Abstract
Knowledge of causal processes through mediation analysis can help improve the effectiveness and reduce costs of public health programs, like HIV prevention and treatment interventions. Advancements in mediation using the potential outcomes framework provide a method for estimating the causal effect of interventions on outcomes via a mediating variable. The purpose of this paper is to provide practical information about mediation and the potential outcomes framework that can enhance data analysis and causal inference for intervention studies. Causal mediation effects are defined and then estimated using data from an HIV intervention randomized trial among people who inject drugs (PWID) in Ukraine. Results from a potential outcomes mediation analysis show that the intervention had a total causal effect on incident HIV infection such that participants in the experimental group were 36% less likely to become infected during the 12-month study than those in the control arm, but that neither self-efficacy nor network communication mediated this effect. Because neither putative mediator was significant, measurement and confounding issues should be investigated to rule out these mediators. Other putative mediators, such as injection frequency, route of administration, or HIV knowledge can be considered. Future research is underway to examine additional, multiple mediators explaining efficacy of the current intervention and sensitivity to confounding effects.
Resumen
El conocimiento de procesos causales a través del análisis de mediación puede ayudar a mejorar la eficiencia y reducir los costos de programas de salud pública, incluyendo la prevención y el tratamiento del VIH. Avances en mediación, utilizando el enfoque de resultados potenciales ofrece un método para estimar los efectos causales de las intervenciones en variables dependientes a través de variables mediadoras. El objetivo de este artículo es ofrecer información acerca del análisis de mediación y del enfoque de resultados potenciales, el cual permite el análisis y la inferencia causal de las intervenciones. Los efectos causales de mediación son definidos y estimados utilizando los datos de un ensayo clínico con asignación aleatoria para disminuir riego del VIH entre usuarios de drogas inyectables (UDI) en Ucrania. Los resultados del análisis de mediación desde el enfoque de resultados potenciales muestran que la intervención tuvo un efecto causal total en la incidencia de infección del VIH, tal que, durante los 12 meses de estudio, fue 36% menos probable que los participantes del grupo experimental se infectaran en comparación con aquellos en el grupo control. Sin embargo, ni la autoeficacia ni la red comunicación mediaron el efecto. Dado que ninguno de los mediadores resultó ser significativo, sería necesario investigar problemas con la medición y efectos de confusión para poder descartarlos. Otros mediadores podrían ser considerados, tales como frecuencia de la inyección, ruta de administración, o el conocimiento acerca del VIH. Futuras investigaciones podrían estudiar diferentes y múltiples mediadores para explicar la eficacia de esta intervención y realizar un análisis de sensibilidad de efectos de confusión.
Similar content being viewed by others
Data Availability
N/A
Code Availability
A SAS program for analysis is available in the appendix, or by contacting the first author.
References
Risher K, Mayer KH, Beyrer C. HIV treatment cascade in MSM, people who inject drugs, and sex workers. Curr Opin HIV AIDS. 2015;10(6):420–9.
Macdonald V, Verster A, Baggaley R. A call for differentiated approaches to delivering HIV services to key populations. J Int AIDS Soc. 2017;20(S4):28–31.
Lall P, Lim SH, Khairuddin N, Kamarulzaman A. Review: An urgent need for research on factors impacting adherence to and retention in care among HIV-positive youth and adolescents from key populations. J Int AIDS Soc. 2015;18(2S1):41–53.
Mathers BM, Degenhardt L, Phillips B, Wiessing L, Hickman M, Strathdee SA, et al. Global epidemiology of injecting drug use and HIV among people who inject drugs: A systematic review. The Lancet. 2008;372(9651):1733–45.
Degenhardt L, Mathers B, Vickerman P, Rhodes T, Latkin C, Hickman M. Prevention of HIV infection for people who inject drugs: Why individual, structural, and combination approaches are needed. The Lancet. 2010;376(9737):285–301.
Pearl J. Interpretation and identification of causal mediation. Psychol Methods. 2014;19(4):459.
Pearl J. The causal mediation formula: A guide to the assessment of pathways and mechanisms. Prev Sci. 2012;13(4):426–36.
VanderWeele TJ. Mediation and mechanism. Eur J Epidemiol. 2009;24(5):217–24.
Hernàn MA, Robins JM. Causal Inference. Boca Raton: Chapman & Hall/CRC; 2019. (forthcoming).
Baron RM, Kenny DA. The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51(6):1173.
MacKinnon DP, Lockwood CM, Hoffman JM, West SG, Sheets V. A comparison of methods to test mediation and other intervening variable effects. Psychol Methods. 2002;7(1):83–104.
Imai K, Keele L, Tingley D. A general approach to causal mediation analysis. Psychol Methods. 2010;15(4):309.
Jo B. Causal inference in randomized experiments with mediational processes. Psychol Methods. 2008;13(4):314.
Liu W, Kuramoto SJ, Stuart EA. An introduction to sensitivity analysis for unobserved confounding in nonexperimental prevention research. Prev Sci. 2013;14(6):570–80.
Stuart EA, Bradshaw CP, Leaf PJ. Assessing the generalizability of randomized trial results to target populations. Prev Sci. 2015;16(3):475–85.
Valeri L, VanderWeele TJ. Mediation analysis allowing for exposure–mediator interactions and causal interpretation: Theoretical assumptions and implementation with SAS and SPSS macros. Psychol Methods. 2013;18(2):137–50.
Ajzen I, Fishbein M. Understanding attitudes and predicting social behavior. New Jersey: Prentice-Hall; 1980.
Albarracin D, Johnson BT, Fishbein M, Muellerleile PA. Theories of reasoned action and planned behavior as models of condom use: a meta-analysis. Psychol Bull. 2001;127(1):142.
Fisher JD, Fisher WA. Changing AIDS-risk behavior. Psychol Bull. 1992;111(3):455.
Fisher JD, Fisher WA, Williams SS, Malloy TE. Empirical tests of an information-motivation-behavioral skills model of AIDS-preventive behavior with gay men and heterosexual university students. Health Psychol. 1994;13(3):238.
Bryan A, Schmiege SJ, Broaddus MR. Mediational analysis in HIV/AIDS research: Estimating multivariate path analytic models in a structural equation modeling framework. AIDS Behav. 2007;11(3):365–83.
O’Rourke HP, MacKinnon DP. Reasons for testing mediation in the absence of an intervention effect: A research imperative in prevention and intervention research. J Stud Alcohol Drugs. 2018;79(2):171–81.
O’Rourke HP, MacKinnon DP. The importance of mediation analysis in substance-use prevention. Prevention of Substance Use: Springer; 2019. p. 233–46.
Kisbu-Sakarya Y, MacKinnon DP, O’Rourke HP. Statistical models of mediation for drug program evaluation. In: Scheier LM, editor. Handbook of adolescent drug use prevention: Research, intervention strategies, and practice. Washington: American Psychological Association; 2015. p. 459–78.
Hardnett FP, Pals SL, Borkowf CB, Parsons J, Gomez C, O’Leary A. Assessing mediation in HIV intervention studies. Public Health Rep. 2009;124(2):288–94.
Pitpitan EV, Kalichman SC, Garcia RL, Cain D, Eaton LA, Simbayi LC. Mediators of behavior change resulting from a sexual risk reduction intervention for STI patients, Cape Town, South Africa. J Behav Med. 2015;38(2):194–203.
Pitpitan EV, Patterson TL, Abramovitz D, Vera A, Martinez G, Staines H, et al. Policing behaviors, safe injection self-efficacy, and intervening on injection risks: Moderated mediation results from a randomized trial. Health Psychol. 2016;35(1):87–91.
Bandura A. Self-Efficacy: The Exercise of Control. New York: Worth Publishers; 1997.
MacKinnon DP, Valente MJ, Gonzales O. The correspondence between causal and traditional mediation analysis: The link is the mediator by treatment interaction. Prev Sci. 2020;21(2):147–57.
VanderWeele TJ. A unification of mediation and interaction: a four-way decomposition. Epidemiology (Cambridge, Mass). 2014;25(5):749–61.
Rijnhart JJM, Valente MJ, MacKinnon DP, Twisk JWR, Heymans MW. The use of traditional and causal estimators for mediation models with a binary outcome and exposure-mediator interaction. Structural Equation Modeling: A Multidisciplinary Journal. 2018. https://doi.org/10.1080/10705511.2020.1811709.
Tofighi D, MacKinnon DP. RMediation: An R package for mediation analysis confidence intervals. Behav Res Methods. 2011;43(3):692–700.
Holland PW. Statistics and causal inference. J Am Stat Assoc. 1986;81(396):945–60.
Cole SR, Frangakis CE. The consistency statement in causal inference: a definition or an assumption? Epidemiology. 2009;20(1):3–5.
Hernán MA. Does water kill? A call for less casual causal inferences. Ann Epidemiol. 2016;26(10):674–80.
VanderWeele TJ. Concerning the consistency assumption in causal inference. Epidemiology. 2009;20(6):880–3.
Hernán MA, Robins JM. Estimating causal effects from epidemiological data. J Epidemiol Community Health. 2006;60(7):578–86.
Cole SR, Hernán MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol. 2008;168(6):656–64.
VanderWeele TJ, Vansteelandt S. Conceptual issues concerning mediation, interventions and composition. Statistics and its Interface. 2009;2(4):457–68.
Valente MJ, Pelham WEI, Smyth HL, MacKinnon DP. Confounding in statistical mediation analysis: What it is and how to address it. J Couns Psychol. 2017;64(6):659–71.
Pirlott AG, MacKinnon DP. Design approaches to experimental mediation. J Exp Soc Psychol. 2016;66:29–38.
Valente MJ, Rijnhart JJM, Smyth HL, Muniz FB, MacKinnon DP. Review and comparison of software for the estimation of causal mediation effects. Structural Equation Modeling. 2020.
Ikram MA, VanderWeele TJ. A proposed clinical and biological interpretation of mediated interaction. Eur J Epidemiol. 2015;30(10):1115–8.
VanderWeele TJ. A three-way decomposition of a total effect into direct, indirect, and interactive effects. Epidemiology (Cambridge, Mass). 2013;24(2):224–32.
Booth RE, Davis JM, Dvoryak S, Brewster JT, Lisovska O, Strathdee SA, et al. HIV incidence among people who inject drugs (PWIDs) in Ukraine: Results from a clustered randomised trial. The Lancet HIV. 2016;3(10):e482–9.
Casey MK, Timmermann L, Allen M, Krahn S, Turkiewicz KL. Response and Self-Efficacy of Condom Use: A Meta-Analysis of this Important Element of AIDS Education and Prevention. Southern Communication Journal. 2009;74(1):57–78.
VanderWeele TJ. Causal Mediation Analysis With Survival Data. Epidemiology. 2011;22(4):582–5.
Hoyle RH, Kenny DA. Sample size, reliability, and tests of statistical mediation. In: Hoyle RH, editor. Statistical strategies for small sample research. Thousand Oaks, CA: Sage; 1999. p. 195–222.
Chen HT. Theory-driven evaluations. Newbury Park, CA: Sage; 1990.
MacKinnon DP. Analysis of mediating variables in prevention and intervention research. NIDA Res Monogr. 1994;139:127–53.
Bandura A. Social cognitive theory: An agentic perspective. Annu Rev Psychol. 2001;52(1):1–26.
Fishbein M. A theory of reasoned action: Some applications and implications. Nebr Symp Motiv. 1979;27:65–116.
Winett RA, Anderson ES, Desiderato LL, Solomon LJ, Perry M, Kelly JA, et al. Enhancing social diffusion theory as a basis for prevention intervention: A conceptual and strategic framework. Appl Prev Psychol. 1995;4(4):233–45.
Valente TW. Network models of the diffusion of innovations. Computational & Mathematical Organization Theory. 1996;2(2):163–4.
Latkin CA, Mandell W, Vlahov D, Oziemkowska M, Celentano DD. The long-term outcome of a personal network-oriented HIV prevention intervention for injection drug users: The SAFE study. Am J Community Psychol. 1996;24(3):341–64.
Latkin CA. Outreach in natural settings: The use of peer leaders for HIV prevention among injecting drug users’ networks. Public Health Rep. 1998;113(Suppl 1):151.
Latkin CA, Knowlton AR. New directions in HIV prevention among drug users: Settings, norms, and network approaches to AIDS prevention (SNNAAP): A social influence approach. In: Latkin CA, editor. Emergent Issues in the Field of Drug Abuse. Bingley: Emerald Group Publishing Limited; 1999. p. 261–87.
Fisher JD, Fisher WA, Misovich SJ, Kimble DL, Malloy TE. Changing AIDS risk behavior: Effects of an intervention emphasizing AIDS risk reduction information, motivation, and behavioral skills in a college student population. Health Psychol. 1996;15(2):114–23.
MacKinnon DP. Introduction to statistical mediation analysis. New York: Routledge; 2008.
Albert JM, Nelson S. Generalized causal mediation analysis. Biometrics. 2011;67(3):1028–38.
Avin C, Shpitser I, Pearl J. Identifiability of path-specific effects. UCLA: Department of Statistics, UCLA. 2005:Retrieved from https://escholarship.org/uc/item/45x689gq.
Daniel R, De Stavola B, Cousens S, Vansteelandt S. Causal mediation analysis with multiple mediators. Biometrics. 2015;71(1):1–14.
Imai K, Yamamoto T. Identification and sensitivity analysis for multiple causal mechanisms: Revisiting evidence from framing experiments. Political Analysis. 2013;21(2):141–71.
Lange T, Rasmussen M, Thygesen LC. Assessing natural direct and indirect effects through multiple pathways. Am J Epidemiol. 2013;179(4):513–8.
Nguyen TQ, Webb-Vargas Y, Koning IM, Stuart EA. Causal mediation analysis with a binary outcome and multiple continuous or ordinal mediators: Simulations and application to an alcohol intervention. Structural equation modeling: a multidisciplinary journal. 2016;23(3):368–83.
Taguri M, Featherstone J, Cheng J. Causal mediation analysis with multiple causally non-ordered mediators. Stat Methods Med Res. 2018;27(1):3–19.
VanderWeele T, Vansteelandt S. Mediation analysis with multiple mediators. Epidemiologic Methods. 2014;2(1):95–115.
Wang W, Nelson S, Albert JM. Estimation of causal mediation effects for a dichotomous outcome in multiple-mediator models using the mediation formula. Stat Med. 2013;32(24):4211–28.
Yu Q, Fan Y, Wu X. General multiple mediation analysis with an application to explore racial disparities in breast cancer survival. Journal of Biometrics and Biostatistics. 2014;5(2):1–9.
Zheng C, Zhou XH. Causal mediation analysis in the multilevel intervention and multicomponent mediator case. Journal of the Royal Statistical Society: Series B (Statistical Methodology). 2015;77(3):581–615.
Mayer A, Thoemmes F, Rose N, Steyer R, West SG. Theory and analysis of total, direct, and indirect causal effects. Multivar Behav Res. 2014;49(5):425–42.
Steen J, Loeys T, Moerkerke B, Vansteelandt S. Medflex: An R package for flexible mediation analysis using natural effect models. J Stat Softw. 2017. https://doi.org/10.18637/jss.v076.i11.
Tingley D, Yamamoto T, Hirose K, Keele L, Imai K. mediation: R package for causal mediation analysis. J Stat Softw. 2014;59(5):38.
Medley A, Kennedy C, O’Reilly K, Sweat M. Effectiveness of peer education interventions for HIV prevention in develo** countries: A systematic review and meta-analysis. AIDS Educ Prev. 2009;21(3):181–206.
Crepaz N, Marshall KJ, Aupont LW, Jacobs ED, Mizuno Y, Kay LS, et al. The efficacy of HIV/STI behavioral interventions for African American females in the United States: A meta-analysis. Am J Public Health. 2009;99(11):2069–78.
LaCroix JM, Pellowski JA, Lennon CA, Johnson BT. Behavioural interventions to reduce sexual risk for HIV in heterosexual couples: A meta-analysis. Sexually transmitted infections. 2013;89(8):620–7.
Eaton LA, Huedo-Medina TB, Kalichman SC, Pellowski JA, Sagherian MJ, Warren M, et al. Meta-analysis of single-session behavioral interventions to prevent sexually transmitted infections: Implications for bundling prevention packages. Am J Public Health. 2012;102(11):e34–44.
Tan JY, Huedo-Medina TB, Warren MR, Carey MP, Johnson BT. A meta-analysis of the efficacy of HIV/AIDS prevention interventions in Asia, 1995–2009. Soc Sci Med. 2012;75(4):676–87.
Albarracin J, Albarracin D, Durantini M. Effects of HIV-prevention interventions for samples with higher and lower percents of Latinos and Latin Americans: a meta-analysis of change in condom use and knowledge. AIDS Behav. 2008;12(4):521–43.
Fishbein M, Pequegnat W. Evaluating AIDS prevention interventions using behavioral and biological outcome measures. Sex Transm Dis. 2000;27(2):101–10.
Lyles CM, Kay LS, Crepaz N, Herbst JH, Passin WF, Kim AS, et al. Best-evidence interventions: Findings from a systematic review of HIV behavioral interventions for US populations at high risk, 2000–2004. Am J Public Health. 2007;97(1):133–43.
Coffman DL. Estimating causal effects in mediation analysis using propensity scores. Structural equation modeling: a multidisciplinary journal. 2011;18(3):357–69.
Shadish WR, Cook TD, Campbell DT. Experimental and quasi-experimental designs for generalized causal inference. Belmont: Wadsworth Cengage Learning; 2002.
Vansteelandt S. Estimating direct effects in cohort and case–control studies. Epidemiology. 2009;20:851–60.
Acknowledgements
This research was supported in part by the National Institute on Drug Abuse (NIDA; R37DA09757) and (R01DA042666). Data are from a NIDA-funded randomized trial (R01DA026739). We acknowledge the staff and directors of the organization who participated in the original intervention and data collection, including Dimitry Kryzhko with Health of Nation in Makeyevka/Donetsk; Olga Kostyuk and Tatiana Semikop with Faith, Hope and Love in Odessa; and Elena Goryacheva with the Charity Foundation Vykhod in Nikolayev. We would also like to thank the participants who gave their time for the project.
Funding
This research was supported in part by the National Institute on Drug Abuse (NIDA; R37DA09757) and (R01DA042666). Data are from a NIDA-funded randomized trial (R01DA026739).
Author information
Authors and Affiliations
Contributions
Heather Smyth was responsible for conducting the statistical analysis. David MacKinnon and Eileen Pitpitan assisted with the analysis and data interpretation. Heather Smyth, Eileen Pitpitan, and David MacKinnon wrote the initial draft of the manuscript. Robert Booth secured funding for and led the original intervention trial. All authors contributed to and approved the final manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare they have no conflict of interest.
Ethical Approval
This study represents a secondary data analysis of a study, including informed consent, that was approved by the Colorado Multiple Institutional Review Board at the University of Colorado Denver and by the Ukrainian Institute on Public Health Policy. As a secondary analysis of de-identified data, the current analysis was not deemed as Human Subjects research by the Institutional Review Board at San Diego State University.
Consent to Participate
N/A
Consent for Publication
N/A
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
PROC CAUSALMED Program
/*The following program performs a causal mediation analysis*/
/*X is treatment (0=control, 1=treatment), M is a continuous mediator measured at time 2, Y is a binary outcome (0=no disease, 1=disease), and C is the baseline measurement of the mediator as a covariate*/
/*The class statement specifies the two categorical variables. The descending option is used to predict the probability Y=1*/
/*A binary distribution for Y and the log link function are specified in the model statement. The log link is used because the outcome is not rare.*/
/*bootstrap confidence intervals are specified using 1000 bootstraps and a specified random seed.*/
Title 'Single mediator with baseline covariate';
proc causalmed data=use.data pall alpha = .05;
class X Y/descending;
model Y = X|M / dist=bin link=log;
mediator M = X;
covar C;
bootstrap CI (all) nboot = 1000 seed = 08012019;
run;
quit;
Rights and permissions
About this article
Cite this article
Smyth, H.L., Pitpitan, E.V., MacKinnon, D.P. et al. Assessing Potential Outcomes Mediation in HIV Interventions. AIDS Behav 25, 2441–2454 (2021). https://doi.org/10.1007/s10461-021-03207-x
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10461-021-03207-x