Article Text
Abstract
Introduction Alcohol withdrawal syndrome (AWS) is a common reason for hospital admission. However a significant number of these patients have co-existent liver disease or other medical problems. There is little information regarding the management of these patients.
Methods Patients were assessed for hazardous drinking using the FAST score. Data was collected prospectively on FAST positive patients with regards to their subsequent treatment using a unified AWS guideline. Patients with known liver disease or presenting with decompensated liver disease were identified (Group 1) and compared with patients admitted with non-hepatic primary alcohol-related diagnoses (Group 2) and patients with admitted primarily non-alcohol related medical diagnoses (Group 3). Results are expressed as median (95% confidence).
Results 53 patients had significant liver disease (Group 1), with 153 in Group 2 and 106 in Group 3. Median MELD score in Group 1 was 15.4 [12.8, 17.6]. The three groups had similar FAST scores: 14 [12, 15], 14 [13, 14] and 13 [12, 14] respectively. Fewer patients in Group 1 and Group 3 required benzodiazepine (BZD) treatment compared with Group 2. When required, the median BZD prescription (mg diazepam equivalent) during admission was greater for Group 2 patients compared with Group 1 and Group 3 (Table). More patients in Group 1 were treated with lorazepam rather than diazepam; 13% cf 5% in Group 2 (p = 0.1, –1, 20.4) and 3% in Group 3 (p = 0.04; 0.7, 22.3). The proportion of patients requiring parenteral treatment for breakthrough symptoms were similar: Group 1: 5.6%, Group 2: 4.6%, Group 3 4.7%.
There were correlations between the FAST score and subsequent amount of diazepam prescribed for Group 2 (p = 0.002; 0.09, 0.4) and Group 3 (p = 0.03; 0.02, 0.41), but not for Group 1 (p = 0.26; –0.12, 0.43). Overall survival until 33 months was less in Group 1 (64%) compared with Group 2 (84%; p = 0.0007 HR 0.29 [0.14, 0.60] and Group 3 (81%; p = 0.016 HR 2.30 [1.17, 4.55]).
Conclusion Compared to patients admitted with primarily AWS or alcohol related seizures, patients with chronic liver disease and those with other medical problems were less likely to require any benzodiazepine therapy and require lesser amounts of such treatment. The expected association between indicators of harmful/dependent drinking and BZD requirement was lost in liver disease patients. Care should be taken to avoid unnecessary over-treatment of these patients.
Disclosure of Interest None Declared