Skip to main content

COVID-19 Research Resources
A curated list of research resources around guidelines, policies, and procedures related to COVID-1, drawn from Harvard University, affiliated academic healthcare centers, and government funding agencies

COVID-19 Research Resources
A curated list of research resources around guidelines, policies, and procedures related to COVID-1, drawn from Harvard University, affiliated academic healthcare centers, and government funding agencies

Calendar

Biostatistics journal club: In-Hospital Mortality by Pump Type for Pediatric Extracorporeal Membrane Oxygenation Using a Multi-Center Registry Data – Is the Propensity Score and Outcome Model Correct? – September 16

Wednesday, September 16, 2020
1:00 pm – 2:00 pm
Zoom

Biostatistics journal club: In-Hospital Mortality by Pump Type for Pediatric Extracorporeal Membrane Oxygenation Using a Multi-Center Registry Data – Is the Propensity Score and Outcome Model Correct?

Edie Weller, PhD, Harvard Medical School, Boston Children’s Hospital, will evaluate the association of pump type with in-hospital mortality using data from a multi-center registry. Discussion will be based around what variables to include in propensity score and outcome model, and does the bootstrap approach to account for estimation of the inverse probability weights seem reasonable in the context of multiple imputation?

Registration required.

Edie Weller, PhD
Associate Professor of Pediatrics, Harvard Medical School; Director, Biostatistics and Research Design (BARD) in the Institutional Centers of Clinical and Translational Research (ICCTR), Boston Children’s Hospital

In-hospital mortality by pump type for pediatric extracorporeal membrane oxygenation using a multi-center registry data: Is the propensity score and outcome model correct?

Abstract
Presentation is a work in progress. Extracorporeal membrane oxygenation (ECMO) is a life support machine used for the management of children with cardiopulmonary failure. In this project, the goal is to evaluate the association of pump type with in-hospital mortality using data from a multi-center registry. Main analysis challenge is accounting for the imbalance of covariates by pump type. We used a) multiple imputation method to address missing data and b) inverse probability weights from a propensity score model in the outcome model with variance estimated using bootstrap. Results are combined across imputation data sets.

Discussion

  • What variables to include in propensity score and outcome model?
  • Does the bootstrap approach to account for estimation of the inverse probability weights seem reasonable in the context of multiple imputation?

Sign up to receive our newsletter: courses, funding, events, and resources.