Considerations for APMS 2022: Suggested content to amend

As well as what new topics to add, participants were asked about how existing questionnaire content from the 2014 survey should be updated or adapted for the 2022 survey. Several participants flagged the need to maintain consistency in questions so that trends over time can be constructed, and how this needed to be balanced against the benefits of updating any section. In some academic events people expressed frustration when a question changed between waves, inhibiting the ability to monitor temporal trends, making it difficult to control in the same way or to test whether a strength of association had changed over time. Any move to improve questions or modules can involve a trade off, and it is not always possible to anticipate where changing an item will cause problems for a future analysis. Changes therefore should be undertaken with caution and where there is a real need.

Government respondents suggested including questions about adverse childhood experiences (ACEs), such as the 10-item ACEs questionnaire developed by Felitti et al., (1998) to measure childhood trauma. They suggested the data could be linked to other APMS information to assess the links between ACES and mental health difficulties in adulthood. These could potentially replace the childhood neglect questions (10 items) asked in 2014.

APMS asks about experience of childhood sexual abuse (CSA). Government participants requested expanding this section to ask about perpetration of CSA as well. They explained that information about perpetration and how many people have a sexual interest in children is unknown and would be helpful for policy decisions. They stated that European and Canadian surveys asked a similar question. Related queries were raised about the low levels of reporting of domestic violence perpetration.

Additionally, there was a request to include people who had experienced CSA, about whether they had accessed and taken-up support for this; as well as the barriers they experienced to accessing support. 

There was a request to expand the 2014 discrimination questions and to gather more information about the nature and location (e.g. at work, on transport/in public, at home) of incidents of discrimination. 

Furthermore, participants with lived experience of mental ill-health suggested the survey may benefit from collecting information of the experience of mental health stigma, both in terms of stigma experienced by people with a mental health condition, but broader public perceptions of those with mental health conditions. They did not suggest a particular scale to use. Stigma has already been explored on other surveys, such as the British Social Attitudes series.

Consultation participants suggested

• sources of debt: the response options will need to be updated for the 2022 survey, to reflect changes in loan companies and include internet-based loans since 2014. One suggestion was to review the Wealth and Asset Survey for possible questions about debt
• money related stress: one suggestion was to include the InCharge Financial Distress/Financial Wellbeing Scale (8-items)
• another suggestion was to ask about the use of foodbanks, as an indicator of financial insecurity and hardship

There was a concern that perpetration of domestic violence is under-reported, due to social desirability bias. Suggestions to enhance these questions included:

• to soften the introduction of these questions by stating that “people in relationships often have difficulties or arguments”, for example
• to consider the order in which the questions appear and potentially ask about perpetration before the experience of violence  
• to determine perpetration, it was suggested to ask if individuals had been granted a restraining order or arrested for coercion and control to identify
• as mentioned above, exploration of domestic violence in lockdowns during the pandemic
• to explore the impact of domestic violence on the household, and not just for the survey respondent
• access to treatment and support following domestic violence and exploration of both motivation for seeking help, but also the barriers to accessing services. There was a suggestion to ask whether people take up statutory or voluntary sector services

The consultation highlighted that the illicit drug use response options will need to be updated for the 2022 survey, to reflect changes in language since the last survey.

Those using the alcohol data for secondary analysis favoured the Alcohol Use Disorders Identification Test (AUDIT, 10-item) and suggested the Severity of Alcohol Dependence Questionnaire (SADQ, 20-item) could be removed and questions on drug and alcohol treatment included instead.

Additionally, there was concern that alcohol consumption is asked about in terms of ‘units’. Respondents questioned whether the general public understand alcohol unit measures and therefore how accurate their responses are. An alternative approach was not proposed.

Respondent’s highlighted that those who did not currently drink alcohol but who may have former or sustained alcohol related problems were not identifiable in the data.

One further suggestion was to ask about family substance use history. And another to ask if individuals had obtained prescribed medication illegally.

There was a suggestion to ask about vaping, alongside smoking.

There was interest in expanding the employment questions to include:

• a review of all employment and benefits questions, to reflect the move to home working for many, and how this may have affected work-related stress, as well as the major changes to the welfare benefit system. As mentioned above, there is also interest in capturing employment status during the pandemic.
• employment contracts such as participation in the gig economy, zero-hour contracts and having multiple jobs at the same time. Participants wanted to be able to explore the implications of insecure employment for people’s mental health.
• occupational health questions: to ask all employed people about reasonable adjustments for mental health problems to support employment.
• items that allow for understanding of workplace absenteeism and presenteeism.
• items that allow for economic costings, such as the EQ-5D

Participants suggested that the Labour Force Survey, Labour Market Survey and Understanding Society Surveys, may have questions that could be used for this purpose.  

Various environmental factors were highlighted: in terms of the immediate housing context, the neighbourhood assets and amenities (including libraries and leisure facilities, access to green and blue space, and noise disturbance), and in terms of wider environmental behaviours and attitudes more generally.

Government representatives and third sector organisations stressed the need to better understand inequalities in mental health prevalence and service use across the population. As such, they recommended expanding information collected about each of the nine protected characteristics in the Equality Act 2010.

• gender identity: The inclusion of gender identity questions was a priority for 23% of survey respondents. Focus group participants stressed the importance of collecting data on both sex and lived gender. One suggestion was to include the proposed gender identification in the Office for National Statistics Census 2021 question. 
• sexual orientation: There was a request to ask all survey respondents about their sexual orientation and not limit it to those aged under 65 years

Consultation participants highlighted that obsessive compulsive disorder (OCD) is a common disorder and can have a limiting impact on people’s lives. As such there was a request for a better measure of OCD symptoms. An alternative measure was not suggested. 

Since the 2014 APMS, there has been an increasing interest in accurate measurement of personality disorder. As such, consultation participants suggested the need to revise and update the personality disorder questions. They suggested that the 2014 APMS questions are limited and too narrow in focus. In 2014 the SCID-II was used to measure antisocial personality disorder (ASPD), however consultation participants questioned how honestly people respond to these items, given their sensitivity. An alternative scale was not suggested. One alternative approach was to conduct an analysis of a cluster of behaviours, symptoms and experiences that are common among those with a personality disorder.

Consultation participants recommended that the APMS should review naming and symptom conventions for mental health conditions to make sure these are up to date with any changes since 2014. For example, there are more personality disorders, a growing interest in complex post-traumatic stress disorder, as well as disassociation disorder.

Those interested in the associations between physical and mental health conditions suggested improvements to the collection of physical health data. These included:

• a review of the physical health condition response options. Consultation participants highlighted that some conditions are very specific (e.g. heart attack) and others are too broad (e.g. cancer).
• they also suggested some additional information about particular conditions would assist secondary analysis.  
• a further improvement suggestion was to ask if the condition is doctor diagnosed and whether the condition is chronic or acute.
• there was also a suggestion to ask about sleep and related disorders such as insomnia, although there is already a sleep complaints symptom section on the CIS-R.

There was a request to expand the 2014 trauma questions to collect information on when incidents happened, (e.g. the year).

Participants also asked if complex-PTSD would be captured in the next survey.  

The 2014 survey included an item on whether the survey respondent had ever been in prison. This item has proved useful, and suggestions were made for follow-up questions to be asked of those reporting having been in prison, including: how long since they most recently left, how many terms served, length of longest sentence, psychiatric service contact in prison, nature of any treatment in prison, and whether referred for psychiatric treatment at point of release.

While the National Survey of Sexual Attitudes and Lifestyles (Natsal) does collect data on sexual behaviour, it has an upper age limit to participation and limited questions on mental health. APMS has no upper age limit to participation and could provide data on sexual behaviour and risk taking in people with range of different mental health conditions.

One suggestion was to change the focus of the 2014 social capital questions, from feeling about the local neighbourhood to ask about community participation, including volunteering and social action groups.

There were also requests to include improved items on loneliness. The current item on this comes from the Social Functioning Questionnaire and is limited by conflating social isolation and loneliness. The UCLA Loneliness Scale (ULS-20) was suggested. This 20-item scale is designed to measure subjective feelings of loneliness as well as feelings of social isolation. There is also a 3-Item Loneliness Scale, designed to be used in interviewer-administered surveys with people aged 18 and over, and developed from the Revised UCLA Loneliness Scale (Hughes, et al. 2004). The ONS harmonised loneliness measure could also be reviewed.

A request was made to keep the 2014 item on ‘trust’ however.

Government and third sector focus group participants suggested small improvements to the suicide and self-harm questions. These included:

• extending the self-harm methods asked about to potentially include other behaviours such as, hitting walls, excessive alcohol consumption, risky behaviours such as not eating.
• review the response options on the motivations for self-harm. Adding in a wider range of known motivators and removal or revision of the wording of the motivation to ‘draw attention’ to a situation. The language of ‘attention’ was considered out-dated.  
• expanding help-seeking response options to reflect the changes since 2014. There was endorsement of including more informal support too, for example, community support and social prescribing, peer support, as well as online information and communities. Similarly, there was a request to include an item about barriers to people seeking help

Participants noted that the data had been used extensively to help understand the veteran population, and to provide control comparison profiles for studies of the current armed service population. It was raised that the survey needed to ensure it was current with the needs of this community. 

The following content was mentioned by one or two survey participants as possible topics to add or amend. Although these are not being prioritised for inclusion in the 2022 survey, we recommend that they be considered when the next survey in the series is being planned.

• Alexithymia – difficulty with identifying and describing emotions
• Caffeine dependence and impact
• Chronic fatigue
• Climate related exposures
• Coping strategies
• Dermatillomania
• Dissociative Identity Disorder
• Electromagnetic radiation exposure
• Head injuries
• Health literacy and public health messaging – sources of information
• Hypermobility conditions
• Iatrogenic trauma – harm caused by medical examination or treatment
• Infertility
• Inflammation
• Mental health awareness in education and training
• Misdiagnosis – differences between self-identified conditions and the views of health professionals, and changes over time in how symptoms are viewed.
• Self-efficacy
• Visions