Kazuki Yoshida1,2,3, Daniel H Solomon1,4, Sebastien Haneuse3, Seoyoung C Kim1,4, Elisabetta Patorno4, Sara K Tedeschi1, Houchen Lyu1, Sonia Hernández-Díaz2, Robert J Glynn3,4. 1. Division of Rheumatology, Immunology and Allergy, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts, USA. 2. Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA. 3. Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA. 4. Division of Pharmacoepidemiology and Pharmacoeconomics, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts, USA.
Abstract
PURPOSE: In observational research, equipoise concerns whether groups being compared are similar enough for valid inference. Empirical equipoise was previously proposed as a tool to assess patient similarity based on propensity scores (PS). We extended this work for multigroup observational studies. METHODS: We modified the tool to allow for multinomial exposures such that the proposed definition reduces to the original when there are only two groups. We illustrated how the tool can be used as a method to assess study design within three-group clinical examples. We then conducted three-group simulations to assess how the tool performed in a setting with residual confounding after PS weighting. RESULTS: In a clinical example based on rheumatoid arthritis, 44.5% of the sample fell within the region of empirical equipoise when considering first-line biologics, whereas 57.7% did so for second-line biologics, consistent with the expectation that a second-line design results in better equipoise. In a simulation where the unmeasured confounder had the same magnitude of association with the treatment as the measured confounders and a 25% greater association with the outcome, the tool crossed the proposed threshold for empirical equipoise at a residual confounding of 20% on the ratio scale. When the unmeasured variable had a twice larger association with treatment, the tool became less sensitive and crossed the threshold at a residual confounding of 30%. CONCLUSION: Our proposed tool may be useful in guiding cohort identification in multigroup observational studies, particularly with similar effects of unmeasured and measured covariates on treatment and outcome.
PURPOSE: In observational research, equipoise concerns whether groups being compared are similar enough for valid inference. Empirical equipoise was previously proposed as a tool to assess patient similarity based on propensity scores (PS). We extended this work for multigroup observational studies. METHODS: We modified the tool to allow for multinomial exposures such that the proposed definition reduces to the original when there are only two groups. We illustrated how the tool can be used as a method to assess study design within three-group clinical examples. We then conducted three-group simulations to assess how the tool performed in a setting with residual confounding after PS weighting. RESULTS: In a clinical example based on rheumatoid arthritis, 44.5% of the sample fell within the region of empirical equipoise when considering first-line biologics, whereas 57.7% did so for second-line biologics, consistent with the expectation that a second-line design results in better equipoise. In a simulation where the unmeasured confounder had the same magnitude of association with the treatment as the measured confounders and a 25% greater association with the outcome, the tool crossed the proposed threshold for empirical equipoise at a residual confounding of 20% on the ratio scale. When the unmeasured variable had a twice larger association with treatment, the tool became less sensitive and crossed the threshold at a residual confounding of 30%. CONCLUSION: Our proposed tool may be useful in guiding cohort identification in multigroup observational studies, particularly with similar effects of unmeasured and measured covariates on treatment and outcome.
Authors: Kazuki Yoshida; Sonia Hernández-Díaz; Daniel H Solomon; John W Jackson; Joshua J Gagne; Robert J Glynn; Jessica M Franklin Journal: Epidemiology Date: 2017-05 Impact factor: 4.822
Authors: Jasvinder A Singh; Kenneth G Saag; S Louis Bridges; Elie A Akl; Raveendhara R Bannuru; Matthew C Sullivan; Elizaveta Vaysbrot; Christine McNaughton; Mikala Osani; Robert H Shmerling; Jeffrey R Curtis; Daniel E Furst; Deborah Parks; Arthur Kavanaugh; James O'Dell; Charles King; Amye Leong; Eric L Matteson; John T Schousboe; Barbara Drevlow; Seth Ginsberg; James Grober; E William St Clair; Elizabeth Tindall; Amy S Miller; Timothy McAlindon Journal: Arthritis Care Res (Hoboken) Date: 2015-11-06 Impact factor: 4.794
Authors: Josef S Smolen; Robert Landewé; Johannes Bijlsma; Gerd Burmester; Katerina Chatzidionysiou; Maxime Dougados; Jackie Nam; Sofia Ramiro; Marieke Voshaar; Ronald van Vollenhoven; Daniel Aletaha; Martin Aringer; Maarten Boers; Chris D Buckley; Frank Buttgereit; Vivian Bykerk; Mario Cardiel; Bernard Combe; Maurizio Cutolo; Yvonne van Eijk-Hustings; Paul Emery; Axel Finckh; Cem Gabay; Juan Gomez-Reino; Laure Gossec; Jacques-Eric Gottenberg; Johanna M W Hazes; Tom Huizinga; Meghna Jani; Dmitry Karateev; Marios Kouloumas; Tore Kvien; Zhanguo Li; Xavier Mariette; Iain McInnes; Eduardo Mysler; Peter Nash; Karel Pavelka; Gyula Poór; Christophe Richez; Piet van Riel; Andrea Rubbert-Roth; Kenneth Saag; Jose da Silva; Tanja Stamm; Tsutomu Takeuchi; René Westhovens; Maarten de Wit; Désirée van der Heijde Journal: Ann Rheum Dis Date: 2017-03-06 Impact factor: 19.103
Authors: Thomas Frisell; Eva Baecklund; Karin Bengtsson; Daniela Di Giuseppe; Helena Forsblad-d'Elia; Johan Askling Journal: Ann Rheum Dis Date: 2017-12-13 Impact factor: 19.103