The investigators propose to conduct a feasibility study of identifying high-risk ICU patients using previously validated integrated prediction model and employing early palliative care intervention. The study will consist of two four-month time periods: usual care in time period one and usual care plus targeted pro-active palliative care intervention within 48 hours of ICU admission in time period two.

The emerging challenge in critical care medicine focuses on improving patient- and family-centered outcomes. (1) The gold standard for patient-centered care is that which aligns with patients' values, preferences, and goals. (2) Palliative care consultants are known for utilizing shared decision making conversations in the intensive care unit (ICU) in which they elicit patients' values and goals. Accurate prognostic information based on the patient's overall condition sets the context for shared decision-making. With the gravity of the decisions made every day in the ICU, prognostic information is invaluable to patients, families, and clinicians. Historically, prognostic models in the ICU have focused on short-term mortality, such as death prior to hospital discharge, but studies have shown that information about post-ICU quality of life outcomes is important to patients and families as is information about survival for six months or more. Researchers have identified that future work in ICU prognostication must focus on accurately predicting the scope of patient-centered outcomes without losing sight of the higher mortality risk faced by ICU patients and survivors. The investigators propose to perform a proof of concept study based on our previously validated integrated prognostic model to identify medical ICU patients with high risk of mortality in six months and determine the utility of this model in terms of improving quality of life and quality of care. The study will consist of two 4-month time periods: usual care and usual care + targeted pro-active palliative care intervention within 48 hours of ICU admission, respectively. For both phases, the investigators will use the previously validated and published integrated prognostic model consisting of the 'surprise question' of "Would you be surprised if this patient died in the next six months?" with a response of "No", the APACHE IV score, and the Charlson Comorbidity Index (CCI) score to identify ICU patients within 24 hours of their ICU admission with a probability estimate of six-month mortality greater than 40 percent.

Six weeks after ICU discharge, we will contact the legal decision maker for the patient (family member or friend who is the medical power of attorney or healthcare surrogate) regarding their satisfaction with critical care services the patient received during the ICU stay. After obtaining verbal consent, questions will be asked from the modified Family Survey-ICU (FS-ICU) and the global performance measure #18 from the Bereaved Family Survey (BFS). The investigators hypothesize that the systematic identification of ICU patients at highest risk of six-month mortality as a prompt for a targeted pro-active palliative care approach will decrease length of ICU and hospital stay, reduce ICU readmissions, lead to earlier treatment limitation orders according to patients' wishes, and increase family satisfaction with the quality of care.

Inclusion Criteria:

• Patients greater than 18 years of age admitted to the medical ICU at West Virginia University Hospital.
• Patients with a predicted six-month mortality greater than 40% on the integrated prognostic model.
• Patients who live longer than 48 hours after medical ICU admission.

Exclusion Criteria:

• Patient who did not meet inclusion criteria will be excluded.
• Pregnant female/incarcerated patients will be excluded.
• Since this is a study in an adult patient population, children will be excluded.