Propylene glycol-associated renal toxicity from lorazepam infusion

Objectives. Using data from patients who developed elevations in serum creatinine concentrations while receiving continuous-infusion lorazepam, we sought to determine the correlations between the magnitude of serum creatinine concentration rise and each of the following variables: serum propylene glycol level, cumulative lorazepam dose, and duration of lorazepam administration. An additional objective was to identify clinical markers for propylene glycol toxicity. Design. Retrospective chart review. Setting. Medical-surgical intensive care unit and burn unit at a university hospital. Patients. Eight patients who developed elevations in serum creatinine concentrations while receiving continuous-infusion lorazepam (range 2-28 mg/hr). Measurements and Main Results. The mean cumulative dose of lorazepam was 4305 mg (range 1200-10,920 mg), and the mean propylene glycol level determined at the time of peak serum creatinine concentration was 1103 mug/ml (range 186-3450 mug/ml). Serum creatinine concentrations increased in all eight patients during lorazepam infusion and decreased in seven within 3 days after stopping infusion. A weak-to-moderate correlation existed between the magnitude of the rise in serum creatinine concentration and propylene glycol level (r=0.53). A weak-to-moderate correlation also was identified between cumulative lorazepam dose and magnitude of serum creatinine concentration rise (r=0.43), and a strong-to-moderate correlation was found between duration of lorazepam infusion and magnitude of serum creatinine concentration rise (r=0.60). Propylene glycol levels were strongly correlated with both serum osmolality and osmol gap. Conclusion. The patients' increased serum creatinine concentrations are likely to have resulted from exposure to propylene glycol as a result of lorazepam infusion. Serum osmolality and osmol gap may be useful markers for propylene glycol toxicity.