Neurological Conditions: estimating the prevalence in Scotland of selected conditions using GP and Hospital Admissions datasets

Annex D: Methodology

General Practice Dataset

Data Request

Data for the specified neurological conditions were requested from general practices through the Scottish Primary Care Information Resource (SPIRE). SPIRE is a system that enables PHS to request and obtain authorised information from general practice clinical systems about specific health conditions. These data can then be released to inform the public and healthcare staff and other professionals about the incidence and prevalence of those conditions.

The Scottish Government analytical team requested via PHS a data extract which was approved by the SPIRE Prioritisation sub-group.

Identification of Diagnostic Codes

General practices in Scotland record patient diagnoses through the use of 'Read codes', a national standard coding system used in Scottish general practices for recording clinical information (signs, symptoms, diagnoses or activities).

The Scottish Government commissioned Primary Care Informatics (PCI) to identify the Read codes associated with each of the neurological conditions included in the data requests. PCI comprises a group of GPs, general practice managers and others with a special interest in health informatics. Their services include guidance on Read Codes.

PCI provided a set of Read codes for the majority of conditions, taking account of some advice requested and received from clinicians on the NACNC. These Read codes were used by PHS to extract the general practice data for the specified conditions. The list of Read codes and an explanatory background document are available as supporting documents accompanying this publication.

For four conditions, PCI found no suitable Read codes (see Annex C) and so data could not be extracted.

Data Extract Provided

The data provided comprise total counts of currently registered patients ever diagnosed for each specified neurological condition as of 8 April 2022, and recorded as alive as of that date (general practice electronic records include data from as far back as the 1940s.) For each condition, a breakdown by NHS Health Board, by Health and Social Care Partnership, and by sex and age are also available. The data provided are aggregate data and do not include any patient-identifiable data. The dataset is available in a spreadsheet as a supporting document accompanying this publication.

Missing and Excluded Data

The general practice data presented in this report do not include data from all NHS Boards and general practices.

NHS Ayrshire & Arran and NHS Highland did not participate in the extraction of data from SPIRE.

916 general practices were recorded on the April 2022 GP Practice Contact Details and List Sizes. 674 of these practices returned data from SPIRE.

Data from 21 practices were removed from the dataset following a quality assurance process undertaken by PHS. Reasons for removing data were as follows:

• 4 practices returned zero counts for all searches and were removed from the dataset.
• 17 practices were removed because of differences in their patient list sizes compared with published data. Practices with a difference greater than 10% of the data published on Scottish Health and Social Care Open Data for April 2022, and that also had a difference of more than 100 patients, were removed from the dataset.

This left data from 653 general practices, representing 71% of the practices in operation at that time (general practices sometimes close or merge).

The 653 general practices whose data were extracted and retained for analysis had a total population of 4,260,960 registered patients, representing 72.7% of the 5,858,622 registered patients at the time of the data extract.

Hospital Admissions Dataset

Data Request

To complement the general practice prevalence dataset, the Scottish Government analytical team also requested estimates of prevalence from PHS based on hospital records. This data extract was undertaken by the Scottish Burden of Disease (SBoD) study team in PHS.

This piece of analysis drew only on hospital data used in the following Scottish Morbidity Records (SMR) datasets, to identify individuals for inclusion in the estimates:

• SMR01 definition: this contains general acute inpatient and daycase records. Data are derived from diagnoses made on hospital discharge.
• SMR04 definition: this contains mental health inpatient admissions and discharges. Data are derived from diagnoses made on both hospital admission and discharge.

It should be noted that this analysis does not include data from attendances at outpatient clinics as it is not mandatory for diagnoses to be recorded on the SMR00 Outpatient Attendance record. Therefore, patients who have attended outpatient settings only (i.e. with no inpatient/daycase admissions) will not be included in the hospital prevalence estimates.

Identification of Diagnostic Codes

Hospital admissions are coded according to the International Classification of Diseases (ICD). This allows the systematic recording, analysis, interpretation and comparison of mortality (deaths) and morbidity (illness) data.

The Read codes provided were shared with the SBoD team in order to map them across to the appropriate ICD10 codes. In addition, the PHS Terminology Services team provided a full list of ICD10 codes for each specific condition. These codes were combined with the mapped Read codes to produce a final code list. This list is available in a spreadsheet as a supporting document accompanying this publication.

Data Extract Provided

A period prevalence approach was used. Estimates cover people admitted to hospital and diagnosed with any of the specified conditions in the 20-year period up to 31 December 2019, and alive and resident in Scotland at any point in 2019. For the purpose of this analysis, it is assumed that those people still have the diagnosed condition. The data provided are aggregate data and do not include any patient-identifiable data.

SBoD used the National Records of Scotland vital events dataset ("deaths") to exclude people known to have died.

Individuals were excluded if either:

• There was no Community Health Index (CHI) number recorded (this is a unique identifier required to link to mortality data).
• They were non-Scottish residents at the time of the original hospital discharge.
• The CHI register showed that patients were no longer registered with a GP in Scotland, as at 31 December 2019, as proxy for no longer resident in Scotland. The February 2020 register was used, to allow for a lag period for updates.

The dataset is available in a spreadsheet as a supporting document accompanying this publication.