Problematic Medication With Benzodiazepines, “Z-drugs”, and Opioid Analgesics
An analysis of national health insurance prescription data from 2006–2016
; ; ; ; ;
Background: An estimated 1.4 to 2.6 million people in German suffer from drug dependence. Most of them are long-term users of benzodiazepines (BZD), Z drugs (ZD), or opioid analgesics (OA).
Methods: This analysis is based on prescription data from patients of the national statutory health insurance system in the German federal states of Schleswig-Holstein, Hamburg, Bremen, and Lower Saxony. Drug-taking trends, duration, dosage, and long-term use of BZD, ZD, and OA in the years 2006 to 2015 are analyzed; prevalences are estimated for the years 2006 to 2016.
Results: In 2006, 7.7% of patients received at least one prescription for a BZD, ZD, or OA; in 2016, 7.0% did. Over the period of analysis, a marked drop was seen in prescriptions of BZD and a slight fall in prescriptions of ZD (2006: BZD 3.5%, ZS 1.1%; 2016: BZD 2.0%, ZS 0.8%), but there was also an increase in prescriptions of OA, from 4.2% to 4.9%. The number of defined daily doses (DDD) prescribed per year fell for both BZD and ZD. For OA, the number of DDD prescribed per year rose from 2006 to 2009 and decreased by a small amount in subsequent years. The proportions of BZD and ZD patients who had long-term prescriptions fell over time, while the corresponding percentage of OA patients rose.
Conclusion: Nearly one-fifth of all prescriptions for BZD were long-term prescriptions for an entire year, in violation of the relevant guidelines. The rising prevalence of OA use was in the expected range in view of the aging population, but the number of prescriptions rose among younger patients as well. This trend toward more common treatment with opioid analgesics should be critically examined.
Cite this as:
Besides their intended effects, medicines often have unwanted side effects, including the development of prescription drug misuse, potentially leading to dependence. In Germany, an estimated 1.4 to 2.6 million people are dependent on prescription drugs (1–3), primarily sleeping drugs and tranquilizers (benzodiazepines [BZD] or Z-drugs [ZD]) or opioid analgesics (OA). According to a study by Buth et al. (4), 5.1% of the German population were taking BZD (including tetrazepam) and a further 1.1% one of the so-called Z-drugs in 2010. These rates were significantly higher among female patients and especially the elderly. Based on a 3-year period (2006 to 2008), 16.7% of patients who were prescribed BZD and/or ZD were found to show a problematic pattern of use (5). The majority were elderly patients taking the drugs in low doses over very long periods of time. This so-called low-dose dependence can have significant adverse effects on the health of these patients; for instance, it often leads to fractures (6, 7), pneumonia (8), or dementia (9).
OA are indispensable for the treatment of severe pain and in palliative care. However, there is a risk of patients developing opioid dependence (10). In the United States and Canada, physicians have prescribed OA in large quantities over prolonged periods of time for less severe or temporary pain. Many patients developed OA dependence and later switched to more potent analgesics or to heroin (11).
Similarly, the amounts of OA prescribed in Germany increased by 21% between 2007 and 2016 (12). According to a study by Schubert et al. (13), 4.5% of all patients covered by statutory health insurance (Gesetzliche Krankenversicherung, GKV) received an OA (excluding codeine-containing preparations) in 2010. Ten years earlier, the prevalence had been only 3.3% (13). Even though, apart from cancer care, there are few indications for the long-term prescription of OA (14, 15), prolonged treatment with these drugs is common (13, 16).
The aim of this paper is to describe trends in the prescription of BZD, ZD, and OA among the GKV-insured population during the years 2006 to 2016—by sex and age—and to investigate how frequently whole-year prescriptions for these medications were issued.
The analyses reported here are based on the prescriptions issued to GKV patients and filled in pharmacies in the German federal states of Schleswig-Holstein, Hamburg, Bremen, and Lower Saxony during the years 2006 to 2016. Data extraction was performed at the North German Pharmacy Data Center (NARZ/AVN), where 65% to 91% of the pharmacies of these states have their prescriptions processed for billing with the statutory health insurers.
In the analyses, these variations in coverage rates were corrected using weighting. The resulting (weighted) case numbers were then related to the total number of persons insured in these states (around 11 million) (17).
The Anatomical Therapeutic Chemical (ATC) classification system of the World Health Organization (WHO) was used to allocate the active ingredients of the prescribed medicines to the three substance groups evaluated in this paper (18). The ATC index applicable in Germany also includes information about the defined daily doses (DDD) assuming the medications are used as intended. Based on the DDD, the prescribed amounts of active substance per patient and patient year (the 12-month period from the first prescription within a calendar year; see eMethods) can be calculated in standardized form, i.e. without the complex and error-prone calculation of equivalent doses (e.g., diazepam or morphine equivalents).
Within the dataset, each prescription was linked to an individual using a unique, anonymized patient code which does not allow tracing of the patient. The statistical software package SPSS, version 25, was used for weighting and descriptive data analysis (19). Given the high case numbers, confidence intervals were not calculated.
The study was financially supported by the innovation fund of the Federal Joint Committee (Gemeinsamer Bundesauschuss, G-BA; the highest decision-making body of physicians and health insurers in Germany) (funding code 01VSF16049).
Prevalence of prescription of medications with dependence potential
In the course of the year 2006, 7.7% of the GKV-insured persons received at least one prescription for a BZD, ZD, or OA (including multiple prescriptions). In 2016, this proportion had declined to 7.0%. Looking at the individual substance groups, the reduction in the prevalence of BZD from 3.5% in 2006 to 2.0% in 2016 is striking (Figure 1). While the proportion of patients receiving a ZD also decreased slightly, an increase in patients receiving an OA was noted, from 4.2% in the first to 4.9% in the last year of the study. The proportion of patients receiving a prescription of the BZD lorazepam more than doubled between 2006 and 2016, while the proportion of diazepam prescriptions hardly changed and the proportions of oxazepam and bromazepam prescriptions decreased. In 2016, slightly more than two thirds of the ZD patients were prescribed zopiclone, an increase of 10 percentage points compared with 2006. The proportion of zolpidem prescriptions decreased by the same amount during this period. With regard to the WHO step-2 OA, a significant fall in the proportion of tramadol patients was noted, while tilidine prescriptions rose continuously over the years. As for WHO step-3 analgesics, the development of oxycodone prescriptions is striking: the number more than doubled during the analysis period.
In 2006, 4.6% of the female GKV-insured patients but less than half as many of the men, only 2.2%, received BZD (eTable 1).
Likewise, ZD were prescribed around twice as often to women as to men.
A trend towards a slight reduction in prevalence during the study period can be discerned for both ZD and BZD.
The proportion of patients with OA prescriptions was also higher among women—about 1 in 20—than men. During the period covered by our analyses, a slight increase in prevalence of 0.7 percentage points was observed for both sexes.
Furthermore, eTable 1 shows that prevalence rates continuously increased with advancing age. For example, in 2006 only 0.6% of the 29-year-olds but 11.8% of the 75+ age group received a BZD prescription. At the same time, the reported decrease in prevalence is attributable predominantly to a change in the prescription pattern for the elderly. For ZD, a similar trend is noted; however, the prevalence rates are lower and the changes during the study period less pronounced than for BZD.
The most noticeable differences among the analyzed age groups were found for OA. For example, in 2006 only 2.0% of the 30- to 44-year-olds were prescribed OA, but 15.4% of those 75 and older. A slight increase is noted in all age groups in the following years.
Amounts of medications prescribed (DDD) and prescription duration
Figure 2 shows the quantity of drugs prescribed per patient year, based on DDD. For both BZD and ZD, a significant reduction is noted. However, this is due solely to a decrease in prescription duration over the years. For example, the average duration of ZD prescriptions declined from 129 days in 2006 to 109 days in 2015. A comparable reduction was observed for BZD. However, the average daily dose prescribed during the prescription period did not change and was approximately 0.7 DDD for both groups.
The continuous reduction in prescribed doses and prescription duration occurred to the same extent in both men and women (Table 2). The absolute values for both sexes exceed the values for the total group, because sex was not stated for all participants (between 88% and 95% per year; for further details see the Limitations section and the eMethods). The decreases in prescription duration and prescribed dose are most pronounced in elderly patients. In the 75+ age group, the average duration of BZD prescription went down by about 38 days and the average prescribed amount of BZD declined by 28 DDD. Notably, the decreases were smaller (−16 days and −15 DDD) among 45- to 59-year-olds. These differences were less pronounced for ZD.
In the BZD group, the average prescribed daily amount of active substance decreased with increasing age. For example, while the 30- to 44-year-olds were prescribed 0.9 DDD on average, the 60- to 74-year-olds were prescribed 0.7 DDD and the oldest patients even less, 0.6 DDD. In contrast, no such age-related differences were found for ZD.
Among the OA patients, the prescribed daily doses showed a steady slight increase between 2006 and 2009, followed by a continuous decrease (Table 2). However, prescription duration increased from 128 days in 2006 to 141 days in 2015. Consequently, the average prescribed amount of medicine varies between 0.7 DDD and 0.8 DDD per day, depending on the observation year.
These trends in OA prescription differed hardly at between male and female patients or among age groups. However, lower amounts of medicine were prescribed to the oldest patients (mean daily dose 0.6 to –0.7 DDD) than, for example, to the 30- to 44-year-olds (mean 0.9 DDD).
Figure 3 shows the proportion of patients who were prescribed a drug belonging to the respective active substance group at least once per quarter during a patient year. While in 2006 the figure was close to 25% in all three substance groups, the groups differed in how they developed over the following years. While for BZD and ZD there were steady decreases to 18.9% and 19.6%, respectively, the prevalence among OA patients increased by about 4% during the same period.
In all substance groups, the prevalence rates of long-term prescriptions were higher for women than for men (eTable 2). However, far larger differences were noted among the various age groups. In 2006, 14% of the 30- to 44-year-old BZD patients were prescribed the medication for a whole year, against one third of the 75+ age group. While in the group of the 30- to 44-year-old BZD patients this proportion fell by only 2% over the following years, a decrease of almost 10% was noted for the oldest BZD patients. The differences were far less pronounced for patients taking ZD.
An increase in the prevalence of long-term prescriptions with advancing age was also noted among OA patients. However, the proportions increased almost continuously in all age groups between 2006 and 2015 (eTable 2).
The findings presented above demonstrate that BZD and ZD have lost some of their importance as prescription drugs in recent years. However, patients in the 60+ age group still commonly use BZD and take them over prolonged periods of time—usually many years. Nevertheless, among these patients the amount of drug administered is usually low and in most cases the dose is not increased. The negative consequences of such low-dose dependence can still be severe, with particular reference to fractures related to falls following intake of the medication (6, 20, 21).
Furthermore, in clinical practice ZD are now often prescribed instead of BZD (22). It is now generally accepted that especially the prolonged use of zolpidem can result in misuse and dependence (23–25). Furthermore, use of ZD—like BZD—significantly increases the risk of falls, and thus of fractures (26). In our study, both the prevalence rates for the (long-term) prescription of ZD and the size of the prescribed dose showed a slightly decreasing trend. There seems to be growing awareness among medical practitioners that ZD are not a risk-free alternative to BZD.
In 2016, about one-fifth of the patients treated with either BZD or ZD received their respective medications over a whole year (or longer). The reasons for such long-term use cited by Olfson et al. (27) include, among others, treatment of persistent anxiety disorders with BZD, the lack of promising alternative treatment options (especially for sleep disturbances), the lack of motivation of older patients to reduce or discontinue their use of this medication, and the time constraints on physicians. Furthermore, Martinsson et al. (28) showed in their analysis of Swedish prescription data for the years 2006 to 2008 that geriatricians and psychiatrists prescribed psychotropic drugs less frequently than primary care physicians. Possibly the latter lacked both specific knowledge of the risks associated with the use of BZD and ZD in older patients and awareness of alternative treatment strategies.
With regard to OA, our study results indicate that epidemic use, as seen in North America (29), is currently not found in Germany.
Nevertheless, the increase in long-term OA prescription arouses concern. While cancer patients may generally require OA treatment over prolonged periods of time, there are few other conditions where treatment for longer than 6 months is indicated (14).
In conclusion, it can be stated that the treatment of pain patients with opioids is certainly effective, but always associated with risks, which need to be assessed and evaluated on an individual basis. The current German clinical practice guidelines on long-term opioid treatment for chronic non-cancer pain provide valuable guidance (14). However, additional measures should be implemented to limit the number of opioid prescriptions to the level which is medically required. In this respect, White (30) and Compton et al. (31) mention, apart from continuing medical education for prescribing physicians, take-back programs for drugs that are no longer needed, increased use of OA containing naloxone as an additional active ingredient, and especially the establishment of drug monitoring programs. Using the methodology we chose for this study, such a monitoring program could be implemented in the short term and at reasonable financial outlay for OA, hypnotics and sedatives, and any other substance group of interest (32).
The prescriptions of BZD and ZD billed with the German national health insurances (GKV) only account for a portion of all prescribed medications. For 2012, for example, it is assumed that the BZD billed to the GKV accounted for about 45% of all dispensed BZD packages (ZD: 50%) (33). Furthermore, it cannot be ruled out that some patients had some of their prescriptions filled in pharmacies billing via the NARZ and some in pharmacies using other settlement agents. In these patients, both the prescribed dose and the prescription duration would be underestimated.
Another limitation is the assumption that all tablets in the prescribed packages are taken. In clinical practice, however, physicians often recommend lower doses or the patients either stop taking the medication early or continue taking it but at a lower dose (34). In such cases, the amount actually taken would be overestimated.
Patients whose sex was not known (between 5% and 12%, depending on the year) were found to have a significantly shorter duration and lower dose for all three substance groups analyzed. Thus, in the sex-related analyses, these parameters are slightly overestimated with regard to their absolute values.
This study was solely funded by the innovation fund of the Federal Joint Committee (G-BA).
Conflict of interest statement
PD Uwe Verthein has received reimbursement of travel expenses from Sanofi Aventis and Molteni Farmaceutici as well as reimbursement of travel expenses and lecture fees from Mundipharma.
The remaining authors declare that no conflict of interest exists.
Manuscript received on 27 December 2018, revised version accepted on 29 May 2019
Translated from the original German by Ralf Thoene, MD.
PD Dr. phil. Uwe Verthein
Zentrum für Interdisziplinäre Suchtforschung
der Universität Hamburg (ZIS)
Klinik für Psychiatrie und Psychotherapie
20246 Hamburg, Germany
Cite this as:
Buth S, Holzbach R, Martens MS, Neumann-Runde E, Meiners O,
Verthein U: Problematic medication with benzodiazepines, “Z-drugs”, and opioid analgesics—an analysis of national health insurance prescription data from 2006–2016. Dtsch Arztebl Int 2019; 116: 607–14. DOI: 10.3238/arztebl.2019.0607
For eReferences please refer to:
eMethods section, eTables:
Department of Psychiatry and Psychotherapy, University Medical Center Hamburg-Eppendorf, Hamburg: Dr. phil. Sven Buth M. A., Dipl.-Psych. Marcus-Sebastian Martens, Dipl.-Psych.
Eike Neumann-Runde, PD Dr. phil. Uwe Verthein
Department of Psychiatry, Psychotherapy, and Psychosomatics, Hochsauerland Hospital Group, Arnsberg: Dr. med. Rüdiger Holzbach
North German Pharmacy Data Center/Pharmacy Billing Center (NARZ/AVN):
Dipl.-Inf. Ommo Meiners
|1.||Glaeske G, Hoffmann F: Medikamente – Psychotrope und andere Arzneimittel mit Missbrauchs- und Abhängigkeitspotenzial. In: Deutsche Hauptstelle für Suchtfragen, ed.: Jahrbuch Sucht 2014. Lengerich: Pabst Science Publishers 2014.|
|2.||Gomes de Matos E, Atzendorf J, Kraus L, Piontek D: Substanzkonsum in der Allgemeinbevölkerung in Deutschland. SUCHT 2016; 62: 271–81 CrossRef|
|3.||Pabst A, Kraus L, Gomes de Matos E, Piontek D: Substanzkonsum und substanzbezogene Störungen in Deutschland im Jahr 2012. SUCHT 2013; 59: 321–31 CrossRef|
|4.||Buth S, Holzbach R, Rosenkranz M, Verthein U: Der Gebrauch von Medikamenten mit Abhängigkeitspotenzial in Deutschland. Eine prospektive Analyse kassenärztlicher Verschreibungen der Jahre 2006 bis 2010. Bundesgesundheitsblatt 2017; 60: 865–72 CrossRef MEDLINE|
|5.||Verthein U, Holzbach R, Martens MS, Buth S: Langzeitverordnungen von Benzodiazepinen und Z-Substanzen – deskriptive Analysen über einen 3-jährigen Zeitraum. Suchttherapie 2018; 19: 99–106 CrossRef|
|6.||Xing D, Ma XL, Ma JX, Wang J, Yang Y, Chen Y: Association between use of benzodiazepines and risk of fractures: a meta-analysis. Osteoporos Int 2014; 25: 105–20 CrossRef MEDLINE|
|7.||Bakken M, Engeland A, Engesæter L, Ranhoff A, Hunskaar S, Ruths S: Risk of hip fracture among older people using anxiolytic and hypnotic drugs: a nationwide prospective cohort study. Eur J Clin Pharmacol 2014; 70: 873–80 CrossRef MEDLINE PubMed Central|
|8.||Chen TY, Winkelman JW, Mao W-C, Liu CL, Hsu CY, Wu CS: The use of benzodiazepine receptor agonists and the risk of hospitalization for pneumonia. CHEST 2018; 153: 161–71 CrossRef MEDLINE|
|9.||Islam MM, Iqbal U, Walther B, et al.: Benzodiazepine Use and Risk of Dementia in the Elderly Population: A Systematic Review and Meta-Analysis. Neuroepidemiology 2016; 47: 181–91 CrossRef MEDLINE|
|10.||Vowles KE, McEntee ML, Julnes PS, Frohe T, Ney JP, van der Goes DN: Rates of opioid misuse, abuse, and addiction in chronic pain: a systematic review and data synthesis. Pain 2015; 156: 569–76 CrossRef MEDLINE|
|11.||Helmerhorst GTT, Teunis T, Janssen SJ, Ring D: An epidemic of the use, misuse and overdose of opioids and deaths due to overdose, in the United States and canada. Bone Joint J 2017; 99-B: 856–64 CrossRef MEDLINE|
|12.||Schwabe U, Paffrath D, Ludwig WD, Klauber J: Arzneiverordnungs-Report 2017. Berlin, Heidelberg: Springer Verlag 2017 CrossRef|
|13.||Schubert I, Ihle P, Sabatowski R: Zunahme der Opioidverordnungen in Deutschland zwischen 2000 und 2010: Eine Studie auf der Basis von Krankenkassendaten. Dtsch Arztebl 2013; 110: 45–51 VOLLTEXT|
|14.||Häuser W, Bock F, Engeser P, Tölle T, Willweber-Strumpf A, Petzke F: Clinical practice guideline: long-term opioid use in non-cancer pain. Dtsch Arztebl Int 2014; 111: 732–40 CrossRef MEDLINE PubMed Central|
|15.||Els C, Jackson TD, Kunyk D, et al.: Adverse events associated with medium- and long-term use of opioids for chronic non-cancer pain: an overview of Cochrane Reviews. Cochrane Database Syst Rev. 2017;10: CD012509 CrossRef PubMed Central|
|16.||Marschall U, L‘Hoest H, Radbruch L, Hauser W: Long-term opioid therapy for chronic non-cancer pain in Germany. Eur J Pain 2016; 20: 767–76 CrossRef MEDLINE|
|17.||Bundesministerium für Gesundheit: GKV-Versicherte nach Alter und Wohnort. GKV-Statistik KM6. 2006–2016.|
|18.||GKV-Arzneimittelindex im Wissenschaftlichen Institut der AOK (WIdO): Das anatomisch-therapeutisch-chemische Klassifikationssystem mit Tagesdosen. Amtliche Fassung des ATC-Index mit DDD-Angaben für Deutschland im Jahre 2016. Berlin 2016.|
|19.||IBM: IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY 2017.|
|20.||Nurminen J, Puustinen J, Lähteenmäki R, et al.: Handgrip strength and balance in older adults following withdrawal from long-term use of temazepam, zopiclone or zolpidem as hypnotics. BMC Geriatr 2014; 14: 121 CrossRef MEDLINE PubMed Central|
|21.||Khong TP, de Vries F, Goldenberg JSB, et al.: Potential impact of benzodiazepine use on the rate of hip fractures in five large European countries and the United States. Calcif Tissue Int 2012; 91: 24–31 CrossRef MEDLINE PubMed Central|
|22.||Hoffmann F: Perceptions of German GPs on benefits and risks of benzodiazepines and Z-drugs. Swiss Med Wkly 2013; 143: w13745 CrossRef MEDLINE|
|23.||Yen CF, Yen CN, Ko CH, et al.: Correlates of dependence and beliefs about the use of hypnotics among Zolpidem and Zopiclone users. Subst Use Misuse 2015; 50: 350–7 CrossRef MEDLINE|
|24.||Victorri-Vigneau C, Gérardin M, Rousselet M, Guerlais M, Grall-Bronnec M, Jolliet P: An update on Zolpidem abuse and dependence. J Addict Dis 2014; 33: 15–23 CrossRef MEDLINE|
|25.||Rousselet M, Feuillet F, Gerardin M, Jolliet P, Hardouin JB, Victorri-Vigneau C: The french addictovigilance network clinical assessment: Z-drugs, true false twins. Expert Opin Drug Saf 2017; 16: 1063–9 CrossRef MEDLINE|
|26.||Treves N, Perlman A, Kolenberg Geron L, Asaly A, Matok I: Z-drugs and risk for falls and fractures in older adults—a systematic review and meta-analysis. Age Ageing 2018; 47: 201–8 CrossRef MEDLINE|
|27.||Olfson M, King M, Schoenbaum M: Benzodiazepine use in the United States. JAMA Psychiatry 2015; 72: 136–42 CrossRef MEDLINE|
|28.||Martinsson G, Fagerberg I, Wiklund-Gustin L, Lindholm C: Specialist prescribing of psychotropic drugs to older persons in Sweden—a register-based study of 188 024 older persons. BMC Psychiatry 2012; 12: 197 CrossRef MEDLINE PubMed Central|
|29.||Rudd RA, Seth P, David F, Scholl L: Increases in drug and opioid-involved overdose deaths – United States, 2010–2015. MMWR Morb Mortal Wkly Rep 2016; 65: 1445–52 CrossRef MEDLINE|
|30.||White ND: Opioid abuse and overdose: prevention strategies. Am J Lifestyle Med 2015; 9: 412–5 CrossRef|
|31.||Compton WM, Boyle M, Wargo E: Prescription opioid abuse: Problems and responses. Prev Med 2015; 80: 5–9 CrossRef MEDLINE|
|32.||Holzbach R, Martens M, Kalke J, Raschke P: Zusammenhang zwischen Verschreibungsverhalten der Ärzte und Medikamentenabhängigkeit ihrer Patienten. Bundesgesundheitsbl 2010; 53: 319–25 CrossRef MEDLINE|
|33.||Hoffmann F, Glaeske G: Benzodiazepinhypnotika, Zolpidem und Zopiclon auf Privatrezept. Nervenarzt 2014: 1–7 CrossRef MEDLINE|
|34.||Knopf H, Grams D: Arzneimittelanwendung von Erwachsenen in Deutschland. Bundesgesundheitsblatt 2013; 56: 868–77. CrossRef MEDLINE|