For complex medical patients, the transition from hospital to home-based care is a vulnerable period, placing the patient at high risk for adverse events. Using a Care Transition conceptual model, the investigators propose developing and evaluating, through a randomized controlled trial, "e-Coach," an Interactive-Voice-Response-supported (IVR) Care Transition coaching intervention, focused initially on patients hospitalized with heart failure or obstructive lung disease. This trial will test the primary hypothesis that the proportion of patients with one or more re-hospitalizations during a 90-day post-discharge follow-up period will be less in an IVRsupported care transition intervention (e-Coach) compared to a "usual care" comparison group.

For complex medical patients, the transition from hospital to home-based care is a vulnerable period, placing the patient at high risk for adverse events, including the experience of a medical error or loss of community tenure. Recent successful studies have used a Care Transition Intervention (CTI), using a nurse who conducts home visits, telephone follow-up, and provides assistance at and after discharge. Although successful, this model is costly and and not feasible in settings serving geographically dispersed populations. We propose a cost-efficient technological solution to the problems presented by the traditional CTI through "e-Coach," an Interactive-Voice-Response-supported (IVR) Care Transition coaching intervention. We propose to develop and evaluate "e-Coach," by performing a randomized controlled trial of this intervention versus a usual care comparison group. Our Specific Aims are to: 1) Randomize 720 patients at high risk of transition-related errors (complex adult patients discharged alive after a hospitalization with congestive heart failure (CHF) or chronic obstructive pulmonary disease (COPD), from a geographically diverse area including many rural areas across Alabama and the South) to an IVR-supported care transition program ("e-Coach") versus a usual care comparison group. The IVR system will actively call patients at multiple intervals after discharge. In a stepped-care approach, the IVR will be further supported by a Care Transition nurse who monitors patient symptoms through the e-Coach IVR and supports patient self management through telephone-based interactions when needed, up to 3 months after discharge; 2) Evaluate use of the e-Coach by patients and healthcare providers; 3) Evaluate the impact of the e-Coach on patient outcomes, including 90 day rehospitalizations, successful community tenure over a 3 month period, medication discrepancies, and patient self-efficacy based on the previously validated Care Transition Measure; and 4) Quantify the cost associated with the e-Coach.

Inclusion Criteria:

• CHF/COPD patients
• English-speaking
• Medicare beneficiaries

Amendment to Inclusion Criteria:

• Recruited non-Medicare eligible beneficiaries

Exclusion Criteria:

• Prognosis of 6 months or less
• Cognitive impairment with no available proxy/caregiver
• No possession of a phone

Amendments to exclusion criteria:

• heart or lung transplant recipients
• dialysis patients
• individuals already in the Cystic Fibrosis program or receiving intensive monitored care
• individuals with a ventricular assist device (LVAD; RVAD; BiVAD)
• individuals utilizing a pre-paid phone service

A few patients were excluded from the final study groups. 6 patients died prior to discharge, 2 patients declined participation, and 4 patients became ineligible. After hospital discharge, 12 patients declined follow-up, we were unable to contact 5 for follow-up, and 4 became ineligible.

Adverse event data were collected throughout the study period (2 years)

There were no serious adverse events (SAE) or other adverse events as a result of the intervention. Any adverse events that occurred were expected outcomes, given the baseline health level of the participants.

Limitations include generalizability. Our single healthcare system likely has unique characteristics. Readmission to the hospital is also a limited quality measure. There is currently no way to deem if readmissions are appropriate or avoidable.