Drug-related hospital statistics
Scotland 2019 to 2020

A National Statistics publication for Scotland

Published: 15 June 2021
Type: Statistical report
Author: Public Health Scotland

Drugs
Hospital care

Methods

SMR01 – General acute inpatient and day case return

Information about stays in general acute hospitals, where drug misuse was diagnosed as a factor in the patient's treatment (select ‘General acute’ from the Hospital type box) are derived from the general acute inpatient and day case return (SMR01).

SMR01 is an episode based patient record relating to all inpatient and day cases discharged from hospitals (including paediatric facilities) in NHS Scotland. It doesn’t include records from mental health, maternity, neonatal and geriatric long stay specialities. The SMR01 basic data set encompasses patient identification and demographic information, episode management information and general clinical information. Items such as waiting time for inpatient or day case admission and length of stay may be derived from the episode management information. A record is generated for each inpatient and day case episode, of which there are about 1,200,000 each year. Attendances at Accident and Emergency Departments that do not result in an admission are not included. Up to six diagnoses are recorded per SMR01 episode.

SMR04 – Mental health inpatient and day case return

Information about stays in psychiatric hospitals, where drug misuse was diagnosed as a factor in the patient's treatment (select ‘Psychiatric’ from the Hospital type box) are derived from the mental health inpatient and day case return (SMR04).

On the SMR04 form, up to six separate diagnoses can be recorded on both the admission and discharge parts of the record. Diagnosis on discharge may differ from diagnosis on admission. Discharge diagnoses are reported in these statistics as they are regarded as more accurate than admission diagnoses. A diagnosis in the first position is regarded as the main diagnosis. A diagnosis in any of the six positions (main and supplementary) is referred to as ‘in any position’.

SMR01 and SMR04 – 'Any hospital type'

The data presented in the combined analysis (select 'Any hospital type' from the Hospital type box) are derived from both general acute (SMR01) and psychiatric (SMR04) drug-related hospital records.

Combined analysis of stays includes all general acute and psychiatric activity. Patients are counted only once per financial year, even though the same patient may have stayed in both general acute and psychiatric hospitals on multiple occasions in that time period.

Analytical definitions

A period of health care in a hospital setting is known as a continuous inpatient stay (CIS). A CIS is made up of individual episodes (where the patient is under the care of an individual consultant). A patient may have more than one stay and hence the number of patients in a specific financial year can be less than the total number of stays for that period. Also, patients may have drug-related stays in multiple geographical areas during a financial year, meaning that the sum of stays across all geographical areas may not equal the Scotland total.

For the purposes of this analysis, a CIS is counted as associated with drug misuse if any of the episodes of which it is made up include a drug misuse diagnosis in any position (main position refers to primary diagnosis and five supplementary positions refer to secondary diagnoses). Drug misuse is recorded using the International Classification of Diseases 10th Revision (ICD10) Codes. The following codes were used in this analysis:

i) To define a drug-related hospital stay (referred to as Any diagnosis in the dashboard):

Table A1.1: Drug-related hospital stay diagnosis codes

ICD 10 Code	Description
F11	Mental and behavioural disorders due to: Opioids
F12	Mental and behavioural disorders due to: Cannabinoids
F13	Mental and behavioural disorders due to: Sedatives/Hypnotics
F14	Mental and behavioural disorders due to: Cocaine
F15	Mental and behavioural disorders due to: Other Stimulants
F16	Mental and behavioural disorders due to: Hallucinogens
F18	Mental and behavioural disorders due to: Volatile Solvents
F19	Mental and behavioural disorders due to: Multiple/Other Drugs
T40.0	Poisoning by narcotics: Opium
T40.1	Poisoning by narcotics: Heroin
T40.3	Poisoning by narcotics: Methadone
T40.5	Poisoning by narcotics: Cocaine
T40.6	Poisoning by narcotics: Unspecified Narcotics
T40.7	Poisoning by narcotics: Cannabis
T40.8	Poisoning by narcotics: LSD
T40.9	Poisoning by narcotics: Unspecified Hallucinogens

For the T-codes listed below, a CIS is counted if there is a presence in the same CIS of at least one of the ICD-10 mental and behavioural disorder codes F11-F16, F18 or F19

T40.2	Poisoning by narcotics: Other opioids
T40.4	Poisoning by narcotics: Other synthetic narcotics
T42.3	Poisoning by antiepileptic, sedative-hypnotic and antiparkinsonism drugs: Barbiturates
T42.4	Poisoning by antiepileptic, sedative-hypnotic and antiparkinsonism drugs: Benzodiazepines
T43.6	Poisoning by psychotropic drugs NEC: Psychostimulants with abuse potential
T52	Toxic effect of organic solvents

ii) To define a drug-related mental and behavioural hospital stay (referred as Mental & Behavioural in the dashboard)

Table A1.2: Mental and behavioural hospital stay diagnosis codes

ICD 10 Code	Description
F11	Mental and behavioural disorders due to: Opioids
F12	Mental and behavioural disorders due to: Cannabinoids
F13	Mental and behavioural disorders due to: Sedatives/Hypnotics
F14	Mental and behavioural disorders due to: Cocaine
F15	Mental and behavioural disorders due to: Other Stimulants
F16	Mental and behavioural disorders due to: Hallucinogens
F18	Mental and behavioural disorders due to: Volatile Solvents
F19	Mental and behavioural disorders due to: Multiple/Other Drugs

iii) To define a drug-related overdose hospital stay (referred as Overdose in the dashboard)

Table A1.3: 'Overdose' hospital stay diagnosis codes

ICD 10 Code	Description
T40.0	Poisoning by narcotics: Opium
T40.1	Poisoning by narcotics: Heroin
T40.3	Poisoning by narcotics: Methadone
T40.5	Poisoning by narcotics: Cocaine
T40.6	Poisoning by narcotics: Unspecified Narcotics
T40.7	Poisoning by narcotics: Cannabis
T40.8	Poisoning by narcotics: LSD
T40.9	Poisoning by narcotics: Unspecified Hallucinogens

*For the T-codes listed below, a CIS is counted if there is a presence in the same CIS of at least one of the ICD-10 Mental and Behavioural Disorder codes F11-F16, F18 or F19*

T40.2	Poisoning by narcotics: Other opioids
T40.4	Poisoning by narcotics: Other synthetic narcotics
T42.3	Poisoning by antiepileptic, sedative-hypnotic and antiparkinsonism drugs: Barbiturates
T42.4	Poisoning by antiepileptic, sedative-hypnotic and antiparkinsonism drugs: Benzodiazepines
T43.6	Poisoning by psychotropic drugs NEC: Psychostimulants with abuse potential
T52	Toxic effect of organic solvents

For data presented on drug type, there is an element of double counting as stays, patients and ‘new patients’ may each be associated with multiple drug types (e.g. diagnoses of both opiate and cocaine misuse). If multiple drugs have been noted in case notes, the advised coding is to record each substance in a separate diagnosis position where possible. Sometimes the coder may be forced to use the unspecific ICD-10 code F19 (‘multiple/other drugs’), for example, if case notes only state ‘multiple/other drugs’ there is no way of identifying which substances were involved. Sometimes the F19 code may be used if the patient has many other diagnoses recorded, leaving insufficient space to record specific drugs separately.

When gathering information from stays, demographic data (age, gender, deprivation quintile) are extracted from the first episode of the stay (thus corresponding most closely to the circumstances of the patient at the point they entered hospital). However, the allocated year is defined by the date of discharge. Therefore, a stay spanning two financial years (e.g. 2012/13 and 2013/14) will be counted as having occurred in the most recent of those years, or when the patient was discharged (2013/14 in this example).

Some caution is necessary when using these data as (a) drug misuse may only be suspected and may not always be recorded by the hospital, and (b) where drug misuse is recorded, it may not be possible to identify which drug(s) may be involved.

In the length of stay analysis, length is measured from the date of initial admission of the CIS to the ultimate date of discharge for that stay.

An inpatient admission is categorised as an emergency, urgent or routine inpatient admission except for maternity and neonatal admissions. The appropriate admission category depends on the clinical condition of the patient as assessed by the receiving consultant. This measure is not standardised. More details can be found in the PHS Data Dictionary.

When figures are broken down by geographical area or age the numbers in some categories can be very small. In these cases both differences between categories and trends over time should be interpreted with caution because they may be misleading.

Statistical disclosure control has been applied to protect patient confidentiality. Therefore, the figures presented in these statistics may not be additive and may differ to those reported in previous publications.

EASR

European Age-sex Standardised Rates (EASRs) are calculated for hospital activity indicators because the overall rate may vary with the age-sex structure of the populations. The direct standardisation method was used, with the age-sex specific rates of the local population applied to the age-sex structure of a standard population. This gives the overall rate that would have occurred in the local population if it had the same age-sex profile as the standard population. It allows valid comparisons to be made between local areas and other countries with differing population age-sex structures. In the dashboard, EASRs are expressed per 100,000 population per financial year.

The latest available National Records of Scotland mid-year population estimates were used in EASR calculations for NHS Board and Alcohol and Drug Partnership (ADP) analysis and for Scottish Index of Multiple Deprivation analysis. At the time of the analysis, the latest estimates were for the year 2019.

The European Standard Population (ESP) is used to calculate EASRs within this publication. The ESP, which was originally introduced in 1976, was revised in 2013. Before publication of 2012/13 data in February 2014, the Drug Related Hospital Statistics publication used ESP1976 to calculate EASRs. Since 2014, the ESP2013 has been used to calculate EASRs for all years (including those before 2012/13). Therefore, findings from publications since February 2014 are not comparable with earlier publications. See Appendix A1 in the 2013/14 report (PDF) for further details.

Deprivation

Information on deprivation is reported by the Scottish Index of Multiple Deprivation (SIMD) quintiles in the dashboard.

Socio-economic deprivation describes a range of individual and environmental factors whose effects can accrue over time. Information describing income, employment, education and other measures of affluence or deprivation are not readily available but an estimate can be made by measuring characteristics of the area in which an individual resides. If an area is identified as deprived, this can relate to the fact that the people who live there have a low income, it can also mean that there are fewer resources and opportunities in that area. SIMD has over 30 indicators in 7 domains (income, employment, education, health, housing, geographical access to services and crime), which are combined into an overall index. Neighbourhoods are ranked on the basis of their SIMD score and assigned to equally-sized groups representing different levels of deprivation (five groups (quintiles) in this instance).

Drug-related general acute hospital data are used as the basis of the indicator ‘Hospital stays (CIS) related to drug misuse: standardised ratio’ within the Health domain used for SIMD calculation. While drug-related hospital admission data contribute to the calculation of SIMD, the weight of these data within the overall SIMD index is minor (0.84%). On this basis, the use of SIMD within deprivation analysis in the DRHS publication is not considered to introduce substantial methodological bias.

See Technical Notes on the Scottish Government website (PDF) for further details about SIMD.

When data are analysed, different levels of SIMD quintiles can be used depending on the aim of the analysis. This dashboard uses SIMD in two different ways:

National SIMD

For the purpose of comparing different locations (for example, Scotland, NHS Board and ADP figures) on an equal basis, ‘within-Scotland SIMD’ quintiles should be used. To compare different locations, select the ‘Location comparison’ option in the Deprivation tab of the Data explorer. The deprivation analysis in the Trend data dashboard is based on National SIMD only. National SIMD analysis was used in all DRHS publications prior to the release of 2018/19 data on 27 October 2020.

Local SIMD

The release of 2018/19 data also includes analysis based on within-NHS Board or within-ADP quintiles. This provides a summary of the deprivation characteristics of a specific geographical area in relation to the population of that location (see the example below for a description of the difference between National SIMD and Local SIMD). The results of this analysis cannot be compared with other locations. To generate a deprivation profile based on Local SIMD quintiles, select the ‘Location profile’ option in the Deprivation tab of the Data explorer. Local SIMD analysis is not shown in the Trend data dashboard. Note that within-NHS Board quintiles are applied to NHS Board locations and within-ADP quintiles are applied to ADPs when ‘Location profile’ is selected.

There are no neighbourhoods in NHS Orkney which are comparable with the most deprived neighbourhoods in Scotland and classified as SIMD1 on a national basis. Therefore, if NHS Orkney is analysed on a ‘within-Scotland’ or ‘National SIMD’ basis, there are no people with drug-related hospital stays from SIMD1 (stays are most common among people from SIMD2 neighbourhoods) (see Figure 1). However, it is possible to assign the most deprived 20% of neighbourhoods within NHS Orkney to SIMD1 on the basis of the Local SIMD (within-NHS Board) analysis as this is based on differences in deprivation within the NHS Orkney population (see Figure 2). The main difference between National SIMD or Local SIMD is the ability to compare either within Scotland or only within the local area.

Figure 1: The bar chart shows stay rates by deprivation quintile for NHS Orkney (financial year 2018/19, Any hospital type, Any diagnosis, Any drug type). While the x-axis has five quintiles from 1 - most deprived (left) to 5 - least deprived (right), stay rates are shown for quintiles 2, 3 and 4 only. There are no entries for most deprived and least deprived quintiles because no people with drug-related hospital stays were assigned to SIMD1 and SIMD5. — Image caption Figure 1: Analysis based on 'national SIMD' (or 'within-Scotland')

Figure 2: The bar chart shows stay rates by deprivation quintile for NHS Orkney (financial year 2018/19, Any hospital type, Any diagnosis, Any drug type). The x-axis has five quintiles from 1 - most deprived (left) to 5 - least deprived (right). Stay rates are shown for all quintiles. — Image caption Figure 2: Analysis based on 'local SIMD' (or 'within-NHS board')

PHS’ SIMD quintiles are constructed using a population weighting method. This method is different from how the Scottish Government (SG) creates SIMD quintiles. For more detail see population weighting section on the PHS SIMD webpage (external website).