Modeling excess zeros and heterogeneity in count data from a complex survey design with application to the demographic health survey in sub-Saharan Africa

Abstract

Purpose

To show a novel application of a weighted zero-inflated negative binomial model in modeling count data with excess zeros and heterogeneity to quantify the regional variation in HIV-AIDS prevalence in sub-Saharan African countries.

Methods

Data come from latest round of the Demographic and Health Survey (DHS) conducted in three countries (Ethiopia-2011, Kenya-2009 and Rwanda-2010) using a two-stage cluster sampling design. The outcome is an aggregate count of HIV cases in each census enumeration area of each country. The outcome data are characterized by excess zeros and heterogeneity due to clustering. We compare scale weighted zero-inflated negative binomial models with and without random effects to account for zero-inflation, complex survey design and clustering. Finally, we provide marginalized rate ratio estimates from the best zero-inflated negative binomial model.

Results

The best fitting zero-inflated negative binomial model is scale weighted and with a common random intercept for the three countries. Rate ratio estimates from the final model show that HIV prevalence is associated with age and gender distribution, HIV acceptance, HIV knowledge, and its regional variation is associated with divorce rate, burden of sexually transmitted diseases and rural residence.

Conclusions

Scale weighted zero-inflated negative binomial with proper modeling of random effects is shown to be the best model for count data from a complex survey design characterized by excess zeros and extra heterogeneity. In our data example, the final rate ratio estimates show significant regional variation in the factors associated with HIV prevalence indicating that HIV intervention strategies should be tailored to the unique factors found in each country.

Keywords

HIV multi-country survey data negative binomial regional variation sub-Saharan Africa zero-inflation

Get full access to this article

View all access options for this article.

References

Mishra V, Medley A, Hong R, et al. Levels and spread of HIV seroprevalence and associated factors: evidence from national household surveys. 2009.

Joint United Nations Programme on HIV/AIDS. Global Fact Sheet UNAIDS 2013, http://www.unaids.org/en/media/unaids/contentassets/documents/epidemiology/2013/gr2013/20130923_FactSheet_Global_en.pdf (2013, accessed 1 July 2014).

Joint United Nations Programme on HIV/AIDS and World Health Organization. Sub-Saharan Africa: AIDS epidemic update regional summary. Geneva, Switzerland, Joint United Nations Programme on HIV/AIDS, 2008. UNAIDS/08.08E/JC1526E.

HIV and AIDS in Sub-saharan Africa. AVERT, 2014, http://www.avert.org/hiv-and-aids-sub-saharan-africa.htm (2014, accessed 14 July 2001).

Joint United Nations Programme on HIV/AIDS. Regional facts sheet 2012 Sub-saharan Africa. Joint United Nations Programme on HIV/AIDS, http://www.unaids.org/en/media/unaids/contentassets/documents/epidemiology/2012/gr2012/2012_FS_regional_ssa_en.pdf (2012, accessed 14 July 2001).

Bradley

Thompson

Byam

et al.

Access and quality of rural healthcare: ethiopian millennium rural initiative. Int J Qual Health Care 2011; 23: 222–230.

Joint United Nations Programme on HIV/AIDS and World Health Organization. Global Report: UNAIDS Report on the Global AIDS Epidemic, Geneva, Switzerland, Joint United Nations Programme on HIV/AIDS. Report no. UNAIDS/10.11E/JC1958E, 2010.

ICF International. Demographic and Health survey sampling and household listing manual: Measure DHS. Calverton, Maryland, USA: ICF International, 2012.

Rutstein

Rojas

. Guide to DHS statistics, Calverton, Maryland: ORC Macro, 2006.

10.

Biomarker Field Manual. Calverton, Maryland, ICF International/Demographic and Health Surveys, 2012.

11.

Lambert

. Zero-inflated Poisson regression, with an application to defects in manufacturing. Technometrics 1992; 34: 1–14.

12.

Ridout

Hinde

DeméAtrio

. A score test for testing a zero-inflated Poisson regression model against zero-nflated negative binomial alternatives. Biometrics 2001; 57: 219–223.

13.

Long

Preisser

Herring

et al.

A marginalized zero-inflated Poisson regression model with overall exposure effects. Stat Med 2014; 33: 5151–5165.

14.

Preisser JS, Das K, Long DL, et al. A marginalized zero-inflated negative binomial regression model with overall exposure effects. The University of North Carolina at Chapel Hill, Department of Biostatistics, Technical Report Series Working paper no. 43. 2014.

15.

SAS Institute Inc. SAS/STAT(R) 9.3 User's guide, Cary North Carolina: SAS Institute Inc, 2011.

16.

Pfeffermann

Skinner

Holmes

et al.

Weighting for unequal selection probabilities in multilevel models. J Roy Stat Soc Ser B, Stat Methodol 1998; 60: 23–40.

17.

Rabe-Hesketh

Skrondal

. Multilevel modelling of complex survey data. J Roy Stat Soc: Ser A 2006; 169: 805–827.

18.

Asparouhov

. General multi-level modeling with sampling weights. Commun Stat Theor Methods 2006; 35: 439–460.

19.

Smyth

. Pearson's goodness of fit statistic as a score test statistic. Lecture notes – monograph series. Hayward, CA: Institute of Mathematical Statistics, 2003, pp. 115–126.

20.

Lupia

Chien

S-C

. HIV and AIDS epidemic in Kenya: an overview. J Exp Clin Med 2012; 4: 231–234.

21.

Tanser

Bärnighausen

Hund

et al.

Effect of concurrent sexual partnerships on rate of new HIV infections in a high-prevalence, rural South African population: a cohort study. The Lancet 2011; 378: 247–255.

22.

Buvé

Caraël

Hayes

et al.

Multicentre study on factors determining differences in rate of spread of HIV in sub-Saharan Africa: methods and prevalence of HIV infection. AIDS 2001; 15: S5–S14.

23.

Magadi

Desta

. A multilevel analysis of the determinants and cross-national variations of HIV seropositivity in sub-Saharan Africa: evidence from the DHS. Health Place 2011; 17: 1067–1083.

24.

Lagarde

Auvert

Caraël

et al.

Concurrent sexual partnerships and HIV prevalence in five urban communities of sub-Saharan Africa. AIDS 2001; 15: 877–884.

25.

Mah

Halperin

. Concurrent sexual partnerships and the HIV epidemics in Africa: evidence to move forward. AIDS Behav 2008; 14: 11–16.

26.

Genberg

Hlavka

Konda

et al.

A comparison of HIV/AIDS-related stigma in four countries: negative attitudes and perceived acts of discrimination towards people living with HIV/AIDS. Soc Sci Med 2009; 68: 2279–2287.

27.

Winskell

Hill

Obyerodhyambo

. Comparing HIV-related symbolic stigma in six African countries: social representations in young people's narratives. Soc Sci Med 2011; 73: 1257–1265.

28.

Garcia-Calleja

. National population based HIV prevalence surveys in sub-Saharan Africa: results and implications for HIV and AIDS estimates. Sex Trans Infect 2006; 82: iii64–iii70.

29.

Angotti

Bula

Gaydosh

et al.

Increasing the acceptability of HIV counseling and testing with three C's: convenience, confidentiality and credibility. Soc Sci Med 2009; 68: 2263–2270.

30.

Bärnighausen

Bor

Wandira-Kazibwe

et al.

Correcting HIV prevalence estimates for survey nonparticipation using heckman-type selection models. Epidemiolgy 2011; 22: 27–35.

31.

Lam

Xue

Bun Cheung

. Semiparametric analysis of zero-inflated count data. Biometrics 2006; 62: 996–1003.

32.

Neelon

O'Malley

Normand

SLT

. A Bayesian model for repeated measures zero-inflated count data with application to outpatient psychiatric service use. Stat Modell 2010; 10: 421–439.

33.

Giardina

Gosoniu

Konate

et al.

Estimating the burden of malaria in senegal: bayesian zero-inflated binomial geostatistical modeling of the MIS 2008 data. PLoS ONE 2012; 7: e32625–e32625.