This multicenter RCT of 200 hospitalized patients and their family members evaluates an "informed assent" approach to discussing cardiopulmonary resuscitation, compared to usual care, in older seriously ill hospitalized patients with severe life-limiting illness or severe functional impairment.

Increasing evidence suggests that high quality palliative care in older patients improves quality of care, patient and family satisfaction, and costs of care. However, while some specific palliative care interventions have been investigated, exactly how we should provide this care remains unclear. One important component of high quality provider-patient communication is discussing cardiopulmonary resuscitation (CPR) within the context of the patient's values and goals of care. Unfortunately, these conversations are often rushed, focused only on the procedure rather than the patient's goals and values, and CPR is often offered as if a choice on a menu. Elderly hospitalized patients are unlikely to have made decisions about CPR prior to hospitalization, and these discussions often leave patients and families feeling burdened, stressed, and concerned. Outcomes after in-hospital CPR in chronically ill older patients continue to be very poor, despite decades of efforts toward improvement.

The investigators aim to change the paradigm of CPR discussions with older adults unlikely to benefit from CPR. This study is a randomized controlled trial comparing an innovative "informed assent" approach to discussing CPR versus usual care with attention control for older hospitalized adults with life-limiting illness or severe functional impairment, enrolling patients and family members at three different study sites. Informed assent includes: 1) eliciting values and goals, 2) explaining CPR in the context of illness, and 3) a caring statement that the patient should not receive CPR if values include avoiding burdensome therapies unlikely to provide benefit. Thus, the provider can elicit values, convey information, and state an assessment, while allowing the patient or family to disagree. The investigators hypothesize that informed assent will improve quality of and satisfaction with communication about CPR; reduce the burden of potentially harmful CPR discussions, including reduced patient and family symptoms of depression and anxiety; and reduce intensity of care and health care utilization. The intervention builds on prior work investigating CPR outcomes in older adults and performing pilot studies of the informed assent intervention in both inpatients and outpatients.

Inclusion Criteria:

1. >65 years old

2. English speaking

3. Must have one or more of the following:

   1. Chronic life-limiting illness with median survival <2 years defined as: 1) metastatic cancer or inoperable lung cancer; 2) COPD requiring oxygen; 3) New York Heart Association Class III or IV heart failure, 4) Child's Class C cirrhosis or MELD score of >20, 5) End-stage renal disease (must be on dialysis and ≥ 75 years old), 6) Advanced pulmonary fibrosis/interstitial lung disease, 7) Advanced pulmonary hypertension
   2. Severe functional impairment defined as dependence with >4 activities of daily living (ADLs) on Katz Index of Independence in ADLs.

Exclusion Criteria:

1. Has already definitively chosen DNR status
2. Unable to provide informed consent
3. Refused consent
4. Currently listed on a transplant list (awaiting transplant)
5. Inappropriate for study enrollment per clinician
6. Known to have a left ventricular assist device (LVAD)
7. Research team unavailable
8. Patient discharged from hospital prior to enrollment

This study includes quantitative data that the investigators plan to share with other researchers with interests in the quality of care provided to older hospitalized seriously ill patients and their families. De-identified questionnaire data, medical record data, and administrative data will be made available in the form of datasets with no personally identifying characteristics retained, to protect the identity of the respondents.