A prospective pharmacokinetic study of chlordiazepoxide in patients admitted to an intensive care unit or a high dependency unit after treatment of alcohol withdrawal symptoms

Introduction Chlordiazepoxide is used to manage alcohol withdrawal symptoms. Owing to its active metabolites with long half-lives, vulnerable patients may be at increased risk of prolonged sedation and respiratory depression, potentially requiring intensive care. This study aimed to characterize chlordiazepoxide dosing and related plasma concentrations in patients admitted to an intensive care unit or high dependency unit after chlordiazepoxide treatment for alcohol withdrawal symptoms.Methods A prospective clinical cohort study of patients admitted to the intensive care unit or high dependency unit after protocolized treatment with chlordiazepoxide ≥200 mg. Patients were classified as either sedated (respiratory insufficiency or somnolence) or agitated (alcohol withdrawal symptoms despite benzodiazepine treatment). Blood samples were collected at enrolment and 12 h later to measure concentrations of chlordiazepoxide, norchlordiazepoxide, and demoxepam.Results Of 26 patients included, 19 patients were sedated, and seven were agitated. At inclusion, 85% had plasma chlordiazepoxide concentrations above 3.5 mg/L. The median cumulative dose was higher in the agitated group (1,350 mg [interquartile range: 900–2,050 mg]) compared to the sedated group (700 mg [interquartile range: 275–1,250 mg]). Plasma chlordiazepoxide concentrations were significantly higher (P = 0.002) in the agitated group compared to the sedated group (20 mg/L versus 8.6 mg/L). Correlation between dose and plasma concentration (r2 = 0.35; P = 0.001) was poor.Discussion Agitated patients were treated with higher doses and expectedly had higher chlordiazepoxide plasma concentrations than sedated patients, indicating that individual sensitivity and other differences play a role in clinical outcome, which emphasizes that therapeutic drug monitoring is not suitable for guiding treatment with chlordiazepoxide.Conclusions Poor correlation between doses, plasma concentrations, and clinical effects limits the ability to use cumulated doses and plasma concentrations to predict the risk of prolonged sedation and respiratory depression, making chlordiazepoxide unreliable in the treatment of alcohol withdrawal symptoms.