Fewer data sources provide information on the breakdown of under-five deaths by month of death required to measure NN, PNN or CHD rates.
Data (147 countries):
1234 VR country-years
526 DHS complete birth history country-years
We estimate a hierarchical model relating the probability of an under-5 death occurring during the NN, PNN or CHD period to the level of under-5 mortality using a random intercept and slope.
We tested multiple specifications for this model and used the model with the highest predictive validity.
Predictive Validity for Neonatal, Postneonatal, Childhood Model Design Age Mean Relative Error Median Relative Error In Sample Neonatal 9.3% 6.2% Post-neonatal 10.9% 8.0% Child 11.4% 7.8% Dropping 20% of data Neonatal 9.8% 6.7% Post-neonatal 11.8% 8.7% Child 11.9% 8.1% Dropping 20% of countries Neonatal 11.4% 7.1% Post-neonatal 12.9% 9.1% Child 14.3% 8.9% Dropped last 10 years Neonatal 14.4% 11.4% Post-neonatal 21.0% 15.1% Child 19.2% 13.3%
Country Examples: Norway
Country Examples: El Salvador
Country Examples: Laos
Country Examples: Liberia
Global Under Five Deaths 1970-2010
Global Neonatal, Postneonatal, and Childhood Mortality Rates 1970-2010
Under-Five Death Rates by Region Group A
Under-Five Death Rates by Region Group B
Under-Five Death Rates by Region Group C
Under-Five Death Rates by Region Group D
Under-Five Mortality Rate in 2010
Annual Rate of Decline in U5 1990-2010
Regional Rates of Decline NN, PNN and CHD 1990-2010
Accelerating Progress on U5
Developing countries on track to achieve MDG 4 Country Rate of decline (%) Country Rate of decline (%) Maldives 9.21 Syrian Arab Republic 5.11 United Arab Emirates 8.43 Algeria 4.99 Oman 6.92 Bhutan 4.92 Turkmenistan 6.65 Cuba 4.88 Viet Nam 6.4 Armenia 4.84 Sri Lanka 6.28 Thailand 4.83 Egypt 6.2 Brazil 4.81 Malaysia 5.85 Bahrain 4.74 Tunisia 5.69 China 4.72 El Salvador 5.64 Mexico 4.63 Lebanon 5.62 Bangladesh 4.57 Peru 5.42 Honduras 4.57 Libyan Arab Jamahiriya 5.34 Barbados 4.55 Nepal 5.33 Costa Rica 4.5 Chile 5.16 Turkey 4.47 Mongolia 5.15
Does not have a great effect on the estimates
Simulation studies suggest that in countries with sero-prevalence of 20% under-5 mortality may be under-estimated by 10 per 1000 to over-estimated by 6 per 1000.
Analysis of 16,174 measurements for 187 countries from 1970 to 2010 show that under-5 deaths have dropped from 16 million in 1970 to 7.7 million in 2010.
Annual rate of decline in developing countries during MDG4 period is 2.1%; below the target of 4.4% per year.
In 1990, 12 countries had under-5 death rates over 200 per 1000; now there are none.
Progress has been slowest in sub-Saharan Africa, but in 34 countries in the region the rate of decline is faster in 2000 to 2010 compared to 1990 to 2000.
Learning from Success Stories
Across and within the 21 regions, there is marked variability in the pace of decline.
Next step for the research agenda should be to examine what are the policy determinants that may have contributed to faster or slower rates of decline in under-5 mortality.
Rapid declines in under-5 death rates are possible; 66 countries in the world have decreased child mortality by 30% in a period of 5 years.
Such rapid declines provide hope that accelerated progress is possible.