BACKGROUND: Cognitive impairment is common in older patients with heart failure (HF), leading to higher 30-day readmission rates than those without cognitive impairment. OBJECTIVES: The aim of this study was to determine whether increased readmissions in older adults with cognitive impairment are related to HF severity and whether readmissions can be modified by caregiver inclusion in nursing discharge education. METHODS: This study used prospective quality improvement program of cognitive testing and inclusion of caregivers in discharge education with chart review. Two hundred thirty-two patients older than 70 years admitted with HF were screened for cognitive impairment using the Mini-Cog; if score was less than 4, nurses were asked to include caregivers in education on 2 cardiovascular units with an enhanced discharge program. Individuals with ventricular assist device, transplant, or hospice were excluded. Measurements include Mini-Cog score, 30-day readmissions, readmission risk score, ejection fraction, brain natriuretic peptide, and medical comorbidities. RESULTS: Readmission Risk Scores for HF did not correlate with Mini-Cog scores, but admission brain natriuretic peptide levels were less abnormal in those with better Mini-Cog scores. Only for patients with cognitive impairment, involving caregivers in discharge teaching given by registered and advanced practice nurses was associated with decreased 30-day readmissions from 35% to 16% (P = .01). Readmission rates without/with cognitive impairment were 14.1% and 23.8%, respectively (P = .09). Abnormal Mini-Cog screen was associated with a significantly increased risk of 30-day readmission (odds ratio, 2.23; 95% confidence interval, 1.06-4.68; P = .03), whereas nurse documentation of education with family was associated with a significantly decreased risk of 30-day readmission (odds ratio, 0.46; 95% confidence interval, 0.24-0.90; P = .02). CONCLUSIONS: Involving caregivers in discharge education significantly reduced 30-day readmission rates for patients with HF and cognitive impairment. The Readmission Risk Score was similar between patients older than 70 years with and without cognitive impairment. We have hypothesis-generating evidence that identification of cognitive impairment and targeted caregiver engagement by nurses may be critical in the reduction of readmission rates for older patients with HF.