MEDBRIDGE-Guided NCM/CHW Post-Discharge Support for High-Risk T2D Patients

This single-arm feasibility pilot evaluates a MEDBRIDGE-guided nurse case manager (NCM) and community health worker (CHW) post-discharge support intervention for high-risk patients with type 2 diabetes (T2D). MEDBRIDGE is an AI-driven risk stratification tool that integrates medication data, clinical factors, and social determinants of health from electronic health records to identify patients at elevated risk of HbA1c elevation, diabetes-related emergency department visits, and hospitalizations within 3 months post-discharge.

The intervention follows a four-phase workflow: (1) Risk Assessment, where the NCM reviews the daily MEDBRIDGE-generated high-risk patient list; (2) Initial Patient Contact, where the NCM/CHW team initiates contact after discharge to review medications and identify discrepancies; (3) Ongoing Support, where the CHW provides monthly check-ins to monitor adherence, address barriers, and coordinate with primary care providers; and (4) Transition to Routine Care, where the team facilitates handoff to the patient's primary care provider with a summary of activities and recommendations.

Primary outcomes include feasibility (recruitment rate, retention rate), patient acceptability (Acceptability of Intervention Measure), and intervention fidelity (NCM/CHW contact log adherence). Secondary outcomes include HbA1c change, diabetes-related emergency department visits, and diabetes-related hospitalizations within 3 months post-discharge, analyzed descriptively and compared with non-intervention high-risk patients identified by MEDBRIDGE.

The study is conducted at the University of Alabama at Birmingham (UAB) Hospital and Cooper Green Mercy Health Services, a county-owned safety-net facility serving predominantly low-income and uninsured residents of Jefferson County, Alabama. The patient population is characterized by higher rates of medication complexity, limited health literacy, and social vulnerability. Approximately 20 UAB hospitalizations involve Cooper Green patients with T2D each month, providing a sufficient recruitment base for the target enrollment of 45 patients over a 6-month recruitment window.

• MEDBRIDGE-Guided NCM/CHW Post-Discharge Support Intervention (Behavioral): A 3-month post-discharge care coordination intervention delivered by a nurse case manager (NCM) and community health worker (CHW) team, guided by MEDBRIDGE AI-driven risk stratification. The intervention follows a four-phase workflow: (1) Risk Assessment, where the NCM reviews the daily MEDBRIDGE-generated high-risk patient list; (2) Initial Patient Contact, where the NCM/CHW team initiates contact after discharge to review medications and identify discrepancies; (3) Ongoing Support, where the CHW provides monthly check-ins to monitor adherence, address barriers, and coordinate with primary care providers; and (4) Transition to Routine Care, where the team facilitates handoff to the patient's primary care provider with a summary of activities and recommendations. Participants receive a minimum of 3 contacts over the intervention period.

Primary outcomes

• Recruitment Rate [6-month recruitment window]: Number of patient enrollments per month compared against threshold of 6 enrollments per month. Enrollment rate calculated as the percentage of eligible patients (based on MEDBRIDGE recommendations) who enrolled in the intervention (completed first interaction); target 50% or higher.
• Retention Rate [3 months post-enrollment]: Percentage of enrolled patients completing each round of the intervention (first, second, and up to the final round). Target completion rate of 60% or higher. Estimated 85% retention per subsequent interaction, yielding approximately 61% final retention.
• Patient Acceptability [End of 3-month intervention]: Patient scores on the Acceptability of Intervention Measure (AIM), assessed using a 5-point Likert scale. Target mean score of 4 or higher. Surveys assess comfort with the intervention process and communication with the NCM/CHW team.
• Intervention Fidelity [Throughout 3-month intervention]: NCM/CHW self-report logs including interaction start and end timestamps and a checklist of core intervention components performed, compared against adherence threshold of 0.85. Bi-weekly supervision sessions review logs to ensure protocol adherence.

Secondary outcomes

• HbA1c Change [3 months post-discharge]: Continuous variable of HbA1c change calculated for each patient using HbA1c self-check at baseline and 3 months. An indicator variable on whether HbA1c reduction exceeds 0.5% will be created. Values descriptively compared with non-intervention high-risk patients identified by MEDBRIDGE.
• Diabetes-Related Emergency Department Visits [3 months post-discharge]: Count of diabetes-related emergency department visits identified through retrospective EHR data review. Values descriptively compared with non-intervention high-risk patients identified by MEDBRIDGE.
• Diabetes-Related Hospitalizations [3 months post-discharge]: Count of diabetes-related hospitalizations identified through retrospective EHR data review. Values descriptively compared with non-intervention high-risk patients identified by MEDBRIDGE.

Sex: All, ages: 18 Years to any

Inclusion Criteria:

* Adults aged 18 years or older
* Diagnosis of type 2 diabetes (HbA1c of 6.5% or higher or relevant ICD-10 codes including E11, E13, E14, R73, L97.509, K31.84)
* Discharged from UAB Hospital or its emergency departments
* Receiving primary care at Cooper Green Mercy Health Services or UAB Post Discharge Clinic
* Identified as high risk by the MEDBRIDGE prediction tool based on elevated risk of HbA1c elevation, diabetes-related emergency department visits, or diabetes-related hospitalizations within 3 months post-discharge

Exclusion Criteria:

* Under age 18
* No indication of type 2 diabetes
* Not affiliated with Cooper Green Mercy Health Services or UAB Post Discharge Clinic for primary care
* Unable to provide informed consent
* Currently enrolled in another post-discharge intervention study

• Seung-Yup Lee, PhD, Principal investigator, University of Alabama at Birmingham