ObjectiveTo develop, deploy, and evaluate an intervention designed to identify and mitigate conflict in decision making in the intensive care unit.DesignNonrandomized, controlled trial.SettingSeven intensive care units at four Boston teaching hospitals.PatientsA total of 1,752 critically ill patients, including 873 study cases analyzed here.InterventionSocial workers interviewed families of patients deemed at high risk for decisional conflict and provided feedback to the clinical team, who then implemented measures to address the problems identified.Measurements and Main ResultsPatient or surrogate satisfaction with intensive care unit care and the probability of choosing a specific plan for treatment in the intensive care unit was studied. Inclusion criteria identified 873 patients at risk for decisional conflict. Thirty-nine percent of the patients in the intervention phase of the study (172 patients) received the intervention. In multivariate analyses, receiving the intervention significantly increased the likelihood of deciding to forgo resuscitation (odds ratio [OR] = 1.81,p= .017), the likelihood of choosing a treatment plan for comfort-care only (OR = 1.94,p= .018), and the likelihood of choosing an aggressive-care treatment plan (OR = 2.30,p= .002). Receiving the intervention did not significantly affect overall satisfaction with the care provided (OR = 0.68,p= .14), satisfaction with the amount of information provided (OR = 0.86,p= .44), or satisfaction with the degree of involvement in decision making (OR = 0.84,p= .54).ConclusionsAlthough there was no impact on patient or surrogate satisfaction with care provided in the intensive care unit, the intervention did facilitate deliberative decision making in cases deemed at high risk for conflict. The lessons learned from the experience with this intervention should be helpful in ongoing efforts to improve care and to achieve outcomes desired by critically ill patients, their families, and critical care clinicians.