Patient portals are digital tools intended to enhance patient and caregiver engagement with healthcare systems. Disparities in portal activation and use among high-risk groups, including hospitalized children, could limit the potential benefits of these tools. The study aimed to identify disparities in portal activation and use among pediatric inpatients and to examine associations between portal activity and measures of acute care utilization.
The investigators conducted a cross-sectional analysis of patients admitted to two children’s hospitals from 2022 to 2024. The cohort included 40,371 hospitalized patients. The authors first described unadjusted patterns of portal activation and use. They then applied multivariable regression models to examine associations between portal activation and use (before, during, and after hospitalization) and a set of predictors that included patient-level characteristics, household-level factors, and population-level variables. Outcomes of interest in the models included critical care admissions, hospital length of stay (LOS), and 30-day readmissions.
In this sample of pediatric inpatients, 93% of patients had activated portal accounts. Despite this high overall activation rate, the analysis identified persistent disparities in both activation and usage across several sociodemographic domains. Specifically, patients who identified as non-Hispanic Black, those with public insurance, patients from households with a preferred language other than English, and those residing in lower-opportunity neighborhoods had significantly lower odds of portal activation and lower likelihood of portal use before, during, and after hospitalization. The study reported these associations after adjustment for other measured factors in multivariable models.
The study examined relationships between timing of portal use and acute care outcomes. Notable findings included:
Prehospital portal use was associated with decreased odds of critical care admission; the odds ratio reported was 0.64 with a 95% confidence interval of 0.58 to 0.70.
Prehospital portal use was also associated with a shorter hospital length of stay; the incident rate ratio reported was 0.74 with a 95% confidence interval of 0.69 to 0.79.
There was no observed difference in 30-day readmissions associated with portal use.
The authors characterize the relationships between portal activity and acute care utilization as complex and note that these associations merit further investigation.
Although overall portal activation in this multicenter pediatric inpatient sample was high (93%), important disparities in both activation and longitudinal use persisted by race/ethnicity, insurance status, household language preference, and neighborhood opportunity. The observed association of prehospital portal use with lower odds of critical care admission and shorter LOS suggests a potential relationship between prior engagement with the portal and downstream acute care measures, but causality cannot be inferred from this cross-sectional analysis.
Based on the study findings, the authors recommend focusing efforts to promote equitable portal usage on families who are publicly insured, who prefer languages other than English in the household, and who live in lower-opportunity neighborhoods. They also emphasize that the links between portal activity and acute care utilization are complex and require further study to understand mechanisms and to guide interventions.
The study provides population-level associations from a large cohort of hospitalized children at two institutions over a defined period. The findings describe disparities in digital tool adoption and correlations between portal engagement and certain in-hospital outcomes. The abstract reports effect estimates for two outcomes (critical care admission and LOS) and reports no difference in 30-day readmissions; additional methodological details, covariates included in models, and potential confounders are reported in the full article but are not included in the abstract. Details not reported in the abstract (for example, precise variable definitions, model specifications beyond the reported estimates, and sensitivity analyses) cannot be inferred and would require review of the full manuscript.
Health systems implementing or promoting patient portal tools should be aware that high activation rates do not eliminate disparities in use. Targeted strategies to improve equitable engagement should prioritize families who are publicly insured, who prefer non-English household languages, and who live in lower-opportunity neighborhoods. Further research is needed to clarify whether increased portal engagement causally influences acute care utilization and to identify effective interventions to reduce disparities in digital engagement.