Certain methods of sedation increase the duration of respiratory failure. Two strategies, a nursing- implemented sedation algorithm and daily interruption of sedatives, decrease length of mechanical ventilation compared to "conventional care" but have not been compared to each other. The reason certain methods of sedation lead to prolonged respiratory failure is unknown but may be related to altered pharmacokinetics and dynamics that are unique to critically ill patients. Critically ill patients receive substantial doses of sedatives over prolonged periods. The impact of these management strategies on short- and long-term psychiatric complications are unknown. The study seeks to test the central hypothesis that sedation practices impact strongly on outcome of respiratory failure and psychiatric complications. The three specific aims are (1) to compare two sedation strategies (protocol directed sedation and daily interruption of sedatives), (2) to examine the prevalence of psychiatric complications, and (3) to compute the pharmacokinetics of commonly used sedatives and narcotics.

These aims will be achieved by enrolling critically ill patients in a prospective randomized trial comparing the above mentioned sedation strategies, and assessing sedation level as well as delirium throughout the duration of respiratory failure. Sedative plasma levels will be measured, and pharmacokinetics computed. Psychiatric morbidity will be assessed by administration of validated questionnaires.