Format

Send to

Choose Destination
J Clin Epidemiol. 2016 Aug;76:183-92. doi: 10.1016/j.jclinepi.2016.01.030. Epub 2016 Feb 4.

The occurrence of Simpson's paradox if site-level effect was ignored in the TREAT Asia HIV Observational Database.

Author information

1
The Kirby Institute, UNSW Australia, Sydney, NSW 2052, Australia. Electronic address: ajiamsakul@kirby.unsw.edu.au.
2
HIV-NAT, The Thai Red Cross AIDS Research Centre, Bangkok, Thailand; Department of Global Health, Academic Medical Center, University of Amsterdam, Amsterdam Institute for Global Health and Development, Amsterdam, The Netherlands.
3
YRGCARE Medical Centre, Chennai, India.
4
Research Institute for Tropical Medicine, Manila, Philippines.
5
Research Institute for Health Sciences, Chiang Mai, Thailand.
6
HIV-NAT, The Thai Red Cross AIDS Research Centre, Bangkok, Thailand.
7
Beijing Ditan Hospital, Capital Medical University, Beijing, China.
8
National Center for HIV/AIDS, Dermatology & STDs, Phnom Penh, Cambodia.
9
Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand.
10
The Kirby Institute, UNSW Australia, Sydney, NSW 2052, Australia.

Abstract

OBJECTIVES:

In multisite human immunodeficiency virus (HIV) observational cohorts, clustering of observations often occurs within sites. Ignoring clustering may lead to "Simpson's paradox" (SP) where the trend observed in the aggregated data is reversed when the groups are separated. This study aimed to investigate the SP in an Asian HIV cohort and the effects of site-level adjustment through various Cox regression models.

STUDY DESIGN AND SETTING:

Survival time from combination antiretroviral therapy (cART) initiation was analyzed using four Cox models: (1) no site adjustment; (2) site as a fixed effect; (3) stratification through site; and (4) shared frailty on site.

RESULTS:

A total of 6,454 patients were included from 23 sites in Asia. SP was evident in the year of cART initiation variable. Model (1) shows the hazard ratio (HR) for years 2010-2014 was higher than the HR for 2006-2009, compared to 2003-2005 (HR = 0.68 vs. 0.61). Models (2)-(4) consistently implied greater improvement in survival for those who initiated in 2010-2014 than 2006-2009 contrasting findings from model (1). The effects of other significant covariates on survival were similar across four models.

CONCLUSIONS:

Ignoring site can lead to SP causing reversal of treatment effects. Greater emphasis should be made to include site in survival models when possible.

KEYWORDS:

Clustering; Cohort; Cox; Human immunodeficiency virus; Simpson's paradox; Yule-Simpson

PMID:
26854260
PMCID:
PMC4974156
DOI:
10.1016/j.jclinepi.2016.01.030
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Elsevier Science Icon for PubMed Central
Loading ...
Support Center