ABSTRACT: Spatial scan statistics with circular or elliptic scanning windows are commonly used for cluster detection in various applications, such as the identification of geographical disease clusters from epidemiological data. It has been pointed out that the method may have difficulty in correctly identifying non-compact, arbitrarily shaped clusters. In this paper, we evaluated the Gini coefficient for detecting irregularly shaped clusters through a simulation study. The Gini coefficient, the use of which in spatial scan statistics was recently proposed, is a criterion measure for optimizing the maximum reported cluster size. Our simulation study results showed that using the Gini coefficient works better than the original spatial scan statistic for identifying irregularly shaped clusters, by reporting an optimized and refined collection of clusters rather than a single larger cluster. We have provided a real data example that seems to support the simulation results. We think that using the Gini coefficient in spatial scan statistics can be helpful for the detection of irregularly shaped clusters.

SUBMITTER: Kim J

PROVIDER: S-EPMC5271318 | biostudies-literature | 2017

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Evaluation of the Gini Coefficient in Spatial Scan Statistics for Detecting Irregularly Shaped Clusters.

Kim Jiyu J Jung Inkyung I

PloS one 20170127 1

Spatial scan statistics with circular or elliptic scanning windows are commonly used for cluster detection in various applications, such as the identification of geographical disease clusters from epidemiological data. It has been pointed out that the method may have difficulty in correctly identifying non-compact, arbitrarily shaped clusters. In this paper, we evaluated the Gini coefficient for detecting irregularly shaped clusters through a simulation study. The Gini coefficient, the use of wh ...[more]

PMID: 28129368

Similar Datasets

Project description:BACKGROUND:Ageing is becoming a considerable public health burden in China, which produces great societal development challenges. Healthy and active longevity could ease the ageing burden on families and communities. To date, most studies of the oldest-old distribution are focused on a simple scale from spatial perspective, and the multi-scale spatio-temporal clusters trend in the oldest-old population has not yet been determined. Thus, the objective in present study is to use a new method to evaluate the spatio-temporal pattern and detect the risk clusters in the oldest-old population from three scales. METHODS:Individuals aged 65 years or older and individuals aged 80 years or older on three scales in China from 2000 to 2010 were used. The exploratory spatial data analysis was performed using Moran's I statistic, and the pattern of the oldest-old clusters among humans was examined by using the spatial scan statistical method. Then, spatial stratified heterogeneity was used to explore the factors affecting the spatial heterogeneity of the oldest-old population. RESULTS:The oldest-old index in the southeast coastal areas is higher than that in the northwest inland areas in China. A three-ladder terrain distribution of the oldest-old index from west to east is obvious. The overall pattern of the oldest-old index evolves from a "concave" shape to an "east-west uplift, and northern collapse" shape. Space-time analysis revealed that high-risk areas were concentrated in five regions: the Yangtze River Delta, the Pearl River Delta, the Southeast Coast, Sichuan and Chongqing, and the Central Plains. The oldest-old cluster at different scales shows a similar pattern, but local differences exist. The risk at the prefecture scale and county scale is greater than at the interprovincial scale; the sublevel can identify clusters that have not been identified at the previous level, especially the bordering areas of prefectures and counties; and more risk units and greater relative risk are found in urban areas than in rural areas. CONCLUSIONS:The results emphasized that spatial scan statistics can be used to estimate the spatial clusters of the oldest-old people. The detection of these clusters might be highly useful in the surveillance of the ageing phenomenon, thus helping local public health authorities measure the population burden at all locations, identifying geographical areas that require more attention, and evaluating the impacts of intervention programs.

Project description:BackgroundPeople with HIV infection in the United States are often affected by chronic viral hepatitis. These coinfected people with either HBV or HCV are at increased risk for serious, life-threatening complications. Coinfections with viral hepatitis may also complicate the delivery of anti-retroviral therapy (ART) by escalating the risk of drug-related hepatoxicity. According to the Centers for Disease Control and Prevention (CDC), approximately 10 percent of people with HIV in the United States also have HBV, and 25 percent also have HCV coinfection. With the advent of highly active antiretroviral therapy (HAART) and the increased life-expectancy of HIV patients, clinicians are more likely to be confronted with issues related to co-infection and the management challenges that they present, especially in resource-limited settings. The purpose of this analysis was to identify geographical clusters of HIV- (HBV/HCV) co-infection and compared to the geographical clusters of not co-infected using DC, Department of Health surveillance data. The results of the analysis will be used to target resources to areas at risk.MethodsHIV and Hepatitis surveillance data were matched among cases diagnosed between 1980 and 2016. HIV-hepatitis co-infected and the not co-infected spatial clusters were detected using discrete Poisson model. Kulldorff's spatial scan statistic method was implemented in the free software tool called SaTScan which has been widely adopted for detecting disease cluster. The analysis was conducted by tracts, but for visualization, ease of interpretation and assist in policy making the tract map was overlaid with the ward map using ArcGIS 10.5.1.ResultsBetween 1980 and 2016, there were 12,965 diagnosed cases of HIV, of which 2,316 HIV/Hepatitis matches were identified. Of the 2316 co-infected people living in DC, 25 percent (N = 590) of people had HBV, and 75 percent (N = 1,726) had HCV. Out of 12,965 diagnosed cases, remaining 10,649 did not have any co-infections (not co-infected). IDU (27.16 percent) and MSM (32.86 percent) were the highest mode of transmission for co-infected population. African-American were reported 83.64 percent (N = 1,937) among co-infection population. Three clusters were identified for both co-infected population in DC. The largest cluster radius for co-infected analysis covers wards 6, 7 and 8 as well as large parts of 2 and 5 (p < 0.001). Multiple clusters were identified for not co-infected population (p < 0.001). IDU (n = 450) was the highest mode of transmission for the co-infected clusters. For all clusters combined of not co-infected population highest mode of transmission were MSM (n = 2,534). This analysis also showed racial disparity, economic deprivation and lack of education were prominent in the co-infected clusters.ConclusionWe identified locations of high risk clusters where enhanced hepatitis and HIV prevention, treatment, and care can help combat the epidemic. The clusters radius expands into the neighboring state of Maryland as well. The findings from this analysis will be used to target area based public health policy and healthcare interventions for HIV-hepatitis. It is recommended based on the analysis that needle exchange programs can successfully control new HIV infections as well as hepatitis co-infections.

Dataset Information

Evaluation of the Gini Coefficient in Spatial Scan Statistics for Detecting Irregularly Shaped Clusters.

Publications

Evaluation of the Gini Coefficient in Spatial Scan Statistics for Detecting Irregularly Shaped Clusters.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets