Methodology

Additional information and links to explain how the audit data is collected and analysed.

This document provides the background, methodology and guidance for the interpretation of the learning disability breakdowns of CVDPREVENT indicators.

152 KB

Severe Mental Illness Data Guidance

This document provides the background, methodology and guidance for the interpretation of the severe mental illness breakdowns of CVDPREVENT indicators.

Sections:

Value definitions

National value: The total proportion, calculated for England. This value shows us the proportion when we consider all patients registered at GP practices in England, that participate in the CVDPREVENT audit (more than 95% of GP practices in England participate).

System median: The median of the area values that have been calculated for all geographical areas/systems that are of the same type as that chosen in the blue bar. For example, if you have selected a PCN, the system median will be a PCN system median and show the middle value from a list of the PCN area values for the indicator, ordered from smallest to largest.

Area value: The local proportion, calculated for the geographical area/system selected in the blue bar. This value shows us the proportion when we consider just those patients registered at GP practices within the selected area/system, that participate in the CVDPREVENT audit (you can view the participation rate for the chosen area/system in the blue bar).

Population and practice coverage

Population coverage: number of patients from the GP list size of contributing practices as a proportion of the total GP patient registered population for the area.

Practice coverage: calculated from the number of practices for which data has been received as a proportion of practices.

Full details available in the Methodology Annex, p7-8.

Beta

The CVDPREVENT Data & Improvement Tool is in the ‘beta’ phase of development. This means you’re looking at the first version of the tool, and it being continually tested and improved.

If you have any feedback, a feature request, or if you have spotted an error, please click here to complete a feedback form. This feedback is monitored, but the team may not be able to respond.

Quintile Highlighting

Quintiles are highlighted to show how the selected area’s value for a particular indicator and inequalities marker compares to the overall data for that area’s system level. For example, values for a CCG will be compared to all available data at CCG level for the same marker split into quintiles.

For some inequalities markers, the data is consistent or the range is small. In these cases, multiple quintiles may be highlighted, as the selected area’s value will fall within multiple system level quintiles.

Reintroduction of CVDP004CKD

The CVDPREVENT indicator CVDP004CKD: Percentage of patients aged 18 and over with GP recorded CKD categories G3a to G5 (previously stage 3 to 5) with a record of a urine albumin:creatinine ratio (or protein:creatinine ratio) test in the preceding 12 months has been re-introduced from the June 2023 update. However, any comparisons with the previous data periods will not be possible and hence, we have removed any previous data points.

Previously published figures are likely to represent an underestimate of the albumin/protein creatinine ratio testing performed on patients with a CKD diagnosis. The CVDPREVENT audit extract was amended to include a more complete record of ACR/PCR testing. This amendment was made in the June 2023 update.

CVDP002AF withheld from September 2023 dataset

The CVDPREVENT indicator CVDP002AF: Percentage of patients aged 18 and over with GP recorded atrial fibrillation and a record of a CHA2DS2-VASc score of 2 or more, who are currently treated with anticoagulation drug therapy was withheld from the September 2023 data update.

A technical issue was identified in the data collection process which has affected the audit values for this indicator in the September 2023 extract. This does not impact any historical data for this indicator and from the December 2023 update the indicator was reinstated.

Ambitions

The ambitions for CVDP007HYP and CVDP003CHOL reference the national ambitions set out in the 2023/24 Priorities and Operational Planning Guidance, NHSE, January 2023.

The ambitions for CVDP009CHOL, CVDP007CHOL, CVDP002AF and CVDP001SMOK all refer to the upper limit of the target for the relevant QOF indicator, Quality and Outcomes Framework guidance for 2023/24, NHSE, 2023. There is not always an exact read across from CVDPREVENT indicators to QOF indicators. This is intended as a guide rather than a mandated target.

Outcomes admissions data for NHS Frimley ICB

The CVDPREVENT indicator CVDP002ADMN: Data on hospital admissions for Frimley ICB has been removed for this period.

Data on hospital admissions outcomes for Frimley ICB has been removed for several time periods. Between July 2022 and March 2023, Frimley Health NHS Foundation Trust were unable to submit data for use in Hospital Episode Statistics (HES), the source for admissions data used in the outcomes analysis. When reviewing initial data, we also observed Frimley ICB as an outlier on these indicators. It is possible that patients from other ICBs were admitted to Frimley NHS Trust and their data affected by this issue, but we understand the effect of this to be minimal and there are no clear impacts on neighbouring ICBs in the outcomes data. We will reinstate data for Frimley ICB as soon as it is possible to do so.

Reporting of the ‘not stated’ ethnic group for outcomes indicators

The ‘not stated’ ethnicity category is removed from the outcomes indicator breakdowns for some periods because there are dependencies between the different data sources used in the recording of ethnic group. As of the June 2024 CVDPREVENT extract, 3.5% of people have a ‘not stated’ ethnicity.

For admissions indicators (CVDP001ADMN and CVDP002ADMN), as ethnic group can be taken from primary care or hospital records, a person admitted to hospital is more likely to have a stated ethnic group than a person not admitted to hospital. This means that admission rates cannot be produced for ‘not stated’ ethnicity, as these would be artificially low.

In the June 2024 CVDPREVENT extract, which relates to the outcomes follow-up period April 2023-March 2024, there was a substantial update of ethnicity information which led to fewer people having a ‘not stated’ ethnicity, but this was not applied to those who were no longer reported within CVDPREVENT at that time (e.g. those who died). This created a bias in reporting of all mortality indicators for ‘not stated’ ethnicity, which will continue until the follow-up period in the outcomes data is July 2024-June 2025 (i.e. the September 2025 CVDPREVENT extract). Other ethnic group categories may have been slightly affected by the same issue, but the impact has been assessed as minimal and so data for these groups will continue to be reported.

Removal of hypertension risk category breakdowns for hypertension outcomes indicators

Prior to the April 2023 – March 2024 outcomes data, patients with hypertension that were included in CVDP001ADMN, CVDP002ADMN, CVDP005MORT and CVDP006MORT, were allocated to a hypertension risk category based on the latest available blood pressure reading prior to the data period start date. In other words, there was no limitation on the time frame for these readings. From the April 2023 – March 2024 data, this has been changed to only include blood pressure readings within 12 months prior to the data period. The hypertension risk category breakdown has been removed for all time periods before the April 2023 – March 2024 period, because the unlimited time frame for the readings does not properly reflect the risk categories.

Correction notice – 20/02/25

Indicator data by deprivation quintile have been re-published for September 2024

We have been made aware that for some areas within the clinical indicator dataset, the denominators for some of the deprivation groupings were lower than would be expected. This is because we have used a new method to assign deprivation groupings to patients. This was meant to take account of changes in geographical mappings that ONS publish. The new method resulted in lower than expected denominators and so for consistency, the data for the September 2024 extract has reverted to the previous method where the denominator size is more in line with expectations.