Managing confounding and effect modification in pediatric/congenital interventional cardiology research

Abstract

Measuring the effect of a treatment on outcomes is an important goal for research in pediatric/congenital interventional cardiology. The breadth of anatomic and physiologic variations, patient ages, and genetic syndromes and noncardiac comorbid conditions all represent sources of potential confounding and effect modification that are major obstacles to this goal. If not accounted for, these factors can obscure the “true” treatment effect and lead to spurious conclusions about the relative efficacy and/or safety of therapies. In this review, we discuss the importance of confounding and effect modification in pediatric/congenital interventional cardiology research. We define these terms and discuss strategies (both in study design and data analysis) to mitigate error introduced by confounding and effect modification. The importance of confounding by indication in pediatric/congenital cardiology is discussed along with specific methods to address it.

Incidence, Management, and Outcomes of Pulmonary Embolism at Tertiary Pediatric Hospitals in the United States

“…However, potentially important clinical covariates are not available in an administrative database such as PHIS, 35 and the potential for unmeasured confounding must be acknowledged. Although there are methods to overcome unmeasured confounding, 36 this was beyond the scope of this study. Similarly, it was difficult to ascertain nosocomial PE, which may represent a clinically distinct entity from PE present on admission.…”

Incorporating economic analysis in interventional cardiology research

“…Finally, cost studies are not immune to issues of bias and confounding (including confounding by indication) 78 . Authors should describe how these issues are addressed (e.g., randomization, propensity scores, etcetera) 78 …”

“…78 Authors should describe how these issues are addressed (e.g., randomization, propensity scores, etcetera). 78…”

“…Confounding by indication is highly likely in interventional cardiology studies, so simple adjustment for measurable confounders may not be sufficient. 78 In these cases, advanced techniques using propensity scores, 15,16,61,79,80 inverse probability weighting, 81,82 or instrumental variables 83 to mimic randomization (so-called "pseudorandomization") may be necessary to overcome these limitations.…”

“…A comprehensive analysis will evaluate the degree to which these decisions introduce bias (usually through sensitivity analyses). Finally, cost studies are not immune to issues of bias and confounding (including confounding by indication) 78 . Authors should describe how these issues are addressed (e.g., randomization, propensity scores, etcetera) 78 …”