JMIR Publications

ABSTRACT

Background:

The US healthcare system faces escalating costs, increasing emphasis on patient autonomy, and a regulatory shift toward patient-centered care and patient-reported outcomes (PROs). Leveraging PROs to support shared decision-making (SDM) has potential to improve outcomes and reduce healthcare utilization for patients with advanced chronic conditions.

Objective:

To evaluate the impact of a PRO-based clinical dashboard on healthcare utilization among patients with advanced cancer and chronic kidney disease (CKD).

Methods:

We conducted a quasi-experimental, propensity score-adjusted difference-in-differences analysis using observational data from June 2020 to January 2022. PRO clinical dashboards were co-designed with patients, clinicians, and stakeholders within the Northwestern Medicine PRO system to enhance SDM. The intervention group included adults with advanced cancer or CKD who used the dashboard, while the comparison group did not. Primary outcomes were unplanned all-cause hospital admissions, potentially avoidable all-cause emergency department (ED) use, all-cause excess days in acute care within 30 days post-discharge, and 7-day readmissions. Secondary outcomes included advanced cancer-specific metrics (e.g., hospital/ED use during outpatient chemotherapy, triage clinic use, advance directive completion, and hospice use) and CKD-specific metrics (e.g., CKD-related hospital/ED use and disease progression).

Results:

For advanced cancer patients, dashboard use was significantly associated with reduced all-cause hospital admissions (b = -0.093; 95% CI, -0.181, -0.014) and decreased hospital admissions and emergency department (ED) visits for those receiving outpatient chemotherapy (b = -0.0879; 95% CI, -0.161, -0.012). However, dashboard use was associated with a lower likelihood of completing advanced directives (b = -0.025; 95% CI, -0.049, -0.003) and higher likelihood of 7-day hospital readmissions (b = 0.044; 95% CI, 0.011, 0.078), with 93% of the post-intervention readmissions being planned. There were no differences in excess acute care days or oncology triage clinic use. Among CKD patients, dashboard use was not associated with significant differences in admissions, excess days in acute care, 7-day readmissions, CKD-related ED or inpatient use, or CKD progression.

Conclusions:

The use of a co-designed, shared decision support dashboard was associated with fewer hospital admissions and ED visits among advanced cancer patients but was not associated with reduced utilization of high-cost services among CKD patients. Further research is needed to better understand the strengths and limitations of SDM interventions to improve outcomes for patients with advanced illness.