On the logic of collapsibility for causal effect measures

Liu et al. (2020) discuss the relation between efficacy measures within subgroups and efficacy measures on the population level, which can be obtained by merging the subgroups. They come to the conclusion that neither odds ratios (for binary endpoints) nor hazard ratios (for time-to-event endpoints)...

Full description

Bibliographic Details
Published in:Biometrical Journal
Main Authors: Didelez, Vanessa, Stensrud, Mats Julius
Language:English
Published: 2021
Subjects:
Online Access:https://repository.publisso.de/resource/frl:6433311
https://doi.org/10.1002/bimj.202000305
id ftleibnizopen:oai:oai.leibnizopen.de:G0CQhIgBdbrxVwz6En8q
record_format openpolar
spelling ftleibnizopen:oai:oai.leibnizopen.de:G0CQhIgBdbrxVwz6En8q 2023-06-18T03:40:48+02:00 On the logic of collapsibility for causal effect measures Didelez, Vanessa Stensrud, Mats Julius 2021 https://repository.publisso.de/resource/frl:6433311 https://doi.org/10.1002/bimj.202000305 eng eng CC BY 4.0 Biometrical journal, 64(2):235-242 Causal inference Survival analysis Collapsibility Confounding 2021 ftleibnizopen https://doi.org/10.1002/bimj.202000305 2023-06-04T23:12:49Z Liu et al. (2020) discuss the relation between efficacy measures within subgroups and efficacy measures on the population level, which can be obtained by merging the subgroups. They come to the conclusion that neither odds ratios (for binary endpoints) nor hazard ratios (for time-to-event endpoints) are suitable measures of efficacy in this context. This insight is not new, and more general settings have been considered previously (Daniel, Zhang, & Farewell, 2020; Greenland & Pearl, 2011; Greenland, Robins, & Pearl, 1999; Huitfeldt, Stensrud, & Suzuki, 2019; Martinussen & Vansteelandt, 2013; Pang, Kaufman, & Platt, 2013; Sjölander, Dahlqwist, & Zetterqvist, 2016). While we largely agree with their conclusion, we do so for different reasons and would like to point out a number of important subtleties that have perhaps not been appreciated by Liu et al. (2020). These should be carefully understood to avoid any further misleading interpretations. In particular, we want to emphasise, like many before, that confounding and non-collapsibility are separate issues (Didelez et al., 2010; Greenland, 1996; Greenland & Pearl, 2011; Greenland et al., 1999; Pand, Kaufman, & Platt, 2013; Pang et al., 2013; Shrier & Pang, 2015); to cite Greenland (2011): ‘confounding may occur with or without non-collapsibility, and non-collapsibility may occur with or without confounding’. Moreover, in view of patients and investigators preferring contrasts in terms of absolute risks (Murray, Caniglia, Swanson, Hernández-Díaz, & Hernán, 2018), we are sceptical about the emphasis on relative median survival time proposed in Liu et al. (2020). Other/Unknown Material Greenland LeibnizOpen (The Leibniz Association) Greenland Hernán ENVELOPE(-62.867,-62.867,-64.900,-64.900) Huitfeldt ENVELOPE(-64.754,-64.754,-65.993,-65.993) Biometrical Journal 64 2 235 242
institution Open Polar
collection LeibnizOpen (The Leibniz Association)
op_collection_id ftleibnizopen
language English
topic Causal inference
Survival analysis
Collapsibility
Confounding
spellingShingle Causal inference
Survival analysis
Collapsibility
Confounding
Didelez, Vanessa
Stensrud, Mats Julius
On the logic of collapsibility for causal effect measures
topic_facet Causal inference
Survival analysis
Collapsibility
Confounding
description Liu et al. (2020) discuss the relation between efficacy measures within subgroups and efficacy measures on the population level, which can be obtained by merging the subgroups. They come to the conclusion that neither odds ratios (for binary endpoints) nor hazard ratios (for time-to-event endpoints) are suitable measures of efficacy in this context. This insight is not new, and more general settings have been considered previously (Daniel, Zhang, & Farewell, 2020; Greenland & Pearl, 2011; Greenland, Robins, & Pearl, 1999; Huitfeldt, Stensrud, & Suzuki, 2019; Martinussen & Vansteelandt, 2013; Pang, Kaufman, & Platt, 2013; Sjölander, Dahlqwist, & Zetterqvist, 2016). While we largely agree with their conclusion, we do so for different reasons and would like to point out a number of important subtleties that have perhaps not been appreciated by Liu et al. (2020). These should be carefully understood to avoid any further misleading interpretations. In particular, we want to emphasise, like many before, that confounding and non-collapsibility are separate issues (Didelez et al., 2010; Greenland, 1996; Greenland & Pearl, 2011; Greenland et al., 1999; Pand, Kaufman, & Platt, 2013; Pang et al., 2013; Shrier & Pang, 2015); to cite Greenland (2011): ‘confounding may occur with or without non-collapsibility, and non-collapsibility may occur with or without confounding’. Moreover, in view of patients and investigators preferring contrasts in terms of absolute risks (Murray, Caniglia, Swanson, Hernández-Díaz, & Hernán, 2018), we are sceptical about the emphasis on relative median survival time proposed in Liu et al. (2020).
author Didelez, Vanessa
Stensrud, Mats Julius
author_facet Didelez, Vanessa
Stensrud, Mats Julius
author_sort Didelez, Vanessa
title On the logic of collapsibility for causal effect measures
title_short On the logic of collapsibility for causal effect measures
title_full On the logic of collapsibility for causal effect measures
title_fullStr On the logic of collapsibility for causal effect measures
title_full_unstemmed On the logic of collapsibility for causal effect measures
title_sort on the logic of collapsibility for causal effect measures
publishDate 2021
url https://repository.publisso.de/resource/frl:6433311
https://doi.org/10.1002/bimj.202000305
long_lat ENVELOPE(-62.867,-62.867,-64.900,-64.900)
ENVELOPE(-64.754,-64.754,-65.993,-65.993)
geographic Greenland
Hernán
Huitfeldt
geographic_facet Greenland
Hernán
Huitfeldt
genre Greenland
genre_facet Greenland
op_source Biometrical journal, 64(2):235-242
op_rights CC BY 4.0
op_doi https://doi.org/10.1002/bimj.202000305
container_title Biometrical Journal
container_volume 64
container_issue 2
container_start_page 235
op_container_end_page 242
_version_ 1769006100759183360