Objective: This study will: 1) determine how patients with serious illnesses value specific, commonly occurring symptoms and outcomes; 2) determine how patients' experience with specific symptoms and outcomes influence the value they place on these symptoms and outcomes; and 3) extend the findings from an ongoing large, multicenter study of prognoses and preferences of seriously ill patients (SUPPORT project) by providing a common metric (utility scale) to compare aggregate outcomes. Design: A composite patient preference model(CPPM) will be derived using conjoint analysis. The CPPM will be a utility scale for 6 common symptoms and outcome states (attributes): being in pain, being on a respirator, being fed through a tube, being unconscious, being confused, living in a nursing home. Conjoint analysis is a decompositional method that derives estimates of """"""""part-worth"""""""" utilities of patient preferences (i.c., weights for selected attributes) based on patients' evaluations of a set of alternatives (scenarios) that represent various combinations and levels of attributes. These attributes will be presented using a confounded factorial design. Setting: Patients will be selected from the Medical and Surgical inpatient services of MetroHealth Medical Center of Cleveland, a large county-owned primary teaching hospital for Case Western Reserve University. Patients: A total of 420 patients will be studied (60 in the pilot study, 360 in the main study). They will be selected from among patients who are screened for the SUPPORT project but do not meet the severity of illness criteria to enter SUPPORT. Thus, study subjects will be adults who will have experience with one or more of the attributes being studied and with serious illness. Patients will be selected in approximately equal numbers from among 8 diagnostic categories used in the SUPPORT project. Measurements: Patients will be asked to rate 21 sets of attributes (scenarios) on a scale ranging from 'very willing to tolerate' to 'would rather die' than to tolerate a given situation. Data will be collected on demographic features and about each patient's prior experience with the attributes. Anticipated Results: The CPPM will delineate how patients with serious illnesses value specific symptoms and outcomes. The study also will determine how experience with these symptoms and outcomes influence the value placed on them by patients. Significance: The development of a CPPM will importantly extend the results of the SUPPORT project by providing a common metric which can be used to compare aggregate outcomes. It also could be used to help inform judgments made for patients who are unable to make judgments for themselves and to inform policy decisions that may influence the care of seriously ill patients.
