Differential Impact of Virtual Family-Centered Rounds in the Neonatal Intensive Care Unit by Social Factors: A Post Hoc Subgroup Analysis

<b>Background:</b> Barriers to attending family-centered rounds (FCR) exist for socially disadvantaged families. Using telehealth to conduct virtual FCR could potentially promote equitable parent/guardian FCR access. The objective of this work was to assess whether the effects of a virtual FCR intervention on parent FCR attendance varied by subgroups defined by social factors. <b>Methods:</b> We conducted a post hoc analysis of a randomized controlled trial of virtual FCR in the neonatal intensive care unit. Parents of intervention arm infants were invited to participate in virtual FCR plus usual care; control arm infants received usual care. Participants were analyzed according to the assigned group and by race/ethnicity, insurance, mother's education, and neighborhood health conditions. We used Poisson regression to estimate and compare FCR parent attendance rates. Heterogeneity of intervention effects was assessed using interaction terms to evaluate the relative benefit of the intervention in increasing parent FCR attendance. <b>Results:</b> We included all enrolled trial subjects (74 intervention, 36 control). Intervention arm infants had 3.36 (95% confidence interval [CI]: 2.66-4.23) times the FCR parent attendance rate of subjects in the control arm. Compared with the corresponding reference subgroup, intervention benefits were 2.15 times (95% CI: 1.30-3.56) better for racial/ethnic minorities, 3.08 times (95% CI: 1.59-5.95) better for those with private insurance, 2.68 times (95% CI: 1.12-6.40) better for those whose mother reported no college education, and 4.14 times (95% CI: 2.07-8.25) better for those from a neighborhood with worse health conditions. <b>Conclusions:</b> Virtual FCR improved parent FCR attendance overall, with even greater benefits for certain subgroups. Further research is needed to mitigate the differential benefit demonstrated for privately insured subjects.