Clusters of cause specific neonatal mortality and its association with per capita gross domestic product: A structured spatial analytical approach

Background Infant mortality rate is a measure of population health and neonatal mortality account for great proportion of these deaths. Underdevelopment might be associated to higher neonatal mortality risk due to assistant related factors. Spatial and temporal distribution of mortality help identifying and developing strategies for interventions. Objective To investigate the cluster areas of asphyxia-associated neonatal mortality and to explore its association with per capita gross domestic product (GDP) in São Paulo State (SP), Brazil. Methods Ecological study including live births residents in SP from 2004–2013. Neonatal deaths (0–27 days) with perinatal asphyxia were defined as intrauterine hypoxia, birth asphyxia or meconium aspiration syndrome written in any line of the Death Certificate. Geoprocessing analytical approach included detection of first order effects through quintiles and spatial moving average maps, followed by second order effects by global and local spatial autocorrelation (Moran and LISA, respectively) before and after smoothing with local Bayesian estimates. Finally, Spearman correlation was applied between asphyxia-associated neonatal mortality and mean per capita GDP rates for the municipalities with significant LISA. Results There were 6,713 asphyxia-associated neonatal deaths among 5,949,267 live births (rate: 1.13/1000) in SP. Spatial moving average maps showed a non-random distribution among municipalities, with presence of clusters (I = 0.048; p = 0.023). LISA map identified clusters of asphyxia-associated neonatal mortality in the south, southeast and northwest. After applying local Bayes estimates, clusters were more pronounced (I = 0.589; p = 0.001). There was a partial overlap of the areas of higher asphyxia-associated neonatal mortality and lower mean per capita GDP. Conclusions Spatial analysis identified cluster areas of high asphyxia-associated neonatal mortality and low per capita GDP rates, with a significant negative correlation. This optimized, structured, and hierarchical approach to identify high-risk areas of cause-specific neonatal mortality may be helpful for guiding public health efforts to decrease neonatal mortality.

Pemodelan dengan Spatial Autoregressive (SAR) pada Angka Putus Sekolah Bagi Anak Usia Wajib Belajar di Provinsi Sulawesi Selatan

Regresi spasial merupakan pengembangan dari regresi klasik. Pengembangan ini berdasarkan adanya pengaruh tempat atau spasial dari data yang dianalisis. Beberapa model regresi spasial adalah Spatial Autoregressive (SAR), Spatial Error Model (SEM) dan Spatial Moving Average (SARMA). Penelitian ini menggunakan analisis model SAR terhadap angka putus sekolah di Sulawesi Selatan. Data yang digunakan merupakan data sekunder dari Badan Pusat Statistik Provinsi Sulawesi Selatan tahun 2018. Penelitian ini dilakukan untuk mengetahui model Spatial Autoregressive (SAR) pada data banyaknya angka putus sekolah yang terjadi di Provinsi Sulawesi Selatan, serta mengenalisis faktor-faktor yang memberikan pengaruh signifikan terhadap pertumbuhan angka putus sekolah. Hasil penelitian ini memperoleh model yaitu ; sehingga faktor-faktor yang berpengaruh secara signifikan terhadap angka putus sekolah di Sulawesi Selatan adalah pengeluaran per kapita, rasio murid terhadap sekolah dan jumlah penduduk miskin.Keywords: Regresi Spasial, Spatial Autoregressive Model (SAR), Angka Putus Sekolah

The Assessment Of Geographical Borders In Economic Research

The length of common border between two geographical units is frequently used as a basic weight in spatial analysis. The newest methodological propositions such as tests for hierarchical relations (Markowska et. al. 2014; Sokołowski et. al. 2013), regional spatial moving average and new spatial correlation coefficient (Markowska et. al. 2015) are using border lengths. In cited references new methods have been illustrated by analyses for EU NUTS2 regions. It is obvious that borders between regions belonging to different countries have different socio-economic impact than borders between regions lying in the same country. A new simple method for assesment the importance of borders is proposed in the paper. It is based on a chosen macroeconomic variable available at NUTS 2 level (e.g. GDP, infant mortality, Human Development Index). For neighboring regions bigger value is divided by smaller value giving the local importance of the given border. These measures of local border importance can be than average for borders within the same country and for borders for each pair of neighboring countries.