A model for predicting alcohol withdrawal delirium

Objective: The aim of this study was to develop a model for identifying pat ients with a high risk of developing alcohol withdrawal delirium after asse ssment in the emergency department. Methods: Patients seeking acute treatme nt for alcohol withdrawal at St. Goran's hospital in Stockholm were evaluat ed for known risk factors for alcohol withdrawal delirium. Ah patients with any risk factor were admitted to the hospital and received:standard treatm ent with benzodiazepines. All patients were evaluated at admission by the p hysician in charge at the psychiatric and dependency emergency unit at the hospital. Treatment and final assessment were conducted at the unit's inpat ient acute-treatment facility. Correlations were determined between risk fa ctors noted at admission and development of alcohol withdrawal delirium, as defined in DSM-IV, after admission. A total of 334 alcohol-dependent patie nts were included in the study Results: Twenty-three patients, or 6.9 perce nt, developed alcohol withdrawal delirium after admission despite benzodiaz epine treatment. Zn a stepwise multiple regression model, five risk factors were significantly correlated with the development of alcohol withdrawal d elirium: current infectious disease; tachycardia, defined as a heart rate a bove 120 beats per minute at admission; signs of alcohol withdrawal accompa nied by an alcohol concentration of more than 1 gram per liter of body flui d; a history of epileptic seizures; and a history of delirious episodes. No patient without these five risk factors developed delirium. Conclusion: As sessment for five easily detectable risk factors can enable the clinician t o make an accurate and quantitative assessment of a patient's risk of devel oping alcohol withdrawal delirium.