BACKGROUND AND AIMS: Little is known about long-term consequences of delirium tremens (DT). This study aimed to compare all-cause and cause-specific mortality and alcohol-related morbidity between patients with: (i) DT, (ii) alcohol withdrawal state (AWS) and (iii) alcohol dependence (AD).

DESIGN: A national longitudinal health registry study with linked data from the Norwegian Patient Registry and the Norwegian Cause of Death Registry. Participants were all patients registered in the Norwegian Patient Registry between 2009 and 2015 with a diagnosis of AD (ICD-10 code F10.2), AWS (F10.3) or DT (F10.4) and aged 20-79 years were included (n = 36 287).

MEASUREMENTS: Patients were categorized into three mutually exclusive groups; those with DT diagnosis were categorized as DT patients regardless of whether or not they had received another alcohol use disorder diagnosis during the observation period or not. Outcome measures were: annual mortality rate, standardized mortality ratios (SMR) for all-cause and cause-specific mortality and proportion of alcohol-related morbidities which were registered in the period from 2 years before to 1 year after the index diagnosis.

FINDINGS: DT patients had higher annual mortality rate (8.0%) than AWS (5.0%) and AD (3.6%) patients, respectively. DT patients had higher mortality than AD patients and AWS patients SMR was particularly elevated for unnatural causes of death, and more so for DT patients than for AD patients or AWS patients. For all comorbidities, we observed a higher proportion among DT patients than among AWS or AD patients (P < 0.001).

CONCLUSIONS: People treated for delirium tremens appear to have higher rates of mortality and comorbidity than people with other alcohol use disorders.