Mark E McGovern1, Till Bärnighausen, Giampiero Marra, Rosalba Radice. 1. From the aHarvard Center for Population and Development Studies, Cambridge, MA; bDepartment of Global Health and Population, Harvard School of Public Health, Boston, MA; cWellcome Trust Africa Centre for Health and Population Studies, University of KwaZulu-Natal, Mtubatuba, South Africa; dDepartment of Statistical Science, University College London, London, UK; and eDepartment of Economics, Mathematics and Statistics, Birkbeck, University of London, London, UK.
Abstract
BACKGROUND: Heckman-type selection models have been used to control HIV prevalence estimates for selection bias when participation in HIV testing and HIV status are associated after controlling for observed variables. These models typically rely on the strong assumption that the error terms in the participation and the outcome equations that comprise the model are distributed as bivariate normal. METHODS: We introduce a novel approach for relaxing the bivariate normality assumption in selection models using copula functions. We apply this method to estimating HIV prevalence and new confidence intervals (CI) in the 2007 Zambia Demographic and Health Survey (DHS) by using interviewer identity as the selection variable that predicts participation (consent to test) but not the outcome (HIV status). RESULTS: We show in a simulation study that selection models can generate biased results when the bivariate normality assumption is violated. In the 2007 Zambia DHS, HIV prevalence estimates are similar irrespective of the structure of the association assumed between participation and outcome. For men, we estimate a population HIV prevalence of 21% (95% CI = 16%-25%) compared with 12% (11%-13%) among those who consented to be tested; for women, the corresponding figures are 19% (13%-24%) and 16% (15%-17%). CONCLUSIONS: Copula approaches to Heckman-type selection models are a useful addition to the methodological toolkit of HIV epidemiology and of epidemiology in general. We develop the use of this approach to systematically evaluate the robustness of HIV prevalence estimates based on selection models, both empirically and in a simulation study.
BACKGROUND: Heckman-type selection models have been used to control HIV prevalence estimates for selection bias when participation in HIV testing and HIV status are associated after controlling for observed variables. These models typically rely on the strong assumption that the error terms in the participation and the outcome equations that comprise the model are distributed as bivariate normal. METHODS: We introduce a novel approach for relaxing the bivariate normality assumption in selection models using copula functions. We apply this method to estimating HIV prevalence and new confidence intervals (CI) in the 2007 Zambia Demographic and Health Survey (DHS) by using interviewer identity as the selection variable that predicts participation (consent to test) but not the outcome (HIV status). RESULTS: We show in a simulation study that selection models can generate biased results when the bivariate normality assumption is violated. In the 2007 Zambia DHS, HIV prevalence estimates are similar irrespective of the structure of the association assumed between participation and outcome. For men, we estimate a population HIV prevalence of 21% (95% CI = 16%-25%) compared with 12% (11%-13%) among those who consented to be tested; for women, the corresponding figures are 19% (13%-24%) and 16% (15%-17%). CONCLUSIONS: Copula approaches to Heckman-type selection models are a useful addition to the methodological toolkit of HIV epidemiology and of epidemiology in general. We develop the use of this approach to systematically evaluate the robustness of HIV prevalence estimates based on selection models, both empirically and in a simulation study.
Authors: A Rogier T Donders; Geert J M G van der Heijden; Theo Stijnen; Karel G M Moons Journal: J Clin Epidemiol Date: 2006-07-11 Impact factor: 6.437
Authors: Sian Floyd; Anna Molesworth; Albert Dube; Amelia C Crampin; Rein Houben; Menard Chihana; Alison Price; Ndoliwe Kayuni; Jacqueline Saul; Neil French; Judith R Glynn Journal: AIDS Date: 2013-01-14 Impact factor: 4.177
Authors: Daniel R Hogan; Joshua A Salomon; David Canning; James K Hammitt; Alan M Zaslavsky; Till Bärnighausen Journal: Sex Transm Infect Date: 2012-12 Impact factor: 3.519
Authors: Jessica R Marden; Linbo Wang; Eric J Tchetgen Tchetgen; Stefan Walter; M Maria Glymour; Kathleen E Wirth Journal: Epidemiology Date: 2018-05 Impact factor: 4.822
Authors: Anton M Palma; Giampiero Marra; Rachel Bray; Suzue Saito; Anna Colletar Awor; Mohamed F Jalloh; Alexander Kailembo; Wilford Kirungi; George S Mgomella; Prosper Njau; Andrew C Voetsch; Jennifer A Ward; Till Bärnighausen; Guy Harling Journal: J Int AIDS Soc Date: 2022-08 Impact factor: 6.707
Authors: Mark E McGovern; Giampiero Marra; Rosalba Radice; David Canning; Marie-Louise Newell; Till Bärnighausen Journal: J Int AIDS Soc Date: 2015-11-26 Impact factor: 5.396
Authors: Mark E McGovern; Kobus Herbst; Frank Tanser; Tinofa Mutevedzi; David Canning; Dickman Gareta; Deenan Pillay; Till Bärnighausen Journal: Int J Epidemiol Date: 2016-12-01 Impact factor: 7.196
Authors: Neema R Mosha; Omololu S Aluko; Jim Todd; Rhoderick Machekano; Taryn Young Journal: BMC Med Res Methodol Date: 2020-03-14 Impact factor: 4.615