Effect Estimates in Randomized Trials and Observational Studies: Comparing Apples With Apples

Sara Lodi (1) , Andrew Phillips (2) , Jens Lundgren (3, 4) , Roger Logan (5) , Shweta Sharma (6) , Stephen Cole , Abdel Babiker (2) , Matthew Law (7) , Haitao Chu (6) , Dana Byrne (8) , Andrzej Horban (9) , Jonathan A. C. Sterne (10, 11) , Kholoud Porter (2) , Caroline Sabin (2) , Dominique Costagliola (12, 13) , Sophie Abgrall (12, 13, 14) , John Gill (15) , Giota Touloumi (16) , Antonio Pacheco , Ard van Sighem (17) , Peter Reiss (17, 18, 19) , Heiner C. Bucher (20, 21) , Alexandra Montoliu Giménez (22) , Inmaculada Jarrin (23) , Linda Wittkop (24, 25, 26) , Laurence Meyer (27, 28, 29) , Santiago Perez-Hoyos , Amy Justice (30) , James D. Neaton (6) , Miguel Hernán (5)

1 BUSM - Boston University School of Medicine
2 UCL - University College of London [London]
3 Rigshospitalet [Copenhagen]
4 UCPH - University of Copenhagen = Københavns Universitet
5 Harvard T.H. Chan School of Public Health
6 University of Minnesota Medical School
7 UNSW - The Kirby Institute for Infection and Immunity in Society
8 CMSRU - Cooper Medical School of Rowan University [Camden]
9 Medical University of Warsaw - Poland
10 UNC - University of North Carolina [Chapel Hill]
11 University of Bristol [Bristol]
12 iPLESP - Institut Pierre Louis d'Epidémiologie et de Santé Publique
13 SU - Sorbonne Université
14 Service de médecine interne, immunologie clinique [Béclère]
15 University of Calgary
16 University of Athens Medical School [Athens]
17 Stichting HIV Monitoring [Amsterdam]
18 VU University Medical Center [Amsterdam]
19 Amsterdam Institute for Global Health & Development [Amsterdam, The Netherlands]
20 University Hospital Basel [Basel]
21 Unibas - Université de Bâle = University of Basel = Basel Universität
22 CEEISCAT - Centre d'Estudis Epidemiològics sobre les Infeccions de Transmissió Sexual i Sida de Catalunya
23 ISC - Instituto de Salud Carlos III [Madrid]
24 Institut de Santé Publique, d'Épidémiologie et de Développement (ISPED)
25 BPH - Bordeaux population health
26 UB - Université de Bordeaux
27 CHU Bordeaux
28 IRIS - Institut de Recherche Interdisciplinaire sur les enjeux Sociaux - sciences sociales, politique, santé
29 UP11 - Université Paris-Sud - Paris 11
30 YSM - Yale School of Medicine [New Haven, Connecticut]

Effect estimates from randomized trials and observational studies might not be directly comparable because of differences in study design, other than randomization, and in data analysis. We propose a 3-step procedure to facilitate meaningful comparisons of effect estimates from randomized trials and observational studies: 1) harmonization of the study protocols (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 for the human immunodeficiency virus from the Strategic Timing of Antiretroviral Therapy (START) randomized trial and the observational HIV-CAUSAL Collaboration.

Mots clés: antiretroviral initiation; causal inference; per-protocol effect; target trial

Domaines: Sciences du Vivant [q-bio]; Médecine humaine et pathologie

Sara Lodi, Andrew Phillips, Jens Lundgren, Roger Logan, Shweta Sharma, et al.. Effect Estimates in Randomized Trials and Observational Studies: Comparing Apples With Apples. American Journal of Epidemiology, 2019, 188 (8), pp.1569-1577. ⟨10.1093/aje/kwz100⟩. ⟨hal-02481150⟩