Accurate estimates of the under-five mortality rate in a developing world context are a key barometer of the health of a nation. This paper describes a new model to analyze survey data on mortality in this context. We are interested in both spatial and temporal description, that is wishing to estimate under-five mortality rate across regions and years and to investigate the association between the under-five mortality rate and spatially varying covariate surfaces. We illustrate the methodology by producing yearly estimates for subnational areas in Kenya over the period 1980–2014 using data from the Demographic and Health Surveys, which use stratified cluster sampling. We use a binomial likelihood with fixed effects for the urban/rural strata and random effects for the clustering to account for the complex survey design. Smoothing is carried out using Bayesian hierarchical models with continuous spatial and temporally discrete components. A key component of the model is an offset to adjust for bias due to the effects of HIV epidemics. Substantively, there has been a sharp decline in Kenya in the under-five mortality rate in the period 1980–2014, but large variability in estimated subnational rates remains. A priority for future research is understanding this variability. In exploratory work, we examine whether a variety of spatial covariate surfaces can explain the variability in under-five mortality rate. Temperature, precipitation, a measure of malaria infection prevalence, and a measure of nearness to cities were candidates for inclusion in the covariate model, but the interplay between space, time, and covariates is complex.
AlkemaLNewJRPedersenJ, et al.Child mortality estimation 2013: an overview of updates in estimation methods by the United Nations inter-agency group for child mortality estimation. PLoS One2014; 9: e101112–e101112.
2.
AlkemaLNewJ. Global estimation of child mortality using a Bayesian B-spline bias-reduction model. Ann Appl Stat2014; 8: 2122–2149.
3.
BarasaEManyaraAMolyneuxS, et al.Recentralization within decentralization: county hospital autonomy under devolution in Kenya. PLoS One2017; 12: e0182440–e0182440.
4.
Dwyer-LindgrenLKakunguFHangomaP, et al.Estimation of district-level under-5 mortality in Zambia using birth history data, 1980–2010. Spat Spatiotemporal Epidemiol2014; 11: 89–107.
5.
RueHMartinoSChopinN. Approximate Bayesian inference for latent Gaussian models using integrated nested Laplace approximations (with discussion). J R Stat Soc Ser B2009; 71: 319–392.
6.
MercerLWakefieldJPantazisA, et al.Small area estimation of childhood of childhood mortality in the absence of vital registration. Ann Appl Stat2015; 9: 1889–1905.
7.
Knorr-HeldL. Bayesian modelling of inseparable space-time variation in disease risk. Stat Med2000; 19: 2555–2567.
8.
Pezzulo C, Bird T, Utazi EC, et al. Geospatial modeling of child mortality across 27 countries in Sub-Saharan Africa. DHS Spatial Analysis Reports; No. 13. Rockville, US: ICF International, 2016.
9.
LerouxBLeiXBreslowNEstimation of disease rates in small areas: a new mixed model for spatial dependence. In: HalloranMBerryD (eds). Statistical models in epidemiology, the environment and clinical trials, New York: Springer, 1999, pp. 179–192.
10.
WakefieldJ. Ecologic studies revisited. Annu Rev Public Health2008; 29: 75–90.
11.
RaoJMolinaI. Small Area Estimation, 2nd ed. New York: John Wiley, 2015.
12.
CongdonPLloydP. Estimating small area diabetes prevalence in the US using the behavioral risk factor surveillance system. J Data Sci2010; 8: 235–252.
13.
YouYZhouQM. Hierarchical Bayes small area estimation under a spatial model with application to health survey data. Survey Methodol2011; 37: 25–37.
14.
PorterATHolanSHWikleCK, et al.Spatial Fay–Herriot models for small area estimation with functional covariates. Spat Stat2014; 10: 27–42.
15.
ChenCWakefieldJLumleyT. The use of sample weights in Bayesian hierarchical models for small area estimation. Spat Spatiotemporal Epidemiol2014; 11: 33–43.
16.
VandendijckYFaesCKirbyRS, et al.Model-based inference for small area estimation with sampling weights. Spat Stat2016; 18: 455–473.
17.
WatjouKFaesCLawsonA, et al.Spatial small area smoothing models for handling survey data with nonresponse. Stat Med2017; 36: 3708–3745.
18.
BurkeMHeft-NealSBendavidE. Sources of variation in under-5 mortality across sub-Saharan Africa: a spatial analysis. Lancet Global Health2016; 4: e936–e945.
19.
LarmarangeJBendaudV. HIV estimates at second subnational level from national population-based surveys. AIDS2014; 28: S469–S476.
20.
HallettTAndersonS-JAsanteCA, et al.Evaluation of geospatial methods to generate subnational HIV prevalence estimates for local level planning. AIDS2016; 30: 1467–1474.
21.
GoldingNBursteinRLongbottomJ, et al.Mapping under-5 and neonatal mortality in Africa, 2000–15: a baseline analysis for the sustainable development goals. Lancet, 390: 2171–2182.
22.
LindgrenFRueHLindströmJ. An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic differential equation approach (with discussion). J R Stat Soc Ser B2011; 73: 423–498.
23.
GBD 2016 Mortality Collaborators. Global, regional, and national under-5 mortality, adult mortality, age-specific mortality, and life expectancy, 1970–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet2017; 390: 1084–1150.
WalkerNHillKZhaoF. Child mortality estimation: methods used to adjust for bias due to AIDS in estimating trends in under-five mortality. PLoS Med2012; 9: e1001298–e1001298.
26.
SkinnerCWakefieldJ. Introduction to the design and analysis of complex survey data. Stat Sci2017; 32: 165–175.
27.
HorvitzDThompsonD. A generalization of sampling without replacement from a finite universe. J Am Stat Assoc1952; 47: 663–685.
28.
WakefieldJSimpsonDGodwinJ. Comment: getting into space with a weight problem. Discussion of, “Model-based geostatistics for prevalence mapping in low-resource settings”, by P.J. Diggle and E. Giorgi. J Am Stat Assoc2016; 111: 1111–1119.
29.
AllisonP. Event History and Survival Analysis. 2nd edn2014; Vol. 46, Los Angeles: SAGE Publications.
30.
BlangiardoMCamelettiM. Spatial and Spatio-temporal Bayesian Models with R-INLA, Chichester: John Wiley and Sons, 2015. .
31.
FuglstadG-ASimpsonDLindgrenF, et al.Constructing priors that penalize the complexity of Gaussian random fields. J Am Stat Assoc. In press, 2018.
32.
SimpsonDRueHRieblerA, et al.Penalising model component complexity: a principled, practical approach to constructing priors (with discussion). Stat Sci2017; 32: 1–28.
33.
LinardCGilbertMSnowRW, et al.Population distribution, settlement patterns and accessibility across Africa in 2010. PLoS One2012; 7: e31743–e31743.
34.
WorldPop. Kenya 1 km births, version 2. Technical report, University of Southampton. DOI: 10.5258/SOTON/WP00349, 2017.
35.
Pesaresi M, Ehrlich D, Ferri S, et al. Operating procedure for the production of the global human settlement layer from Landsat data of the epochs 1975, 1990, 2000, and 2014. Publications Office of the European Union, 2016, http://publications.jrc.ec.europa.eu/repository/handle/JRC97705.
36.
LiuLOzaSHoganD, et al.Global, regional, and national causes of under-5 mortality in 2000–15: an updated systematic analysis with implications for the sustainable development goals. Lancet2017; 388: 3027–3035.
37.
BesagJYorkJMolliéA. Bayesian image restoration with two applications in spatial statistics. Ann Inst Stat Math1991; 43: 1–59.
38.
MosleyWHChenLC. An analytical framework for the study of child survival in developing countries. Popul Dev Rev1984; 10: 25–45.
39.
BalkDPullumTStoreygardA, et al.A spatial analysis of childhood mortality in West Africa. Popul Space Place2004; 10: 175–216.
40.
TottrupCTersbolBLindeboomW, et al.Putting child mortality on a map: towards an understanding of inequity in health. Trop Med Int Health2009; 14: 653–662.
41.
Gething P, Tatem A, Bird T, et al. Creating spatial interpolation surfaces with DHS data. Technical report, ICF International. DHS Spatial Analysis Reports No. 11, 2015, USA: ICF International.
42.
BhattSCameronEFlaxmanS, et al.Improved prediction accuracy for disease risk mapping using Gaussian process stacked generalization. J R Soc Interf2017; 14: 20170520–20170520.
43.
Wolpert D. Stacked generalization. Neural Netw 1992; 5: 241–259.
44.
Van der LaanMJPolleyECHubbardAE. Super learner. Stat Appl Genet Mol Biol2007; 6. Article 25. Epub ahead of print 16 September 2007.
45.
Nelson A. Estimated travel time to the nearest city of 50,000 or more people in year 2000. Technical report, Global Environment Monitoring Unit – Joint Research Centre of the European Commission, Ispra, Italy, 2008.
46.
Zomer RJ, Bossio DA, Trabucco A, et al. Trees and water: smallholder agroforestry on irrigated lands in Northern India. International Water Management Institute, 2007. CGSpace: A Repository of Agricultural Research Outputs, http://hdl.handle.net/10568/39909.
47.
ZomerRJTrabuccoABossioDA, et al.Climate change mitigation: a spatial analysis of global land suitability for clean development mechanism afforestation and reforestation. Agric Ecosyst Environ2008; 126: 67–80.
48.
FickSEHijmansRJ. WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. Int J Climatol2017; 37: 4302–4315.
49.
Didan K. MOD13A3 MODIS/Terra vegetation Indices Monthly L3 Global 1 km SIN Grid V006. NASA EOSDIS Land Processes DAAC, 2015.
50.
BhattSWeissDCameronE, et al.The effect of malaria control on plasmodium falciparum in Africa between 2000 and 2015. Nature2015; 526: 207–211.
51.
RootG. Population density and spatial differentials in child mortality in Zimbabwe. Soc Sci Med1997; 44: 413–421.
52.
RootGP. Disease environments and subnational patterns of under-five mortality in Sub-Saharan Africa. Popul Space Place1999; 5: 117–132.
53.
SpiegelhalterDFreedmanLParmarM. Bayesian approaches to randomized trials (with discussion). J R Stat Soc Ser A1994; 157: 357–416.
54.
HeldLSchrödleBRueHPosterior and cross-validatory predictive checks: a comparison of MCMC and INLA. In: KneibTTutzG (eds). Statistical Modeling and Regression Structures – Festschrift in Honour of Ludwig Fahrmeir, Heidelberg, Germany: Physica-Verlag, 2010, pp. 91–110. .
55.
WatanabeS. A widely applicable Bayesian information criterion. J Mach Learn Res2013; 14: 867–897.
56.
LeeRCarterL. Modeling and forecasting US mortality. J Am Stat Assoc1992; 87: 659–671.
57.
SharrowDJClarkSJRafteryAE. Modeling age-specific mortality for countries with generalized HIV epidemics. PLoS One2014; 9: e96447–e96447.
58.
AlexanderMZagheniEBarbieriM. A flexible Bayesian model for estimating subnational mortality. Demography2017; 54: 2025–2041.
59.
RieblerASørbyeSHSimpsonD, et al.An intuitive Bayesian spatial model for disease mapping that accounts for scaling. Stat Methods Med Res2016; 25: 1145–1165.
60.
NychkaDBandyopadhyaySHammerlingD, et al.A multiresolution Gaussian process model for the analysis of large spatial datasets. J Comput Graph Stat2015; 24: 579–599.
61.
CressieNJohannessonG. Fixed rank kriging for very large spatial data sets. J R Stat Soc Ser B2008; 70: 209–226.
62.
BanerjeeSGelfandAEFinleyAO, et al.Gaussian predictive process models for large spatial data sets. J R Stat Soc Ser B (Stat Methodol)2008; 70: 825–848.
63.
BradleyJRCressieNShiT, et al.A comparison of spatial predictors when datasets could be very large. Stat Surveys2016; 10: 100–131.
64.
HeatonMJDattaAFinleyA, et al.Methods for analyzing large spatial data: a review and comparison. arXiv preprint arXiv:1710.050132017.
65.
FonsecaTCFerreiraMA. Dynamic multiscale spatiotemporal models for Poisson data. J Am Stat Assoc2017; 112: 215–234.
66.
TakahashiSMetcalfCJEFerrariMJ, et al.The geography of measles vaccination in the African great lakes region. Nat Commun2017; 8. Epub ahead of print 25 May 2017.
67.
FosterSShimadzuHDarnellR. Uncertainty in spatially predicted covariates: is it ignorable?J R Stat Soc Ser C2012; 61: 637–652.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
