Adult Psychiatric Morbidity Surveys

The Adult Psychiatric Morbidity Surveys (APMS) are a series of surveys which provide data on the prevalence of both treated and untreated psychiatric disorders in the adult population (aged 16 and over). The previous surveys were conducted by the Office for National Mental Health and Wellbeing Surveys Directions specification statistics on behalf of the Department of Health (now known as the Department of Health and Social Care), The Scottish Health Executive and the National Assembly for Wales.

The surveys took place in 1993 covering 16 to 64-year olds and in 2000 covering 16 to 74-year olds and included respondents living in England, Scotland and Wales. From 2007 the surveys were carried out by NHS Digital on behalf of DHSC, and include people aged over 16 (no upper age limit) living in England. NHS Digital contract a survey provider to conduct the survey on their behalf, for 2007 and 2014 this was NatCen Social Research. The surveys capture information on: common mental disorders, mental health treatment and service use, post-traumatic stress disorder, psychotic disorder, autism, personality disorder, attention-deficit/hyperactivity disorder, bipolar disorder, alcohol, drugs, suicidal thoughts, suicide attempts, self-harm, and comorbidity.

Changes to the surveys will involve consultation with the wider user group and the public and guidance from a dedicated APMS Steering Group. The APMS Steering Group is chaired by NHS Digital with representatives from key stakeholders such as:

• the Department of Health and Social Care (DHSC)
• Public Health England (PHE)
• NHS England/NHS Improvement
• Department of Experimental Psychology
• representatives from the research community
• the Well-Being Programme, London School of Economics
• service user representatives
• the survey contractors responsible for delivering the survey

The role of the Steering Group is advisory, and the aim is to guide the delivery of the surveys and help ensure they meet the needs of those involved in commissioning, providing or using mental health services. As part of this role members will be asked to:

• provide direction and recommendations on priorities relating to the survey
• receive regular progress reports on milestones, risk and quality issues, including response rates and offer guidance where necessary
• provide advice and guidance on changes in data to be collected, methodology, questionnaires, survey materials, sample design, processing and presentation of results
• provide advice and guidance on statistical issues relating to the survey
• resolve conflict and escalate concerns where necessary through NHS Digital governance paths

Following recommendations from DHSC and the APMS steering group, this Specification will be reviewed and updated. This is expected to be completed under the relevant wok package without the need for the Directions to be amended, provided any such changes are in scope of the Directions.

Scope

The scope of the surveys includes all adults in the general population, aged 16 and over (no upper limit), living in England. The scope may change over time, for example new topic areas, inclusion of devolved nations (subject to approvals and governance arrangements). However, consideration will always be given to maintaining the time series to allow the monitoring of changes over time. The scope of the surveys will be reviewed periodically, and feedback obtained from stakeholders to ensure that they meet current needs. Any changes will be reviewed by the APMS Steering Group and work packages will be updated to reflect the agreed changes.

Source

Historical data

There will be a one-off collection of historical APMS data held by Office for National Statistics for which DHSC was the data controller, this is the data for the surveys for respondents living in England that took place in 1993 covering 16 to 64-year olds and in 2000 covering 16 to 74-year olds. The APMS data previously collected by NHS Digital on behalf of DHSC for the surveys carried out between 2007 and 2020 are also brought within the scope of these Directions

Subsequent surveys

The mode of the surveys will be chosen depending upon survey content, timing, length and complexity of questionnaire, participant criteria and available budget. Different modes may also be chosen within the same survey year, for example face to face interview, paper, web or app self-completion. The 2007 and 2014 surveys were conducted as a face to face interview with some self-completion questions.

The data collection takes place by surveying a large sample of the general population, for example covering people living in private households. These Directions allow NHS Digital to use any appropriate dataset that it holds or freely accessible to provide a sampling framework to identify adults that are currently alive and living in England. The sampling frame uses the best available source for providing a list of residences in England and adults aged over 16; for example, the 2007 and 2014 survey used the small user Postcode Address File (PAF) which covers private households in England. A small proportion of households are currently not covered by the PAF file (less than 3%). The 2021 survey will use a combination of the PAF file and NHS Digital datasets for example, the personal demographic service, this combination will allow a better coverage of participants. From the sampling datasets, participants are ordered by a number of different criteria (for example. region, age of participants and census data) and a systematic random sample of individuals selected from the ordered list. This helps ensure the selected population is correctly representative and increases the precision of survey estimates.

All selected individuals are asked to take part in the first interview. A sub-sample are then selected to take part in a second interview; selection is based on responses to specific screening questions for certain mental health disorders in the first interview. The chosen disorders and the criteria for selection are to be decided at the time of the survey development. In 2007 the disorders included borderline disorder, antisocial personality disorder, psychosis and autism whereas in 2014 this was limited to psychosis and autism.

As part of the survey participants are also consented for their contact details to be kept on file to allow for future contact for further research. Where consent has been received, follow up studies of the survey cohorts will be carried out at appropriate intervals, this will allow for longitudinal analysis to show any changes in mental health over time and allow for the identification of risk factors. NHS Digital will use all appropriate means to ensure that to the best of their ability, the identified follow up cohort contains the latest contact details for the participants and that all participants are currently alive for example checking contact details against the Personal Demographic Service and updating for mortality information. The follow up study can take the format of an online, telephone or face to face survey, however the content of the survey should be tailored to the chosen methodology. The survey may include any aspect of the original survey content and any new content that will enable monitoring of mental health and improve care for patients at that time.

Any change to this methodology will be managed through the change management process and the relevant work package.

Category

For the basis of sampling the following personal details will be required from the Personal Demographic Service (or similar NHS Digital tracing service):

NHS Number (not shared with the survey contractor, held by NHS Digital for validation purposes for follow up and data linkage)

Serial ID (NHS Digital assign unique ID to cohort) 

• name
• address 
• postcode
• telephone numbers
• email address
• Date of Birth (DOB)
• gender

Where a person consents to future contact the survey contractor will share the serial ID, name, address, postcode, telephone number and email address with NHS Digital. NHS Digital will validate these details against the personal demographic service and will provide any updates. A flag will also be placed on the cohort for anyone that has since died.

Where a person consents to data linkage the survey contractor will share the serial ID, name, DOB, postcode with NHS Digital. NHS Digital will validate the participants details and provide the serial ID and DOB along with health information to the survey contractor.

The APMS dataset including linked asset will include personal data as follows:

• name
• address
• postcode
• telephone number
• email address
• date of birth
• gender 
• date of death

And special category data

• ethnicity
• religion
• sexual orientation
• cause of death 
• health and mental health related information based on survey responses and linkage

The full list of survey questions asked can be found as part of the relevant National Statistics Publication for the individual survey years.

Frequency

The timing of surveys will be specified by the sponsors, DHSC, depending upon the requirement for updated mental health information. So far surveys have been carried out in 1993, 1994, 1997, 2000, 2007 and 2014.

Processing

NHS Digital contract a survey contractor to carry out the survey on its behalf and produce the analysis as specified by NHS Digital. The survey contractor acts as a data processor on behalf of NHS Digital.

The survey contractor designs a sampling methodology for the survey involving the postcode address file and/or the personal demographic service and ensures that chosen participants are representative of the general population; this involves ordering residences by region, postcode and census data. NHS Digital review and approve the sampling methodology. Where the personal demographic service is used National Data Opt Outs will be applied prior to sampling. The survey contractor will design sampling criteria which will be applied and the resultant sampling dataset will be transferred securely to the survey contractor. The information about the sample is held electronically. Chosen residences are contacted and asked to take part in the survey, where a residence contains more than one resident, one adult is randomly selected to take part in the interview.

The survey contractor collects, processes and holds the APMS data on behalf of NHS Digital. NatCen currently hold the 2007 and 2014 data on behalf of NHS Digital. During the face to face interview the data is either captured electronically or via a paper self-completion questionnaire. Paper responses are transferred to an electronic copy and are retained as per the survey contractor’s retention policy before being destroyed. The created dataset is held electronically by the survey contractor.

Survey responses are validated for completeness and data quality and are carried out by the survey contractor. This includes checking that any information captured on paper has been transferred to electronic copy correctly, answers are valid for example coding consistently where a respondent answers yes to one question but no to a similar question later, and any non-responses are coded the same. Checks are carried out to ensure the responses are representative of the general population of adults and that there is no evidence of bias, for example, age, season and region. Derived variables are also created; this is when a new answer is created from the respondent’s answers to one or more questions for example  calculating an overall well-being score. Weights are also created to allow the answers to be compared for example adjusting for ethnicity, age, region, presence of a mental health condition. Population trends are created to look at changes over time. This involves processing responses for past survey years to look at trends over time. The survey contractor prepares responses to Freedom of Information (FOI) requests and Parliamentary Questions on behalf of NHS Digital. They also fulfil data requests from external customers for bespoke analysis following and in line with NHS Digital approval through NHS Digital’s Data Access Request Service (DARS).

For the Follow up study, contact details for survey respondents are updated by cross checking against relevant NHS Digital datasets, including latest address details and information about deaths, this is to ensure that information is not sent to the wrong address or unnecessary stress is caused to a bereaved family. Information for the follow up study is sent to the survey respondent asking them to take part in the follow up study. The survey responses are collated and held electronically, completeness and data quality checks are carried out as outlined above for the main survey and derived variables and weights created. The survey responses for the follow up study are linked by a serial number to the answers from the main survey to allow longitudinal analysis and measure change over time for the individual. This will be carried out by the survey contractor who at that time has been contracted to carry out the follow up study, N.B this maybe a different survey contractor than the one that did the original survey.

On completion of a survey, a number of different data files will be transferred to NHS Digital via secure encrypted file transfer. The data files will be stored in the NHS Digital Data Management Environment (DME). The data files for transfer are the same type of file for the main survey and the follow up and include the complete data file (direct identifiers are held separately and a pseudo key is created to be able to identify individuals) and the data that has been anonymised in accordance with the ICO Anonymisation Code of Practice that can be made available to researchers. The direct identifiers will allow future linkage to other datasets that NHS Digital hold. The datasets will be used by internal NHS Digital staff for future analysis if commissioned to do so by the sponsor, linkage to other NHS Digital dataset, and on-boarding to NHS Digital’s Data Access Request Service to allow government, researchers, charities to access the data. Currently the survey contractor holds these data files on behalf of NHS Digital. The reduced anonymised dataset is also deposited with the UK Data Service. The UK Data Service is funded by the Economic and Social Research Council (ESRC) and provides access and support to researchers to a number of different data collections including major UK government-sponsored surveys, cross-national surveys, longitudinal studies and UK census data. The UK Data Service act as a data processor for NHS Digital and following instruction from NHS Digital grant access to the survey data to named researchers following approval through the NHS Digital Data Access Request Service (DARS).

For each survey, the analysis and reporting requirements will be defined by NHS Digital. This will be reviewed and agreed by the APMS Steering group. The agreed analysis and reporting requirements will enable reporting against current policy priorities.

Data linkage

Individuals are asked for their consent to use their NHS number to link their survey data with other health information about them. In previous surveys the linkage has been limited to certain datasets  for example Hospital Episode Statistics data, Cancer Registration administrative data and civil registration mortality data. However, in future surveys there is a requirement to widen the capacity for linkage to include any NHS Digital dataset, where individuals consent, these datasets will include health data, hospital in-patient care, hospital out-patient care, hospital emergency care, maternity, mental health, social care, primary care, diagnostic care, mortality, cancer and diabetes

Data linkage involves using the individual’s contact details to validate their NHS number from the personal demographic service. This can then be used to link to other health information that NHS Digital hold.

Respondents are also asked whether they will give permission for NHS Digital or the survey contractor (acting on behalf of NHS Digital), to hold their contact details on file and approach them in the future to take part in follow up research. Those respondents that agree to this will be contacted to take part in the follow up studies.

In 2016 a consultation took place with users of the statistics to understand who uses the surveys and for what purpose and any improvements to future surveys and outputs. This was a public consultation, and 205 responses were received. Respondents covered a range of areas including academic, clinical, public, private and voluntary sectors and members of the general public.

Responses were received from the following organisations (this list does not include all respondents but a selection of the type of organisations that participated); DHSC, Public Health England (PHE), NHS England (NHSE), Association of Directors for Public Health, Local Authorities, NHS Trusts, NHS Clinical Commissioning Groups, Universities, Royal College of Psychiatrists, Making Every Adult Matter, MIND, Rethink, Samaritans, MQ Transforming Mental Health, Action on Smoking and Health, Addiction Dependency Solutions, Alcohol Health Alliance UK.

Future changes to the survey will involve consultation with the wider user group and the public and guidance from the dedicated APMS Steering Group (see section 3.1). The Mental Health and Wellbeing Surveys Directions Specification v 1.1 Published 08/02/2021 Copyright © 2020 NHS Digital Page 12 of 21 Steering Group will guide the delivery of the survey and help ensure it meets the needs of those involved in commissioning, providing or using mental health services.

Regular dissemination/sharing analysis

Analysis will be undertaken ad-hoc to respond to Parliamentary Requests and FOI requests. The APMS steering group will provide guidance and support with the layout of reports and the types of analysis, this will involve the sharing of proposed outlines for the survey methodology and reporting, however no results will be shared ahead of publication.

Data Access Request Service (DARS)

At the end of each survey year, data anonymised in accordance with the ICO Anonymisation Code of Practice will be shared with the sponsor, DHSC, subject to approval through DARS.

Data anonymised in accordance with the ICO Anonymisation Code of Practice, which does not include information which could directly identify the respondent is created; name, date of birth, address and contact details are removed from it, and the dataset is deposited with the UK Data Service. Researchers, analysts, charities, local government and national government organisations can apply to NHS Digital DARS to use the data for research and statistical purposes providing that the purpose is of benefit to health and/or adult social care and they have a legal basis to do so. If NHS Digital approve, the UK Data Service, as NHS Digital’s processor, will disseminate the data on NHS Digital’s behalf.

Customers can also apply to NHS Digital for bespoke data requests, for data items that have not been included in the anonymised dataset, subject to an appropriate legal basis. If the application is approved via DARS, the survey contractor, subject to funding from the customer, will create the bespoke extract and disseminate to the customer. In the future, the APMS dataset will be available for customers to access via the Data Access Environment (subject to approval from DARS).

Data to be published

A National Statistics report and subsequent tables and charts will be published on the NHS Digital website following each commissioned survey. The Code of Practice for Statistics will be followed for each publication.

Data prohibited from publication

NHS Digital must not publish information that has been anonymised in accordance with the ICO Anonymisation Code of Practice1 for release by DARS.

Changes to this Specification will be managed and agreed with the Department of Health and Social Care (on behalf of the Secretary of State) to ensure that any such change is within scope of the Mental Health and Wellbeing Surveys Directions 2020. Any changes to the survey methodology or questionnaire content will be discussed with the APMS Steering Group and the work package and this Specification will be updated to reflect agreed changes.