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

Sara Lodi; Andrew Phillips; Jens Lundgren; Roger Logan; Shweta Sharma; Stephen R Cole; Abdel  Babiker; Matthew Law; Haitao Chu; Dana Byrne; Andrzej Horban; Jonathan Sterne; Kholoud Porter; Caroline A Sabin; Dominique Costagliola; Sophie Abgrall; Michael Gill; Giota Touloumi; Antonio Guilherme Pacheco; Ard van Sighem; Peter Reiss; Heiner C Bucher; Alexandra Giménez; Inmaculada Jarrin; Linda Wittkop; Laurence Meyer; Santiago Pérez-Hoyos; Amy Justice; James D.  Neaton; Miguel A Hernán

doi:10.1093/aje/kwz100

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

Sara Lodi, Andrew Phillips, Jens Lundgren, Roger Logan, Shweta Sharma, Stephen R Cole, Abdel Babiker, Matthew Law, Haitao Chu, Dana Byrne, Andrzej Horban, Jonathan Sterne, Kholoud Porter, Caroline A Sabin, Dominique Costagliola, Sophie Abgrall, Michael Gill, Giota Touloumi, Antonio Guilherme Pacheco, Ard van SighemPeter Reiss, Heiner C Bucher, Alexandra Giménez, Inmaculada Jarrin, Linda Wittkop, Laurence Meyer, Santiago Pérez-Hoyos, Amy Justice, James D. Neaton, Miguel A Hernán

Research output: Contribution to journal › Article (Academic Journal) › peer-review

86 Citations (Scopus)

133 Downloads (Pure)

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 (ART) in HIV-positive individuals from the START randomized trial and the observational HIV-CAUSAL Collaboration.

Original language	English
Number of pages	9
Journal	American Journal of Epidemiology
Early online date	7 May 2019
DOIs	https://doi.org/10.1093/aje/kwz100
Publication status	E-pub ahead of print - 7 May 2019

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1093/aje/kwz100

Full-text PDF (accepted author manuscript)
This is the author accepted manuscript (AAM). The final published version (version of record) is available online via Oxford University Press at https://academic.oup.com/aje/advance-article/doi/10.1093/aje/kwz100/5486454 . Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 347 KB

Handle.net

Persistent link

Embargoed Document

Full-text PDF (proof manuscript)
Proof, 1.11 MB
Request copy

Cite this

Lodi, S., Phillips, A., Lundgren, J., Logan, R., Sharma, S., Cole, S. R., Babiker, A., Law, M., Chu, H., Byrne, D., Horban, A., Sterne, J., Porter, K., Sabin, C. A., Costagliola, D., Abgrall, S., Gill, M., Touloumi, G., Pacheco, A. G., ... Hernán, M. A. (2019). Effect Estimates in Randomized Trials and Observational Studies: Comparing Apples With Apples. American Journal of Epidemiology. Advance online publication. https://doi.org/10.1093/aje/kwz100

@article{8c77389de81443f08d6d1ecb2b558b33,

title = "Effect Estimates in Randomized Trials and Observational Studies: Comparing Apples With Apples",

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 (ART) in HIV-positive individuals from the START randomized trial and the observational HIV-CAUSAL Collaboration. ",

author = "Sara Lodi and Andrew Phillips and Jens Lundgren and Roger Logan and Shweta Sharma and Cole, {Stephen R} and Abdel Babiker and Matthew Law and Haitao Chu and Dana Byrne and Andrzej Horban and Jonathan Sterne and Kholoud Porter and Sabin, {Caroline A} and Dominique Costagliola and Sophie Abgrall and Michael Gill and Giota Touloumi and Pacheco, {Antonio Guilherme} and {van Sighem}, Ard and Peter Reiss and Bucher, {Heiner C} and Alexandra Gim{\'e}nez and Inmaculada Jarrin and Linda Wittkop and Laurence Meyer and Santiago P{\'e}rez-Hoyos and Amy Justice and Neaton, {James D.} and Hern{\'a}n, {Miguel A}",

year = "2019",

month = may,

day = "7",

doi = "10.1093/aje/kwz100",

language = "English",

journal = "American Journal of Epidemiology",

issn = "0002-9262",

publisher = "Oxford University Press",

}

Lodi, S, Phillips, A, Lundgren, J, Logan, R, Sharma, S, Cole, SR, Babiker, A, Law, M, Chu, H, Byrne, D, Horban, A, Sterne, J, Porter, K, Sabin, CA, Costagliola, D, Abgrall, S, Gill, M, Touloumi, G, Pacheco, AG, van Sighem, A, Reiss, P, Bucher, HC, Giménez, A, Jarrin, I, Wittkop, L, Meyer, L, Pérez-Hoyos, S, Justice, A, Neaton, JD & Hernán, MA 2019, 'Effect Estimates in Randomized Trials and Observational Studies: Comparing Apples With Apples', American Journal of Epidemiology. https://doi.org/10.1093/aje/kwz100

TY - JOUR

T1 - Effect Estimates in Randomized Trials and Observational Studies

T2 - Comparing Apples With Apples

AU - Lodi, Sara

AU - Phillips, Andrew

AU - Lundgren, Jens

AU - Logan, Roger

AU - Sharma, Shweta

AU - Cole, Stephen R

AU - Babiker, Abdel

AU - Law, Matthew

AU - Chu, Haitao

AU - Byrne, Dana

AU - Horban, Andrzej

AU - Sterne, Jonathan

AU - Porter, Kholoud

AU - Sabin, Caroline A

AU - Costagliola, Dominique

AU - Abgrall, Sophie

AU - Gill, Michael

AU - Touloumi, Giota

AU - Pacheco, Antonio Guilherme

AU - van Sighem, Ard

AU - Reiss, Peter

AU - Bucher, Heiner C

AU - Giménez, Alexandra

AU - Jarrin, Inmaculada

AU - Wittkop, Linda

AU - Meyer, Laurence

AU - Pérez-Hoyos, Santiago

AU - Justice, Amy

AU - Neaton, James D.

AU - Hernán, Miguel A

PY - 2019/5/7

Y1 - 2019/5/7

N2 - 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 (ART) in HIV-positive individuals from the START randomized trial and the observational HIV-CAUSAL Collaboration.

AB - 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 (ART) in HIV-positive individuals from the START randomized trial and the observational HIV-CAUSAL Collaboration.

U2 - 10.1093/aje/kwz100

DO - 10.1093/aje/kwz100

M3 - Article (Academic Journal)

C2 - 31063192

SN - 0002-9262

JO - American Journal of Epidemiology

JF - American Journal of Epidemiology

ER -

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

Abstract

UN SDGs

Access to Document

Handle.net

Embargoed Document

Fingerprint

Cite this