Format

Send to

Choose Destination
Am J Epidemiol. 2019 May 7. pii: kwz100. doi: 10.1093/aje/kwz100. [Epub ahead of print]

Effect estimates in randomized trials and observational studies: comparing apples with apples.

Author information

1
Department of Biostatistics, Boston University School of Public Health, Boston, Massachusetts.
2
Institute for Global Health, University College London, United Kingdom.
3
Department of Infectious Diseases, Rigshospitalet, University of Copenhagen, Denmark.
4
Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts.
5
Division of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, Minnesota.
6
Medical Research Council, Clinical Trials Unit in University College London, London, United Kingdom.
7
The Kirby Institute, Sidney, Australia.
8
Division of Infectious Diseases, Department of Medicine, Cooper University Hospital, Cooper Medical School at Rowan University, New Jersey.
9
Medical University of Warsaw, Department for Adult's Infectious Diseases, Warsaw, Poland.
10
Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, North Carolina.
11
Department of Population Health Sciences, University of Bristol, Bristol, United Kingdom.
12
INSERM, Sorbonne Université, Institut Pierre Louis d'Épidémiologie et de Santé Publique (IPLESP), Paris, France.
13
AP-HP, Hôpital Antoine Béclère, Service de Médecine Interne, Clamart, France.
14
Southern Alberta Clinic, Calgary, Canada.
15
Department of Medicine, University of Calgary, Canada.
16
National and Kapodistrian University of Athens, Faculty of Medicine, Dept. of Hygiene, Epidemiology and Medical Statistics, Greece.
17
Programa de Computação Científica, Fundacao Oswaldo Cruz, Rio de Janeiro, Brasil.
18
Stichting HIV Monitoring, Amsterdam, the Netherlands.
19
Amsterdam University Medical Centres, University of Amsterdam, Department of Global Health and Division of Infectious Diseases, Amsterdam, the Netherlands.
20
Amsterdam Institute for Global Health and Development, and Amsterdam Public Health Research Institute, Amsterdam, the Netherlands.
21
Basel Institute for Clinical Epidemiology and Biostatistics, University Hospital Basel, University of Basel, Switzerland.
22
Centre for Epidemiological Studies on HIV/STI in Catalonia (CEEISCAT), Agència de Salut Pública de Catalunya (ASPC), Badalona, Spain.
23
Centro Nacional de Epidemiología, Instituto de Salud Carlos III, Madrid, Spain.
24
Univ. Bordeaux, ISPED, Inserm, Bordeaux Population Health Research Center, team MORPH3EUS, UMR 1219, CIC-EC 1401, F-33000 Bordeaux, France.
25
CHU de Bordeaux, Pôle de santé publique, Service d'information médicale, F-33000 Bordeaux, France.
26
Université Paris Sud, UMR 1018, le Kremlin Bicêtre, France.
27
Vall d'Hebrón Research Institute; Barcelona, Spain.
28
Yale University School of Medicine, New Haven, Connecticut.
29
Department of Biostatistics, Harvard T.H. Chan School of Public Health.
30
Harvard-MIT Division of Health Sciences and Technology, Boston, Massachusetts.

Abstract

Effect estimates from randomized trials and observational studies may not be directly comparable because of differences in study design, other than randomization, and in data analysis. We propose a three-step procedure to facilitate meaningful comparisons of effect estimates from randomized trials and observational studies: 1) harmonization of the study protocol (eligibility criteria, treatment strategies, outcome, start and end of follow-up, causal contrast) so that the studies target the same causal effect, 2) harmonization of the data analysis to estimate the causal effect, and 3) sensitivity analyses to investigate the impact of discrepancies that could not be accounted for in the harmonization process. To illustrate our approach, we compared estimates of the effect of immediate with deferred initiation of antiretroviral therapy in individuals positive to the human immunodeficiency virus from the START randomized trial and the observational HIV-CAUSAL Collaboration.

KEYWORDS:

antiretroviral initiation; causal inference; per-protocol; target trial

PMID:
31063192
PMCID:
PMC6670045
[Available on 2020-08-01]
DOI:
10.1093/aje/kwz100

Supplemental Content

Full text links

Icon for Silverchair Information Systems
Loading ...
Support Center