Prescription of Sedative Hypnotics in Psychiatric and Somatic Inpatients
Abstract
Abstract: Benzodiazepines (BZD) and Z-drugs (ZD) are often prescribed to treat psychiatric disorders. Despite the risk of dependency and serious adverse events, they are increasingly prescribed to outpatients. Yet little is known about the prescription to inpatients. In this article, we analyze the BZD and ZD prescription patterns in the routine data of 913 psychiatric and somatic inpatients treated in Bochum, Germany. BZD were prescribed more often as regular medication in psychiatry and ZD in somatic hospitals. ZD were more often used on demand, by female patients, and combined with other substances. Both were frequently prescribed to older adults, especially as part of somatic treatment and in combination with opioids. Prescription of sedative hypnotics is frequent in psychiatric and somatic inpatients. Their use seems to be partially independent of recorded psychiatric indications.
Benzodiazepines (BZD) and the so-called Z-drugs (ZD; i.e., zopiclone, zaleplon, and zolpidem) are sedative prescription drugs that inhibit neuronal excitability by increasing the affinity of the γ-aminobutyric acid-A (GABAA) receptor complex. They elicit sedative, hypnotic, anxiolytic, anticonvulsant, and muscle relaxant effects (Drover, 2004). BZD are used to treat several indications, including anxiety, insomnia, epilepsy, and alcohol withdrawal, whereas ZD were introduced in the late 1980s solely for treating insomnia (Donoghue & Lader, 2010). Insomniac symptoms are common in industrial nations and are accompanied by increased morbidity and mortality (Levenson et al., 2015). Although BZD and ZD are efficacious in treating insomnia, they are associated with severe adverse effects, such as impaired driving performance (Gunja, 2013), cognitive deficits (Boeuf-Cazou et al., 2011; Stranks & Crowe, 2014), and negative effects on body balance (Mets et al., 2010). These effects especially concern adults above the age of 65 years (Cusack, 2004). Combining both BZD and ZD with opioids, which is also most pronounced in older adults (Gerlach et al., 2017), even increases the risk of serious adverse events (Boon et al., 2020; Cho et al., 2020). BZD and ZD have been shown to bear a risk of dependency with tolerance, withdrawal symptoms, and rebound insomnia (Lader, 2011; Yen et al., 2015). Nevertheless, prescriptions of sedatives in outpatient treatment have substantially increased (Agarwal & Landon, 2019), especially in older adults (Marra et al., 2015), and concomitant with increased overdose mortality (Bachhuber et al., 2016). Also, the proportion of patients with BZD long-term use increases with age (Jacob et al., 2017). In Germany, about 1.9 million people are addicted to prescription drugs (mainly sedatives; cf. Soyka et al., 2005), which prompted discussion about prescription behavior (Bundesärztekammer, 2007). Although outpatient prescriptions in Germany are regularly assessed using routinely collected data, many prescriptions of sedative hypnotics fail to be reported to the statutory health insurance companies (Hoffmann & Glaeske, 2014), and data on inpatient prescriptions remain scarce. Previous studies point toward the frequent prescription of sedative hypnotics in a UK psychiatric hospital (Haw & Stubbs, 2007) and German general hospitals (Lücke et al., 2018), but no comparison has been made. Thus, the present study compared the prescription of BZD and ZD in psychiatric and somatic settings regarding age, gender, and diagnosis.
Methods
The study was approved by the Ethical Committee of the Department of Medicine at the Ruhr University Bochum (4633-13).
Data Collection
Data of 913 inpatients (accommodated in double-bed rooms) were collected on randomized reference dates throughout 1 year in a psychiatric (n = 447, 52.6% female, 46.91 ± 16.56 years) and somatic university hospital (n = 466, 60.1% female, 67.7 ± 16.85 years) in Bochum. The wards pertained to orthopedic or general internal medicine and major psychiatric entities (personality, psychotic, affective, substance use disorders, and psychogeriatrics). The collected data comprised sociodemographic information (date of birth, gender, migration background, legal career, current legal status) and disease-related information (ward specialization, length of stay, duration of the disease, psychiatric and somatic diagnoses, tobacco use, medication) drawn from the patient’s medical record. We classified the diagnoses according to the International Classification of Diseases (ICD-10-GM Version 2013).
Statistical Analyses
We conducted the statistical analyses using IBM SPSS Statistics Version 26.0. Standard descriptive analyses of clinical and sociodemographic variables are presented as mean (M) with standard deviation (SD), number of subjects (n), and percentage. We used the Mann-Whitney U-test to determine statistically significant differences between psychiatric and somatic cases regarding age, length of stay, number of psychiatric/somatic diagnoses, and number of medications. We used the chi-square statistic to test the relationships between psychiatric/somatic cases and categorical variables like gender, psychiatric/somatic diagnoses, medication, etc. If the conditions for performing Pearson’s chi-square test were not met, we used the exact Fisher’s test. p-values < .05 (*), < .01 (**), and < .001 (***) were interpreted as significant.
Results
Sample Characteristics
Table 1 shows the demographic and illness characteristics. Psychiatric and somatic subgroups differed significantly in age, gender, and length of stay. Somatic diseases were present in 42.5% of psychiatric patients, mostly cardiovascular diseases. Their most common psychiatric diagnoses were mood (F3; 72%), substance-induced disorders (F1; 24.6%), neurotic stress-related (F4; 24.1%), and psychotic disorders (F2; 18.6%). One-third of the somatic patients (30.5%) were diagnosed with a psychiatric disorder, mainly organic mental and substance-induced disorders. The mean number of prescribed drugs was significantly lower in psychiatric patients than in somatic patients.
The Prescription Pattern of Benzodiazepines (BZD)
Table 2 shows the characteristics of BZD (and ZD) prescriptions. There were no gender differences between psychiatric and somatic BZD patients. Patients with BZD were younger than those without BZD in the whole sample (U = 51,081, Z = −2.749, p = .006), but age did not differ in the psychiatric or somatic subgroup. In somatic treatment, adults over 65 received significantly more often BZD (two-thirds of ZD recipients were elderly). In psychiatric treatment, BZD were mostly prescribed for mood, substance-induced and psychotic disorders, whereas in somatic treatment, mainly substance-induced and organic mental disorders were treated with BZD. However, only every other somatic patient receiving BZD (55%) had a psychiatric diagnosis to begin with. There was no difference between BZD prescription as regular medication for treating patients with substance-induced disorders (psychiatric treatment: 18.18%; somatic treatment: 16.39%; χ2(1) = 0.029, p = 1.000, Cramer’s V = 0.022). However, BZD were more often prescribed as medication on demand for the treatment of substance-induced disorders in psychiatric (16.4%) than in somatic (3.3%) treatment (χ2(1) = 4.395, p = .036, Cramer’s V = 0.160), which is explained by a more frequent use of clomethiazole for patients with substance-induced disorders in somatic treatment (88.4%) than in psychiatric treatment (18.0%) (χ2(1) = 69.525, p < .001, Cramer’s V = 0.638). About one-quarter of psychiatric (25.1%) but only 10% of somatic inpatients received BZD, which differed significantly. Whereas in psychiatry, the ratio between regular and on-demand BZD prescription was nearly balanced, BZD in somatic treatment were mainly prescribed as regular medication (87%). Regarding the course of treatment, psychiatric patients received significantly more often regular BZD within their first week, but regular BZD prescription did no longer differ in the subsequent weeks between somatic and psychiatric inpatients.
The Prescription Pattern of Z-Drugs (ZD)
ZD were prescribed significantly more often to female than to male somatic inpatients (χ2(1) = 4.247, p = 0.039, Cramer’s V = 0.095), but not in psychiatry (χ2(1) = 1.469, p = 0.226, Cramer’s V = 0.057). Overall, patients receiving ZD were older than those without (U = 42,679, Z = −3.058, p = .002), but age only differed significantly within the psychiatric subgroup (U = 8,602, Z = −2.083, p = .037). In somatic treatment, adults over 65 received significantly more often ZD (two-thirds of ZD recipients were elderly). Only a minority of somatic patients receiving ZD had a psychiatric diagnosis (33%). In both psychiatric and somatic treatment, ZD were mostly prescribed to patients with mood, substance-induced, and organic mental disorders. ZD were prescribed to about 12% of psychiatric and 17% of somatic inpatients, which was significantly more frequent both as regular as well as on-demand medication. Similar to BZD prescription, the ratio between regular and on-demand ZD prescription was nearly balanced in psychiatric treatment. In contrast, ZD were prescribed mainly on demand (82%) to somatic cases. Regarding the course of treatment, somatic patients received significantly more ZD on demand within their second week, but ZD prescriptions on demand no longer differed in the subsequent weeks between somatic and psychiatric inpatients.
Figure 1 shows the prescription of sedative hypnotics (BZD + ZD) throughout treatment, which only differed significantly between psychiatric and somatic treatment within the first week (20.43% vs. 10.6%; χ2(1) = 6.193, p = 0.013, Cramer’s V = 0.122). 57 out of 162 psychiatric inpatients and 1 out of 5 somatic inpatients were already in treatment longer than 5 weeks and were still receiving sedative hypnotics. Because of the small and different sample sizes, a valid statistical comparison was impossible.
Combination of Sedative Hypnotics with Other Medication
Polypharmacy, most commonly defined as the prescription of five or more drugs to a patient (Masnoon et al., 2017), was frequently observed in both psychiatric (43.4%) and somatic (76.4%) cases (Table 3). It was associated with ZD (χ2(1) = 22.302, p < .001, Cramer’s V = 0.223), but not BZD prescription (χ2(1) = 0.094, p = 0.759, Cramer’s V = 0.014) in psychiatric inpatients. Similarly, somatic patients undergoing polypharmacy significantly more often received ZD (χ2(1) = 12.675, p < .001, φ = 0.165), than BZD (χ2(1) = 15.391, p = 1.000, φ = 0.005). The combination of both BZD and ZD was prescribed to only a small number of patients without significant differences between specializations. A combination of sedative hypnotics with opioids was significantly more common in somatic inpatients, and groups differed significantly in age. There was no gender effect on sedative and opioid comedication in psychiatric (χ2(1) = 1.44, p = .397, Cramer’s V = 0.177) or somatic inpatients (χ2(1) = 0.27, p = .605, Cramer’s V = 0.040).
Discussion
The present study revealed higher BZD prescriptions in psychiatric than in somatic inpatients, both regarding regular and on-demand medication within their first week of treatment. It can be hypothesized that this correlates to the presence of acute psychiatric symptoms leading to hospitalization itself, such as agitation, severe anxiety, or suicidal ideation. The nearly constant BZD prescription rate in somatic treatment suggests a less symptom-led approach. In both specializations, BZD were more frequently prescribed as regular medication, whereas in psychiatric treatment an on-demand prescription was more common. This might partly result from the treatment of substance-induced disorders, which was typically performed with BZD in psychiatry but clomethiazole in somatic treatment. Although the use of BZD may seem unavoidable in severe cases of anxiety, insomnia, and suicidal excitation, there is no single psychiatric symptom or syndrome for which the guidelines recommend them as first choice. BZD quickly alleviate anxiety or sleep problems in the short run but are only marginally effective in the long run and bear the risk of dependency (Lader, 2014). Occasional use can also close the door to further diagnostics and alternative treatments. BZD should thus be avoided whenever possible in favor of alternative pharmacological and nonpharmacological interventions tailored to the respective disorder (e.g., sleep hygiene, cognitive behavior therapy, phytotherapeutics, sedative antidepressants, and low-potency antipsychotics). BZD should be prescribed at the lowest effective dose and for the shortest possible period (Lader, 2014).
In contrast to BZD, ZD were prescribed more frequently as regular medication in psychiatry and more frequently on demand in somatic treatment. The widespread use of ZD on demand in somatic inpatients aligns with a BZD decrease and ZD increase in German prescription rates over the last decades (Hoffmann & Glaeske, 2014; Schwabe & Paffrath, 2015). Only a minority of somatic inpatients being prescribed BZD or ZD had a clinical mental health diagnosis or a diagnosed sleep disorder, which aligns with prior findings (Nobis et al., 2014). The significantly higher ZD prescription to female patients observed in our study might mirror the fact that insomnia is twice as common in women than men (Schlack et al., 2013). Like BZD, the widespread use of ZD should be critically evaluated. Although subject to controversy, a considerable addictive potential of ZD should be assumed as well (Hoffmann & Glaeske, 2014; Yen et al., 2015), and long-term cognitive impairments have been described (Boeuf-Cazou et al., 2011). Considering that acute sleep deprivation during hospital treatment is associated with worse health outcomes, more attention should be paid to nonpharmacological interventions improving sleep conditions (e.g., earplugs, eye masks, calming music, daytime bright light exposure) (DuBose & Hadi, 2016). In cases of severe insomnia, ZD should be preferentially employed after all other interventions have been exhausted, as has been argued similarly for BZD (Lader, 2014).
Our data show that sedative hypnotics are commonly prescribed to adults older than 65 years, also found in outpatients (Jacob et al., 2017) and in other industrial nations (Kurko et al., 2015). However, BZD and ZD proved to have an unfavorable risk-benefit ratio in the treatment of insomnia in older adults (Glass et al., 2005). An age-related increase of BZD receptor sensitivity (Bogunovic & Greenfield, 2004), a decreased metabolism with subsequently increased peak plasma levels (Kinirons & O’Mahony, 2004), and pre-existing balance problems cause an especially pronounced risk of imbalance and falls in the first minutes after wake-up among older adults (Bartlett et al., 2009). Thus, avoiding short-acting BZD and ZD has been strongly recommended in older adults by the German PRISCUS list (Holt et al., 2010) and the US Beers criteria (American Geriatrics Society, Beers Criteria Update Expert Panel, 2015). We also found that ZD on demand in combination with regular opioids was more frequent in older somatic inpatients, which should be critically reflected.
Some limitations must be considered when interpreting the results of the current study. First, we analyzed prescription patterns as part of this study though we have no data on the actual intake of medication. This should be taken into account, especially for “on-demand” medication. Also, because of the cross-sectional study design, our data did not include the duration of use, which might have provided further important insights into prescription patterns. Furthermore, considering the trend of increasing outpatient treatment and decreasing hospitalization, it can be assumed that mainly severely ill, elderly, or socially impaired people are treated on an inpatient basis, although the severity of illness was not assessed. So, somatic cases had a higher mean age and comprised more women than men. This limits the means of statistical comparison between geriatric patients in the different cohorts as well as the overall validity of age and gender effects. However, because mental disorders accompany increased somatic comorbidity and mortality (Caneo, 2018), it can be assumed that the mean biological age of psychiatric cases exceeds their mean nominal age. The fact that the data were acquired in the internal medicine and orthopedic wards of a somatic hospital leads to an overrepresentation of cardiovascular and musculoskeletal disease in the somatic cohort. The latter might be responsible for a potentially higher rate of opioid prescriptions in somatic cases compared to an average patient in other somatic hospitals. No representativeness can be claimed for our sample of 913 patients in two German hospitals in the same city, who were assessed on randomly selected reference dates throughout 1 year.
In light of the side effects discussed, we recommend future studies investigating the reasons why prescription rates of sedative hypnotics remain relatively high in clinical practice and also exploring further pharmacological as well as nonpharmacological alternatives.
Keypoints
- –Sedative hypnotics such as benzodiazepines and Z-drugs are prescribed to outpatients throughout the disciplines, but little is known about prescriptions to inpatients.
- –Prescription rates increased throughout the recent decades, despite the risk of dependency and many serious adverse events.
- –A cross-sectional analysis of 913 inpatients of university hospitals revealed frequent prescriptions of sedative hypnotics on both psychiatric and somatic wards.
- –Psychiatric patients more often received BZD during their first treatment week; female somatic inpatients received more ZD on demand.
- –Both drugs were frequently prescribed to adults older than 65 years, especially in somatic treatment.
The authors gratefully acknowledge the collaborative partners in the somatic and psychiatric university hospitals.
References
(2019). Patterns in outpatient benzodiazepine prescribing in the United States. JAMA Network Open, 2(1), Article
e187399 . https://doi.org/10.1001/jamanetworkopen.2018.7399. (2015). American Geriatrics Society 2015 updated Beers criteria for potentially inappropriate medication use in older adults. Journal of the American Geriatrics Society, 63(11), 2227–2246. https://doi.org/10.1111/jgs.13702
(2016). Increasing benzodiazepine prescriptions and overdose mortality in the United States, 1996–2013. American Journal of Public Health, 106(4), 686–688. https://doi.org/10.2105/AJPH.2016.303061
(2009). Association between risk factors for injurious falls and new benzodiazepine prescribing in elderly persons. BMC Family Practice, 10(1). https://doi.org/10.1186/1471-2296-10-1
(2011). Impact of long-term benzodiazepine use on cognitive functioning in young adults: The VISAT cohort. European Journal of Clinical Pharmacology, 67(10), 1045–1052. https://doi.org/10.1007/s00228-011-1047-y
(2004). Practical geriatrics: Use of benzodiazepines among elderly patients. Psychiatric Services, 55(3), 233–235. https://doi.org/10.1176/appi.ps.55.3.233
(2020). Combining opioids and benzodiazepines: Effects on mortality and severe adverse respiratory events. Annals of Cardiothoracic Surgery, 9(2), 542–557. https://doi.org/10.21037/apm.2019.12.09
. (2007). Medikamente – schädlicher Gebrauch und Abhängigkeit. Leitfaden für die ärztliche Praxis
[Medication – Harmful use and dependence. Guide for medical practice] . Deutscher Ärzte-Verlag. https://www.bundesaerztekammer.de/fileadmin/user_upload/_old-files/downloads/LeitfadenMedAbhaengigkeit.pdf(2018). Managing cardiovascular disease risk in patients with severe mental illness. The Lancet. Psychiatry, 5(2), 97–98. https://doi.org/10.1016/S2215-0366(18)30005-1
(2020). Risk of overdose with exposure to prescription opioids, benzodiazepines, and non-benzodiazepine sedative-hypnotics in adults: A retrospective cohort study. Journal of General Internal Medicine, 35(3), 696–703. https://doi.org/10.1007/s11606-019-05545-y
(2004). Pharmacokinetics in older persons. The American Journal of Geriatric Pharmacotherapy, 2(4), 274–302. http://www.ncbi.nlm.nih.gov/pubmed/15903286
(2010). Usage of benzodiazepines: A review. International Journal of Psychiatry in Clinical Practice, 14(2), 78–87. https://doi.org/10.3109/13651500903447810
(2004). Comparative pharmacokinetics and pharmacodynamics of short-acting hypnosedatives. Clinical Pharmacokinetics, 43(4), 227–238. https://doi.org/10.2165/00003088-200443040-00002
(2016). Improving inpatient environments to support patient sleep. International Journal for Quality in Health Care, 28(5), 540–553. https://doi.org/10.1093/intqhc/mzw079
(2017). Opioids and other central nervous system-active polypharmacy in older adults in the United States. Journal of the American Geriatrics Society, 65(9), 2052–2056. https://doi.org/10.1111/jgs.14930
(2005). Sedative hypnotics in older people with insomnia: Meta-analysis of risks and benefits. BMJ, 331(7526), Article
1169 . https://doi.org/10.1136/bmj.38623.768588.47(2013). In the Zzz zone: the effects of Z-drugs on human performance and driving. Journal of Medical Toxicology, 9(2), 163–171. https://doi.org/10.1007/s13181-013-0294-y
(2007). Benzodiazepines - A necessary evil? A survey of prescribing at a specialist UK psychiatric hospital. Journal of Psychopharmacology, 21(6), 645–649. https://doi.org/10.1177/0269881106072386
(2014). Benzodiazepinhypnotika, Zolpidem und Zopiclon auf Privatrezept
[Benzodiazepine hypnotics, zolpidem and zopiclone on private prescriptions] . Der Nervenarzt, 85(11), 1402–1409. https://doi.org/10.1007/s00115-014-4016-8(2010). Potentially inappropriate medications in the elderly: The PRISCUS list. Deutsches Arzteblatt International, 107(31–32), 543–551. https://doi.org/10.3238/arztebl.2010.0543
(2017). Long-term use of benzodiazepines in older patients in Germany: A retrospective analysis. Therapeutic Advances in Psychopharmacology, 7(6–7), 191–200. https://doi.org/10.1177/2045125317696454
(2004). Drug metabolism and ageing. British Journal of Clinical Pharmacology, 57(5), 540–544. https://doi.org/10.1111/J.1365-2125.2004.02096.X
(2015). Long-term use of benzodiazepines: Definitions, prevalence and usage patterns. A systematic review of register-based studies. European Psychiatry, 30(8), 1037–1047. https://doi.org/10.1016/j.eurpsy.2015.09.003
(2011). Benzodiazepines revisited: Will we ever learn? Addiction, 106(12), 2086–2109. https://doi.org/10.1111/j.1360-0443.2011.03563.x
(2014). Benzodiazepine harm: How can it be reduced? British Journal of Clinical Pharmacology, 77(2), 295–301. https://doi.org/10.1111/j.1365-2125.2012.04418.x
(2015). The pathophysiology of insomnia. Chest, 147(4), 1179–1192. https://doi.org/10.1378/chest.14-1617
(2018). Off-label prescription of psychiatric drugs by non-psychiatrist physicians in three general hospitals in Germany. Annals of General Psychiatry, 17(1), Article
7 . https://doi.org/10.1186/s12991-018-0176-4(2015). Benzodiazepine prescribing in older adults in US ambulatory clinics and emergency departments (2001–10). Journal of the American Geriatrics Society, 63(10), 2074–2081. https://doi.org/10.1111/jgs.13666
(2017). What is polypharmacy? A systematic review of definitions. BMC Geriatrics, 17(1), 1–10. https://doi.org/10.1186/s12877-017-0621-2
(2010). Effect of hypnotic drugs on body balance and standing steadiness. Sleep Medicine Reviews, 14(4), 259–267. https://doi.org/10.1016/j.smrv.2009.10.008
(2014). Therapie mit Psychopharmaka bei internistischen Patienten einer Universitätsklinik
[Psychopharmacological treatment of patients in a university department of internal medicine] . Deutsche Medizinische Wochenschrift, 139(09), 417–422. https://doi.org/10.1055/s-0034-1369817(2013). Häufigkeit und Verteilung von Schlafproblemen und Insomnie in der deutschen Erwachsenenbevölkerung
[Frequency and distribution of sleep problems and insomnia in the adult population in Germany] Bundesgesundheitsblatt – Gesundheitsforschung – Gesundheitsschutz, 56(5–6), 740–748. https://doi.org/10.1007/s00103-013-1689-2Schwabe, U.Paffrath, D. (Eds.). (2015). Arzneiverordnungs-Report 2015
[Prescription drug report 2015] . Springer. https://doi.org/10.1007/978-3-662-47186-9(2005). Wo verstecken sich 1, 9 Millionen Medikamentenabhängige?
[Where are the 1.9 million patients dependent on legal drugs hiding?] Der Nervenarzt, 76(1), 72–77. https://doi.org/10.1007/s00115-004-1828-y(2014). The acute cognitive effects of zopiclone, zolpidem, zaleplon, and eszopiclone: A systematic review and meta-analysis. Journal of Clinical and Experimental Neuropsychology, 36(7), 691–700. https://doi.org/10.1080/13803395.2014.928268
(2015). Correlates of dependence and beliefs about the use of hypnotics among zolpidem and zopiclone users. Substance Use & Misuse, 50(3), 350–357. https://doi.org/10.3109/10826084.2014.980955
