Admission and end of life sedation and analgesia preferences in the ICU

Purpose: To investigate physician and nurse preferences for intensive care unit (ICU) sedation at admission and during subsequent withdrawal of life support. Methods: There is a paucity of data describing variability in physician and nursing preferences with analgesia and sedation in the ICU. Prior to a wider study, a single clinical scenario was sent out to 8 intensivists. They were asked to write admission orders for analgesia and sedation for an 81-year-old postoperative patient and to indicate bolus and/or infusion doses to be prescribed in the first hour of admission. The same scenario was sent out to 144 ICU nurses who were asked to record how much analgesia and sedation they would provide based on a standardized medical order. Subsequently the patient deteriorated with ARDS and dialysis dependency. The family requested withdrawal of life support. The 8 intensivists were asked to record analgesia and sedation orders in association with withdrawal of life support. Nurses were asked to record the amount of medication they would provide over 1 hour in association with a standardized physician order. Physicians and nurses also recorded their preferences for the mode of ventilator withdrawal and FiO₂ reduction. Results: Physicians: All 8 responded. On admission there was a ten-fold variation in morphine dosage in the first hour. With withdrawal of life support again there was a 10-fold variation in morphine and also a 7-fold variation in chosen benzodiazepine dosages. When adjusting the ventilator 4/8 physicians opted for a change from CMV to T-piece, with 4/8 requesting single step reduction to FiO₂ 21% and 4/8 a graded reduction, ensuring comfort at each stage. Nurses: Responses were received from 75/144. On admission bolus doses of opiates and benzodiazepines were chosen by 62% and 88% of nurses respectively with infusion chosen by 3% and 9% respectively. Bolus plus infusion choices for opiates and benzodiazepines were made by 35% and 9%. At withdrawal of life support there was considerable variation in morphine dosing (2-60 mg in 30 mins). Request for a T-piece was made by 45%, with 64% wanting single and 31% a graded reduction to FiO₂ 21%. Conclusions: There is considerable variation in both physician and nursing opiate and benzodiazepine practices at admission and at end of life and in choices for modes of ventilator and FiO₂ withdrawal. Clinical Implications: This degree of practice variation within a single unit underlines the need for (a) improved surveillance of practices within the ICU, especially at the end of life, and (b) more effective educational initiatives to effect more consistent practices within the ICU.