TY - JOUR
T1 - Stopping randomized trials early for benefit and estimation of treatment effects
T2 - Systematic review and meta-regression analysis
AU - Bassler, Dirk
AU - Briel, Matthias
AU - Montori, Victor M.
AU - Lane, Melanie
AU - Glasziou, Paul
AU - Zhou, Qi
AU - Heels-Ansdell, Diane
AU - Walter, Stephen D.
AU - Guyatt, Gordon H.
AU - Flynn, David N.
AU - Elamin, Mohamed B.
AU - Murad, Mohammad Hassan
AU - Abu Elnour, Nisrin O.
AU - Lampropulos, Julianna F.
AU - Sood, Amit
AU - Mullan, Rebecca J.
AU - Erwin, Patricia J.
AU - Bankhead, Clare R.
AU - Perera, Rafael
AU - Culebro, Carolina Ruiz
AU - You, John J.
AU - Mulla, Sohail M.
AU - Kaur, Jagdeep
AU - Nerenberg, Kara A.
AU - Schünemann, Holger
AU - Cook, Deborah J.
AU - Lutz, Kristina
AU - Ribic, Christine M.
AU - Vale, Noah
AU - Malaga, German
AU - Akl, Elie A.
AU - Ferreira-Gonzalez, Ignacio
AU - Alonso-Coello, Pablo
AU - Urrutia, Gerard
AU - Kunz, Regina
AU - Bucher, Heiner C.
AU - Nordmann, Alain J.
AU - Raatz, Heike
AU - Da Silva, Suzana Alves
AU - Tuche, Fabio
AU - Strahm, Brigitte
AU - Djulbegovic, Benjamin
AU - Adhikari, Neill K J
AU - Mills, Edward J.
AU - Gwadry-Sridhar, Femida
AU - Kirpalani, Haresh
AU - Soares, Heloisa P.
AU - Karanicolas, Paul J.
AU - Burns, Karen E A
AU - Vandvik, Per Olav
AU - Coto-Yglesias, Fernando
AU - Chrispim, Pedro Paulo M
AU - Ramsay, Tim
PY - 2010/3/24
Y1 - 2010/3/24
N2 - Context: Theory and simulation suggest that randomized controlled trials (RCTs) stopped early for benefit (truncated RCTs) systematically overestimate treatment effects for the outcome that precipitated early stopping. Objective: To compare the treatment effect from truncated RCTs with that from metaanalyses of RCTs addressing the same question but not stopped early (nontruncated RCTs) and to explore factors associated with overestimates of effect. Data Sources: Search of MEDLINE, EMBASE, Current Contents, and full-text journal content databases to identify truncatedRCTsupto January2007;search ofMEDLINE,Cochrane Database of Systematic Reviews, and Database of Abstracts of Reviews of Effects to identify systematic reviews from which individualRCTswere extractedupto January 2008. Study Selection: Selected studies were RCTs reported as having stopped early for benefit and matching nontruncated RCTs from systematic reviews. Independent reviewers with medical content expertise, working blinded to trial results, judged the eligibility of the nontruncated RCTs based on their similarity to the truncated RCTs. Data Extraction: Reviewers with methodological expertise conducted data extraction independently. Results: The analysis included 91 truncated RCTs asking 63 different questions and 424 matching nontruncated RCTs. The pooled ratio of relative risks in truncated RCTs vs matching nontruncated RCTs was 0.71 (95% confidence interval, 0.65-0.77). This difference was independent of the presence of a statistical stopping rule and the methodological quality of the studies as assessed by allocation concealment and blinding. Large differences in treatment effect size between truncated and nontruncated RCTs (ratio of relative risks <0.75) occurred with truncated RCTs having fewer than 500 events. In 39 of the 63 questions (62%), the pooled effects of the nontruncated RCTs failed to demonstrate significant benefit. Conclusions: Truncated RCTs were associated with greater effect sizes than RCTs not stopped early. This difference was independent of the presence of statistical stopping rules and was greatest in smaller studies.
AB - Context: Theory and simulation suggest that randomized controlled trials (RCTs) stopped early for benefit (truncated RCTs) systematically overestimate treatment effects for the outcome that precipitated early stopping. Objective: To compare the treatment effect from truncated RCTs with that from metaanalyses of RCTs addressing the same question but not stopped early (nontruncated RCTs) and to explore factors associated with overestimates of effect. Data Sources: Search of MEDLINE, EMBASE, Current Contents, and full-text journal content databases to identify truncatedRCTsupto January2007;search ofMEDLINE,Cochrane Database of Systematic Reviews, and Database of Abstracts of Reviews of Effects to identify systematic reviews from which individualRCTswere extractedupto January 2008. Study Selection: Selected studies were RCTs reported as having stopped early for benefit and matching nontruncated RCTs from systematic reviews. Independent reviewers with medical content expertise, working blinded to trial results, judged the eligibility of the nontruncated RCTs based on their similarity to the truncated RCTs. Data Extraction: Reviewers with methodological expertise conducted data extraction independently. Results: The analysis included 91 truncated RCTs asking 63 different questions and 424 matching nontruncated RCTs. The pooled ratio of relative risks in truncated RCTs vs matching nontruncated RCTs was 0.71 (95% confidence interval, 0.65-0.77). This difference was independent of the presence of a statistical stopping rule and the methodological quality of the studies as assessed by allocation concealment and blinding. Large differences in treatment effect size between truncated and nontruncated RCTs (ratio of relative risks <0.75) occurred with truncated RCTs having fewer than 500 events. In 39 of the 63 questions (62%), the pooled effects of the nontruncated RCTs failed to demonstrate significant benefit. Conclusions: Truncated RCTs were associated with greater effect sizes than RCTs not stopped early. This difference was independent of the presence of statistical stopping rules and was greatest in smaller studies.
UR - http://www.scopus.com/inward/record.url?scp=77949879943&partnerID=8YFLogxK
U2 - 10.1001/jama.2010.310
DO - 10.1001/jama.2010.310
M3 - Review article
C2 - 20332404
AN - SCOPUS:77949879943
SN - 0098-7484
VL - 303
SP - 1180
EP - 1187
JO - JAMA - Journal of the American Medical Association
JF - JAMA - Journal of the American Medical Association
IS - 12
ER -