Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Dec;54(6):2025-2041.
doi: 10.1007/s13524-017-0618-7.

A Flexible Bayesian Model for Estimating Subnational Mortality

Monica Alexander  1 , Emilio Zagheni  2 , Magali Barbieri  3   4
Affiliations

A Flexible Bayesian Model for Estimating Subnational Mortality

Monica Alexander et al. Demography. 2017 Dec.

Abstract

Reliable subnational mortality estimates are essential in the study of health inequalities within a country. One of the difficulties in producing such estimates is the presence of small populations among which the stochastic variation in death counts is relatively high, and thus the underlying mortality levels are unclear. We present a Bayesian hierarchical model to estimate mortality at the subnational level. The model builds on characteristic age patterns in mortality curves, which are constructed using principal components from a set of reference mortality curves. Information on mortality rates are pooled across geographic space and are smoothed over time. Testing of the model shows reasonable estimates and uncertainty levels when it is applied both to simulated data that mimic U.S. counties and to real data for French départements. The model estimates have direct applications to the study of subregional health patterns and disparities.

Keywords: Bayesian hierarchical model; France; Mortality; Principal components; Subnational estimation.

PubMed Disclaimer

Figures

Fig. 1
Fig. 1
Example data and principal components of (logged) U.S. state mortality schedules: Males, 1980–2010
Fig. 2
Fig. 2
Illustrating the effect of geographic pooling. Model fitted without (left) and with (right) geographic pooling. Data are simulated assuming a total population of 5,000
Fig. 3
Fig. 3
Illustrating the effect of smoothing over time. The solid line shows median estimates for μβ1;t from a model without smoothing imposed. The dashed line shows median estimates for μβ1,t from a model with smoothing imposed according to Eq. (7)
Fig. 4
Fig. 4
Female life expectancy at birth estimates for France, 2008 (years): Constrained versus unconstrained model
Fig. 5
Fig. 5
True, simulated, and estimated mortality rates for three hypothetical counties
Fig. 6
Fig. 6
Observed and estimated mortality rates, Lozère: Males, 1975 and 2008
Fig. 7
Fig. 7
Observed and estimated mortality rates, Somme: Males, 1975 and 2008
Fig. 8
Fig. 8
Life expectancy estimates for males, 2008 (e0, years)
Fig. 9
Fig. 9
Standard deviation of β1 and β2 over time
Fig. 10
Fig. 10
Mortality rates in areas with low (left) and high (right) values of β2
See this image and copyright information in PMC

References

    1. Alkema L, New JR. Global estimation of child mortality using a Bayesian B-spline bias-reduction method. Annals of Applied Statistics. 2014;8:2122–2149.
    1. Alkema L, Raftery AE, Gerland P, Clark SJ, Pelletier F, Buettner T, Heilig GK. Probabilistic projections of the total fertility rate for all countries. Demography. 2012;48:815–839. - PMC - PubMed
    1. Barbieri M. Mortality in France by département. Population. 2013;68:375–417.
    1. Bijak J. Bayesian methods in international migration forecasting. In: Raymer J, Willekens F, editors. International migration in Europe: Data, models and estimates. Chichester, UK: John Wiley & Sons; 2008. pp. 253–281.
    1. Bravo JM, Malta J. Estimating life expectancy in small population areas. In: EROSTAT–European Commission, editor. Work session on demographic projections: EUROSTAT methodologies and working papers. 2010. Luxembourg City, Luxembourg: Publications Office of the European Union; 2010. pp. 113–126.

Publication types

MeSH terms