570 - Spatial distribution of bacterial sepsis associated neonatal deaths in a middle-income country
Monday, April 25, 2022
3:30 PM – 6:00 PM US MT
Poster Number: 570 Publication Number: 570.427
Daniela T. Costa-Nobre, Universidade Federal de São Paulo - UNIFESP, São Paulo, Sao Paulo, Brazil; Maria Fernanda B. Almeida, Universidade Federal de Sao Paulo, Sao Paulo, Sao Paulo, Brazil; Ana Silvia Marinonio, Universidade Federal de São Paulo, São Paulo, Sao Paulo, Brazil; Adriana Sanudo, Federal University of São Paulo, São Paulo, Sao Paulo, Brazil; Rita Balda, Universidade Federal de Sao Paulo, São Pailo, Sao Paulo, Brazil; MILTON H. MIYOSHI, unifesp, São Paulo, Sao Paulo, Brazil; Kelsy C. Areco, Federal University of São Paulo, São Paulo, Sao Paulo, Brazil; Mandira D. Kawakami, Federal University of Sao Paulo, Sao Paulo, Sao Paulo, Brazil; Tulio Konstantyner, Universidade Federal de São Paulo, São Paulo, Sao Paulo, Brazil; Paulo Bandiera-Paiva, Universidade Federal de São Paulo, São Paulo, Sao Paulo, Brazil; Rosa V. Freitas, Fundação Seade, São Paulo, Brasil, Sao Paulo, Brazil; Lilian C. Morais, Fundação Seade, São Paulo, Sao Paulo, Brazil; Monica La Porte Teixeira, Fundação Seade, São Paulo, Sao Paulo, Brazil; Bernadette C. Waldvogel, UFMG, São Paulo, Sao Paulo, Brazil; Ruth Guinsburg, Universidade Federal de São Paulo, São Paulo, Sao Paulo, Brazil; Carlos Roberto Kiffer, UNIFESP EPM, São Paulo, Sao Paulo, Brazil
Associate Professor Universidade Federal de Sao Paulo Sao Paulo, Sao Paulo, Brazil
Background: Almost half of all under-five deaths occur during the neonatal period. The 3rd Sustainable Development Goal (SDG) aims, by 2030, ending preventable deaths of newborns and children under 5 years of age. Spatial analysis can be useful to analyze cause-specific neonatal mortality clusters.
Objective: To explore the spatial distribution patterns and clusters of neonatal mortality associated with bacterial sepsis between 2004 -2015, in São Paulo (SP) State, Brazil through a structured spatial analytical approach.
Design/Methods: Population-based study including all live births ≥22 weeks’ gestation without congenital anomalies, from 2004-2015, of mothers living in SP State. Data was retrieved by deterministic linkage from Death and Birth Certificates. Neonatal deaths associated with bacterial sepsis were defined as deaths (0-27days) with bacterial sepsis (ICD10: A32.7, A40, A41.0-A41.5, P36 and P37.2) written in any line of the Death Certificate. Bacterial sepsis associated mortality rate and the overall neonatal mortality rate by 1000 live births were calculated for each of the 645 State municipalities. After applying a smoothing approach for reducing impact of outliers with local Bayes estimates, global spatial autocorrelation (Moran Index, I) was used to test for spatial random distribution. Clusters were identified by Local Indicators of Spatial Autocorrelation (LISA) maps.
Results: During 2004-2015, 41,715 neonatal deaths and 15,088 bacterial sepsis associated deaths occurred among 7,165,209 live births in SP State. Both the neonatal mortality rate associated with bacterial sepsis and the overall neonatal mortality rate spatial distributions showed a non-random distribution among counties, with presence of clusters (I=0.61; p=0.001, and I=0,55; p=0.001, respectively) (Fig 1). LISA cluster maps allowed the identification of well-delimited high-risk areas for deaths with bacterial sepsis and for neonatal mortality distribution (Fig 2). There was an important overlap of the clusters areas for both rates, confirming the importance of the bacterial sepsis as a cause of neonatal deaths. On the other hand, the differences between clusters are not irrelevant.Conclusion(s): Public health efforts should focus on the overlapping areas for cause specific and overall mortality to decrease neonatal mortality and to achieve the 3rd SDG. High-risk areas for bacterial sepsis associated mortality should be further studied, but current findings may point to reference cluster areas with higher capacity for complex care.
FAPESP Grant: # 2017/03748-7; Agreement SEADE/EPM-Unifesp: #23089.004297/2008-11/#23089.000057/2014-95 Figure 1: Bacterial sepsis associated mortality rates (A), and overall mortality rates (B) per 1000 live birth, from 2004 to 2015, São Paulo State, Brazil. Figure 2: LISA cluster map for bacterial sepsis associated mortality (A), and overall mortality (B) in São Paulo State after smoothing with local Bayes estimate, 2004-2015.