Public perceptions of non-pharmaceutical interventions for reducing transmission of respiratory infection: systematic review and synthesis of qualitative studies

We sought papers whose primary focus was public perceptions (general public or patient groups) of non-pharmaceutical measures to reduce transmission of acute respiratory infections and/or interventions that aimed to promote non-pharmaceutical respiratory infection control. To be eligible for the review, studies had to use qualitative methods of data collection and data analysis. Studies that used a mixed methodology were included if they comprised a substantive qualitative component (i.e. the depth and breadth of the qualitative data was sufficient to form a stand-alone qualitative paper). Studies whose primary focus was the views of healthcare professionals, views of children or views about respiratory infections with no data on reduction of transmission were excluded (Table 2).

Five electronic databases were searched: Medline (1946 to January Week 3 2013), PsycINFO (1887 to February week 1 2013), Embase (1980 to 2013 Week 05) CINAHL (1982 to February week 3 2013), and Web of Science (1981 to February week 4 2013). The last search was conducted on 26th February 2013. We conducted scoping exercises to develop comprehensive search strategies for each database (Additional file 1). We adopted a pre-planned, comprehensive search strategy, rather than using iterative theoretical sampling seeking conceptual saturation, to enhance the transparency and reproducibility of the review [30]. As many electronic databases do not yet have gold standard indexing of qualitative research our search terms for qualitative research were adopted from lists of qualitative search terms from key literature [31, 32]. No language restrictions or year limits were applied to the searches. Potentially relevant studies not published in English were translated by a native speaker. Although the focus of this review was non-pharmaceutical respiratory infection control, the search term ‘vaccination’ was initially included to allow for papers that included both pharmaceutical and non-pharmaceutical measures to be identified. Studies whose primary focus was vaccination alone were then excluded.

One author (ET) screened titles and abstracts of all identified papers against the inclusion criteria. Those that did not meet the inclusion criteria or were duplicates were excluded. A second author (MS) screened 10% of the title and abstracts to check that potentially relevant studies had not been missed. Full papers for the remaining identified records were retrieved and assessed for eligibility. Multiple papers from a single study were included if each paper presented unique data. Reference lists of all potentially relevant papers were reviewed and corresponding authors of included papers contacted to allow additional relevant studies to be identified.

Prior to synthesis, 3 authors (ET, MS & AG) extracted data and appraised the quality of included papers. Details about study design, participants (number and characteristics), type of non-pharmaceutical intervention and study context (timing, location, type of respiratory infection) were extracted into a Microsoft Word template. Papers were appraised using the Critical Appraisal Skills Programme (CASP) quality assessment tool for qualitative studies [33]. This tool enabled independent critical review of the quality of our identified studies. We employed this tool in order to systematically examine and document the strengths and weaknesses of the studies but not necessarily to exclude them. We acknowledge the potential risk that valuable new insights, grounded in data, might be missed when studies are excluded from a synthesis due to methodological flaws or lack of reporting. However, through discussion and consensus, one paper was excluded following appraisal as it reported no details about data collection, management and analysis. As such we could not be confident that this paper met our inclusion criteria of using qualitative methods of data collection and data analysis.

We used thematic synthesis [34] and meta-ethnographic [35] methods to synthesise findings. Meta-ethnography uses the notion of first, second and third order constructs to synthesise qualitative papers, where first order constructs reflect data on participant views, second-order constructs are the original researchers’ interpretations of themes arising from data, and third-order constructs are the new, common themes or interpretations derived by the synthesis of second-order constructs from multiple papers. Many of the papers included in this synthesis were applied studies that offered more descriptive analyses rather than conceptually rich analyses so a meta-ethnography was not possible and thematic synthesis was employed to identify and organise the synthesis data into themes. Components were adopted from meta-ethnography (reciprocal translational analysis and refutational synthesis) in order to facilitate synthesis across heterogeneous circumstances (different respiratory infections, different non-pharmaceutical interventions and groups with very different demographic, sociocultural and illness-related characteristics). Each paper was read repeatedly to become familiar with the data. The findings from each paper (participants’ views and author interpretation of findings) were then extracted verbatim and imported into Nvivo version 9.2. Data for each paper were coded line by line by ET according to meaning and content. Coding was inductive in nature (i.e. codes were grounded in the data, reflecting the language present in original studies). The list of codes generated for each paper was then systematically compared and developed as subsequent papers were coded (i.e. codes translated into one another or new codes added to the list). A coding manual was produced by ET to ensure transparent coding of the data. Initial themes and sub-themes were developed and refined through discussion with MS, AG and LY. ET then carried out consistency checks on the text coded to each theme and sub theme and organised themes into grids and tables to be compared and juxtaposed to examine variability between studies and by infection context, populations and intervention type. ET, MS, AG and LY then reviewed and discussed the evolving coding manual further and more abstract themes were developed based on the authors’ inferences and interpretation of the data. Diagramming was then employed to explore links and illustrate the key themes within the data.

The characteristics of the 16 included studies are shown in Table 3. Studies were published between 2005 and 2012 and represented the views of 1,022 participants from 9 different countries. The majority of studies (n = 12) recruited members of the general public with varying demographics and in different contexts. Two studies focused on specific ethnic groups and two on people with a chronic illness. In terms of infection context, 10 studies focused on an actual respiratory infection outbreak/pandemic, namely SARS and the H1N1 2009 pandemic. Typically, the included studies explored multiple non-pharmaceutical interventions or combined mitigation strategies involving pharmaceutical and non-pharmaceutical interventions. Personal protective measures were most commonly studied (n = 10).

The synthesis findings comprised three parts: 1) the ways in which the public evaluate non-pharmaceutical interventions, 2) public beliefs about respiratory infections and emerging outbreaks, 3) presentation of advice on adopting non-pharmaceutical respiratory infection control i.e. public preferences for how and by whom non-pharmaceutical respiratory infection control advice is presented. This paper focuses on the first two parts as they were relevant to the original aim of the review of understanding public perceptions of non-pharmaceutical interventions rather than views about how advice about adopting these behaviours is presented. Seven key themes were identified: perceived benefits of non-pharmaceutical interventions, perceived disadvantages of non-pharmaceutical interventions, personal and cultural beliefs about infection transmission, diagnostic uncertainty in emerging respiratory infections, perceived vulnerability to respiratory infection, anxiety about emerging respiratory infections and communications about emerging respiratory infections (Figure 2). Table 4 presents the definitions of each theme and an index of which themes were present in each paper. The themes and corresponding quotes are described below. Quotes are labelled with study reference, population, infection context and location.

Some personal protective measures, such as hand-washing and respiratory hygiene, appeared to be widely acceptable methods for preventing respiratory infection transmission across several studies. Studies conducted prior to and during the H1N1 2009 pandemic indicated that people were generally very familiar with hand and respiratory hygiene behaviours (e.g. hand washing, cough/sneeze etiquette) and viewed them as usual, acceptable, common-sense actions to reduce infection transmission.

“Mainly they tell you to wash your hands....Cover your mouth when you cough.......And don’t share hankies, they say, yeah.” (Pacific Peoples, H1N1 2009, New Zealand)[50]

“I already do the necessary, like washing my hands after the loo, which is, well, common practice and basic hygiene and being aware of it” (General public, non-pandemic, UK)[44]

For people who perceive themselves to be at high risk of catching and spreading respiratory infections, mask wearing was seen not only as an effective but also as a visibly demonstrative method of respiratory infection control, which could reduce social stigma. Chronic kidney disease (CKD) patients in Hong Kong saw mask wearing as a useful way to demonstrate their desire to be socially responsible and protect themselves and others during H1N1 2009 pandemic. This was related to social stigma experienced by CKD patients in Hong Kong during SARS, where they were perceived publically as high risk ‘super spreaders’ of infection.

“I wore facemasks all the time to tell others that I was responsible about my own and about others’ health. This strategy really worked as my friends and colleagues did not isolate me too much since then. I think I can use the same strategy in this swine flu to make others feel more comfortable with me, because I can show that even if I carry a lot of flu virus, I have already tried my best to protect their health by wearing facemasks.” (CKD patient, H1N1 2009, Hong Kong)[45]

Similarly, isolation and social distancing behaviours were viewed as acceptable in some contexts as a way of being socially responsible (i.e. protecting wider society). For example, being isolated or quarantined during SARS was accepted as necessary in order to protect others from infection. Studies conducted in a non-pandemic context indicated that the general public viewed home isolation, closure of schools and public gathering places (e.g. religious centres), and travel restrictions during an emerging respiratory infection outbreak as important for the protection of society as a whole, and would adopt these behaviours in principle for the greater good.

“We’re all trying to be good citizens. And we’re all trying to help, you know, other people by making sacrifices like being in quarantine.” (General public, SARS 2005, Canada)[36]

“. . . you know, if you can’t go to church for one week because everyone’s sick, then you know, call everybody on the phone or something. Do something different for the good of everybody else; you may have to suffer a little bit here and there.” (General public, non-pandemic 2009, USA)[42].

Non-pharmaceutical respiratory infection control was seen as problematic for some in terms of its perceived irrelevance. In particular, studies conducted pre and during H1N1 2009 pandemic suggested that advice to carry out additional hand washing or to focus on hand washing solely or intentionally for respiratory infection control was not seen as necessary or relevant to members of the general public who perceived themselves to be ‘regular hand washers’ (i.e. already regularly wash their hands for basic hygiene). This highlights the habitual, ingrained nature of hand washing behaviour and the firmly established social norms around when is acceptable to wash hands (e.g. after going to the toilet), which are likely to influence the perceived acceptability of hygiene behaviours [53].

“I don’t think that anybody washes their hands more than what they already do. You only wash your hands at normal intervals that I think you would normally, like if you're eating, after you've been to the toilet, etcetera.” (General public, non-pandemic, UK)[41].

“I can’t say I’ve changed anything cos I already do my hand washing, so others might benefit from this advice more really.” (General public, H1N1 2009 pandemic, UK)[46].

Potential to attract social stigma and cause embarrassment or discrimination was another perceived downside of adopting personal protective measures. Concerns about frequent hand-washing attracting social stigma (being perceived as overly fastidious or obsessive) or causing offence (insisting on hygienic behaviours in others) were evident in various infection contexts. Likewise, it appeared that mask wearing, although seen by some as an effective precaution, could generate concern about attracting discrimination. This was due to the presence of a mask being seen to explicitly indicate infection/illness in the wearer.

”If I’d said [when meeting friends for lunch] ah, before we touch any food we must all go and wash our hands, I’m not sure what everybody’s reaction would’ve been … I think people would’ve looked at me as if I’m slightly mad. ”(General public, H1N1 2009, UK)[44].

“Going out to the streets with a mask, people would stare at you as if you were contagious, but what they don’t know is that the mask is also protecting us from them” (Hispanic female, non-pandemic, USA)[47].

.. .like the old leprosy people in Europe.. .” (Domestic student, H1N1 2009, Australia)[52].

Other perceived drawbacks revolved around the potential negative impact of adopting non-pharmaceutical respiratory infection control. Common perceived barriers to carrying out personal protective measures appeared to relate to personal impacts such as the perceived physical discomfort of mask wearing (e.g. difficulties breathing whilst wearing masks), and the perceived discomfort and impracticalities of hand and respiratory hygiene (e.g. dry, sore hands and noses from frequent hand washing and tissue use; not having tissues when you need them and inconvenience of frequent hand-washing).

“It is hard to breathe with the mask. It is uncomfortable around chin area because I am sweating and the mask feels damp” (Hispanic female, non-pandemic, USA)[47].

“People don’t wash their hands after sneezing and coughing. Is it possible to wash hands frequently? If you sneeze 100 times, will you wash your hands 100 times? But we should wash our hands before taking a meal.”(General public, non-pandemic, Bangladesh)[48].

Common perceived barriers to social isolation and personal distancing from those who were symptomatic seemed to relate to perceived emotional costs. For example, people who were quarantined during SARS reported feeling segregated and stigmatised for being infected.

“Well I didn’t know I was going to get into this, but I actually feel like crying just to think about it, because I’m sure you saw the movie Ben Hur. I thought of that movie all the time while I was in quarantine because I remember the part of him going and looking for his sister and his mother, where they had that . . . sickness, leprosy. And they could not be with the rest of the people. They were down in a valley where all these people were and that’s how I felt. I was separate from the world.” (General public, SARS 2005, Canada)[37].

Furthermore, personal distancing was viewed as unacceptable within households and some cultural groups as it may limit social interactions which were perceived as socially and culturally necessary. For Maori people in New Zealand during H1N1, concerns about being able to continue to observe specific cultural practices and greeting protocols seemed to outweigh the perceived need to adopt personal distancing behaviours. Similarly, studies conducted pre-H1N1 and during SARS, suggested that the perceived need or wish to care for sick (isolated) loved ones can override any concerns about self-protection and personal distancing.

“My older sister said something about me having a really bad cold that weekend. And that’s only when I realized I could have got it. And I could have been very, very sick. I could have died. But it never, ever came to mind. We were so focused on her.” (General public, SARS 2005, Canada)[36].

What if it’s a baby you've got to look after you can't do it can you?” (General public, non-pandemic 2009, UK)[41].

People also seemed to consider the feasibility of distancing behaviours (social isolation and social distancing) in terms of the economic impact, both on a personal and societal level. Common perceived obstacles to staying home if sick and social distancing (such as school closures) during the H1N1 2009 pandemic were economic pressures to continue to work and concerns about familial and workplace commitments, and the wider adverse socioeconomic economic impact.

“The girl downstairs who got swine flu, works in my office, came to work three times with swine flu because she was bored at home, felt that there were things that needed doing in the office, and felt guilty at being away for so long. She came in, not being able to find a manager, hang around for three hours and then got sent home again” (General public, H1N1 2009, UK)[49].

“. . if they keep your kids home from school so you can’t work, people are going to go, ’I can’t do that, you know, I have to go to work, I have to have somebody take care of my kids’ . . . some people might choose to keep their kids home from school if they had that luxury, but too many people now don’t.” (General public, non-pandemic, USA)[42].

“If you shut down the schools though, you’ve basically shut down the economy because you’d have to have, then people would have to stay home so you’re affecting a lot more than people getting sick, you’ve just affected a huge financial working to the bulk of the country. That’s a big decision.” (General public, non-pandemic, USA)[42].

Common personal and cultural beliefs about how respiratory infections are caught and spread were evident across various infection contexts. Typically, respiratory viruses were seen as transmitted by air, caught via proximity to symptomatic others and more likely to be spread in cold ambient and water temperatures. Such beliefs are likely to influence the perceived efficacy of non-pharmaceutical interventions, particularly social isolation.

“I’ve always thought that on an aeroplane, because they re-circulate the air, there’s a chance that viruses and things are perhaps recycled through the air or ventilation system.” (General public, H1N1 2009, UK)[44]

“Just because like you haven’t got any symptoms doesn’t mean you are not carrying the germ. You know, everyone thinks that umm you know if someone’s got a cold and you stand, don’t stand near me, but just by talking to them you’ve probably picked up the germ. So I don’t really see the point in staying at home if you’ve got the symptoms because the chances are everyone has already come into contact with it anyway” (General public, H1N1 2009, UK)[49].

“A person should not be in contact with cold water for a long time to avoid getting cold/cough.” (General public, non-pandemic, Bangladesh)[48].

People seemed to acknowledge that a degree of uncertainty is to be expected in an emerging respiratory infection such as SARS. However, during the H1N1 2009 pandemic, they expressed doubts about their ability to identify symptoms (e.g. distinguishing pandemic flu symptoms from seasonal flu symptoms) and concerns at having to self-diagnose or make their own judgement about the presence of infection. Such diagnostic uncertainty seemed to be exacerbated for people who had other health concerns. This suggests that people do not see it as their role to self-diagnose in an emerging respiratory infection outbreak, which could influence the likelihood of implementing non-pharmaceutical interventions such as social isolation and using remote healthcare when symptomatic.

“I mean, the big thing is what are the symptoms, particularly what are the unique symptoms to whatever the pandemic is, that differentiates it from regular flu, or a cold? And how infectious is it, and what’s the mechanism of infection” (General public, H1N1 2009, New Zealand)[50].

“I think the vagueness of the symptoms could be confused with perhaps ordinary flu or just your condition really. You know, if you’ve got COPD, then it’s not necessarily swine flu at all. And I don’t really know how you can say it’s swine flu without having any tests [others in group agreeing].” (Chest patient, H1N1 2009, UK)[43].

“I think probably the most difficult part is to decide whether you have swine flu or not. So lots of people will be confused, when they have symptoms, what should they do” (General public, H1N1 2009, UK)[49].

In general the public appeared to accept the existence of risk to the community in an emerging respiratory infection outbreak. However, some may view themselves as less vulnerable or more capable than ‘others’ who they identify as high risk. During the H1N1 2009 pandemic the public seemed to evaluate their vulnerability to respiratory infection in terms of perceived health status and their proximity to the origin of outbreak (both in terms of geographical distance and perceived differences in living environments). A common belief expressed was that ‘others’ were at increased risk of infection, including people with chronic health problems, those in particular occupations (e.g. teachers), and those with impaired immune systems. This reflects a typical way of dealing with a health risk and its negative impact by “othering” i.e. blaming or differentiating the self from ‘the other’ or other groups of people and consequently denying personal risk [26]. This may result in people viewing non-pharmaceutical interventions to minimise the spread of infection in an emerging outbreak as more applicable to others than to themselves.

P1: “I’m personally not worried. I think my immune system is working well and I’m not in the situation of having an illness. P2: Same. I know like thousands of people have got it, but…I don’t personally feel at risk.” (General public, H1N1 2009 pandemic, UK)[46].

Similarly, in non-pandemic contexts the public appeared to focus on their lack of geographical proximity to the perceived origins of an emerging respiratory infection outbreak. A common belief was that a novel respiratory virus was unlikely to originate in a ‘modern, developed country’ so they would have more time to prepare and cope better if and when it reached them. For some, the perceived lack of geographical proximity to the outbreak seemed to result in a perceived lack of immediacy of risk, which suggests that risk of infection would need to be locally imminent (evident) before advice messages would be considered and behaviours adopted.

P1: We can get a vaccine off of them, sounds nasty but it's true, it's how the world works” (General public, non-pandemic, UK)[41].

“. . . that right now it’s in some third world country and it may come here. I don’t think that’s going to be good enough. I think there’s going to have to be some indication that it is actually in your own community before you take steps as drastic as shutting down anything.” (General public, non-pandemic, USA)[42].

In a pandemic context, the concept of geographical proximity was challenged. Instead, it was recognised that respiratory viruses could spread worldwide quite rapidly and to more geographically remote locations due to air travel.

“Well we think that we’re different because we’re far away. But actually, if you think of how people travelled here, it’s the biggest factor for it always, because everyone who comes here comes in an aeroplane, pretty much. And they come from everywhere.” (General public, H1N1 2009, New Zealand)[50].

During the H1N1 2009 pandemic the public also appeared to evaluate their susceptibility to a new respiratory infection in terms of their own living environments compared to the living circumstances where they believed a novel respiratory infection was likely to emerge (e.g. low hygiene levels, high population density, poor border control and health systems).

“It won’t be like ones in the past, years ago. There’s so much medical research been done. And things are cleaner, better housing, people are cleaner and the streets are cleaner so a pandemic wouldn’t be so bad now” (General public, H1N1 2009 pandemic, UK)[46].

“.. ..some cultures where they have let say more respect for traditional medicine than modern medicine.. .. …are also going to be a problem.. .that’s why lots of pandemic in say Asia and Africa, and not so much in Europe or America” (International postgraduate student, H1N1 2009, Australia)[52].

This seems to reflect a belief that it ‘won’t happen to me, it happens to others, elsewhere’ and suggests that the public do not rationally evaluate their risk of infection but actively try to distance themselves from the threat by delineating themselves from others and other circumstances which are associated with risk of infection. This may lead to underestimating personal risk in an emerging respiratory infection outbreak and feeling that advice to adopt non-pharmaceutical interventions is not relevant.

The novelty factor (shock of the new) appeared to affect public anxiety in the early stages of an emerging outbreak. During SARS and the H1N1 2009 pandemic public anxiety typically evolved over time from an initial state of anxiety gradually decreasing to the point of people making light of it as the public’s familiarity with the outbreak or respiratory infection increased through personal experience or knowing someone who had contracted it. This could be a further example of people distancing themselves from threat in order to cope with the novel health threat.

“I think everyone has taken it a lot less seriously now than when it was first like, when it was first thing, then everyone was like ahhh, we have this swine flu and everybody is going to die but not now, it’s like if someone coughs, “ha swine flu!”, you know, no one cares now” (General public, H1N1 2009, UK)[49].

“There is so much joking about it, like I was on the train and this guy just coughed, and he was like, ’it’s ok, it’s not mutated, I don’t have swine flu’ and everyone just laughed. No, I don’t think people are worried about it” (General public, H1N1 2009 pandemic, UK)[46].

For some, particularly those who were regarded to be more vulnerable to infection (e.g. people with a chronic illness, pregnant women, mothers of young children), the novelty of the virus and in particular their perceived lack of protection from it seemed to contribute to their worry.

“[with seasonal flu] we all feel quite safe because we’ve got a protection and we know ordinary seasonal flu can be serious. But we’ve got our jab and it’s protected us. And suddenly there’s a flu out there what there hasn’t been a jab for, and we can catch it as quick as anybody else. And nobody quite knows really what effect it’s going to have on us and I think this has been some of it, because right at this time we’re vulnerable, we’ve no protection given us. And we all feel as we need that protection to get through this … And I think that’s making us worry.” (Chest patient, H1N1 2009, UK)[43].

In contrast, some people with other health issues expressed less worry about becoming infected as they saw their current health issues as more pressing.

“I can’t eat properly and while I’m eating I’m gasping for breath … so swine flu is the least of my worries, if you know my meaning … this [chest problem] is the priority. If I can get this right, if I can at least walk a little bit more than I can do now, I’d be happy.” (Chest patient, H1N1 2009, UK)[43].

Diminishing anxiety over the course of an outbreak seemed to be influenced by people’s views about communications about an emerging respiratory infection outbreak. Whist some people felt that they were not given all the facts during the H1N1 2009 pandemic or that its severity was prematurely downplayed, generally the public were quite sceptical about the way information on the new respiratory infection was presented to them (especially by the media). Typically, communication efforts are seen as unreliable, premature, inconsistent, sensationalist and unduly alarmist.

P1: they’ve [the media] set about and managed to get everyone, or the majority of people into quite a panic about the whole thing. P2: They do on purpose whip up panic and anxiety in people” (General public, H1N1 2009 pandemic, UK)[46]

“At the beginning sensationalism -a blast of information-, followed by information in small doses and poor, bad information” (Chronically ill patient, H1N1 2009, Spain)[51]

“…the minister said one thing, someone else said another thing, Sarkozy said he was going to get everyone vaccinated, the German [leader] would say whatever, hence, we were getting so much contradictory information” (General public, H1N1 2009, Spain)[51]

“I feel they’ve a lot of hype with a lot of things, not just the swine flu, and particularly the media, they like to blow things up, don’t they? They like to scare people really. On the other hand I suppose scaring people is only one way to get them to move.” (Chest patient, H1N1 2009, UK)[43]

A common belief was that media reporting in an emerging outbreak is over-hyped or can amplify the risks. It appeared that the public evaluated their personal risk by comparing personal experiences with the ‘official’ information they have been given. When this didn’t match up to actual experiences, the public doubted the credibility of the information being presented and further doubted the media as a reliable information source. This inconsistency may lead to public fatigue about respiratory infection communications and a blunting of advice messages on non-pharmaceutical interventions. Doubts about the perceived credibility and trustworthiness of information about a new respiratory infection outbreak are likely to influence public behavioural responses to an emerging respiratory infection outbreak and may also lead to people to disregard future advice.

“It became reasonably clear reasonably quickly last time that hundreds and thousands and millions weren’t dying. Even when they kept on sort of saying things were happening, and then you saw the numbers, it just didn’t add up.” (General public, H1N1 2009, New Zealand)[50]

“Before that we had avian flu, and before that SARS and they were all, this is the end of the world. Don’t travel, don’t eat chickens, watch out for the dead duck in the street. And yeah, it raises your expectations and then nothing happens, and then next time it comes along you are just more cynical” (General public, H1N1 2009, UK)[49]

Some aspects of non-pharmaceutical respiratory infection control were viewed favourably by participants in these studies due to their familiarity, potential to reduce social stigma and capacity to demonstrate social responsibility or community mindedness. Hand hygiene and respiratory hygiene, in particular, appeared to be well-established and accepted concepts in the minds of these participants. Doubts and concerns existed about the perceived relevance of non-pharmaceutical respiratory infection control, its potential to attract social stigma and its perceived adverse impact (physically, emotionally and socioeconomically). There were particular concerns about personal distancing and the wearing of masks in some contexts. The synthesis also suggested that the perceived necessity, efficacy and feasibility of adopting non-pharmaceutical respiratory infection control may be influenced by personal and cultural beliefs about respiratory infection transmission, perceptions and feelings about vulnerability to respiratory infections, and concerns around self-diagnosis of respiratory infections and the communication of reliable information about respiratory infections in emerging respiratory infection outbreaks.

Synthesising the qualitative literature highlighted the ways in which the public evaluate the feasibility of non-pharmaceutical respiratory infection control. Instead of passively accepting non-pharmaceutical interventions that are recommended to reduce the transmission of respiratory infections, the public actively consider the perceived costs and benefits of adopting non-pharmaceutical interventions and reflect on their beliefs about and feelings towards respiratory infections and emerging outbreaks.

Widespread public endorsement of non-pharmaceutical respiratory infection control was highlighted by the synthesis. However, such endorsement of non-pharmaceutical interventions, particularly hygiene, may not accurately and transparently reflect people’s beliefs about these behaviours. Agreeing with the socially desirable position of taking action to protect oneself and others from respiratory infection may instead reflect a more perfunctory or socio-linguistic function. Indeed, by demonstrating that you know and subscribe to what is accepted as ’common-sense’ correct behaviour could serve the purpose of affirming membership of a particular sociocultural group e.g. ‘good citizens’ [54, 55]. Given the inconsistency between the socially desirable assertions about hygiene and social distancing behaviours and the numerous perceived barriers and costs of the behaviours that are subsequently raised, this seems quite probable.

One reported barrier was diagnostic uncertainty in emerging respiratory infections. In particular, public concern seems to exist about identifying symptoms and making a judgement about the presence of a respiratory infection. The public do not see it as their role to self-diagnose and feel uncomfortable about taking on the perceived role of the doctor. This finding is consistent with previous research suggesting that members of the general public do not feel comfortable with making medical decisions [56‐58]. In an emerging respiratory infection outbreak, promotion of self-diagnosis and accessing remote healthcare is vital to reduce spread of respiratory infection and avoid overstretching healthcare services. Further research is required to explore how public confidence in remote healthcare could be improved.

Socio-economic barriers to adopting social distancing behaviours were also highlighted in the synthesis. This is consistent with previous research that has shown lack of access to child care and financial barriers with regards to social isolation exist, particularly for those on low incomes [59‐61]. Difficulties in following non-pharmaceutical interventions faced by those with fewer resources has the potential to exacerbate the socio-economic differences in the public health impact of a respiratory infection, particularly in an emerging outbreak [62].

Another potential barrier to adopting non-pharmaceutical interventions appears to be the denial of personal risk. Although everyone is potentially at risk of respiratory infection, the public seem to deal with the threat of infection by attributing vulnerability to ‘other, less good’ groups of people. This process of ‘othering’ is consistent with previous qualitative research on how the public make sense of emerging infectious diseases [27, 28]. Such denial could be seen literally as public ignorance or erroneous evaluations of risk of infection. However, denial of personal risk also means asserting that one is not in the group of people who are vulnerable or in danger due to either their risky, unhygienic behaviour or being weak (e.g. due to illness or age) - so rather than a misunderstanding of risk this can reflect a means of coping with the threat of infection. This interpretation suggests an important role of emotional and sociocultural factors in shaping individual ways to perceive and react to respiratory infection risk, rather than assuming that people’s perceptions of respiratory infection risk are purely an individual, rational, cognitive process [26, 63].

Our systematic review and synthesis of qualitative studies on non-pharmaceutical respiratory infection control is novel. It provides helpful insight into the ways in which the general public process the risk of respiratory infection and evaluate the feasibility of non-pharmaceutical respiratory infection control. We used rigorous methods of systematic review and referred to the ENTREQ statement [64] to facilitate clear reporting of our synthesis of qualitative research. Although we contacted all authors to obtain any further studies, we did not comprehensively search the grey literature, which may have excluded some relevant literature. The review is clearly limited to the perceptions of participants in the included primary studies, which were typically conducted in more developed countries and predominantly explored perceptions of personal protective measures. However, the synthesis did incorporate people’s perceptions of different non-pharmaceutical respiratory infection control behaviours in varying respiratory infection contexts and can offer a higher level of conceptual thinking about public responses to non-pharmaceutical respiratory infection control across different contexts. Nevertheless, it is acknowledged that the process of synthesising qualitative studies is inherently interpretive. Our synthesis is one possible interpretation of the data. It is entirely possible that another research team may generate another interpretation of this set of studies.

Adoption of non-pharmaceutical public health measures is likely to be improved by addressing common public beliefs and concerns about the necessity, efficacy, acceptability and feasibility of non-pharmaceutical respiratory infection control. To maximise adoption of non-pharmaceutical interventions to reduce transmission of respiratory infections it may be necessary to find ways that allow people to associate non-pharmaceutical respiratory infection control with a positive identity rather than a negative or vulnerable identity, i.e. viewing non-pharmaceutical respiratory infection control as behaviours to be adopted by all, not just as actions for those perceived as more vulnerable to infection. For example, positive framing of advice messages around maintaining well-being rather than avoiding infection might improve the perceived relevance of non-pharmaceutical respiratory infection control to those who do not acknowledge that they could be at risk of infection. Further research is needed to clarify how best to reframe advice.