Adjusting for sampling variability in sparse data: geostatistical approaches to disease mapping

International Journal of Health Geographics

William C Miller¹  Christopher D Pilcher⁵  Dionne C Gesink²  Marc L Serre³  Kristen H Hampton⁴ 

[1] Department of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA;Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada;Department of Environmental Sciences and Engineering, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA;Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA;HIV/AIDS Division, San Francisco General Hospital, University of California-San Francisco, San Francisco, CA, USA

关键词: epidemiological methods;    spatial analysis;    spatial distribution;    sampling variability;    disease mapping;   

Others  :  812700
DOI  :  10.1186/1476-072X-10-54

 received in 2011-05-24, accepted in 2011-10-06,  发布年份 2011

【 摘 要 】

Background

Disease maps of crude rates from routinely collected health data indexed at a small geographical resolution pose specific statistical problems due to the sparse nature of the data. Spatial smoothers allow areas to borrow strength from neighboring regions to produce a more stable estimate of the areal value. Geostatistical smoothers are able to quantify the uncertainty in smoothed rate estimates without a high computational burden. In this paper, we introduce a uniform model extension of Bayesian Maximum Entropy (UMBME) and compare its performance to that of Poisson kriging in measures of smoothing strength and estimation accuracy as applied to simulated data and the real data example of HIV infection in North Carolina. The aim is to produce more reliable maps of disease rates in small areas to improve identification of spatial trends at the local level.

Results

In all data environments, Poisson kriging exhibited greater smoothing strength than UMBME. With the simulated data where the true latent rate of infection was known, Poisson kriging resulted in greater estimation accuracy with data that displayed low spatial autocorrelation, while UMBME provided more accurate estimators with data that displayed higher spatial autocorrelation. With the HIV data, UMBME performed slightly better than Poisson kriging in cross-validatory predictive checks, with both models performing better than the observed data model with no smoothing.

Conclusions

Smoothing methods have different advantages depending upon both internal model assumptions that affect smoothing strength and external data environments, such as spatial correlation of the observed data. Further model comparisons in different data environments are required to provide public health practitioners with guidelines needed in choosing the most appropriate smoothing method for their particular health dataset.

【 授权许可】

   
2011 Hampton et al; licensee BioMed Central Ltd.

【 图 表 】

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

【 参考文献 】

• [1]Best N, Richardson S, Thomson A: A comparison of Bayesian spatial models for disease mapping. Statistical Methods in Medical Research 2005, 14(1):35-59.
• [2]Leyland AH, Davies CA: Empirical Bayes methods for disease mapping. Statistical Methods in Medical Research 2005, 14(1):17-34.
• [3]Wakefield J, Elliott P: Issues in the statistical analysis of small area health data. Statistics in Medicine 1999, 18(17-18):2377-2399.
• [4]Wakefield JC, Best NG, Waller L: Bayesian approaches to disease mapping. In Spatial Epidemiology: Methods and Applications. Edited by Elliott P, Wakefield J, Best N, Briggs D. Oxford: Oxford University Press; 2000:104-127.
• [5]Waller LA, Gotway CA: Applied Spatial Statistics for Public Health Data. Hoboken, NJ: John Wiley & Sons, Inc.; 2004.
• [6]Pascutto C, Wakefield JC, Best NG, Richardson S, Bernardinelli L, Staines A, Elliott P: Statistical issues in the analysis of disease mapping data. Statistics in Medicine 2000, 19(17-18):2493-2519.
• [7]Lawson AB: Statistical Methods in Spatial Epidemiology. Chichester: John Wiley & Sons Ltd.; 2001.
• [8]Liao H-H, Laymon P, Shull K: Automated Process for Accessing Vital Health Information at Census Tract Level. In Geographic Information Systems in Public Health: Proceedings of the Third National Conference: 17-18 August 1998; San Diego. Edited by Williams RC, Howie MM, Lee CV, Henriques WD. Atlanta: Centers for Disease Control and Prevention; 2000:119-136. retrieved September 2005 from http://www.atsdr.cdc.gov/gis/conference98/index.html webcite
• [9]Bithell JF: A classification of disease mapping methods. Statistics in Medicine 2000, 19(17-18):2203-2215.
• [10]Clayton D, Kaldor J: Empirical Bayes estimates of age-standardized relative risks for use in disease mapping. Biometrics 1987, 43(3):671-681.
• [11]Devine OJ, Louis TA, Halloran ME: Empirical Bayes methods for stabilizing incidence rates before mapping. Epidemiology 1994, 5(6):622-630.
• [12]Goovaerts P: Geostatistical analysis of disease data: estimation of cancer mortality risk from empirical frequencies using Poisson kriging. International Journal of Health Geographics 2005, 4:31. BioMed Central Full Text
• [13]Goovaerts P, Gebreab S: How does Poisson kriging compare to the popular BYM model for mapping disease risks? International Journal of Health Geographics 2008, 7:6. BioMed Central Full Text
• [14]Ali M, Goovaerts P, Nazia N, Haq MZ, Yunus M, Emch M: Application of Poisson kriging to the mapping of cholera and dysentery incidence in an endemic area of Bangladesh. International Journal of Health Geographics 2006, 5:45. BioMed Central Full Text
• [15]Christakos G, Li X: Bayesian Maximum Entropy Analysis and Mapping: A Farewell to Kriging Estimators? Mathematical Geology 1998, 30(4):435-462.
• [16]Serre ML, Christakos G: Modern geostatistics: computational BME analysis in the light of uncertain physical knowledge - the Equus Beds study. Stochastic Environmental Research and Risk Assessment 1999, 13(1-2):1-26.
• [17]Christakos G, Bogaert P, Serre ML: Temporal GIS: Advanced Functions for Field-Based Applications. New York: Springer-Verlag; 2002.
• [18]Choi K-M, Serre ML, Christakos G: Efficient mapping of California mortality fields at different spatial scales. Journal of Exposure Analysis and Environmental Epidemiology 2003, 13(2):120-133.
• [19]Law DCG, Serre ML, Christakos G, Leone PA, Miller WC: Spatial analysis and mapping of sexually transmitted diseases to optimise intervention and prevention strategies. Sexually Transmitted Infections 2004, 80(4):294-299.
• [20]Christakos G, Olea R, Serre ML, Yu H-L, Wang L: Interdisciplinary Public Health Reasoning and Epidemic Modelling: The Case of Black Death. New York: Springer-Verlag; 2005.
• [21]Gesink Law DC, Bernstein KT, Serre ML, Schumacher CM, Leone PA, Zenilman JM, Miller WC, Rompalo AM: Modeling a syphilis outbreak through space and time using the Bayesian Maximum Entropy approach. Annals of Epidemiology 2006, 16(11):797-804.
• [22]Lee S-J, Yeatts KB, Serre ML: A Bayesian Maximum Entropy approach to address the change of support problem in the spatial analysis of childhood asthma prevalence across North Carolina. Spatial and Spatio-temporal Epidemiology 2009, 1(1):49-60.
• [23]Monestiez P, Dubroca L, Bonnin E, Durbec J-P, Guinet C: Geostatistical modelling of spatial distribution of Balaenoptera physalus in the Northwestern Mediterranean Sea from sparse count data and heterogeneous observation efforts. Ecological Modelling 2006, 193(3-4):615-628.
• [24]Rothman KJ, Greenland S: Measures of Disease Frequency. In Modern Epidemiology. 2nd edition. Edited by Rothman KJ, Greenland S. Philadelphia: Lippincott-Raven; 1998:29-46.
• [25]MathWorks, Inc: MaTLab, the language of technical computing. In using MATLAB version 6.1. Natick, MA: The MathWorks, Inc; 2001.
• [26]The Bayesian Maximum Entropy software for Space/Time Geostatistics and temporal GIS data integration [http://www.unc.edu/depts/case/BMELIB/] webcite
• [27]Gesink DC, Sullivan AB, Miller WC, Bernstein KT: Sexually transmitted disease core theory: roles of person, place, and time. American Journal of Epidemiology 2011, 174(1):81-89.
• [28]Lin LI: A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989, 45(1):255-268.
• [29]Rothenberg RB: The geography of gonorrhea. Empirical demonstration of core group transmission. American Journal of Epidemiology 1983, 117(6):688-694.
• [30]Becker KM, Glass GE, Brathwaite W, Zenilman JM: Geographic epidemiology of gonorrhea in Baltimore, Maryland, using a geographic information system. American Journal of Epidemiology 1998, 147(7):709-716.
• [31]Zenilman JM, Ellish N, Fresia A, Glass GE: The geography of sexual partnerships in Baltimore: applications of core theory dynamics using a geographic information system. Sexually Transmitted Diseases 1999, 26(2):75-81.
• [32]North Carolina Department of Health & Human Services, Division of Public Health: North Carolina Epidemiologic Profile for HIV/STD Prevention & Care Planning, July 2005. Raleigh, NC; 2005.
• [33]Geographic Data Technology, ESRI: U.S. ZIP Code Points. In ESRI Data & Maps 2004. Redlands, CA: ESRI; 2004.
• [34]Stern HS, Cressie N: Posterior predictive model checks for disease mapping models. Statistics in Medicine 2000, 19(17-18):2377-2397.
• [35]Marshall EC, Spiegelhalter DJ: Approximate cross-validatory predictive checks in disease mapping models. Statistics in Medicine 2003, 22(10):1649-1660.
• [36]Yu H-L, Yang S-J, Yen H-J, Christakos G: A spatio-temporal climate-based model of early dengue fever warning in southern Taiwan. Stochastic Environmental Research and Risk Assessment 2011, 25(4):485-494.
• [37]Deutsch CV: Direct assessment of local accuracy and precision. In Geostatistics Wollongong '96. Volume 1. Edited by Baafi EY, Schofield NA. Dordrecht, The Netherlands, Kluwer Academic Publishers; 1997::115-125.