Commentary: Excitement at the interface of disciplines: The mean cumulative function

Inverse Probability Weighting to Estimate Exposure Effects on the Burden of Recurrent Outcomes in the Presence of Competing Events

Recurrent events-outcomes that an individual can experience repeatedly over the course of follow-up-are common in epidemiologic and health services research. Studies involving recurrent events often focus on time to first occurrence or on event rates, which assume constant hazards over time. In this paper, we contextualize recurrent event parameters of interest using counterfactual theory in a causal inference framework and describe an approach for estimating a target parameter referred to as the mean cumulative count. This approach leverages inverse probability weights to control measured confounding with an existing (and underutilized) nonparametric estimator of recurrent event burden first proposed by Dong et al. in 2015. We use simulations to demonstrate the unbiased estimation of the mean cumulative count using the weighted Dong-Yasui estimator in a variety of scenarios. The weighted Dong-Yasui estimator for the mean cumulative count allows researchers to use observational data to flexibly estimate and contrast the expected number of cumulative events experienced per individual by a given time point under different exposure regimens. We provide code to ease application of this method.

Recurrent Pulmonary Artery Interventions Following the Norwood Procedure Are Not Associated With Conduit Type

Given pulmonary artery interventions following the Norwood procedure can recur, the average number of occurrences per patient over time is likely more informative than the crude percentage of patients who required an intervention. Pulmonary artery intervention was defined as any surgical or catheter-based procedure after the Norwood procedure. The number of pulmonary artery interventions for patients with hypoplastic left heart syndrome were compared between patients with modified Blalock-Taussig Shunts (MBTS) and right ventricle-to-pulmonary artery conduits (RVPA) at a single institution from 2011 to 2018. The comparison was replicated using data from the Single Ventricle Reconstruction Trial (SVR), a nonoverlapping dataset. The mean number of pulmonary artery interventions per patient over time (mean cumulative function, MCF) is described using Nelson-Aalen estimates and compared using the pseudo-score test. The number of patients requiring intervention was compared using the chi-square test. Using our institutional dataset, the Norwood operation was performed on 117 patients (59 MBTS, 58 RVPA). In total, 73 patients had a pulmonary artery intervention, including 32 of 58 (55%) after MBTS and 41 of 59 (69%) after RVPA (P= 0.11). The MCF did not vary between cohorts (P = 0.55). Using the SVR trial dataset, 140 of 549 patients required pulmonary artery intervention, including 55 (21%) after MBTS and 85 (30%) after RVPA (P = 0.0090). The MCF did not vary between cohorts (P = 0.067). Although more patients with RVPA than MBTS require pulmonary artery interventions after the Norwood procedure, the MCFs are not different, which may be of greater importance to patients and families.