The dataset included 10,332 patients with a primary diagnosis of AUD or acute intoxication with a total of 38,780 drug prescriptions. AUD inpatients comprised 7.3% of all 143,436 psychiatric inpatients assessed on the reference days in the study period.

The majority of AUD patients were male (69.4%) and aged 30–60 years (77.0%). The mean age was 48.6 (± 12.1) years. Males were slightly younger than females (48.1 ± 11.9 vs. 49.4 ± 12.6 years, (t(10,330) = 5.0621, p < 0.001). The proportion of female patients was greater among patients aged ≥ 60 years (Table 1).

The most common diagnosis was dependence syndrome (F10.2, 65.8%), followed by withdrawal states (F10.3, 11.2%) and acute intoxication (F10.0, 9.7%) (Table 1). The subdiagnoses F10.1, F10.2, F10.3 and F10.4 were significantly more common in men than in women (p < 0.05 each) (Table 1).

In young patients (< 31 years), subdiagnoses F10.0 and F10.1 were more common, while F10.4, F10.5, F10.6 and F10.7 affected older (> 60 years) patients more often (Table 1).

Data on secondary and tertiary psychiatric diagnoses were available for 6,968 AUD-patients for 2007–2016. Other substance use-related diagnoses (F1, 51.8%) and therein additional alcohol-related diagnoses (F10.x, 33.7%) were prevalent, followed by mood and affective disorders (F3, 18.4%) (Table 2). Only approximately one-third (34.6%) of patients did not have an additional psychiatric diagnosis. Almost half of the patients with a main diagnosis of acute intoxication had at least one secondary diagnosis of F10, and only 4% have no additional psychiatric diagnosis.

Overall, 92.7% (n = 9,575) of patients were treated with at least one drug. Most patients were treated with psychotropic drugs (76.2%, n = 7,874) with an average of 1.9 ± 1.1 psychotropic drugs per patient. More women than men received psychotropic drugs (79.5% vs. 74.8%, p < 0.001). Among non-psychotropic drugs, vitamins, and proton pump inhibitors (PPIs) were used most frequently (42.2% and 28.4%, respectively).

Antidepressant (31.2%), antipsychotic (29.7%), anticonvulsant (26.4%) and tranquilizing drugs (24.3%) were frequently used, followed by other psychotropic (9.4%, mostly clomethiazole) and hypnotic drugs (5.3%). Antiparkinsonian drugs (2.6%), lithium (1.0%) and nootropic drugs (0.9%) were rarely applied (Table 3).

Antidepressant, antipsychotic and hypnotic drugs were more commonly used in women, only anticonvulsant drugs were used more often in men. There was no difference in the use of tranquilizing drugs (Table 3).

Among antidepressant drugs, SSRIs (12.8%) and NaSSAs (8.5%) were most frequently used, followed by SNRIs, TCAs and other antidepressant drugs (6.1%, 5.4% and 3.7%, respectively). Among antipsychotic drugs, SGAs (14.4%) and low-potency FGAs (13.2%) were used in nearly equal proportions, while high-potency FGAs were used less frequently (7.0%) (Table 3). Benzodiazepines prevailed among tranquilizing drugs (23.6%), while Z-drugs were most common among hypnotics (3.0%).

Because states of withdrawal (F10.3), delirium (F10.4) and psychotic disorder (F10.5) associated with AUD tend to require distinct treatment, we performed a separate analysis of the drug utilization patterns of these diagnostic subgroups. Patients with F10.3, F10.4 and F10.5 received more drugs in general and more psychotropic drugs but considerably fewer antidepressant drugs than the total group of patients with AUD. All patients and in particular patients with F10.3 were less often treated with antipsychotic drugs than patients with F10.4 or F10.5. Tranquilizing drugs were given more often to patients with F10.3 and F10.4, while the use of hypnotic drugs was lower in these groups than in the whole population. The opposite was found among patients with F10.5. The utilization rate of hypnotic drugs was greater and that of tranquilizing drugs was lower than the rates for all patients, especially for patients with primary diagnoses of F10.3 and F10.4. Anticonvulsant drugs were given relatively often to patients with F10.3 and less often to patients with F10.5. One of the greatest differences was detected for clomethiazole, which was more commonly used for patients with F10.3 and F10.4 (Table 4).

Furthermore, we analyzed the utilization rates for patients with additional non-AUD psychiatric diagnoses (Table 5). For this, we divided patients into two groups: patients with at least one additional psychiatric diagnosis other than F10.x and patients without.

Overall, patients with additional psychiatric diagnoses were treated with any kind of psychotropic drug more often than patients with only an F10 diagnosis (86.6% vs. 71.4%) We found significantly higher utilization rates for patients with additional psychiatric diagnoses than for patients without an additional psychiatric diagnosis for almost all kinds of drugs, with the exception of tranquilizing, anticonvulsant and antiparkinson drugs and nootropics. Anticonvulsant drugs were more commonly used for patients with only an F10 diagnosis than for patients with any other additional diagnoses, but the difference was not significant. Antidepressant drugs were given more than twice as often to patients with additional psychiatric diagnoses (50.9% vs. 20.5%).

We further analyzed differences in drug use according to age (i.e., ≤ 30, 31–60, > 60 years). Both the frequency of overall drug use and the use of psychotropic drugs increased with age. The proportion of patients receiving at least one psychotropic drug increased from 71.3% among those ≤ 30 to 77.2% among those > 60 years. A total of 27.9% of patients aged 31–60 years received an antipsychotic drug, while the frequency of using this group of drugs was greater among younger (< 31 years) and older (> 60 years) patients (36.4% and 35.7%, respectively). Tranquilizing and hypnotic drugs were given more frequently with increasing age (Table 6).

Regarding the use of individual drugs, carbamazepine was the most common drug (11.1%), followed by diazepam (10.1%), mirtazapine (8.5%) and oxazepam (8.2%) (Table 7).

Commonly used antipsychotic drugs were quetiapine (6.4%) and haloperidol (5.4%). Clomethiazole was given to 5.8% of all patients. Lorazepam was more commonly used in women (women: 7.1% vs. men: 4.2%, p < 0.001), while diazepam was given more often to men (men: 10.9% vs. women: 8.3%, p < 0.005). The biggest differences between the sexes were found for the antidepressant drugs venlafaxine (women: 7.1% vs. men: 3.5%, p < 0.001), escitalopram (women: 5.8% vs. men: 3.0%, p < 0.001), citalopram (women: 5.5% vs. men: 3.2%, p < 0.001) and trazodone (women: 4.0% vs. men: 1.0%, p < 0.001) and the antipsychotic drug quetiapine (women: 8.3% vs. men: 5.5%, p < 0.001), all of which were given to women more often. On the other hand, men were treated more often with clomethiazole (men: 6.7% vs. women: 4.1%, p < 0.001) (Table S1).

Drugs to support relapse prevention were used in 3.5% of all patients, mainly acamprosate (2.2%), followed by disulfiram (0.9%), naltrexone (0.4%), and nalmefen (0.1%).

As described above, we analyzed drug use in patients with additional psychiatric diagnoses for all drugs used in ≥ 2.5% of patients. Among patients with additional psychiatric diagnoses, mirtazapine (14.2%) and quetiapine (12.7%) were the most commonly used drugs, whereas mirtazapine (6.2%) and quetiapine (5.4%) were used less often in patients without additional psychiatric diagnoses. The use of diazepam and oxazepam was similar in both groups (11.9% vs. 11.4%), 9.4% vs. 10.3% resp.). Antidepressant drugs were given more often to patients with additional psychiatric diagnoses than to those without, e.g., venlafaxine (9.0% vs. 2.4%), escitalopram (6.8% vs. 2.9%), citalopram (5.7% vs. 2.5%) and sertraline (5.0% vs. 1.8%). Anticonvulsant drugs such as carbamazepine (6.3% vs. 10.6%) and oxcarbazepine (3.0% vs. 4.6%) were more often used in patients without additional psychiatric diagnoses. Clomethiazole was given approximately twice as often to patients with only AUD (6.2% vs. 3.3%). Furthermore, haloperidol was the only antipsychotic drug that was more commonly used in patients without additional psychiatric diagnoses (7.0% vs. 2.8%). In contrast, all other antipsychotic drugs were used more often in patients with additional psychiatric diagnoses (Table S2).

Next, we calculated the average prescribed dose in relation to the drug’s DDD, which represents its assumed average maintenance dose per day for its primary indication in adults. We found mean dose equivalents significantly exceeding one DDD for diazepam (1.9 ± 1.6 times DDD, median dose 15 mg/day), venlafaxine (1.7 ± 0.8 times DDD, median dose 150 mg/day), and sertraline (1.7 ± 0.9 times DDD, median dose 75 mg/day). In contrast, relatively low dose equivalents were observed for quetiapine (0.5 ± 0.5 times DDD, median dose 125 mg/day), pipamperone (0.3 ± 0.2 times DDD, median dose 40 mg/day), carbamazepine (0.5 ± 0.2 times DDD, median dose 600 mg/day), prothipendyl (0.3 ± 0.1 times DDD, median dose 80 mg/day), and trazodone (0.5 ± 0.2 times DDD, median dose 150 mg/day) (Table 7).

The use rates of antidepressant (23.8% to 34.9%) and antipsychotic drugs (24.4% to 33.8%) increased, and the use of tranquilizing drugs nearly doubled from the first (2000–2004: 15.7%) to the last period (2013–2016: 30.1%) (Fig. 1). Anticonvulsant (28.4% to 22.9%) and hypnotic drugs (6.0% to 4.2%) as well as the proportion of patients without any drugs (13.2% to 3.0%) decreased (Fig. 1A).

Changes in the utilization spectrum of psychoactive drugs in 4–5-year periods from 2000 to 2016: A main classes of psychotropic drugs, B antidepressant drugs, C antipsychotic drugs. N: 2000–2004: 2168, 2005–2008: 2533, 2009–2012: 2774, 2013–2016: 2857

Both the average number of drugs per patient (3.5 to 4.3) and the average number of psychotropic drugs per patient (1.8 to 2.0) increased over time. With regard to age, we found no significant differences or trends in the number of psychotropic drugs. Among the non-psychotropic drugs, the utilization rates for PPIs increased drastically from the first (2000–2004: 13.2%) to the last period (2013–2016: 39.0%). Additionally, vitamins were more frequently used (35.1% to 48.9%).

Among antidepressant drugs, the use of SNRIs (3.1% to 8.1%), NaSSAs (5.8% to 10.0%) and “other antidepressant drugs” (1.0% to 5.6%) increased steeply from 2000–2004 to 2013–2016, primarily due to the increasing use of venlafaxine (3.1% to 6.0%), mirtazapine (5.2% to 10.0%) and trazodone (0.6% to 3.7%). While the use of citalopram remained stable (3.0% and 2.9%), the use of its enantiomer escitalopram increased from 1.1% in the first period to 4.1% in the last period. The only decreasing utilization rate among the antidepressant drugs was found for TCAs (6.5% to 5.3%) (Fig. 1B).

The large increase in the use of antipsychotic drugs is mainly caused by the increasing utilization rate of SGAs (8.3% to 22.5%) and therein quetiapine (0.9% to 11.9%). The use of low-potency FGAs increased as well (11.8% to 15.3%) (Fig. 1C), among which pipamperone (3.0% to 4.7%) and prothipendyl (1.3% to 3.5%) were used more often.

Tranquilizing drug utilization rates increased over time. For example, oxazepam rates increased from 4.2% to 11.4%, and diazepam rates nearly doubled from the first to the last period (6.6% to 13.0%).

The utilization of hypnotic drugs decreased slightly. In this group, the benzodiazepine analogues zolpidem and zopiclone were given most often, although their use decreased from 2000–2004 to 2013–2016 (zopiclone 2.2% to 1.1%; zolpidem 1.8% to 1.0%).

The utilization rates of carbamazepine (17.5% to 6.1%) and clomethiazole (8.3% to 3.3%) decreased steeply from the first to the last period.