Physical Health Checks for Severe Mental Illness Supporting Information

This page provides the background, technical information, and guidance on key performance indicators for the quarterly physical health checks for people wth severe mental illness (PHSMI) publication.

People with severe mental illness (SMI) are at increased risk of poor physical health. NHS England's commissioning guidance document Improving the physical health of people living with severe mental illness outlines that:

• Adults living with SMI have an almost five times increased risk of dying prematurely compared to those living without SMI, and premature deaths of people living with SMI are increasing.
• An estimated 50% of deaths in people living with SMI are attributable to smoking.
• The most impactful single modifiable candidate for increasing life expectancy of people living with schizophrenia is smoking, and for people living with bipolar disorder it is sedentary behaviour.

This publication series replaces the previous aggregate SMI physical health checks SDCS collection which was retired at the end of 2023-24. Historic data from this collection can be found here.

Data for this publication comes from the General Practice Extraction Service (GPES) which gathers information from GP clinical systems. The PHSMI GPES dataset is extracted quarterly. 

More information about the GPES service can be found here and business rules for the PHSMI GPES extract can be found here.

The total number of people on the General Practice SMI registers (on the last day of the reporting period) excluding patients recorded as 'in remission'.

The PHSMI register includes all patients with a diagnosis of schizophrenia, bipolar disorder, and other psychoses. These are identified using the MH_COD SNOMED code cluster. To find a list of all relevant SNOMED codes under the MH_COD cluster visit the Primary Care Domain Reference Set Portal. The content of clusters is dynamic and can be updated several times throughout the year. For the most up to date content it is therefore recommended that the codes listed in the reference set portal are referred to.

Note: The SMI GPES extract will only include checks delivered to those with a diagnosis of schizophrenia, bipolar disorder, and other psychoses (MH_COD) and will not include those without such a diagnosis but who have been prescribed lithium.

Annual validation of SMI registers and remission codes

General Practices are responsible for keeping an up to date ('validated') SMI register, including the appropriate use of 'in remission' codes. Validating SMI registers on an annual basis should mean that registers are approximately accurate and up to date unless a practice has a high patient turnover necessitating more frequent validation. Validation is important for ensuring that the right people are invited for an annual physical health check.

People who are in remission will remain on the SMI register in case of relapse but are not expected to be invited for an annual physical health check. A remission code (under the SNOMED cluster MHREM_COD) should only be applied if for at least five years, the patient has all of the following:

• No record of anti-psychotic or mood stabiliser medication
• No mental health in-patient episodes
• No secondary or community care mental health follow-up.

Core Metric

Out of the denominator, the number of people who have received a full physical health check in the last 12 months to the end of the reporting period.

A person is counted as having had the core physical health check if they have received all of the six component parts at any point in the 12 monhs to the end of the reporting period. The six elements of the core health check are:

• Alcohol status
• Blood glucose or HbA1c (as clinically appropriate)
• Blood pressure
• Body Mass Index (BMI)
• Lipid profile
• Smoking status

A person is identified as having had a component of the check if one or more of the SNOMED codes for that element of the check is recorded in the patient record. The six elements of the core health check are identified through code clusters listed below. To find a list of all relevant SNOMED codes under the clusters listed, please visit the Primary Care Domain Reference Set Portal.

For the blood pressure, blood glucose, blood lipids, and BMI clusters both the code and attached value must be recorded to count towards the physical health check.

Data in the publication is sensitive as it relates to people with severe mental illness. Tables in this publication at GP Practice and PCN level contain counts of small numbers of patients on the severe mental health register. These small counts represent a risk of disclosure and therefore disclosure control measures have been applied as follows:

• Sub-national counts between 1 and 7 (inclusive) are displayed as "*"
• Zeroes are unchanged
• All other sub-national counts are rounded to the nearest 5.
• No percentages are shown where the numerator and denominator is between 1 and 7
• Where the numerator is zero, the percentage is zero.
• Where counts are rounded, the percentage is calculated using the rounded numerator and denominator.

Note: At national level there are no counts of below 8 and therefore no data are suppressed.

Users should be aware that percentages calculated using the rounded numerator and denominator will differ from the percentage that would be calculated using unrounded numbers, and that this difference is quite large for smaller denominators. Therefore percentages are not shown for the GP and PCN tables as many GPs and PCNs have relatively small numbers of people on their SMI registers (the denominators).

To demonstrate data completeness, the number and percentage of ICBs, sub-ICBs, PCNs and GPs included in the data extract are reported in the Summary tab of the Excel tables. A comparison of the number of people on the PHSMI register in the data extract is also given against an ‘expected’ number. The ‘expected’ number of people on the SMI register is calculated from the latest Quality and Outcomes Framework (QOF) mental health register information. The incompleteness of data should be considered when interpreting the statistics.

Reasons for differences between the 'expected' SMI register size and the number on the SMI register in the data extract include:

• Not all GP practices are included in the PHSMI GPES extract
• The PHSMI GPES extract isn't for direct patient care and so does not include patients who have opted out of sharing their data. QOF includes these patients as it is an aggregate collection.