The Hospital to Home Outcomes (H2O) trials examined the effectiveness of postdischarge nurse support on reuse after pediatric discharge. Unexpectedly, children randomly assigned to a nurse visit had higher rates of reuse than those in the control group. Participants in randomized control trials are heterogeneous. Thus, it is possible that the effect of the intervention differed across subgroups (ie, heterogeneity of treatment effect [HTE]). We sought to determine if different subgroups responded differently to the interventions.
The H2O trial is a randomized controlled trial comparing standard hospital discharge processes with a nurse home visit within 96 hours of discharge. The second trial, H2O II, was similar, except the tested intervention was a postdischarge nurse phone call. For the purposes of the HTE analyses, we examined our primary trial outcome measure: a composite of unplanned 30-day acute health care reuse (unplanned readmission or emergency department or urgent care visit). We identified subgroups of interest before the trials related to (1) financial strain, (2) primary care access, (3) insurance, and (4) medical complexity. We used logistic regression modeling with an interaction term between subgroup and treatment group (intervention or control).
For the phone call trial (H2O II), financial strain significantly modified the effect of the intervention such that the subgroup of children with high financial strain who received the intervention experienced more reuse than their control counterparts.
In HTE analyses of 2 randomized controlled trials, only financial strain significantly modified the nurse phone call. A family’s financial resources may affect the utility of postdischarge support.