Forgotten Hazards of Sedatives

Abstract & Commentary

By Saadia R. Akhtar, MD, MSc, Idaho Pulmonary Associates, Boise. Dr. Akhtar reports no financial relationship to this field of study.

Synopsis: This case series and prospective observational study describe propylene glycol toxicity in patients receiving IV benzodiazepines. The authors estimate the incidence of this important but unrecognized complication to be 19%.

Source: Wilson KC, et al. Chest. 2005;128:1674-1681.

Propylene glycol is used as the carrier vehicle for a number of drugs including lorazepam and diazepam. It may cause metabolic abnormalities such as anion gap metabolic (usually lactic) acidosis and hyperosmolality. Case reports also describe it causing sepsis-like symptoms, cardiac arrhythmias, and neurological changes (agitation, seizures or coma).1 Toxicity may be more common in patients with renal dysfunction.

A prospective, observational study was performed to determine the incidence of propylene glycol toxicity in a single medical ICU. All admissions over a 3-month period were screened. Two groups of patients were enrolled: those receiving benzodiazepines with propylene glycol vehicle (lorazepam and diazepam, 21 patients) and those receiving an alternative benzodiazepine (midazolam, 23 patients). Usual clinical data were collected by medical record review. Propylene glycol toxicity was defined as hyperosmolality or anion gap metabolic acidosis not explained by another cause and reversed by cessation of the benzodiazepine. Standard statistical methods were used to compare the 2 groups.

Patients receiving benzodiazepines with propylene glycol vehicle were more likely to have a history of heavy alcohol intake. Otherwise there were no significant differences in age, gender, co-morbid conditions, admitting diagnoses or clinical data between the 2 groups. Four (19%) of the 21 patients receiving the benzodiazepines with propylene glycol vehicle had evidence of propylene glycol toxicity. All did well after cessation of the infusions and were discharged from the ICU to the floor in stable condition.

Propylene glycol levels between 58 and 127 mg/L were measured in patients with only metabolic abnormalities. Based on their prior case reports, the authors note that levels ranging from 104 to 144 mg/dL were seen in patients with clinical deterioration felt to be secondary to propylene glycol toxicity. Toxicity almost always occurred in patients receiving greater than the usual recommended daily doses of lorazepam (0.01-0.1 mg/kg/hr) and diazepam (5-10 mg IV every 3-4 hours). However, there was at least 1 person with toxicity after as little as 68 mg of IV lorazepam by continuous infusion.


Wilson et al's report—the largest study of this topic in the published English literature—serves as an important reminder of a serious potential adverse effect of lorazepam or diazepam use in the ICU. It also suggests that at least metabolic evidence of propylene glycol toxicity may be much more common than previously realized.

The study clearly has numerous limitations. It is a small, single-center, unblinded observational study. Thus, considerable bias (in patient selection and management and data collection and interpretation) may exist, and certainly the incidence results may not be generalizable. It is unclear exactly what criteria were used to determine whether there was an alternate explanation for acidosis or hyperosmolality in patients classified as having propylene glycol toxicity. It is difficult from these data to define specific threshold levels of benzodiazepines that may lead to toxicity or to determine a threshold level of propylene glycol beyond which toxicity occurs. Further investigation is indicated to address these issues. More information is also needed to understand when and why propylene glycol toxicity may result in clinical deterioration. Long-term outcomes of toxicity (if any) ought to be investigated. Finally, the mechanism(s) of toxicity must be more clearly elucidated.

Despite its limitations, this remains an important report. Until the specifics of propylene glycol toxicity are better defined, at least awareness of and vigilance for this condition are warranted. Potential propylene glycol toxicity is yet another reason to consider limiting sedative use in the ICU.


1. Arroliga AC, et al. Relationship of continuous infusion lorazepam to serum propylene glycol concentration in critically ill adults. Crit Care Med. 2004;32:1709-1714.