Approach
Benzodiazepine (BZD) overdose is usually determined by the patient's history, either from his or her own report, or by circumstantial evidence, such as BZD prescriptions or a history of psychiatric illness or previous overdose.
The key factors in making the diagnosis are signs and symptoms of central nervous system (CNS) inhibition, particularly drowsiness with unremarkable vital signs and no focal signs on neurologic exam.
Clinical evaluation
Ataxia and altered mental state are the most common features, but symptoms and signs may be nonspecific and highly variable. Typical signs of CNS depression include slurred speech, incoordination, unsteady gait, and impaired attention or memory, especially the loss of anterograde memory (patients should be asked about recent events since exposure, such as who brought them to hospital and how). Other physical signs may include nystagmus and decreased deep tendon reflexes. More subtle signs of impaired mental status include inappropriate behavior or judgment, and labile mood.
Overdose may also present with paradoxical stimulation, particularly in patients with psychiatric, hyperactive, or aggressive disorders.
Laboratory investigations
Initial tests to exclude other causes include a rapid bedside determination of glucose, pulse oximetry, CBC, serum chemistry panel, serum ethanol level, urine toxicology screen, and ECG. ECG can show abnormalities including transient first and second degree block and QT prolongation but investigations are otherwise usually normal in pure BZD overdose. Any abnormalities should prompt further investigations as appropriate. CNS depression can mirror respiratory depression; therefore, patients with deteriorating levels of consciousness should be closely observed for any deterioration in oxygen saturation.
Immunoassay screening tests for BZD that are typically used in emergency care settings have many false-positive and false-negative results; thus their use should be discouraged.[30] Most screening tests for BZD detect only certain metabolites of BZD (namely, desmethyldiazepam or oxazepam); hence, many BZDs are not routinely detected. Negative screening tests do not therefore exclude BZD overdose. Furthermore, because of the long half-life of many BZDs and their metabolites, a positive screening test may fail to differentiate between BZD use at some time in the past (sometimes even weeks previously) and acute overdose.
Healthcare professionals should take into consideration that fentanyl (an opioid analgesic) is an increasingly common cause of overdose, both with and without BZDs, but is not generally included on urine toxicology screens. In 2021, 70% of all BZD-involved deaths in the US also involved fentanyl.[14]
Flumazenil trial
Previously used in first-time or infrequent BZD users presenting with CNS depression and suspected BZD overdose. When used as a trial, resolution of CNS depression following flumazenil administration confirmed BZD overdose.
Flumazenil should not be used in patients known to be, or suspected of, taking long-term BZDs, or in patients known to be taking drugs that can increase the risk of seizures, such as tricyclic antidepressants. Flumazenil can induce seizures in patients who have a history of seizures, are dependent on BZDs, or who take tricyclic antidepressants.[31][32] It can also cause dysrhythmias in patients taking prodysrhythmic medication such as tricyclic antidepressants or carbamazepine.
It is worth noting that the risks associated with use of flumazenil, outweigh any potential benefits for most patients and the UK National Poisons Information Service strongly advises against a flumazenil trial, or use of flumazenil as a diagnostic test.[33] However, the American Heart Association recommends that flumazenil can be effective in select patients with respiratory depression or respiratory arrest caused by pure BZD poisoning who do not have contraindications to flumazenil.[34] It may also be safe in some low-risk presentations (e.g., pediatric exploratory ingestions and iatrogenic overdoses during procedural sedation).[34]
Use of this content is subject to our disclaimer