Limits...
Spatio-temporal analysis of mortality among children under the age of five in Manhiça (Mozambique) during the period 1997-2005.

Escaramís G, Carrasco JL, Aponte JJ, Nhalungo D, Nhacolo A, Alonso P, Ascaso C - Int J Health Geogr (2011)

Bottom Line: The results showed that childhood mortality in all the area was modified from year to year describing a convex time-trend but the spatial pattern described by the neighbourhood-specific underlying mortality rates did not change during the entire period from 1997 to 2005, where neighbourhoods with highest risks are situated in the peripheral side of the district.The results of this study suggest that the health intervention programmes established in Manhiça to alleviate the effects of flooding on child mortality should cover a period of around five years and that special attention might be focused on eradicating malaria transmission.These outcomes also suggest the utility of suitably modelling space-time trend variations in a region when a point effect of an environmental factor affects all the study area.

View Article: PubMed Central - HTML - PubMed

Affiliation: Bioestadística, Departament de Salut Pública, Universitat de Barcelona, Barcelona, Spain.

ABSTRACT

Background: Reducing childhood mortality is the fourth goal of the Millennium Development Goals agreed at the United Nations Millennium Summit in September 2000. However, childhood mortality in developing countries remains high. Providing an accurate picture of space and time-trend variations in child mortality in a region might generate further ideas for health planning actions to achieve such a reduction. The purpose of this study was to examine the spatio-temporal variation for child mortality rates in Manhiça, a district within the Maputo province of southern rural Mozambique during the period 1997-2005 using a proper generalized linear mixed model.

Results: The results showed that childhood mortality in all the area was modified from year to year describing a convex time-trend but the spatial pattern described by the neighbourhood-specific underlying mortality rates did not change during the entire period from 1997 to 2005, where neighbourhoods with highest risks are situated in the peripheral side of the district. The spatial distribution, though more blurred here, was similar to the spatial distribution of child malaria incidence in the same area. The peak in mortality rates observed in 2001 could have been caused by the precipitation system that started in early February 2000, following which heavy rains flooded parts of Mozambique's southern provinces. However, the mortality rates at the end of the period returned to initial values.

Conclusions: The results of this study suggest that the health intervention programmes established in Manhiça to alleviate the effects of flooding on child mortality should cover a period of around five years and that special attention might be focused on eradicating malaria transmission. These outcomes also suggest the utility of suitably modelling space-time trend variations in a region when a point effect of an environmental factor affects all the study area.

Show MeSH

Related in: MedlinePlus

First plot: dots indicate annual log standardised mortality rates for Manhiça district, while the smoothed line is the estimated second-order polynomial time trend derived from the Poisson regression model. The same information is shown in subsequent plots per each of the 115 neighbourhoods that comprise Manhiça. Last plot: boxplots showing the variability among the 115 intercept, first- and second-order polynomial estimates derived from the 115 neighbourhood-specific Poisson regression models.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3050678&req=5

Figure 1: First plot: dots indicate annual log standardised mortality rates for Manhiça district, while the smoothed line is the estimated second-order polynomial time trend derived from the Poisson regression model. The same information is shown in subsequent plots per each of the 115 neighbourhoods that comprise Manhiça. Last plot: boxplots showing the variability among the 115 intercept, first- and second-order polynomial estimates derived from the 115 neighbourhood-specific Poisson regression models.

Mentions: Exploratory results for our data showed a mean time-trend convex curvature among southern Manhiça neighbourhoods associated with the crude logSMR and revealing a marked increase followed by decreased shape (Figure 1, first plot in the grid). The estimated second-order polynomial time-trend confirmed this result as the sign of the quadratic term is negative (Table 1 and Figure 1).


Spatio-temporal analysis of mortality among children under the age of five in Manhiça (Mozambique) during the period 1997-2005.

Escaramís G, Carrasco JL, Aponte JJ, Nhalungo D, Nhacolo A, Alonso P, Ascaso C - Int J Health Geogr (2011)

First plot: dots indicate annual log standardised mortality rates for Manhiça district, while the smoothed line is the estimated second-order polynomial time trend derived from the Poisson regression model. The same information is shown in subsequent plots per each of the 115 neighbourhoods that comprise Manhiça. Last plot: boxplots showing the variability among the 115 intercept, first- and second-order polynomial estimates derived from the 115 neighbourhood-specific Poisson regression models.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3050678&req=5

Figure 1: First plot: dots indicate annual log standardised mortality rates for Manhiça district, while the smoothed line is the estimated second-order polynomial time trend derived from the Poisson regression model. The same information is shown in subsequent plots per each of the 115 neighbourhoods that comprise Manhiça. Last plot: boxplots showing the variability among the 115 intercept, first- and second-order polynomial estimates derived from the 115 neighbourhood-specific Poisson regression models.
Mentions: Exploratory results for our data showed a mean time-trend convex curvature among southern Manhiça neighbourhoods associated with the crude logSMR and revealing a marked increase followed by decreased shape (Figure 1, first plot in the grid). The estimated second-order polynomial time-trend confirmed this result as the sign of the quadratic term is negative (Table 1 and Figure 1).

Bottom Line: The results showed that childhood mortality in all the area was modified from year to year describing a convex time-trend but the spatial pattern described by the neighbourhood-specific underlying mortality rates did not change during the entire period from 1997 to 2005, where neighbourhoods with highest risks are situated in the peripheral side of the district.The results of this study suggest that the health intervention programmes established in Manhiça to alleviate the effects of flooding on child mortality should cover a period of around five years and that special attention might be focused on eradicating malaria transmission.These outcomes also suggest the utility of suitably modelling space-time trend variations in a region when a point effect of an environmental factor affects all the study area.

View Article: PubMed Central - HTML - PubMed

Affiliation: Bioestadística, Departament de Salut Pública, Universitat de Barcelona, Barcelona, Spain.

ABSTRACT

Background: Reducing childhood mortality is the fourth goal of the Millennium Development Goals agreed at the United Nations Millennium Summit in September 2000. However, childhood mortality in developing countries remains high. Providing an accurate picture of space and time-trend variations in child mortality in a region might generate further ideas for health planning actions to achieve such a reduction. The purpose of this study was to examine the spatio-temporal variation for child mortality rates in Manhiça, a district within the Maputo province of southern rural Mozambique during the period 1997-2005 using a proper generalized linear mixed model.

Results: The results showed that childhood mortality in all the area was modified from year to year describing a convex time-trend but the spatial pattern described by the neighbourhood-specific underlying mortality rates did not change during the entire period from 1997 to 2005, where neighbourhoods with highest risks are situated in the peripheral side of the district. The spatial distribution, though more blurred here, was similar to the spatial distribution of child malaria incidence in the same area. The peak in mortality rates observed in 2001 could have been caused by the precipitation system that started in early February 2000, following which heavy rains flooded parts of Mozambique's southern provinces. However, the mortality rates at the end of the period returned to initial values.

Conclusions: The results of this study suggest that the health intervention programmes established in Manhiça to alleviate the effects of flooding on child mortality should cover a period of around five years and that special attention might be focused on eradicating malaria transmission. These outcomes also suggest the utility of suitably modelling space-time trend variations in a region when a point effect of an environmental factor affects all the study area.

Show MeSH
Related in: MedlinePlus