ABSTRACT
Background: Gestational diabetes mellitus (GDM) is defined as carbohydrate intolerance of variable severity with onset or first recognition during the present pregnancy. It affects 7% of all pregnancies worldwide and in India it ranges from 6 to 9% in rural and 12 to 21% in urban area. The aim of this study was to compare the DIPSI criteria with the two-step method (Carpenter and Couston criteria.) and to study merits and demerits of one step and two step tests for GDM.Methods: A total 400 pregnant women of gestational age between 24-28 weeks attending antenatal clinic at this study tertiary care center were enrolled in this study. 200 pregnant women were enrolled in each of the study group (Group I OGTT and Group II DIPSI).Results: In Group I (OGTT) screening 47 (23.5%) were tested positive. In Group II cases, screening test results were found positive among 44 (22%). Out of 95 high-risk pregnant women 38 (40%) were positive for GDM by OGTT and 34 (35.78%) were positive by DIPSI. Out of 305 non high-risk pregnant women, 9 (2.95%) were positive for GDM by OGTT and 10 (3.27%) were positive by DIPSI.Conclusions: Present study concludes that DIPSI is the test which can predict GDM in population comparable to another test like OGTT. Also, India’s major population reside in rural areas, ANC are mostly conducted by ANM, therefore screening test should be easy to perform and interpret.
ABSTRACT
Los métodos de clasificación permiten explorar y analizar grandes conjuntos de datos visualmente, lo cual es de gran utilidad para tomar decisiones rápidas. El objetivo fue comparar dos métodos de análisis de clúster para big data en variables demográficas de las provincias del Ecuador. Se hizo uso de un estudio observacional de tipo comparativo mediante la representación simultanea del HJ-Biplot y el método Two Step (clúster bietápico), a través del software MultBiplot y SPSS. Los datos corresponden a variables demográficas de interés sociosanitarias tasa de mortalidad general, tasa de mortalidad infantil, tasa de natalidad, densidad poblacional, porcentaje urbano y esperanza de vida, medidas en las provincias del Ecuador. Se utilizaron datos provenientes del Instituto de Estadísticas y Censos INEC. Se analizó la asociación entre variables y se identificaron clústeres de las provincias del Ecuador según estas variables demográficas. Según la representación simultánea del HJ-Biplot se identificaron 3 clústeres, el clúster 1 son provincias con mayor densidad poblacional y tasas de mortalidad general, pero valores bajos de tasas de natalidad, el clúster 2 agrupa provincias con mayor esperanza de vida y tasas de mortalidad infantil pero bajos valores de tasa de natalidad y el clúster 3 están las provincias con valores altos de tasas de natalidad y valores bajos de densidad poblacional, esperanza de vida, tasas de mortalidad general y mortalidad infantil, distintos resultados se obtuvieron con el método Two Step. Se pudo concluir que estos métodos son de utilidad para explorar las similitudes entre las provincias según variables demográficas.
The classification methods allow to explore and analyze big data sets visually, which is very useful for making quick decisions. This work aimed to compare of two methods of cluster analysis for big data in demographic variables of the provinces of Ecuador. An observational study of comparative type was carried out through the simultaneous representation of the HJ/Biplot and the Two Step method (two-stage cluster), through the MultBiplot and SPSS software. The data correspond to demographic variables of socio-health interest, general mortality rate, infant mortality rate, birth rate, population density, urban percentage and life expectancy, measured in the provinces of Ecuador. Data from Statistics and Census Institute were used. The association between variables was analyzed and clusters of the provinces of Ecuador were identified according to these demographic variables. According to the simultaneous representation of the HJBiplot, 3 clusters were identified, cluster 1 are provinces with higher population density and general mortality rates, but low birth rates values, cluster 2 are provinces with higher life expectancy and mortality rates infantile but low birth rate values and cluster 3 are the provinces with high birth rates values and low population density, life expectancy, general mortality and infant mortality rates, different results were obtained with the Two Step method. It was concluded that these methods are useful for exploring the similarities between provinces according to demographic variables.