Outcomes Associated With Remote Patient Monitoring for Poor Weight Gain

Remote patient monitoring (RPM) uses digital technology to transmit patient data to a health care team.1 RPM empowers caregivers to participate in care at a convenient time and place, minimizing barriers related to work, childcare, and transportation.We launched an RPM program for infants after hospitalization for poor weight gain (PWG),2 providing a hospital-grade scale and application (app) to submit weekly weights and questionnaire responses related to feeding difficulties and food insecurity. RPM nurses conduct in-person education and enrollment at bedside before discharge home. The program allows for facilitated communication between the caregiver, a centralized team of RPM nurses, and the primary care provider. We have described this program in detail in prior work.2,3 Our objective was to examine 12-month outcomes of infants enrolled in RPM for PWG.We conducted a retrospective cohort study at a free-standing children’s hospital with a mature RPM infrastructure.3 The RPM group included infants younger than age 2 years who attended one of our primary care clinics (90% non-Hispanic, 70% Black, and 89% publicly or uninsured), were hospitalized for PWG, had an English-speaking caregiver, and were enrolled in RPM (January 2021 to February 2023). The comparator pre-RPM group included all of the same inclusion criteria but before the RPM program existed (January 2016 to December 2020). Infants who were never followed up with our primary care after discharge or had their feeds managed by a specialty clinic (gastroenterology or neonatology) were excluded from both groups. Data were collected via structured chart review. We reported days enrolled in RPM. Use outcomes included 12-month primary care visits, percentage of recommended well-child visits (WCVs) attended using the American Academy of Pediatrics schedule,4 and 12-month readmissions related to PWG. We examined a 12-month change in z score for weight, calculated using admission z score and z score at approximately 12 months after hospitalization.5 Child Protective Services (CPS) involvement was a dichotomous yes or no for any involvement documented in the electronic health record during the 12 months after hospitalization. Descriptive and bivariate statistics were used to describe differences between the 2 groups. The Institutional Review Board deemed the study exempt.There were 39 infants in RPM and 40 infants in pre-RPM (Table 1). Three eligible infants had enrollment failure during admission (Figure 1). Both groups were predominantly Black (∼75%) and hospitalized for nonorganic causes of PWG with similar admission weight z scores and age.RPM infants attended more recommended WCVs (80%) compared with pre-RPM (50%) (P < .01). Readmissions and change in weight z score were similar in both groups. In the RPM group, 28% of infants had CPS involvement in the 12 months after hospitalization compared with 53% of the pre-RPM group (P = .03).A postdischarge RPM program for infants with PWG was associated with improved attendance of WCVs, fewer outpatient ill visits (eg, weight checks), and a more than 40% reduction in CPS involvement. To our knowledge, no other pediatric RPM programs have been associated with similar outcomes. This RPM program was used as a tool to improve care for families at risk of disconnect from the medical home after hospitalization for PWG. Given the known association between Black maternal race and CPS involvement,6 the findings in this study underscore the potential value of leveraging RPM to improve health inequities.In addition to having more frequent, convenient touchpoints with the care team via texting in the RPM app, we hypothesize that by decreasing the burden of follow-up weight checks, families were able to shift their health care bandwidth to higher-value visits, like WCVs. Still, many enrolled families struggled with engagement with RPM, and despite frequent, structured contacts from the RPM nurses, several patients were discharged for lack of engagement, which we have described in an earlier manuscript.2 We suspect that methods such as user-centered design and coproduction for future RPM programs may help overcome challenges with engagement. One study limitation is that this was conducted at a single site with a mature pediatric RPM infrastructure and might not be easily replicated. Additionally, we used a comparator group from a different time period, did not control for confounders, and cannot infer causality; however, the characteristics between the cohorts were similar and support the comparison. Future work should prioritize digital health interventions that provide more direct access to the health care team for populations at risk for health inequities.