Benzodiazepine use - current trends: evidence review

This paper presents an overview of current knowledge of trends around benzodiazepine-related deaths, hospital admissions, police seizures and prescribing practices in Scotland.

Annex A[55]: Scotland specific literature

Table 1. Scotland specific literature
Author/ Year Aim Methods Results Conclusions Observations/Recommendations for practice
Prevalence and drug seizure
Deeb et al. 2020 [1] Evaluate prevalence and abuse potential of anti-epileptic drugs (AEDs) among prison populations in Scotland Analysed 904 urine samples of admitted and released prisoners over 1 month period 81% of samples found illicit/non-prescribed drugs. Most frequently detected non-prescribed drugs were Benzodiazepines (61%); opiates (54%) and cannabis (47%). High prevalence of illicit substances, particularly benzodiazepines, and AEDs, with majority of AED positive samples containing at least one other illicit substance. Authors note that there is often over-prescribing of gabapentin to individuals who do not meet prescribing criteria.
Information Services Division (ISD) 2019 [2] Provide estimates of national and local/national prevalence of problem drug use in Scotland, for ages 15-64. (Defined as opiate/benzodiazepine use). Estimates drawn from clients in drug treatment services; drug-related hospital admissions; criminal justice social work (CJSW) reports in 2015/16. Estimate range of 55,800 – 58,900 (giving rate of 1.62%). 94% admitted to hospital related to opioids, compared with 82% of all those identified through CJSW, which observed much higher use of benzodiazepines.[56] N/A N/A
Information Services Division (ISD) 2021 [3] Report on hospital stays in relation to drug use diagnosis, including information on number of stays; age; sex; deprivation profile; substances involved and geographical variations across Scotland. Data obtained from general acute inpatient and day case records, and psychiatric inpatient and day case records, from 1996/97 to 2020/21

In 2020/21: 14,310 drug-related hospital stays among 10,437 patients. 270 stays per 100,000 people (decreased from 284 per 100,000 in 2019/20).

Sedative/hypnotic stay rate was 54 per 100,00 (increase from 50 stays 2019/20). Highest rate among 35-44 year olds, although still decreased from previous year. Approx half live in most deprived areas in Scotland. 71% male.

In 2020/21 sedative/hypnotic stays highest across the time series. Opioid-related stays were lower than average, but still accounted for almost half of stays each month. Reductions in number of stays likely influenced by COVID-19 pandemic.
National Records of Scotland 2021 [4] Provide statistics of drug-related deaths in 2020, broken down by age, sex, substances implicated, underlying cause of death and NHS Board. N/A 1339 drug-related deaths in 2020. 93% had more than one substance implicated. Broken down by following: Opioids (implicated in 89% of deaths); benzodiazepines (73%); gapapentin/pregabalin (37%); cocaine (34%); Sharp rise in benzodiazepine-related deaths in past 5 years, from fewer than 200 deaths per year prior to 2016 to nearly 1000 in 2020. 'Street' benzodiazepines implicated in 66% of deaths. N/A
Scottish Government 2019 [5] Presents Official Statistics on drug seizures made by the police in Scotland and the characteristics of those found in possession of drugs. N/A Approx. 354,000 diazepam tablets seized in 2017-18, compared to 2.2 million in 2016-17. Approx. 326,000 etizolam tablets were seized in 2017-18. 6% of drug supply crimes involved diazepam/etizolam. N/A N/A
Scottish Government 2021 [6] Presents Official Statistics on drug seizures made by the police in Scotland and the characteristics of those found in possession of drugs, 2018/19 and 2019/20. N/A

Main class C drugs seized in 2019/20 were 4.9 million benzodiazepine tablets, majority of which were etizolam (compared with 1.8 million in 2018/19).

Main Class A drugs seized were heroin (23kg); cocaine (131kg); ecstasy-type tablets (approx. 28,800 tablets). Main Class B were herbal cannabis (921kg); cannabis resin (649kg) and amphetamines (143kg).

Patterns and motivations of use
McAuley et al. 2021 [7] Commentary on role of benzodiazepines in Scotland's drug-related deaths. N/A [commentary] N/A Authors note market now dominated by NPS-type benzodiazepines/ 'street benzos' as opposed to diverted prescription medications, which have increased risk environment for PWUD. Implement safer supply, drug testing and drug consumption rooms, in addition to addressing socioeconomic inequalities.
MacLeod et al. 2016 [8] Present findings on patterns of NPS use, motivations for use, treatment/legislative consequences. Mixed methods, incl. qualitative interviews and focus groups; two surveys (424 NPS users and 184 service staff). Survey data, SDMD and needle exchange data for Glasgow/Lothian. [technical report]. 41% of survey respondents used benzo NPSs, cited positive effects on sleep and mental health. People who used benzodiazepines emphasised prescription substitution. High prevalence of NPS use and high poly-substance use, with reasons relating to pleasure, price, potency, ease of access, curiosity and influence of peers. Workforce development may benefit from systems approach, involving improving training on NPS and other issues such as mental/sexual health, homelessness etc., along with embedding NPS in development strategies. Improve service engagement of vulnerable populations.
Roe 2020 [9] Explore lived experiences of substance use in Scotland; understand how experiences of time/memory impact on substance use, relationships, treatment engagement. 12 month ethnography using participant observation; qualitative interviews with 15 people; unstructured conversations [PhD thesis]. Frequent use of non-prescribed benzodiazepines among small group of people actively injecting drugs, most commonly with heroin, methadone, alcohol, and cocaine. Motivations for 'street' benzodiazepine use included anxiety management; trauma and PTSD symptom management; management of opiate and/or prescription benzodiazepine withdrawal; pleasure; boredom; price; ease of access and availability. Improve trauma-informed care, including definitions of trauma-informed and its application in practice; strengthen harm reduction approaches not focused on abstinence; embed socio-historical perspectives of drug use/addiction in treatment and recovery services.
Prescribing practices
Counter et al. 2016 [10] Assess prevalence of potentially inappropriate medication (PIMs) in population of community-based multicompartment compliance aid (MCA) users in north-east Scotland. Recorded data for MCAs dispensed by 48 of 50 community pharmacies between June-Oct 2014, with concurrently prescribed medications, patient demographics and Carstairs index of multiple deprivation. 25% of patients prescribed over 10 medications. Significant increase in the risk for at least one PIM associated with female sex; those aged under 80; those living in deprived areas (prescription of >10 medications and prescription of long-acting benzodiazepine). Signficant proportion of MCA users were prescribed PIMs, including drug-drug interactions, with those younger than 80 and those living in poorest areas at greatest risk. Authors state need for more aggressive multidisciplinary approach (involving prescriber, dispending pharmacist and patient) to the review of medications prescribed to MCA users.
Hughes et al. 2016 [11] Explore impact of questions around clinical efficacy, dependence, tolerance and adverse effects of hypnotic and anxiolytic medications on clinical prescribing practice. Review of community‑dispensed prescribing data for patients on the East Practice Medical Center list in Arbroath, Scotland, in 2007, 2011 and 2015. Proportion of patients prescribed benzodiazepines decreased between 2007 and 2015: 83.8% (n = 109) in 2007, 70.5% (n = 122) in 2011, and 51.7% (n = 138) in 2015 (P = 0.006). Proportion of these patients prescribed a non- benzodiazepine drug increased between 2007 and 2015: 30% (n = 39) in 2007, 46.2% (n = 80) in 2011, and 52.4% (n = 140) in 2015 (P = 0.001). Changes in this prescribing practice may reflect the medicalization of insomnia, local changes in prescribing practice and alongside national recommendations. Authors note that locally available Scottish prescribing data can be utilized to look in more detail in primary care prescribing practice, at a single practice level.
Johnson et al. 2018 [12] Identify pattern of benzodiazepine and z-hypnotic prescribing in psychiatric inpatients at discharge and 12 months post-discharge. Retrospective observational longitudinal cohort study of two adult psychiatric wards prescribed benzodiazepine/z-hypnotic drug, June-November 2012, 30% of 74 (n=22) patients were prescribed at discharge, 14 of whom were prescribed long-term. One in three patients prescribed a benzodiazepine or z-hypnotics at discharge, with 1 in 5 receiving continuous long-term treatment (prescriptions) for 12 months post-discharge. Authors note that future strategies using routine patient-level prescribing data may support prescribers to review and minimise inappropriate long-term prescribing.
Scottish Drug Deaths Taskforce 2021 [13] Provide MAT Standards informed response for Benzodiazepine harm reduction N/A [guidance document] N/A N/A Guidance recommends more discussion of benzodiazepine use; empathetic listening; needs based-assessment; establishing a zone of accepted risk; improved benzodiazepine harm reduction (including supported self-reduction, medication assisted detox/stabilisation); prescribing where appropriate; establishment of shared goals.
Torrance et al. 2018 [14] Examine variations in national opioid prescription rates, indicators of prescribing quality, co-prescribing of benzodiazepines and relationship with pain severity. Electronic linkages of opioid prescribing determined from (i) national data from ISD, NHS Scotland and (ii) individual data from Generation Scotland: Scottish Family Health Study. Descriptive analyses performed on national data, multilevel modelling to examine factors associated with prescribing variation. Patients in most deprived areas 3.5 times more likely to receive a strong opioid than least deprived. Significant variation in prescribing rates between geographical areas related to deprivation. Of women aged 25-40 prescribed a strong opioid, 40% were also prescribed a benzodiazepine. Increasing primary care prescriptions of opioids (which may be appropriate for pain management) and common co-prescribing of opioids and benzodiazepines. Authors highlight importance of guidance/protocols for safe/person-centred prescribing; and of continued development of alternative interventions to long-term opioids in chronic pain management.
Torrance et al. 2020 [15] Examine national and regional prescribing rates (2006-2016); identify sociodemographic factors, co-prescriptions and mortality. Analysed data from ISD, NHS Scotland, Health Informatics Centre, National Records of Scotland and Tayside Drug Death Databases. Across period, 4-fold increase of.; 16-fold increase of pregabalin. In 2016 'recurrent users' (three or more prescriptions) had mean age 58.1 yr, were mostly females (62.5%), and were more likely to live in deprived areas. Of these, 60% were co-prescribed an opioid, benzodiazepine, or both (opioid 49.9%, benzodiazepine 26.8%, both 17.1%). Significant increase in gapapentinoid prescribing between 2006-2016, alongside mortality. Most likely to be prescribed were older people, women, and those living in deprived areas. Contribution to DRDs may be more related to illegal use with diversion of prescribed medication. More attention needed to gabapentinoid prescribing, and public health approaches to common factors underlying the aetiology of chronic pain, substance misuse, and drug-related deaths.
Weatherburn et al. 2019 [16] Evaluate impact of protocol to reduce prescribing of benzodiazepines and non-benzodiazepine hypnotics in Dundee practices. Analysed quarterly prescribing data from ISD, NHS Scotland between 2015-2018. Data split into four clusters and standardised to Defined Daily Dose (DDDs) per 1000 registered patients. Interrupted Time Series analysis performed to assess prescribing one year after protocol introduced. Reduction in prescribing of BZDs and Z-drugs across all GP practice clusters, but this related to an ongoing downward trend in prescribing. No significant reduction in prescribing as a result of the protocol. Key point is attention to past/contemporary trends in data. And attention to variables that might be hidden such as variation in intervention application by individuals (GPs here). Future studies could look at impact of utilising electronic decision support at time of prescribing and measure workload.
Harms and mortality
Johnson et al. 2016 [17] Systematic review by ScotPHO to investigate role of benzodiazepines in drug-related mortality. Systematic review of literature from 1970 – 2015, with following inclusion criteria: benzodiazepine supply issues; pharmacokinetic and pharmaco dynamic effects and interactions; adverse drug effects; polydrug use and interactions; toxicology; comorbidity; mortality; fatality and non-fatality information; and acute pharmacological overdose treatment. Increasing availability of unregulated benzodiazepines of unknown content and quality. Evidence of self-medication for anxiety with prescribed and/or illicit benzodiazepine-type drugs. Difficult to understand role in DRDs due to supratherapeutic megadoses; use and polypharmacy further complicating adverse drug effects; low and high drug blood concentrations being present in DRDs; and benzodiazepine-type drugs being associated with increased mortality when used at routine doses. Authors highlight areas for further investigation, including: increasing availability of unregulated benzodiazepines of unknown content/quality; self-medicating use of benzodiazepine (prescribed and non-prescribed); unclear role in DRDs; risk to short/long-term mental health/cognitive problems.
McAuley et al. 2015 [18] Explore circumstances surrounding, and characteristics of individuals involved in, NPS deaths at population level. Exploratory descriptive analysis of DRDs involving NPS recorded by Scottish National Drug Related Death Database (NDRDD). Majority of NPS-implicated DRDs involved benzodiazepine type drugs, mainly Phenazepam. Most involved males, older adults, and polydrug use (mostly opioids. No stimulant drugs were co-implicated). Prominent role of unlicensed benzodiazepines in mortality. A range of harm reduction strategies are needed to prevent future deaths. Consistency of how different pathologists/labs interpret drugs as implicated merits further investigation.
McAuley et al. 2019 [19] Conduct epidemiological analyses of association of NPS injecting (including benzodiazepines) and Hepatitis C virus (HCV) among population level sample of PWID to explore increased risk of BBV to NPS injectors. 5 cross sectional surveys of approx. 13,000 PWID attending services providing equipment, between 2008 and 2016. Logistic regression applied to determine associations of NPS injecting and HCV. Proportion of PWID reporting injecting NPS in previous six months increased from 0.2% to 11% in 2015/16. High prevalence in Lothian NHS Board area and recent experience of homelessness. Significant likelihood of NPS injectors to be HCV positive. NPS injecting likely to increase risk of HCV.  
Van Amsterdam et al. 2021 [20] Systematic review to attempt to explain difference in opioid overdose deaths between Scotland and England/Wales. Systematic review Relatively higher polydrug use, particularly benzodiazepines (esp. etizolam). Etizolam manufactured domestically, but benzodiazepines and gabapentinoids are often prescribed together with opioids and partly diverted to the black market. No concrete explanation in literature for higher polydrug use in Scotland. Recommendations are to carefully restrict prescription of opioids and other depressants where not needed for essential medication.
O'Conner et al. 2016 [21] Evaluate whether the Immunalysis® Benzodiazepines ELISA kit could detect phenazepam, etizolam, pyrazolam, flubromazepam, diclazepam and its metabolite delorazepam. Cross-reactivity was assessed by comparing the absorbance of the drug with that of oxazepam, the reference standard. Uncontrolled benzodiazepines cross-react sufficiently to produce a positive result with the Immunalysis® Benzodiazepine ELISA kit. Cross-reactivity ranged from 79 to 107% for phenazepam, etizolam, pyrazolam, flubromazepam, diclazepam and delorazepam fortified into blood. It is possible to detect these newer benzodiazepines with traditional forensic toxicology laboratory tools Authors emphasise importance of including these benzodiazepines in confirmation tests
O'Conner et al. 2020 [22] Explore scale of designer benzodiazepine use in different Scottish sub-populations, and detection of drugs in post-mortem cases for both drug-related and non-drug related deaths. Two Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) methods developed.
Total of 2,582 samples analysed. Of these, 893 were urine samples from living participants and 1691 were blood samples from deceased individuals. All blood samples were from the post-mortem (PM) cohort and 369 (22%) of the cases were positive for the designer benzodiazepines tested [PhD thesis].
Designer benzodiazepines detected: Diclazepam (212 cases); Delorazepam (339 cases); Lormetazepam (144 cases); Flubromazepam (18 cases); Pyrazolam (9 cases). Benzodiazepines and designer benzodiazepines are widely used in the Scottish population. he Immunalysis® Benzodiazepine ELISA kit can positively identify phenazepam, etizolam, diclazepam, delorazepam, pyrazolam and flubromazepam in blood. Results not reflective of wider general inmate population and not every post-mortem case was tested for designer benzodiazepines.



Back to top