Emergency department toxicology: ASSIST

Between August and November 2023, the 'A Surveillance Study of Illicit Substance Toxicity' (ASSIST) emergency department project made 715 detections of 36 different illicit drugs, in samples from 131 patients. The most detected drug category was depressants (66% of detections), followed by opioids (15%). The most detected individual drug was desmethyldiazepam (11%), followed by cocaine (11%), temazepam (10%) and bromazolam (10%).

Background

The ASSIST study is conducted by the emergency department (ED) at the Queen Elizabeth University Hospital (QEUH) in NHS Greater Glasgow and Clyde (GGC). This project has been funded by the Scottish Government since August 2022.

The aim of the study is to monitor drug trends and associated clinical features through the use of prospective surveillance of ED attendances due to acute illicit drug toxicity. The study uses full toxicological analysis through the biorepository-approved surplus sampling.

The study allows drug profiling and the identification of emerging drugs or changing trends to inform appropriate harm reduction measures and public health responses.

Data collection

The study collects anonymised data through the analysis of standard-of-care clinical data for patients attending QEUH ED due to illicit drug toxicity and surplus serum sample toxicology testing. Surplus serum samples are leftover blood samples, which were taken as part of usual care.

Each ED attendance is counted once in these data. If the same person presented more than once, each attendance would be a separate data point.

Drug detections presented here are of 'illicit drugs' only and were not prescribed to the individual.

Illicit drug definition

The use of the term 'illicit drug' encompasses any substance that is controlled. It excludes:

• legal substances, such as alcohol, nicotine, caffeine and paracetamol
• medications recently prescribed to the individual (e.g. if methadone is detected but the individual is prescribed methadone, it will not be included)
• drugs administered to the individual as part of treatment (by ambulance staff or in hospital).

Toxicology analysis of surplus serum samples

Detections presented are for the substance only. If a metabolite (compound produced when a drug breaks down in the body) is detected, it will be presented as the substance only. If the metabolite and substance are detected, it will also be presented as the substance only.

However, if a drug and a metabolite are both detected but the metabolite could also be a drug, we have included both as it is not possible to say which has been used, if not both.

Data from the study are presented in quarters, based on the date of ED presentation, and are provided in the table below:

Results

The first chart shows the most common drug categories detected in toxicology analysis of surplus serum samples between 17 August 2022 and 16 November 2023 (Q1 to Q5).

The results are shown as the total number of detections. They are broken down by the top five drug categories and presented by study quarter.

This stacked bar chart shows the percentage of detections of drug types within QEUH ED across five quarters - Q1 to Q5. The drugs are grouped into five categories (depressant, opioid, stimulant, cannabinoid and other (including dissociative and psychedelic drugs)). Percentage of detections by category is on the y axis and the time period is on the x axis. The most common category was depressants with 57%, 62%, 60%, 64% and 66% across the time series. The second most common was opioids with 17%, 18%, 16%, 16% and 15% across the time series.

Close

ASSIST hospital toxicology analysis: drug categories by study quarter

Summary

Historic trend

Between 17 August 2022 and 16 August 2023:

• The most commonly detected drug type was depressants, making up 61% of all detections, followed by opioids (17%). The most commonly detected individual drug was cocaine (11%).
• 73% of attendances were male and 27% were female.
• 50% of attendances were aged 34 and under. The most common age category was 25 to 34 years (31%), followed by 35 to 44 (26%) and 16 to 24 (19%).
• The most common outcomes were discharged to home (39%), ward (29%) and police custody (10%).

Update

For the most recent period (17 August to 16 November 2023):

• 332 individual ED attendances related to illicit drug use were identified.
• 131 attendances qualified for surplus sampling toxicology testing (as they were categorised as having a higher clinical severity of toxicity as per research inclusion criteria).

Drug type and category

Of the 131 samples analysed:

• there was a total of 715 detections of 36 substances (found through the biological detection of the drug or its metabolite).
• the average number of drugs detected per sample was six.

Of the 715 detections, the following drugs were the most common:

• desmethyldiazepam: 80 (11% of detections, 61% of samples)
• cocaine: 79 (11% of detections, 60% of samples)
• temazepam: 73 (10% of detections, 56% of samples)
• bromazolam: 71 (10% of detections, 54% of samples)
• oxazepam: 59 (8% of detections, 45% of samples)
• diazepam: 47 (7% of detections, 36% of samples)
• etizolam: 42 (6% of detections, 32% of samples)
• cannabis (tetrahydrocannabinol): 36 (5% of detections, 27% of samples)
• morphine: 33 (5% of detections, 25% of samples)
• gidazepam: 32 (4% of detections, 24% of samples)

Please note: desmethyldiazepam, temazepam and oxazepam are all benzodiazepine drugs but can also be found as a metabolite of diazepam and other benzodiazepines.

The chart below shows the most common depressant drug detected in toxicology analysis of surplus serum samples between 17 August 2022 and 16 November 2023 (Q1 to Q5). The percentages are of all detections.

This line chart shows the percentage of detections of specific depressant drugs within QEUH ED across five quarters - Q1 to Q5. Percentage of detections is on the y axis and the time period is on the x axis. The most common depressant drug across the time series with the exception of Q4 was desmethyldiazepam (Q1; 6%, Q2; 5%, Q3; 8%, Q4: 7% and Q5; 8%). This was the second quarter in which xylazine was detected.

Close

ASSIST hospital toxicology analysis: depressant drugs by study quarter

Depressants were the most common drug category, making up 66% of detections (473).

• Benzodiazepines were detected 422 times (59% of all detections):
  • In total, 14 different types of benzodiazepines were detected, including bromazolam (10%) and etizolam (6%).
  • Desmethyldiazepam made up 11% of all detections, up from 8% in Q5.
• Gabapentinoids were detected 36 times (5% of all detections, from 7% in Q4). Of these detections, 11 were gabapentin and 25 were pregabalin.
• There were seven detections of xylazine (1%).

This line chart shows the percentage of detections of specific opioid drugs within QEUH ED across five quarters - Q1 to Q5. Percentage of detections is on the y axis and the time period is on the x axis. The most common opioid drug across the time series was heroin/morphine (Q1: 4%, Q2: 5%, Q3: 6%, Q4: 5% and Q5: 5%). Nitazenes were detected for the first time in Q4 (seven detections) and increased to 2% of detections (16 detections) in Q5. Detections of noscapine decreased from 3% in Q1 to 0% in Q5.

Close

ASSIST hospital toxicology analysis: opioid drugs by study quarter

Opioids were the second most common drug category, making up 15% of detections (108).

• There were 34 detections of heroin/morphine (5%), 22 of methadone and 19 of codeine (3% of detections respectively).
• The presence of noscapine (a compound found in opium) can suggest exposure to opium-derived substances. There were no detections in Q5, down from 11 in Q4.
• There were 16 detections of nitazenes (2%).

Stimulants were the third most common drug category, making up 13% of detections (90).

• The most common stimulant was cocaine, making up 11% of detections (79).
• The second most common stimulant was MDMA, which was detected in seven samples (1%).

Further findings

Complete clinical data are available for 332 attendances for the most recent period (17 August to 16 November 2023):

• 70% of attendees were male (234) and 30% were female (98).
• 18% (60) of attendees were aged 16 to 24 years and 27% (90) were aged 25 to 34.
• 54% of attendees were aged 35 years and over, with 30% (98) aged 35 to 44, 16% (53) aged 45 to 54 and 8% (28) aged 55 years or older.

This stacked bar chart shows the percentage of QEUH ED ASSIST attendees by age group across five time periods - Q1 to Q5. Percentage of attendees is on the y axis and the time period is along the x axis. The most common age group in Q1, Q2 and Q4 was 25 to 34 years (34%, 34% and 29% respectively). In Q3, patients aged 25 to 34 years and 35 to 44 were equally common (both 29%). The most common age group in Q5 was 35 to 44 years (30%).

Close

Attendances by age

ED outcome records show:

• 134 (40%) patients were discharged home
• 82 (25%) were admitted to a ward
• 44 (13%) were taken into police custody
• 28 (8%) self-discharged
• 28 (8%) were admitted to an intensive care unit, high-dependency unit, critical care unit or died
• 16 (5%) were recorded as 'unknown' or 'other'

Clinical severity outcome (after 28 days) recorded:

• 313 patients were discharged following the attendance
• seven patients either died or remained an inpatient following the attendance
• outcomes for 10 patients were 'unknown'

Additional information

Public Health Scotland (PHS) was provided with these data by QEUH, NHS GGC.

The ASSIST trial is registered with Clinical Trials UK (ID: NCT05329142).

Ethical approval has been granted by the West of Scotland Research Ethics Service (IRAS ref: 313616, REC ref: 22/WS/0047) and surplus sampling methodology through Biorepository Ethics (ref: 22/WS/0020).

The West of Scotland Safe Haven research database hosts the electronic clinical data under IRAS ref: 321198 or REC ref: 22/WS/0163.

This study is sponsored by NHS GGC Research and Innovation and is funded by the Scottish Government.

The testing has been carried out by the LGC group (external website). LGC analyse pseudonymised samples using mass spectrometry and screen against a database of over 3,500 analytes. This testing can detect drugs and metabolites, but this analysis does not imply that specific drugs were implicated in harms.

Further information on the study can be found at Clinical Trials (external website).

Post-mortem toxicology testing for controlled substances

In provisional data for Q3 of 2023, the most common drug types detected in post-mortem toxicology were opioids (67%) and benzodiazepines (59%). The percentage of deaths where cocaine was detected continued to increase to 38% (from 36% in Q2 of 2023), making it the most commonly detected individual substance, followed by heroin/morphine (35%), diazepam (30%), methadone (25%) and bromazolam (24%).

Background

All sudden or unexpected deaths in Scotland are subject to investigation by the Crown Office and Procurator Fiscal Service (COPFS) to determine the cause of death and the need for criminal proceedings. In order to inform these decisions, the COPFS commissions post-mortem toxicology and pathology services across Scotland.

This analysis is based on data from toxicology testing conducted at post-mortem for sudden and unexplained deaths, or where there was suspicion of involvement of controlled drugs.

Post-mortem toxicology testing is carried out by two services in Scotland:

• The Scottish Police Authority Forensic Services (SPA FS) covers deaths occurring in the west, east and parts of the north of Scotland.
• The Department of Clinical Biochemistry at NHS Grampian covers deaths in the far north and north-east of Scotland.

The inclusion of data on deaths in the far north and north-east of Scotland (the areas covered by NHS Grampian) from January 2022 onwards, means that the figures presented after that point cover the whole of Scotland and are different from publications released before October 2023.

This report includes two new periods of data:

• 1 June to 30 June 2023, completing data for Q2 of calendar year 2023; and
• July to September (complete period for 1 July to 31 August 2023, with inclusion of available partial data for September 2023), representing data for Q3 of 2023. This period is the most recent data discussed below and is referred to as 'provisional Q3 2023' throughout.

Remaining outstanding data for deaths occurring in September, which will complete the period Q3 of 2023, will be included in the next RADAR publication. More recent data are not yet available due to the overall time it takes to complete toxicology testing and the transition of testing to SPA FS.

The range of substances routinely analysed is extensive and includes the detection of alcohol, prescribed medicines and controlled drugs. The data within this report will develop further as other new or emerging drugs are added to toxicology screening.

The first chart below provides an indication of controlled drugs detected at post-mortem, in deaths occurring between 1 January 2020 and the most recent time period for which data are available (provisional Q3 2023). As indicated by the line on the chart, data for 2020 and 2021 are based on deaths in the west, east and parts of the north of Scotland. From January 2022, the data include all areas in Scotland.

The chart shows the percentage of forensic toxicology cases testing positive for controlled substances (opioids, benzodiazepines, gabapentin and pregabalin, cocaine). The percentage of cases is on the y axis and the calendar year and quarter of death, between Q1 of 2020 (1 January to 31 March) and provisional Q3 of 2023, is on the x axis. The percentage of deaths positive for opioids increased, from 78% in Q1 of 2020 to 82% in Q2 of 2020, but gradually decreased to 67% in Q3 of 2023. A sharp increase in the percentage of deaths where benzodiazepines were present occurred between Q1 and Q2 of 2020. This percentage remained high throughout 2020, before gradually falling from 73% in Q4 of 2020 to 51% in Q3 of 2022. Since then, benzodiazepine detections have increased, to 59% in Q3 of 2023. Following an isolated increase to a peak of 42% in Q2 of 2021, the percentage of deaths involving gabapentin and pregabalin followed a decreasing trend to Q2 of 2022 (27%), before gradually increasing again to 35% in Q3 of 2023. The percentage of deaths where cocaine was present increased markedly during Q2 of 2020 (to 38%) and also more recently in Q3 of 2023 (to 38%), but otherwise remained relatively stable over the time series (averaging 30%).

Close

Forensic toxicology cases testing positive for controlled substances

The following charts provide an indication of specific opioids and benzodiazepines detected at post-mortem, in deaths where controlled drugs were detected, occurring between 1 January 2020 and the most recent time period for which data are available (provisional Q3 2023).

The chart shows the percentage of forensic toxicology cases testing positive for specific individual opioids (heroin/morphine, fentanyls, methadone, buprenorphine and nitazenes). The percentage of cases is on the y axis and the calendar year and quarter of death, between Q1 of 2020 (1 January to 31 March) and provisional Q3 of 2023, is on the x axis. Heroin/morphine has been the most common substance since Q3 of 2021 – the percentage of deaths testing positive for heroin/morphine was relatively stable over the time series (averaging 37%). Methadone detections increased sharply to 45% in Q2 of 2020 but have been gradually decreasing since then (25% in Q3 of 2023). There was a gradual increase in the number of cases where fentanyls were detected from 1% in Q1 of 2022 to 5% in Q3 of 2023. Deaths with buprenorphine present remained relatively low and stable throughout the time series (averaging 6%). Nitazene-type opioid substances (detected for the first time in Q1 of 2022) have increased slightly but remained uncommon, detected in 2% of deaths in Q3 of 2023.

Close

Forensic toxicology cases testing positive for specific opioids

The chart shows the percentage of forensic toxicology cases testing positive for specific individual benzodiazepines (diazepam, etizolam, clonazolam and bromazolam). The percentage of cases is on the y axis and the calendar year and quarter of death, between Q1 of 2020 (1 January to 31 March) and provisional Q3 of 2023, is on the x axis. Diazepam has been the most commonly detected benzodiazepine since Q2 of 2022. Diazepam detections remained relatively stable for most of the time series, averaging 24% until Q3 of 2022. A marked increase (to 35%) was observed in Q3 of 2022 and, in spite of fluctuations, detections have remained high (30% in Q3 of 2023). Bromazolam (detected for the first time in Q1 of 2022) has increased sharply and was detected in 24% of deaths in Q3 of 2023. Etizolam was the most common benzodiazepine detected for some time, averaging 51% between Q2 of 2020 and Q2 of 2021. Since then, there has been a gradual reduction in etizolam detections to 8% in Q3 of 2023. Deaths with clonazolam present increased in Q2 of 2020, peaking at 12% in Q3 of 2021, before falling to low levels from early 2022.

Close

Forensic toxicology cases testing positive for specific benzodiazepines

Summary

Historic trend

• The most commonly detected drug types were opioids and benzodiazepines.
• The percentage of deaths where opioids were detected have been gradually decreasing from a peak of 82% in Q2 of 2020, to 72% in Q2 of 2023:
  • Heroin/morphine and methadone were the most common opioids detected.
  • Methadone detections increased sharply from 28% in Q1 of 2020, to 45% in Q2 of 2020, before decreasing to 30% in Q2 of 2023.
  • There has been a small but gradual increase in the number of cases where fentanyl-type opioids were detected, from around 1% prior to Q2 of 2022, to 4% in Q2 of 2023.
  • Deaths with buprenorphine present remained relatively low and stable throughout the time series (averaging 6%).
  • Nitazene-type opioids (detected for the first time in Q1 of 2022) increased but remained uncommon, with 7 (1%) cases detected in Q2 of 2023.
• The percentage of deaths where benzodiazepines were detected increased sharply between Q1 and Q2 of 2020, remaining high before gradually falling from 73% in Q4 of 2020, to 51% in Q3 of 2022. A gradual increase (to 61% in Q2 of 2023) has been observed since then.
  • Diazepam has been the most commonly detected benzodiazepine since Q2 of 2022, remaining relatively stable until Q4 of 2022, when detections increased markedly (to 35%) and remained high (35% in Q2 of 2023).
  • Bromazolam (detected for the first time in Q1 of 2022) has increased sharply (21% in Q2 of 2023), replacing etizolam as the second most common benzodiazepine detected from Q1 of 2023.
  • Etizolam was the most common benzodiazepine detected for some time, averaging 51% between Q2 of 2020 and Q2 of 2021. Since then, there has been a gradual reduction in etizolam detections (13% in Q2 of 2023).
  • Clonazolam detections peaked at 12% in Q3 of 2021, decreasing to 1% in Q2 of 2023.
• The percentage of deaths involving other opioids and benzodiazepines have remained relatively stable over time
• Following an isolated increase to a peak of 42% in Q2 of 2021, the percentage of deaths involving gabapentin and pregabalin followed a decreasing trend to Q2 of 2022 (27%), before gradually increasing again to Q2 of 2023 (34%).
• The percentage of deaths with cocaine present was relatively stable over the time series (average 30%), with the exception of an isolated increase in Q2 of 2020 (38%) and an increase to 36% in Q2 of 2023.

Update

For the most recent time period (complete period for 1 July to 31 August 2023, with inclusion of available partial data for September 2023):

• The total number of deaths testing positive for controlled substances was 603. Many of these deaths involved multiple positive detections, therefore the total number of detections is greater than the number of deaths.
• The following drugs or drug types were most commonly detected:
  • opioids: 402 (67%)
  • benzodiazepines: 357 (59%)
  • cocaine: 227 (38%)
  • heroin/morphine: 211 (35%)
  • gabapentin and pregabalin: 210 (35%)
  • diazepam: 183 (30%)
  • methadone: 149 (25%)
  • bromazolam: 142 (24%)
• The percentage of deaths where opioids were detected decreased to 67% (from 72% in Q2):
  • Heroin/morphine detections remained stable at 35% (also 35% in Q2).
  • Methadone detections decreased to 25% (from 30% in Q2).
  • Fentanyl-type opioids remained relatively stable, detected in 5% of deaths (4% in Q2).
  • Nitazene-type opioids increased, but remained uncommon, detected in 2% of deaths (11) (from 1% in Q2 (7)).
• The percentage of deaths where benzodiazepines were detected decreased slightly to 59% (from 61% in Q1):
  • Diazepam detections decreased to 30% (from 35% in Q2).
  • Bromazolam continued to increase and was detected in 24% of deaths (from 21% in Q2).
  • Etizolam has continued to decrease and was detected in 8% of deaths (from 13% in Q2).
• The percentage of deaths where cocaine was detected continued to increase to 38% (from 36% in Q2), making it the most commonly detected individual substance.
• The percentage of deaths involving gabapentin and pregabalin was relatively stable at 35% (from 34% in Q2).
• For the first time in post-mortem toxicology testing, there were detections of xylazine (a veterinary medicine with sedative, analgesic and muscle relaxant properties). This substance was detected in 1% of deaths (5).

Additional information

PHS was provided with post-mortem toxicology testing data for deaths occurring in the west, east and parts of the north of Scotland by Forensic Medicine and Science at the University of Glasgow and SPA FS.

In late 2022, post-mortem toxicology services for the west, east and parts of the north of Scotland were transferred from the University of Glasgow to the Scottish Police Authority Forensic Services (SPA FS). During the period of transition, tests were completed by other laboratory testing sites in the UK. Although testing has now been moved to SPA FS, these testing sites continued to provide support in 2023 and data from both SPA FS and outsourced sites have been included in this report.

Data on deaths occurring in the far north and north-east of Scotland from January 2022 onwards, was supplied by the Department of Clinical Biochemistry at NHS Grampian.

The table shows the total number of deaths testing positive for controlled substances, for each calendar year and quarter.

New drugs (bromazolam, desalkylgidazepam, nitazene-type opioids and xylazine) were detected for the first time when screening was expanded or testing was outsourced to other laboratories. These drugs may have been present before this time but were not being tested for, as they have only recently emerged in drug markets. These data will develop further as new or emerging drugs are added to routine toxicology screening by the SPA FS and NHS Grampian.

Detailed interpretation of the levels of drugs found present, drug interactions, co-morbidities, or other factors relating to death, are outside the scope of this analysis. This analysis does not imply that specific drugs were implicated in deaths nor that deaths were classified as ‘drug-related’, and it does not include consideration of wider causes of death.

It should be noted that increases observed in specific substances within this report may be due to differences in toxicology test approaches (e.g. detection of concentration levels of a particular drug) between outsourced laboratories and previous screening. This may result in increases in substance detection. Further data will be required and monitored to determine the impact of any differences in toxicology screening across laboratories.

Additionally, it is important to note that small numbers of detections for specific substances may result in large percentage differences between quarters. Where it is felt that data should be interpreted with caution due to small numbers, the number of detections has also been provided for context.

As some of the data received from other laboratory testing sites did not include date of death, other date variables have been used as a proxy to improve data completeness and enable the inclusion of these deaths within this report. Two separate date variables have been used to approximate date of death information, where this information was unavailable:

1. Date of the case being received or sent to other labs for toxicology testing.
2. Date of toxicology test being completed.

Similarly, as date of death was unavailable for those tests conducted by the Department of Clinical Biochemistry at NHS Grampian, the date when the case was received from the Crown Office and Procurator Fiscal Service has been used instead.

These dates listed above have been used in the analysis as they are considered to provide a close approximation to the month and year of death. It is anticipated that missing information on date of death will be improved over time, as further information becomes available.

Drug seizures in Scottish prisons

The Scottish Prisons Non-Judicial Drug Monitoring Project is a collaboration between the Scottish Prison Service and the Leverhulme Research Centre for Forensic Science at the University of Dundee. The project tests drug seizures made across the Scottish prison estate in order to understand the changing characteristics of synthetic drugs, including synthetic cannabinoids, often referred to as 'spice'.

Analysis of the drug seizures in Scottish prisons from August to October 2023 is ongoing and is unable to be included in this report.

We anticipate the updated data will be available for the next release of this report in April 2024. Data to July 2023 are available in the previous quarterly report.