Clinical sedation scores as indicators of sedative and analgesic drug exposure in intensive care unit patients

Abstract

Background: It is unclear how best to measure sedative/analgesic drug exposure in the clinical care of critically ill patients. Large doses and prolonged use of sedatives and analgesics in the intensive care unit (ICU) may lead to oversedation and adverse effects, including delirium and long-term cognitive impairment. These issues are of particular concern in elderly patients (aged > or =65 years), who account for at least half of all ICU admissions and nearly two thirds of ICU days.

Objective: This pilot study explored the relationships between clinical sedation scores, sedative/analgesic drug doses, and plasma drug concentrations in critically ill patients, the majority of whom were elderly.

Methods: This was a prospective, observational study conducted in a 500-bed, tertiary care community hospital. Study patients included a cohort of mechanically ventilated, medical ICU patients who were admitted to the hospital between April and June 2004 who required use of fentanyl, lorazepam, or propofol. Sedative/analgesic medications were administered according to clinical guidelines. Patients' sedation levels were measured twice daily using the Richmond Agitation-Sedation Scale (RASS). Dosing of fentanyl, lorazepam, and propofol was recorded. Blood was sampled twice daily for up to 5 days to analyze plasma drug concentrations. To specifically evaluate the effects of acute, extended (rather than chronic) sedative and analgesic exposure, the study focused on an ICU population receiving these agents for at least 48 hours but <2 weeks.

Results: Eighteen medical ICU patients (11 females, 7 males; mean [SD] age, 66.1 [18.1] years) on mechanical ventilation comprised the study cohort. Fifteen patients were aged >62 years, and 11 of those were aged > or =71 years. Plasma drug concentrations (median and interquartile range) were as follows: fentanyl--2.1 ng/mL, 0.9-3.4 ng/mL; lorazepam--266 ng/mL, 112-412 ng/mL; and propofol--845 ng/mL, 334-1342 ng/mL. Maximum concentrations were 3- to 12-fold higher than medians (fentanyl, 7.3 ng/mL; lorazepam, 3108 ng/mL; propofol, 10,000 ng/mL). Medication doses were only moderately correlated with RASS scores (Spearman rho): fentanyl--rho = -0.39, P = 0.002; lorazepam--rho = -0.28, P = 0.001; and propofol--rho = -0.46, P < 0.001. Plasma drug concentrations of fentanyl and lorazepam demonstrated moderate correlations with sedation scores (fentanyl--rho = -0.46, P = 0.002; lorazepam: rho = -0.49, P < 0.001), while propofol concentrations correlated poorly with sedation scores (rho = -0.18, P = 0.07). Associations between interval drug doses and plasma concentrations were as follows: fentanyl, rho = 0.84; lorazepam, rho = 0.76; and propofol, rho = 0.61 (all, P < 0.001). Instructive examples of discrepant dose versus plasma concentration profiles and drug interactions are provided, including 3 cases with patients aged > or =64 years.

Conclusions: Elderly patients are commonly encountered in the ICU setting. Only moderate correlations existed between clinical sedation levels and dose or plasma concentration of sedative/analgesic medications in this study population. Further work is needed to understand appropriate and feasible measures of exposure to sedatives/analgesics relating to clinical outcomes.