Sengwee Toh1, Susan Shetterly, John D Powers, David Arterburn. 1. *Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA †Institute for Health Research, Kaiser Permanente Colorado, Denver, CO ‡Group Health Research Institute, Seattle, WA.
Abstract
BACKGROUND: For privacy and practical reasons, it is sometimes necessary to minimize sharing of individual-level information in multisite studies. However, individual-level information is often needed to perform more rigorous statistical analysis. OBJECTIVES: To compare empirically 3 analytic methods for multisite studies that only require sharing of summary-level information to perform statistical analysis that have traditionally required access to detailed individual-level data from each site. RESEARCH DESIGN, SUBJECTS, AND MEASURES: We analyzed data from a 7-site study of bariatric surgery outcomes within the Scalable Partnering Network. We compared the long-term risk of rehospitalization between adjustable gastric banding and Roux-en-y gastric bypass procedures using a stratified analysis of propensity score (PS)-defined strata, a case-centered analysis of risk set data, and a meta-analysis of site-specific effect estimates. Their results were compared with the result from a pooled individual-level data analysis. RESULTS: The study included 1327 events (18.1%) among 7342 patients. The adjusted hazard ratio was 0.71 (95% CI, 0.59, 0.84) comparing adjustable gastric banding with Roux-en-y gastric bypass in the individual-level data analysis. The corresponding effect estimate was 0.70 (0.59, 0.83) in the PS-stratified analysis, 0.71 (0.59, 0.84) in the case-centered analysis, and 0.71 (0.60, 0.84) in both the fixed-effect and random-effects meta-analysis. CONCLUSIONS: In this empirical study, PS-stratified analysis, case-centered analysis, and meta-analysis produced results that are identical or highly comparable with the result from a pooled individual-level data analysis. These methods have the potential to be viable analytic alternatives when sharing of individual-level information is not feasible or not preferred in multisite studies.
BACKGROUND: For privacy and practical reasons, it is sometimes necessary to minimize sharing of individual-level information in multisite studies. However, individual-level information is often needed to perform more rigorous statistical analysis. OBJECTIVES: To compare empirically 3 analytic methods for multisite studies that only require sharing of summary-level information to perform statistical analysis that have traditionally required access to detailed individual-level data from each site. RESEARCH DESIGN, SUBJECTS, AND MEASURES: We analyzed data from a 7-site study of bariatric surgery outcomes within the Scalable Partnering Network. We compared the long-term risk of rehospitalization between adjustable gastric banding and Roux-en-y gastric bypass procedures using a stratified analysis of propensity score (PS)-defined strata, a case-centered analysis of risk set data, and a meta-analysis of site-specific effect estimates. Their results were compared with the result from a pooled individual-level data analysis. RESULTS: The study included 1327 events (18.1%) among 7342 patients. The adjusted hazard ratio was 0.71 (95% CI, 0.59, 0.84) comparing adjustable gastric banding with Roux-en-y gastric bypass in the individual-level data analysis. The corresponding effect estimate was 0.70 (0.59, 0.83) in the PS-stratified analysis, 0.71 (0.59, 0.84) in the case-centered analysis, and 0.71 (0.60, 0.84) in both the fixed-effect and random-effects meta-analysis. CONCLUSIONS: In this empirical study, PS-stratified analysis, case-centered analysis, and meta-analysis produced results that are identical or highly comparable with the result from a pooled individual-level data analysis. These methods have the potential to be viable analytic alternatives when sharing of individual-level information is not feasible or not preferred in multisite studies.
Authors: Xiaojuan Li; Bruce H Fireman; Jeffrey R Curtis; David E Arterburn; David P Fisher; Érick Moyneur; Mia Gallagher; Marsha A Raebel; W Benjamin Nowell; Lindsay Lagreid; Sengwee Toh Journal: Am J Epidemiol Date: 2019-04-01 Impact factor: 4.897
Authors: Christine Y Lu; Marc S Williams; Geoffrey S Ginsburg; Sengwee Toh; Jeff S Brown; Muin J Khoury Journal: Genet Med Date: 2017-08-10 Impact factor: 8.822
Authors: Kathleen M Mazor; Allison Richards; Mia Gallagher; David E Arterburn; Marsha A Raebel; W Benjamin Nowell; Jeffrey R Curtis; Andrea R Paolino; Sengwee Toh Journal: J Comp Eff Res Date: 2017-08-14 Impact factor: 1.744
Authors: Rosa Gini; Martijn Schuemie; Jeffrey Brown; Patrick Ryan; Edoardo Vacchi; Massimo Coppola; Walter Cazzola; Preciosa Coloma; Roberto Berni; Gayo Diallo; José Luis Oliveira; Paul Avillach; Gianluca Trifirò; Peter Rijnbeek; Mariadonata Bellentani; Johan van Der Lei; Niek Klazinga; Miriam Sturkenboom Journal: EGEMS (Wash DC) Date: 2016-02-08
Authors: Sengwee Toh; Laura J Rasmussen-Torvik; Emily E Harmata; Roy Pardee; Rosalinde Saizan; Elisha Malanga; Jessica L Sturtevant; Casie E Horgan; Jane Anau; Cheri D Janning; Robert D Wellman; R Yates Coley; Andrea J Cook; Anita P Courcoulas; Karen J Coleman; Neely A Williams; Kathleen M McTigue; David Arterburn; James McClay Journal: JMIR Res Protoc Date: 2017-12-05
Authors: Julie M Donohue; Marian P Jarlenski; Joo Yeon Kim; Lu Tang; Katherine Ahrens; Lindsay Allen; Anna Austin; Andrew J Barnes; Marguerite Burns; Chung-Chou H Chang; Sarah Clark; Evan Cole; Dushka Crane; Peter Cunningham; David Idala; Stefanie Junker; Paul Lanier; Rachel Mauk; Mary Joan McDuffie; Shamis Mohamoud; Nathan Pauly; Logan Sheets; Jeffery Talbert; Kara Zivin; Adam J Gordon; Susan Kennedy Journal: JAMA Date: 2021-07-13 Impact factor: 56.272