[Skip to Content]
[Skip to Content Landing]
Comment & Response
December 2014

Surgery and Neurodevelopmental Impairment: Need for Time-Dependent Covariates to Correct for Confounding by Indication

Author Affiliations
  • 1Department of Newborn and Developmental Pediatrics, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
  • 2Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada
  • 3Department of Pediatrics, Mt Sinai Hospital, Toronto, Ontario, Canada
JAMA Pediatr. 2014;168(12):1168-1169. doi:10.1001/jamapediatrics.2014.1781

To the Editor We read with interest the study by Morriss et al1 and the accompanying editorial recently published in JAMA Pediatrics.2 We applaud their use of a propensity score modeling approach to answer the following question facing neonatologists and surgeons: is it the surgery or the underlying condition/status of the infant that is associated with neurodevelopmental impairment? However, we wish to highlight an important shortcoming in the authors’ attempt to reduce bias owing to confounding by indication.