Mpox knowledge, attitudes, beliefs, and intended behaviour in the general population and men who are gay, bisexual, and who have sex with men

INTRODUCTION

Mpox (also known as monkeypox) is an orthopox virus that causes fever, headache, exhaustion, swollen glands and aches (joint, muscle, back), followed by a rash with blisters.(1) It spreads from person to person through touching clothing, bedding or towels used by someone with mpox rash, touching mpox skin blisters or scabs, and through the coughs or sneezes of someone with mpox rash (droplet transmission).(1) Since May 2022, there has been a multi-country outbreak of mpox in non-endemic countries.(2) Estimates from the World Health Organization (WHO; data up to 1 December 2022), indicate that there have been over 81,000 cases in 110 countries, resulting in 59 deaths.(3) Most cases have been in men who are gay, bisexual, or have sex with men, with close human skin-to skin contact (including sexual) being the primary driver of transmission.(4) The UK is the sixth country most affected by this mpox outbreak with 3725 cases.(3) Within the UK, most cases have been identified in England (with 69% of English cases in London). Almost all (99%) cases were men; English cases had a median age of 37 years.(5) The peak of the epidemic was seen in June and July 2022, with case numbers falling since the end of July.(6)

In the UK, people who thought they might have mpox were asked to call a sexual health clinic, to stay at home (self-isolate) and avoid close contact with other people.(1) Suspected cases were tested for mpox, and confirmed cases were asked to self-isolate for up to 21 days and engage with contact tracing.(7) Cases, their close contacts, and those most likely to be exposed to mpox were advised to be vaccinated with modified vaccinia Ankara vaccine – which offers cross-protection – to reduce transmission and prevent severe illness.(8, 9) While similar public health actions are now familiar to the public as a result of the COVID-19 pandemic, research conducted during the pandemic indicates that engagement with uptake of testing and self-isolation was sub-optimal in some groups.(10–12)

Various theories have guided research into the psychological factors associated with the uptake of health behaviours. One such theory is the Protection Motivation Theory (PMT), which states that people’s intention to carry out a protective behaviour is influenced by their appraisal of the threat (perceived susceptibility to and severity of, e.g. mpox) and the coping response (perceived effectiveness of and self-efficacy for, e.g. testing, self-isolation, contact tracing, vaccination).(13) During the COVID-19 pandemic, psychological, and socio-demographic factors associated with testing and self-isolating included: higher perceived risk of COVID-19, knowledge of transmission modes, higher perceived effectiveness of protective behaviours, believing that your behaviour had an impact on transmission, and belief that others in the same position would also self-isolate, being female, and having less financial hardship.(10–12, 14) Generally speaking, engagement with protective behaviours was associated with being older, apart from uptake of lateral flow testing, which was higher in younger people.(10, 11) Factors associated with COVID-19 vaccine uptake (completed and intended) included perceiving vaccination to be safe and necessary, perceiving COVID-19 to be more severe, and thinking that others would also be vaccinated.(15–18) Perception of side effects is one of the most common reasons given for refusing vaccination.(19) Historically, smallpox vaccines have been associated with severe adverse effects.(8) While vaccines currently licensed are associated with fewer severe adverse effects, this may affect people’s intention to be vaccinated.

At the time of writing, few scientific studies have investigated behaviour during the 2022 mpox epidemic. While most have investigated vaccination acceptability, few have investigated engagement with a contact tracing system. Those studies that have been done suggest that knowledge of transmission modes is incomplete.(20–22) Greater agreement with vaccination for mpox was associated with perceiving the virus as more dangerous and virulent and higher worry about mpox in a survey of members of the Saudi Arabian general public.(23) In Dutch GBMSM, willingness to accept a vaccine was associated with being more worried about getting mpox, perceiving a higher risk of mpox, perceiving mpox to be more severe, thinking that vaccination is important, thinking that the vaccine was effective, and greater social norms for vaccination.(24) Another study also conducted in the Netherlands found that vaccination intention in GBMSM was associated with higher worry about mpox, knowing someone who had mpox, and being single but dating or in an open/polyamorous relationship.(25) This study also investigated self-isolation intention for 21 days, finding that higher intentions were associated with thinking that mpox had more problematic consequences, and lower education.(25) A study conducted in the UK found that agreement with self-isolation was associated with not having completed a higher education degree, not being employed, and identifying as having a disability.(26) There was no difference in agreement between GBMSM and those who were not GBMSM. Further research is needed to investigate how psychological factors may affect engagement with public health measures (isolating, testing, contact tracing, vaccination) put in place to control the spread of the mpox outbreak in the UK.

Official communications are vital during new and emerging outbreaks, and serve to inform the public about the threat, the public health response, and behaviours that people should engage with in order to protect themselves and others.(27) Messages based on theories of health behaviours, such as the PMT,(13) may therefore increase engagement with protective behaviours. For example, research suggests that COVID-19 vaccination intention increased when communications emphasised the safety and effectiveness of the vaccine and using social norms interventions (e.g. asking people to “join the millions” being vaccinated).(28) While findings relating to messages emphasising the benefits of vaccination to oneself and others were mixed, there is some evidence that the influence of these messages may be most evident in strongly hesitant groups.

In this study, we recruited a general population sample and three GBMSM samples from: a market research company, Grindr, and Meta (Facebook and Instagram). We investigated knowledge, attitudes and beliefs about mpox, and intentions for key behaviours that could prevent the spread of mpox (self-isolation, help seeking [as the advised route into testing], sexual contact behaviour when symptomatic, contact sharing, vaccination). We used an experimental approach to investigate the impact of different brief communication approaches (promoting perceived susceptibility to and severity of illness / necessity and efficacy of the response / benefits of the response / low perceived costs of response) on intended behaviour. Psychological and socio-demographic factors associated with intended uptake of behaviours were also investigated.

METHOD

RESULTS

Top-line results for all survey materials, by sample, are shown in supplementary materials 1.

For regression analyses, we report imputed values. Results using imputed values were compared with non-imputed data. There were no substantial differences in results with and without imputed values.

DISCUSSION

We investigated mpox attitudes, beliefs, and intended behaviours in a general population sample and three GBMSM samples. Samples differed by socio-demographic characteristics. This was reflected in intended behaviours, with GBMSM recruited from Grindr or Meta being more likely to intend to seek help immediately for mpox symptoms, completely stop sexual contact when symptomatic, and be vaccinated for smallpox if advised. High rates of smallpox vaccination in these samples perhaps reflects that they were mostly educated and working, and thus more likely to be health literate and engaged with services. This was also reflected in knowledge about mpox, with Grindr and Meta samples being more likely to correctly identify mpox transmission modes and the symptoms of mpox. Within the general population, rates of “don’t know” answers were high, with 18.4% selecting “don’t know” when asked what the main symptoms of mpox were, and 18.7% to 24.2% selecting “don’t know” for transmission modes (except for having sex with someone who has mpox, where 13.9% selected “don’t know”). This is similar to other surveys of the general population, where 24% of people were not sure if mpox usually spreads by close contact with an infected person.(41)

There were no effects of motivational messaging on our behavioural outcomes, except for in the Meta sample for intention to self-isolate, share details of all recent sexual contacts, and be vaccinated for smallpox if advised. For these outcomes, intention to engage with protective behaviours was higher in the control group. In practice, however, intentions were very similar (means changed by 0.1 to 0.4) and this is unlikely to represent a meaningful difference. Another study investigating gonorrhoea reinfection in young adults has also found that those in the control group (who received monthly texts reminding them to update their contact details) were less likely to be reinfected than those in the intervention group (who received a series of texts that were educational and used behaviour change techniques).(42) These messages were specifically designed to decrease stigma, and partner numbers were higher in the intervention arm. A systematic review of intervention communications during the COVID-19 pandemic found that messages about the personal and collective benefits of vaccination had mixed effects on vaccination intention, with some suggestions that this approach may be more effective in more strongly hesitant individuals.(28)

Generally, higher intentions were seen in the Grindr and Meta samples, except for self-isolation. However, rates should be interpreted with caution, as intended behaviour does not always translate to enacted behaviour.(43) This was shown to be the case for intended and actual engagement with the UK contact tracing programme during the COVID-19 pandemic.(11) Rates of having already been vaccinated for smallpox in 2022 were especially high (32% Grindr, 41% Meta), again suggesting our sample may have been particularly interested in the topic of the survey, especially given issues with access to vaccination.(44) In the UK, the public health agency (UK Health Security Agency) has been working in conjunction with community-based organisations and charities to raise awareness of mpox and how to prevent the transmission in GBMSM as the population most affected. Higher rates of knowledge and behavioural intentions suggest that this targeted messaging has been effective in increasing knowledge and driving protective behaviours. Evidence suggests that while one-off sexual encounters may make up only a minority of sexual interactions, they could account for a large proportion of mpox transmission.(45, 46) Furthermore, if encounters are anonymous, this will affect contact tracing efforts. Therefore, very high rates of intending to completely stop sexual contact with others if symptomatic (91% to 93% in Grindr and Meta samples respectively) are encouraging.

It is notable that 19% (Grindr, Meta) to 30% of people (general population, Savanta GBMSM) agreed that is it best to avoid physical contact with GBMSM because of the mpox outbreak, perhaps also reflecting a change in behaviour in this population. Recent decreases in some sexually transmitted infections in GBMSM, and the decline in mpox transmission in England, suggest this may be the case.(6)

Within the general population, women, people who were older, and those with lower financial hardship were more likely to intend to carry out protective behaviours. This pattern was also widely seen during the COVID-19 pandemic (11, 47, 48) and previous outbreaks.(49) This highlights the importance of considering health equity issues to ensure effective outbreak control. Offering financial support for protective behaviours, especially those that may affect people’s ability to earn an income such as self-isolation, is likely to increase engagement, especially for those in lower income settings. Few socio-demographic characteristics were associated with outcomes in the Grindr sample. This could be because the Grindr sample differed from the general population sample on key characteristics (more educated, higher socio-economic grade, less financial hardship), or be a function of the way analyses were conducted with all variables being entered into a single regression model. The Grindr sample, being smaller, also had about half the statistical power of the general population sample.

Intention to engage in protective behaviours in the general population was also associated with psychological factors such as greater worry about mpox, perceived risk of mpox to others (but there was little evidence of an association with perceived risk to oneself), perceived susceptibility and severity of mpox, and greater knowledge about transmission. This pattern of results was also seen during the COVID-19 and aH1N1 influenza pandemics.(12, 47, 50–52) COVID-19 vaccination intention was also associated with perceived risk to others, but not to oneself.(33) Few associations were significant in the Grindr sample. In addition to the decreased statistical power, this may also reflect the different context and experience that participants in this group had. While for our general population sample the questions about mpox risk were somewhat hypothetical, the Grindr sample were GBMSM and thus more likely to be impacted by mpox (as illustrated by the high vaccination rates). Factors relating to their personal experience of the outbreak may have superseeded any effect of the psychological variables that we assessed. Taken together, these results suggest that clear communications about the level of risk of infection may encourage people to enact protective behaviours when appropriate.

Agreeing that the risks of mpox were being exaggerated was associated with lower intention to engage with protective behaviours. This was also the case in previous pandemics.(47, 53) Other psychological factors associated with decreased intention to engage with protective behaviours were believing that you were already immune to mpox and that your life had been negatively affected by changes made in response to mpox. Attending an important event was one of the reasons given for breaking self-isolation during the aH1N1 pandemic.(54) The peak of the mpox outbreak in the UK occurred in July 2022, a period coinciding with the summer holiday and festival season, including Pride events. It is important to be conscious of asking people to self-isolate or quarantine over periods encompassing key events that may be happening.

Behaviour-specific beliefs were also associated with intentions. For self-isolation, greater intention was associated with greater social norms (thinking that others would also self-isolate). Greater social norms were associated with fewer outings during lockdown (14) and increased vaccination uptake and intention (in oneself and one’s child) (15, 55) during the COVID-19 pandemic. Social norms were also associated with enacting protective behaviours, including self-isolation, during previous outbreaks.(49, 56) Lower self-isolation intention was associated with thinking that self-isolation would have a negative impact on your work. In this study, as in previous research,(11) the main reasons for not being able to self-isolate were needing to go out for essentials (food / medicines), for a walk or some other exercise, and for work. In the UK, people with suspected mpox were directed to call sexual health services, likely explaining this association. Some people may have found this stigmatising as mpox is not a sexually transmitted infection.(57) Stigma surrounding having other sexually transmitted infections is associated with not seeking help.(58) In our study, immediate help seeking was associated with being willing to contact a sexual health clinic. These findings suggest that widening recommended points of contact with health services for suspected mpox cases or contacts may be beneficial. Taken together, results suggest that communications emphasising that others are also engaging with the protective behaviour and recommending non-stigmatising routes into health systems may improve engagement with self-isolation and help seeking.

In the Grindr sample, having been vaccinated for smallpox in 2022 was associated with having ever taken PrEP for HIV. This is probably due to how the vaccine was rolled out in the UK. GBMSM identified by sexual health services as being at highest risk of exposure – using markers similar to those used to assess eligibility for PrEP – were invited to be vaccinated.(59) Being vaccinated was also strongly associated with thinking that you were already immune to mpox in this survey. As this is a cross-sectional survey, we cannot infer the direction of results, but it is likely that respondents believed they were immune to mpox because of their recent vaccination. This is interesting given limited information about the effectiveness of the vaccine at that time.

Strengths of the study include the collection of data from four large samples, including the general population and the population most affected by the recent mpox outbreak (GBMSM). We would also caution readers about several caveats for this study, including that participant socio-demographic characteristics differed by the recruitment method. Participants recruited through Grindr and Meta were more likely to be highly educated, higher socio-economic grade, and have less financial hardship. High rates of smallpox vaccination in these samples suggest that these groups are likely to be interested in and have personal experience relating to the outbreak and may have been more likely to engage with protective behaviours. We do not know if survey respondents are representative in terms of their beliefs, knowledge and intended behaviours with reference to the general population sample, GBMSM generally, and GBMSM who use Grindr and Meta. While overall rates of beliefs, knowledge and intended behaviours may be affected by sampling, associations within the data are likely to hold true.(60)

Other limitations relate to the survey and statistical measures used. The use of a cross-sectional survey means that answers may have been influenced by social desirability and recall bias, although the anonymity of a written survey may have mitigated these effects to some degree. Rates of intended behaviour are often higher than enacted behaviours.(11, 43) Therefore, rates of intended behaviours should be taken with caution. Our motivational messages were very brief and only repeated once, before measuring intended behaviours and behaviour-specific attitudes. A fully developed public communications campaign where people are exposed to co-produced and repeated messages may have more influence. There were high rates of actual and intended vaccination in the Grindr sample (94%). This may have affected our ability to detect associations. Variables were entered into regression models in blocks, with results reported for the final block entered (essentially all variables entered together). Therefore, we investigated the independent effect of a variable, accounting for all other variables included.

This study investigated mpox beliefs, knowledge and intended behaviours in a general population sample and in GBMSM (those most affected by the current outbreak). Intended uptake of protective behaviours differed by behaviour. GBMSM generally had higher intention to engage with protective behaviours, apart from self-isolation. This may have been a function of sampling. Higher knowledge about mpox symptoms and transmission in GBMSM samples suggests that the public health messaging carried out by multiple stakeholders including the UK Health Security Agency, charities and via grass-roots community efforts has been successful. There was no impact of additional motivational messaging on intended uptake of protective behaviours. Associations between increased financial hardship and lower intention to enact protective behaviours suggest that providing financial support to those affected in future outbreaks may increase uptake.

FUNDING SOURCES

LS, JB, RA and GJR are supported by the National Institute for Health and Care Research Health Protection Research Unit (NIHR HPRU) in Emergency Preparedness and Response, a partnership between the UK Health Security Agency, King’s College London and the University of East Anglia. Additionally, IO, LY, TM and RA are funded by the NIHR HPRU in Behavioural Science and Evaluation, a partnership between UKHSA and the University of Bristol. The views expressed are those of the authors and not necessarily those of the NIHR, UKHSA, or the Department of Health and Social Care. The study was funded by NIHR. For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) licence] to any Author Accepted Manuscript version arising.

DATA SHARING STATEMENT

Anonymised data are available online.(61)

AUTHOR CONTRIBUTION STATEMENT

All authors conceptualised the study and contributed to survey materials. LS completed analyses with guidance from HWWP, JB and GJR. LS wrote the first draft of the manuscript. All authors contributed to, and approved, the final manuscript. LS is guarantor. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

DISSEMINATION DECLARATION

Dissemination of survey results to participants is not possible due to the anonymous nature of data collection.

COMPETING INTERESTS STATEMENT

HWWP has received additional salary support from Public Health England and NHS England and receives funding from the National Institute of Health Research. HWWP receives consultancy fees to his employer from Ipsos MORI and has a PhD student who works at and has fees paid by AstraZeneca. GJR has acted as an expert witness in an unrelated case involving Bayer PLC, supported by LS. IO and RA are employees of UKHSA.

Supplementary materials 1. Full survey materials and top-line results for general population and GBMSM samples separately

For categorical data, n (%) are shown. Percentages are column percentages. Where totals do not add up to 100%, it is due to rounding errors.

For continuous data, n (where different to total sample), mean (M), standard deviation (SD), and range are shown.

Total ns are: General population, n=3050; Savanta GBMSM, n=247; Grindr, n=831; Meta [Facebook and Instagram], n=1036.

Supplementary materials 2. Data preparation

Some participants (3.8%, n=196) did not provide their postcode and were therefore missing data about region and index of multiple deprivation. While individual items about financial hardship had small amounts of missing data (2.0% to 3.8%, n=105 to 197), as items were summed to give a scale, missingness on the resulting scale would have been higher (7.9%, n=410). There were no apparent patterns to the missing data. We used multiple imputation to impute missing values for region, index of multiple deprivation and financial hardship items. Socio-demographic variables (age, occupation of highest earner, number of people in household, having a dependent child, employment status, education, ethnicity, marital status, and region, index of multiple deprivation and financial hardship where available) were used to predict missing values using a Markov Chain Monte Carlo (MCMC) method using linear regressions with ten iterations, to give five imputations.

We computed a single variable to indicate perceived susceptibility and severity from three items. For these items, “don’t know” was coded as the mid-point on the scale. We then multiplied responses together (range 1 to 125) and took the cubic root to give a scale from 1 (lowest perceived susceptibility and severity) to 5 (highest perceived susceptibility and severity). We also asked participants how much they agreed that they were already immune to mpox. This item was recoded into a binary item to show perceived immunity to mpox (strongly agree and agree, vs neither agree nor disagree, disagree, strongly disagree, and don’t know).

To create a single variable for perceived knowledge, we used three items (Cronbach’s α=0.74). Responses for these items were given on a five-point scale from “strongly disagree” to “strongly agree” We recoded each item as a binary item (strongly agree and agree [1], vs neither agree nor disagree, disagree, strongly disagree, and don’t know [0]), then summed responses to give a scale from 0 (lowest perceived knowledge) to 3 (highest perceived knowledge).

For knowledge about mpox symptoms, for each symptom selected, we coded participants as being correct (1, symptom listed on NHS mpox website) or incorrect (0, symptom not listed on NHS mpox website). Scores were then summed to give a scale from 0 (no selected symptoms were mpox symptoms) to 4 (all selected symptoms were mpox symptoms); responses of “don’t know” were coded as 0. A single variable denoting understanding of transmission was computed. It is unknown whether mpox can be caught from pet animals (but pet animals can catch mpox from humans (1)), therefore we did not include this item in analyses. We coded answers as correct (1) or incorrect (0; answers of “don’t know” coded as incorrect), and summed items to give a score from 0 (lowest knowledge about transmission) to 6 (highest knowledge about transmission).

To reduce the number of variables included in regression models, we used dimension reduction techniques. We conducted principal components analysis, using direct oblimin rotation as items may have been correlated, on items potentially associated with self-isolation (ten items), help seeking (six items), and vaccination (eight items) separately. The number of factors was determined using a scree plot. The item with the highest loading on to a component was then included in regression analyses. Results of the principal components analyses are reported in the supplementary materials 3.

Supplementary materials 3. Results of principal components analyses

Supplementary materials 4. Influence of motivational messages on behavioural outcomes

Supplementary materials 5. Full results of regressions with intention to seek help immediately

Supplementary materials 6. Full results of regressions with smallpox vaccination