The purpose of this study is to validate a set of proposed clinical criteria that have been designed to identify patients who will die rapidly following the elective removal of life sustaining treatments.

Our study is a simple, observational study. We intend to record the events that occur between the time withdrawal of life sustaining treatment occurs, and the time of death. Our preliminary data, reported in the journals Critical Care Medicine and Pediatric Critical Care Medicine, show that the median time to death is about 25 minutes, and few survive over an hour. We would directly observe no patient for longer than one hour.

This is a straightforward study to characterize the Receiver Operator Characteristic (ROC) curves and Relative Risk of a variety of clinical data elements for predicting early death following discontinuation of life sustaining treatments. This is a multi-center study currently being conducted at: University of Cincinnati Hospital, Case-Western Reserve University Hospital and the University of Pittsburgh Medical Center with the University of Pittsburgh acting as the Coordinating Center. (a) This is a prospective cohort study of patients electively withdrawn from life sustaining treatment. The investigators and the study have no impact on the decision or timing of withdrawal of support. The study will merely record what and what time support is withdrawn, the medications that are provided for patient comfort, and the time of death. Within this group, we will determine whether patients who "meet" the UNOS criteria (UNOS cohort) have a short time to death, enabling us to validate the current, unvalidated UNOS criteria. In addition, we will analyze data from all patients, perform logistic regression to determine other candidate criteria for determining short time to death. (b) Data extractors will collect the data from all identified patients using chart review following procedures approved by their institutional review board. Data forms will be de-identified at each institution prior to transmission to our central database to ensure patient confidentiality.

The results from the project we are proposing will enable more accurate identification of appropriate candidates for NHBOD. We plan to do this by examining pre-mortem data of patients that have died after withdrawal of LST to see if any factors or combination of factors are predictive of death in less than 30 minutes and 60 minutes after LST is removed.

Our research questions are:

1. Do the UNOS criteria predict early death following discontinuation of LST? 2. In addition we will collect data to enable us to address the secondary questions:
2. a. What are the best predictors of early death following discontinuation of LST?, and 2.b. How accurate are they?

We intend to test the following two hypotheses:

H1: The UNOS criteria are predictive of death in less than 30 minutes; H2: The UNOS criteria are predictive of death within 60 minutes.

The researchers selected the following outcomes for the study. Outcome 1: death within 30 minutes of removal of life sustaining treatment. Outcome 2: death within 60 minutes of removal of life sustaining treatment. Outcome 3: number of potential organ donors following cardiac death

Inclusion Criteria:

• all patients who have life sustaining treatments withdrawn

Exclusion Criteria:

• confirmed brain death