Solid fuel users’ perceptions of household solid fuel use in low- and middle-income countries_ A scoping review

Almost half of the global population is exposed to household air pollution (HAP) from the burning of biomass fuels primarily for cooking, and this has been linked with considerable mortality and morbidity. While alternative cooking technologies exist, sustained adoption of these is piecemeal, indicating that there is insufficient knowledge of understandings of HAP within target communities. To identify potential gaps in the literature, a scoping review was conducted focused on solid fuel users' perceptions of HAP and solid fuel use in lowand middle-income countries. From the initial 14,877 search returns, 56 were included for final analysis. An international multi-disciplinary workshop was convened to develop the research question; six key domains: health; family and community life; home, space, place and roles; cooking and cultural practices, environment; and policy and practice development, were also identified using a Social Ecological Model framework. The review showed a series of disconnects across the domains which highlighted the limited research on perceptions of HAP in the literature. Reviewed studies showed that participants emphasized short-term health impacts of HAP as opposed to longer-term health benefits of interventions and prioritized household security over improved ventilation. There was also a socio-demographic gendered disconnect as although women and children generally have most exposure to HAP, their decision-making power about use of solid fuels is often limited. In the domain of policy and practice, the review identified the importance of community norms and cultural traditions (including taste). Research in this domain, and within the environment domain is however limited and merits further attention. We suggest that interventions need to be locally situated and community-led and a deeper understanding of perceptions of HAP could be obtained using participatory and innovative research methods. Bridging the disconnects and gaps identified in this review is essential if the global disease burden associated with HAP is to be reduced.


Introduction
Approximately 3 billion people worldwide rely on solid fuels such as wood, charcoal, animal dung and crop residues to cook, heat and/or light their homes (WHO, 2014). Household air pollution (HAP), that is pollution of the indoor air caused by the incomplete combustion of the solid fuels in the domestic setting, results in an estimated 4 million premature deaths every year from diseases including pneumonia, stroke, ischemic heart disease, chronic obstructive pulmonary disease (COPD) and lung cancer (WHO, 2019). In terms of global health burden, HAP leads to mortality broadly equivalent to a lack of malaria control and is one of the primary environmental health risks (WHO, 2014).
Most of the exposure and consequent disease burden associated with HAP occurs in low-and middle-income country (LMIC) settings (Stanaway et al., 2018), where the use of solid fuel is strongly associated with poverty (e.g. Bonjour et al., 2013). The 'energy ladder', though much critiqued (e.g. Alfaro and Jones, 2018), highlights an idealized progression from solid fuels to cleaner fuels like LPG (liquefied petroleum gas), natural gas and electricity with increasing wealth (Smith, 1987). Ascending the energy ladder, however, is primarily driven by income levels, often with the poorest unable to reach the 'bottom rung' of the ladder with no access to alternatives to reduce their exposure and no opportunity to ascend the ladder. There is also a significant gender dimension to use of solid fuels in LMICs. For example, recent measurements of exposure to HAP from burning solid fuels in households in Uganda and Ethiopia show that women and young girls experience daily concentrations of HAP that are typically seven times higher than men and young boys (Okello et al., 2018). This is reflected in much of the epidemiological evidence reported from HAP studies where women are at greater risk of COPD, lung cancer and eye cataracts (Kurmi et al., 2012).
Many intervention studies have attempted to move people up the energy ladder either by using 'improved' cookstoves (ICS) that still use solid fuels but aim to reduce HAP production during cooking (Alexander et al., 2015) or by moving people to more modern fuels (i.e. electricity or LPG) (Quinn et al., 2019). These approaches place insufficient emphasis on gaining an understanding of the historical, cultural and behavioral complexities that underpin solid fuel use, thereby failing to create sustainable changes that improve people's health. Indeed, despite over 80 million ICS having been distributed since 2010 (GACC, 2017), their sustained use remains a challenge (Das et al., 2018). The urgent need for sustainable solutions facilitating the adoption of cleaner cooking practices has been highlighted by the WHO, who report that without a substantial change in policy, the number of people lacking access to cleaner fuels and technologies will remain static by 2030 (IEA, 2017). Such policy changes and design of sustainable solutions should have individuals at its core and be led by an in-depth understanding of the socio-cultural context of solid fuel usage for cooking, including the influence of non-cooking factors (Gould and Urpelainen, 2018), which can only really be understood when perceptions are considered at the core of solutions.
For instance, introducing new cooking practices needs to respect the cooking-culture relationship, which is particularly relevant in LMICs, with many social and religious occasions being marked by festive meals prepared in traditional ways (Quintero-Angel et al., 2019). Furthermore, beyond the high costs limiting the take-up of clean cooking alternatives, such as ICS, Dickinson et al. (2018) also highlighted the need to recognize the importance of peers in the ICS decision-making process. The role of community leaders in influencing change is an important one. Miller and Mobarak (2013) found that the role of "opinion leaders" were important in the adoption of ICS, however peerto-peer, or social learning had a negative impact on ICS adoption. Enclosed within this socio-cultural context, but often overlooked, is the importance of understanding individuals' views and opinions towards traditional and alternative cooking practices and associated potential impact on their health.
Whilst academics can deliberate policy, any change has to come from individual users. Consequently, interventions which aim to target individuals or communities must first have an understanding of the perceptions of both exposure to HAP and the associated health risk (Egondi et al., 2013). Many studies have assessed exposures to HAP and evaluated implementation of ICS, yet few have considered behavior change as a fundamental component of the success of such interventions (Barnes et al., 2015;Goodwin et al., 2015). Perceptions on the risk of HAP play a critical role in behavior change, with limited awareness or concerns about the negative health-related impact of HAP representing one of the many barriers to sustainable behavioral change to the uptake of less harmful cooking practices (Jewitt et al., 2020). These findings indicate that without an understanding of users' experiences, views, and motivations for change (or lack of), policy and practice have little chance of success.
In summary, while there is considerable evidence that exposure to HAP resulting from domestic solid fuel use is linked to poor respiratory (Tamire et al., 2020) and cardiovascular health (Yang et al., 2018), significantly less research has examined perceptions in LMIC communities in relation specifically to solid fuel use and collection, and how this influences HAP. The aim of this paper is to explore and summarize knowledge of solid fuel users' perceptions of solid fuel use and collection for cooking, heating and lighting in LMICs. To achieve this, this paper reviews evidence from the scientific literature across a variety of domains and disciplines that looks at perceptions of solid fuel use in terms of: • Health • Family and community life • Home, space, place and roles • Cooking and cultural practices • Environment • Practice and policy development.

Methods
To explore solid fuel users' perceptions of solid fuel use for cooking, heating and lighting in LMICs, a scoping review was undertaken. A scoping review provides an overview of a vast topic (Moher et al., 2015) and is a suitable approach for exploring a body of literature and to identify knowledge gaps (Munn et al., 2018). Arksey and O'Malley's (2005) influential work on the 5-step scoping review framework was the basis for the protocol for conducting this scoping review:

Identifying the scope of the review
Owing to the potentially vast number of research questions that could arise from the chosen topic of solid fuel usage, a workshop (conducted at The University of Stirling on 4th December 2018) was used to develop the research question and key domains ( Table 1). The workshop was comprised of academics from and working in the UK, South Africa and Malawi from multiple disciplines including geography, psychology, maternal health, lung health, environmental health, social work and housing to allow us to develop a well-considered research question from a more global perspective. In the final selection of the key domains, the information and opinions gathered in the workshop were considered within the Social Ecological Model (UNICEF, 2016), which suggests that people's perceptions, behaviors and experiences are shaped by the interaction of individuals with their community, environments, interpersonal relationships, and policy and political environments (Fig. 1).

Identifying relevant studies: Eligibility criteria, information sources and searching
Five key databases from multiple disciplines were electronically searched (CINAHL, JSTOR, ScienceDirect, PubMed and Web of Science (Core Collection)) using the search terms detailed in Table 2.
Filtering methods included the publication date range from 2000 to 21st March 2019, published in the English-language only, including those papers where the subject country was LMIC-only (for this purpose defined as those receiving Official Development Assistance (ODA)), and excluding grey-literature. The focus on articles published after 2000 was linked to the start of the Millennium Development Goals (MDGs) (United Nations, 2000). Limiting the scope of the analysis by focusing only on articles published in English may mean that some potentially relevant papers have not been included here but was necessary from a time and resources perspective. Whilst the value of grey literature has been recognized (Rothstein and Hopewell, 2009), due to a lack of methodological guidance on the inclusion of grey literature for scoping studies (Adams et al., 2017), this was excluded from the search. Articles were only included if they referred to (although did not have to be the main focus) on solid fuel users' perceptions of fuel use in the domestic setting. Articles which only explored perceptions of non-domestic users (e.g. traders, construction workers) were excluded.

Study selection
The initial search returned 14,877 articles, of which many were duplicates, which highlighted the thoroughness of the search approach. Level 1 screening (title and abstracts) was used to exclude articles that did not address the study's research question. Level 2 screening (fulltext) was completed by five reviewers independently (AM, LC, HP, SL, IU) following discussion and agreement of the inclusion/exclusion eligibility criteria (outlined in Section 2.2). Each article was checked for fit with the domains outlined in Table 1. To ensure consistency of the selection across all reviewers, AM double checked all the level 2 screening decisions. For a detailed overview of the article selection see Fig. 2.

Data extraction
After level 1 and 2 screening, the full-text study characteristics were tabulated and data extracted into a spreadsheet, organized by domain (Table 1) for included studies. Descriptive characteristics of the included studies, as well as micro-data (specifically that which detailed perceptions or experiences in each key domain), were tabulated. The purpose of this exercise was not to review the included articles, but instead to determine which data to extract based upon the agreed criteria. The primary aim was to determine whether studies directly explored solid fuel users' perceptions, followed then by ensuring each study related to at least one of the six domains of the overarching research question. As the focus of this review was not to explore perceptions on all potential issues around solid fuel use and collection, but rather to scope the genuine perceptions identified by those using solid fuels in LMICs, not all domains were equally populated. Data extraction was undertaken by one reviewer (AM) and double-checked independently by a second reviewer (SL) to ensure all relevant data were extracted and organized within the correct domain.

Data analysis
The data were analyzed by four team members (HP, IU, AM, LC), each analyzing the domain most closely related to their own area of expertise (e.g. health, environment). We used a broad thematic analysis The key domains of interest developed using the Social Ecological Model theory-based framework to explore the study's research question from individual-level (inner oval) to societal (outer oval) (Sallis et al., 2008).  A. McCarron, et al. Environment International 143 (2020) 105991 to develop the main themes from each domain. Subsequently, the four researchers came together to discuss the themes and key findings across all and within each domain.

Results
Our initial search yielded a total of 14,877 articles, which was reduced to 8917 after duplicates were removed electronically (Fig. 2). A further 88 duplicates were removed manually having not been detected electronically using Endnote. Level 1 title and abstract screening excluded 8,490 articles based upon factors including the study area being non-LMIC or studies which did not explore participants' experiences and/or perceptions. The remaining 339 articles were full-text screened for eligibility. Of these, 283 articles were excluded because they did not meet the requirements of the study (i.e. they did not focus on perceptions or they did not explore one of the outlined domains ( Fig. 1;  Fig. 2)). This left 56 articles that met all study criteria and were included in the analysis.
The majority of the 56 included articles were focused on evaluating the impact of interventions such as the introduction of improved cookstoves (ICS)-which, in general, burn less solid fuels thus reduce HAP production when used correctly-or interventions introducing new stoves that use more modern fuels, e.g. electricity or LPG. Although interventions were not deliberately scoped in this review, solid fuel perceptions are not frequently considered without mention of interventions, such as the introduction of ICS (see Appendix). Questionnaires and interviews were the methods most frequently used by researchers to explore perceptions (Table 3), with observations and focus groups also used. Of the six outlined domains (Fig. 1), health and cooking & cultural practice were particularly well documented in the literature, with more than half of all included articles referring to these ( Table 3). Perceptions of solid fuel use and collection were studied across 26 different countries in the included articles, of which were most frequently undertaken in India, Bangladesh, Mexico and Kenya (Fig. 3).
The findings of the scoping review are outlined below in relation to the six domains (health; family and community life; home, space, place and roles; cooking and cultural practices; environment; and practice and policy development; Fig. 1).

Health
Studies exploring perceptions of long-term health impacts related to solid fuel collection (e.g. lower back pain (Borah, 2015), cancer and COPD (Firdaus and Ahmad, 2011), were rarer than studies that focused on short-term health impacts. The use of solid fuels for cooking was associated with household smoke and was considered to impact comfort as well as being linked to a range of health symptoms (e.g. Akolgo et al., 2018;Devakumar et al., 2018;Troncoso et al., 2007). The symptoms most often reported and linked to HAP exposure included respiratory difficulties, runny eyes, runny nose, throat irritation, coughing, breathing difficulties, wheezing, chest pains, dizziness and nausea (Burwen and Levine, 2012;Chandna and Honney, 2017;Devakumar et al. 2018;Gordon et al., 2007;Jin et al., 2006;Miah et al., 2009;Rouse, 2002;Tamire et al. 2018;Thompson et al. 2007, Thurber et al. 2013. Reviewed studies highlighted that women are particularly at risk of health harms from HAP, given that they are the primary cooks and spend more time in the kitchen and the home (Devakumar et al. 2018). Concern was also frequently expressed for children within the household who were considered particularly at risk from breathing in the smoke or from burns (Hooper et al., 2018;Pailman et al., 2018) because they are often near the mother when she cooks (Alam et al., 2006;Devakumar et al., 2018). Some studies reported that participants did not always see a connection between either their use of the stove, fuel type and/or smoke and their health symptoms (Akintan et al., 2018;Cundale et al., 2017, Hollada et al., 2017, Sesan, 2012, rather some saw the smoke as dirty (Wang and Bailis, 2015). Furthermore, several studies highlighted how participants did not necessarily see ash and smoke as a bad thing; in some cases the smoke was perceived as positive because it protected against insects , and thus during the rainy season protecting newborns from insect-borne disease (Tamire et al. 2018). Smoke exposure was seen by some as an unavoidable part of daily life (Rhodes et al. 2014).
Many of the studies (n = 37) in this scoping review investigated the health impacts of interventions to reduce exposure to HAP either through switching to cleaner fuels or through the use of ICS. In those studies, interventions participants frequently reported perceived health benefits such as reduced smoke in the home, reduced respiratory problems, fewer burns, fewer headaches, coughs, back pain or runny noses or eye irritation (Alam et al. 2006;Gebreegziabher et al., 2018;Hessen et al., 2001;Khushk et al. 2005;Rehfuess et al., 2014). However, participants were often less clear or less concerned about the long-term *Some articles investigated more than one study area, deployed several qualitative methodologies, and covered various domains hence counts are greater than 56.^revealing qualitative choice behaviors. In total, 26 different countries spanning 4 continents were studied in included articles and 7 papers explored perceptions in more than one country. A. McCarron, et al. Environment International 143 (2020) 105991 health benefits that the interventions provided (Akintan et al., 2018;Hollada et al., 2017;Stanistreet et al., 2014). Several studies reported that health improvements or benefits were not in fact raised by the participants unless they were prompted by researchers and that participants did not necessarily relate the alleviation of their symptoms to the use of ICS (Cundale et al., 2017, Hollada et al., 2017Jürisoo et al., 2018;Thurber et al., 2013;Thurber et al., 2014). In other studies, participants had concerns over the safety and fumes produced by cleaner fuels also such as LPG (Gould and Urpelainen, 2018;Hollada et al., 2017;Malakar et al., 2018;Thompson et al., 2018). LPG and other fuels were sometimes highlighted to lack some of the wider health benefits that cooking using solid fuels provided, e.g. insect control due to household smoke .

Family and community life
Many of the articles (n = 19) highlighted the important role of the community in shaping thoughts and actions, norms and beliefs, relating to solid fuel use and/or the uptake of ICS. Indeed, neighbors and families had a strong impact on whether an ICS was adopted, and this influence can be both positive and negative. For instance, current practices (e.g. gathering wood) are shaped by community tradition, and therefore give a sense of belonging Sunikka-Blank et al., 2019). Cooking with cleaner fuels (i.e. LPG) when the rest of the community are using solid fuels can be divisive and disassociating from the rest of the community. Conversely, if key members of the community adopt an ICS (e.g. teachers, ministers), they could be a very convincing force in ensuring the entire community adopts the new way of cooking (Jürisoo et al., 2018;Stanistreet et al., 2014). Consequently, new solutions to reduce exposure to HAP through changing cooking habits need to be community-led, including solutions that are approved by the community leaders and meeting the needs and priorities of the local community (Chandna and Honney, 2017;Tamire et al., 2018;Tigabu, 2017).
Several studies highlighted that adopting an ICS was associated with benefits for the whole family, such as acquiring a higher social status (Alam et al., 2006;Diaz et al., 2008;Jürisoo et al., 2018;Thompson et al., 2018) and allowing for more family socializing due to improved conditions (i.e. less smoke) inside the home Stanistreet et al., 2014). Such benefits were particularly pronounced for the children, as the smoke created by cooking with solid fuels is perceived to prevent them from concentrating on their education through not being able to play and study indoors (Devakumar et al., 2018;Rehfuess et al., 2014) and through spending time fetching firewood (Asante et al., 2018).
Studies stressed that despite the benefits and even if ICS are perceived as acceptable by a community, there may still be issues of affordability of the ICS for some within the community (Takama et al., 2012). The household's wealth was an important factor influencing the willingness to adopt a new stove, with high costs being identified as a key barrier among poorer households . This role of cost is considered further in Section 3.6.

Home, space, place and roles
The findings from this review revealed that cooking with solid fuels produces smoke and soot (e.g. Baquié and Urpelainen, 2017), which sticks to walls, appliances and cookware (Yonemitsu et al. 2014), making the home look less attractive and requiring more frequent cleaning (Asante et al., 2018;Wang and Bailis, 2015). Some studies highlighted that this more frequent cleaning can cause painful roughness to the women's hands (Alam et al., 2006). Avoiding using solid fuels was associated with more appealing, cleaner and warmer homes (Chen et al. 2016;Diaz et al., 2008;Granderson et al., 2009;Khushk et (Gordon et al., 2007). A more appealing-looking home could, in fact, improve the household's social status (Wang and Bailis, 2015).
Across the included articles it was clear that gender roles had a key part to play in the fuel decision-making process. In most cases, women are the primary cooks, clean and oversee childcare, whilst men are often seen as the decision-makers (e.g. about fuel usage). In this review, men were generally not accepted as cooks within the home Stanistreet et al., 2014). In the context of uptake of ICS or cleaner fuels, gender roles remained important. The studies stressed that cooking methods need to be time-saving (Cundale et al., 2017;Stanistreet et al., 2014) and suitable for cooking large family meals more efficiently (Hollada et al., 2017). Although the time-saving component would mostly influence the woman's time (i.e. reducing time to collect cooking material, reduced cooking time and reduced time spent on cleaning and illness management (Cundale et al., 2017;Devakumar et al., 2018), several articles also reported the importance of the husband's satisfaction with the meal preparation time (Rhodes et al., 2014;Thompson et al., 2018). However, it is important to consider that while improved satisfaction with cooking times could convince husbands to change practice, they did not always believe the smoke produced during cooking was harmful to health (Thompson et al., 2018). These findings indicated that different aspects of ICS need to be addressed depending on the target audience.
The findings from this review also revealed other competing priorities with regards to home safety rather than HAP exposure (Devakumar et al., 2018). For instance, it was acknowledged that modifications made to the home to protect from other issues (e.g. keeping windows closed to avoid crime or plastic bags in the ceiling to avoid malaria) can worsen HAP, but avoiding the risk of crime or malaria were seen are bigger priorities to individuals than HAP prevention. Lack of ventilation was, in addition to increasing exposure to HAP, linked to bad smells (Devakumar et al., 2018;Qian et al., 2016). Although a questionnaire study did reveal that most participants are aware of the benefits of better ventilation, using a chimney and separating the kitchen from the home (Ali et al., 2011), the issue of proper ventilation is complex and influenced a variety of factors. These factors range from safety concerns (due to insecurity of the community and fears of malaria), the poor nature of houses (e.g. no windows or chimneys available), and avoiding the cold, through to mixed knowledge about the benefits of ventilation (Muindi et al., 2014;van Gemert et al., 2013;Weaver et al., 2017). This stressed the importance of the socio-spatial environment to any intervention to reduce HAP from cooking or heating the home with solid fuels.

Cooking and cultural practices
Our scoping review showed that cultural traditions, norms and habits underpin household cooking practices (e.g. Malakar et al., 2018;Rhodes et al., 2014;Tigabu, 2017;Troncoso et al., 2007). Frequently, the use of solid fuels for cooking has been passed down through the generations and is therefore considered as something to be valued and abided by (e.g. Bielecki and Wingenbach, 2014;Malakar et al., 2018;Rehfuess et al., 2014). Religious beliefs (Wang and Bailis, 2015) and myths (Akintan et al., 2018;Tamire et al., 2018) were found in some cultures to affect the design of the kitchen and the choice of fuel or the stove used. There was often a resistance to change the type of fuel or stove used for cooking because people were satisfied with the traditional methods they used (Akolgo et al., 2018;Stanistreet et al., 2014;Troncoso et al., 2011;Wijayatunga and Attalage, 2003), or because some members of the household were thought not to approve of the change (Troncoso et al., 2007). In Guatemala, Bielecki and Wingenbach (2014) found generational differences also in the uptake of cleaner cooking fuels, with younger generations more willing to move away from solid fuel use. Indeed, younger generations expressed less fear of the ICS exploding and causing fires (Hollada et al., 2017) and seemed more easily convinced to adopt new practices (Wang and Bailis, 2015).
Another important factor underpinning cooking practices was the taste of the food when cooked using different fuels (e.g. Asante et al., 2018;Rhodes et al., 2014), with solid fuels generally considered to give a preferential taste to foods Taylor et al., 2011;van Gemert et al., 2013). For example, roti, the traditional Indian flatbread, was found to be unpalatable when cooked using LPG instead of using traditional cooking methods (i.e. using solid fuels) (Wang and Bailis, 2015). In addition to taste, the cooking fuel used was in some cases also linked to the nutritional content of food (Hollada et al., 2017). Those cultural aspects are key to understanding the complex issue of fuel use for cooking and heating, particularly in LMICs.
Finally, the review revealed that the choice of cooking fuel is underpinned by the availability of fuel types (or conversely the lack of choice) within proximity to the household (Gupta & Köhlin, 2006;Stanistreet et al., 2014;Akolgo et al., 2018;Pailman et al., 2018). Where multiple fuels were available for people, the cost of fuel then commonly influenced cooking fuel choice, with cleaner fuels more expensive than solid fuels (Taylor et al., 2011;van Gemert et al., 2013). The use of cleaner fuels, e.g. LPG, were often linked to convenience (Gould and Urpelainen, 2018), since they were generally perceived to cook food quicker than traditional biomass-based fires (Asante et al., 2018;Wang and Bailis, 2015), and were often used when short of time or when people were tired (Hollada et al., 2017). While some perceived solid fuel collection as a time-consuming task and therefore favored cooking methods which reduced or eliminated time spent on this task (Gebreegziabher et al., 2018;Granderson et al., 2009;Hooper et al., 2018), others perceived time spent collecting fuel as an opportunity for social interactions (Hollada et al., 2017). The versatility and flexibility of open solid fuel fires were valued by some (Akintan et al., 2018), given that fires could be moved to different locations as required and were made on the floor, thereby reducing the amount of lifting required for heavy cooking pots (Troncoso et al., 2007). However, some found that keeping a fire going required constant tending, meaning people could not get on with other tasks at the same time as cooking (Granderson et al., 2009). In comparison to traditional solid fuel-based fires, there were often concerns regarding durability, maintenance and repair with regards to improved biomass cook stoves and stoves that used cleaner fuels (Cundale et al., 2017;Stanistreet et al., 2014). Furthermore, the fuel used for cooking was often also required to meet multiple household energy needs, for example, seasonal space heating (e.g. Benka-Coker et al., 2018;Bielecki and Wingenbach, 2014;Hollada et al., 2017;Rehfuess et al., 2014), heating water for bathing  and lighting the home (Diaz et al., 2008). However, the multiple services provided by the cooking fuel were not always desired, for example the use of solid fuels for cooking indoors during periods of warmer weather was highlighted as an issue because of overheating (Akolgo et al., 2018;Wang and Bailis, 2015). Consequently, it was relatively common for households to rely on multiple fuels or stoves for cooking (e.g. Benka-Coker et al., 2018;Cundale et al., 2017;Pailman et al., 2018), with decisions about which to use dependent upon several factors. The type of food being cooked was frequently a determining factor for which fuel or stove to use (Hollada et al., 2017;Rhodes et al., 2014;Stanistreet et al., 2014). Cooking capacity was also an important factor which determined the stove/fuel type used for cooking particular meals in terms of the number of people that were being fed (Bielecki and Wingenbach, 2014;Rhodes et al., 2014), the size of cooking pot that a particular stove could heat (Diaz et al., 2008;Hooper et al., 2018;Taylor et al., 2011), the variety of household cooking pots that could be used with that method (Bhogle, 2003) and the number of cooking pots that could be used simultaneously (Khushk et al., 2005;Rehfuess et al., 2014;Rouse, 2002). The weather and season also played a role in determining the fuel source or stove type relied upon, with wet conditions reducing reliance on solid fuels for cooking due to difficulties in keeping fuel dry before use (Troncoso et al., 2007) and difficulties lighting (and maintaining) solid fuel-based fires during adverse weather conditions (Hollada et al., 2017;Pailman et al., 2018;Rhodes et al., 2014;Tamire et al., 2018).

Environment
Only two papers were identified under the domain of environment (in this case referring to the wider, outdoor environment) in the review process (Akolgo et al. 2018;Muindi et al. 2014), and so it is difficult to be clear whether people are unconcerned about the environmental aspects of solid fuel collection and use, or simply it is not a current research focus. Deforestation, for example, is a substantial environmental impact of solid fuel harvesting, however its lack of occurrence as such in the scoped literature indicates that it is not something that participants perceive as an environmental issue. While smoke inside the home was frequently identified as a problem associated with solid fuel use (Akolgo et al., 2018;Devakumar et al., 2018;Troncoso et al., 2007), the link between HAP and outdoor air pollution was rarely considered by participants in the reviewed literature. Where outdoor air pollution was considered, this was generally regarded as a normal part of life and participants were resigned to it (Muindi et al., 2014). This may be linked to a lack of knowledge of the wider environmental impacts caused by traditional cooking methods (Akolgo et al., 2018) and the interactions between indoor and outdoor air.

Practice and policy development
A key theme to emerge from the review surrounding practice and policy development was microeconomic constraints as a barrier in the prevention of changes in behavior away from solid fuels (Gordon et al., 2007). The high cost of cleaner fuels or technologies (in this case ICS), was well documented (Devakumar et al. 2018;Gordon et al., 2007;Hollada et al., 2017;Jeuland et al., 2015;Stanistreet et al., 2014). Devakumar et al. (2018) report of participants citing insufficient money to deal with the problem of HAP, which has been linked to the high-cost of cleaner fuels, such as LPG (Hollada et al., 2017), as well as high-costs in the production of ICS (Gordon et al., 2007). However, participants did not always view high cost as a preventative factor and found ways around it. For example, one article found brokering agreements between village members, local government and NGOs allowed people to take out loans or pay in instalments, which supported increased uptake of ICS .
However, incentives to aid the adoption of cleaner fuels and technologies to combat air pollution were not always well received. Hollada et al. (2017) report of participants not using vouchers to lower the cost of LPG issued by the government because it was felt this was not enough, owing to the large investment to purchase the LPG tank. Similarly, although some ICS have been subsidized, these are still seen as largely inaccessible to the poorest families (Gordon et al., 2007). Incentives are not always capital in nature, for example the Maharashtra State Government offer "Clean Village" awards . However, as participants note, adoption of cleaner cookstoves as a result of such incentives are "not because of the smoke", but rather to win the award .
Government involvement was also not always well-regarded (Asante et al., 2018;Muindi et al., 2014). Participants reported experiences of bias in the selection of beneficiary districts for rural LPG programs, with those affiliated with certain political parties more likely to receive a cookstove (Asante et al., 2018). Participants have also reported a lack of voice to petition leaders to address issues surrounding air pollution (Muindi et al., 2014). Moreover, those who have complained in the past have experienced being threatened with eviction and forced to keep quiet (Muindi et al., 2014).
This scoping review highlighted that challenges linked to the introduction of new technologies to reduce exposure to HAP are manifold. Awareness-raising campaigns including stove demonstrations in the home and at markets, at community meetings, and through door-to-door promotion have been attempted by health promoters , and yet despite this, lack of education and information is still being blamed as a factor in preventing dealing with the problem of HAP by study participants (Devakumar et al. 2018;Gordon et al., 2007).

Discussion
A household's decision on which fuel (or combination of fuels) to use to meet their needs is complex and multi-dimensional, incorporating several 'near' and 'visible' factors (e.g. affordability, availability and convenience) and more 'hidden' factors (e.g. cultural traditions and gender roles). This is further complicated by the variety of different services which fuel(s) in the home are required to provide, i.e. cooking, heating and lighting (Sovacool, 2011). The aim of this scoping review was to explore and summarize knowledge of solid fuel users' perceptions of solid fuel use and collection for cooking, heating and lighting in LMICs. Beyond illustrating how perceptions within each of the six key domains (health; family and community life; home, space, place and roles; cooking and cultural practices; environment and practice and policy development; Fig. 1) play an important role in decision-making regarding fuel, two domain-overarching topics were also observed: the role of senses and disconnects. With respect to the role of senses it was clear how these, whether this be sight, smell or taste, were important in shaping perceptions of solid fuel use. Due to the mostly invisible nature of air pollution, people's sensory faculty is vital for giving air pollution a physical and viable parameter, and it is often this physical experience with air pollution that shapes public understanding and risk perception (Bickerstaff, 2004). The theme of "disconnects" was observed in a variety of ways such as a disconnect between solid fuel usage and perceived health impact, temporal disconnect, gender-related disconnect and disconnect between the different relevant stakeholders. Each of these will be explored in more detail below.
Though in most studies participants linked solid fuel use with adverse health impacts manifested most commonly as respiratory irritation (e.g. Chandna and Honney, 2017;Tamire et al. 2018;Thompson et al. 2007), for some there was a disconnect between solid fuel use and health impacts, with either no link made between the two (Akintan et al., 2018;Cundale et al. 2017;Hollada et al., 2017) or smoke perceived to have positive health impacts, e.g. protection against insects . Even where a link was made, exposure to HAP was sometimes considered an unavoidable part of daily life (Muindi et al., 2014;Rhodes et al., 2014), and participants rarely considered the connection between HAP and ambient air pollution. There was also frequently a gap between knowledge of health impacts of solid fuel use and household solid fuel practices. For example, understanding of health impacts of HAP contrasted with the lack of action to reduce HAP through strategies including ventilation of the home. This reflects the complexity of the issue and the wider factors that control household behaviors, i.e. in this instance prioritizing household safety and security (Muindi et al., 2014;van Gemert et al., 2013;Weaver et al., 2017), and the sense of not only belonging in the community, but the esteem that having access to more modern, cleaner cooking alternatives brought (Jürisoo et al., 2018;Kenrick et al. 2010). Importantly, these findings highlight that consideration to apply theoretical frameworks on (health) behavior change could be of relevance to further understand how to develop sustainable interventions to reduce HAP (Morrison and Bennett, 2012). These theoretical frameworks, such as the Theory of Planned Behavior (Ajzen, 1985) and the Health Belief Model (Champion and Skinner, 2008), postulate that knowledge is not the only factor determining behavior change and highlight how changing behavior is influenced by a range of factors. Indeed, our findings align with these postulations and highlight how beyond knowledge on HAP, attitudes towards the new behavior (e.g. favorable evaluation of a solid fuel alternative), subjective norm (i.e. pressure from the community to a cleaner fuel), behavioral control (i.e. the perception of sufficient resources, opportunities and skills to use alternatives to solid fuels) as well as perceived susceptibility and severity (e.g. individual perception of how susceptible they are to and the severity of the health impacts of HAP). While attempts to apply these behavioral change frameworks to the context of cooking are limited, the recently developed CI-CHANGE framework (Kar and Zerriffi, 2018), a cookstove specific behavior change framework which integrates the Theory of Planned Behavior and the Transtheoretical Model, is a promising development and warrants further investigation.
A temporal disconnect in relation to household solid fuel use manifested itself in a number of ways. Fuel collection and/or purchase happens regularly within a household, with decisions being made about which fuels to use depending on the energy services required, the season, the food being cooked, the number of people in the household, among other factors (e.g. Asante et al., 2018;Stanistreet et al., 2014;van Gemert et al., 2013). These short-term decisions contrast with the longer-term changes that affect energy use in the home, including shifts in cultural traditions, beliefs and practices and policy changes (e.g. Malakar et al., 2018;Rhodes et al., 2014;Tigabu, 2017). However, the taste and smell of food cooked in a traditional and culturally appropriate way arguably has no temporal variation, with traditional living valued and preferred by many (Akolgo et al. 2018;Hollada et al. 2017;Rhodes et al., 2014;Taylor et al. 2011;van Gemert et al. 2013). A secondary temporal disconnect was found in terms of health, with many of the health impacts relating to the daily use of solid fuels (e.g. chronic back pain from fuel collection, COPD and lung cancer) not manifesting until later in life (Grigsby et al., 2016;Morgan et al., 2019). Yet, people more frequently referred to the near-term health impacts of solid fuel use (e.g. coughing and wheezing) (Devakumar et al., 2018;Gordon et al., 2007;Miah et al., 2009) than long-term health impacts.
Within the household, socio-demographic factors were important in shaping perceptions of solid fuel use (Kar and Zerriffi, 2018). There was a frequent disconnect between those that were most affected by using solid fuels in the household (i.e. women and children; e.g. Devakumar et al. 2018;Okello et al., 2018;Schlag & Zuzarte, 2008) and those that were making decisions about which fuels to use in the home and budgets (i.e. males; e.g. Rehfuess et al., 2014;Thompson et al., 2018). Age was also an important factor when considering perceptions of solid fuel use, with frequent generational differences in knowledge, attitudes and beliefs (e.g. Bielecki and Wingenbach, 2014). For example, younger people were more likely to be accepting of trying new practices and did not attach the same value to traditions as older generations (Wang and Bailis, 2015). Although children were frequently considered to be most at risk from HAP and burns (Hooper et al., 2018;Pailman et al., 2018), limited attention is given to their role in, and their perception of, cooking practices. Consequently, in designing interventions to reduce HAP it might be of importance to have a clear and specific target audience, with respect to gender and generation or to include flexibility in the approach to account for gender and generational differences in perceptions and responsibilities within the household.
The final level of disconnects was between the different stakeholder groups outlined in our adapted social ecological model (Fig. 1). In particular, community norms and cultural traditions were strong influences on household uptake of new energy sources (e.g. Akintan et al., 2018;Wang and Bailis, 2015). Key community members, e.g. community leaders, teachers and ministers, were also repeatedly highlighted as guardians of cultural norms and important influencers (Jürisoo et al., 2018;Stanistreet et al., 2014). Consequently, energy solutions need to fit the local community context and ideally be community-led (Chandna and Honney, 2017;Tamire et al., 2018;Tigabu, 2017). There were also disconnects between the outer ring of the adapted socialecological model (Practice and policy development; Fig. 1), and households and communities, with policies and interventions designed to reduce reliance on solid fuel use (e.g. subsidies) often not fit for purpose, for example subsidizing the cost of the fuel, but not subsidizing the one-off cost of a new stove (Hollada et al., 2017). Such A. McCarron, et al. Environment International 143 (2020) 105991 high-level consideration of availability and access to alternative fuel sources or cooking technologies may be too distant from the individuals who rely on solid fuels, again, highlighting the need for policies and interventions to be community-led, with a strong understanding of the community's priorities, perceptions and preferences. A full understanding of community priorities will help policymakers and researchers adapt the key messages they are providing to increase uptake of new technologies and/or fuels and ultimately improve health. Lastly, the findings highlighted that there is limited research that focuses on people's perceptions of the societal domains within our social ecological model (Fig. 1), both in terms of the environment (e.g. deforestation and the relationship between HAP and ambient air pollution) and policy and practice (e.g. roles of stakeholders, legislation). These are areas where further research is warranted.

Conclusions and recommendations
We have conducted a scoping review to explore solid fuel users' perceptions of solid fuel use in LMICs. This has highlighted that most previous work exploring peoples' perceptions of solid fuel use for cooking, heating and lighting in LMICs has been undertaken in relation to ICS interventions and not on solid fuel use for open fire or basic cookstoves. Further research that is grounded in and starts at the community-level, which identifies the priorities and perceptions of solid fuel users, and which aims to better understand the lived experience and knowledge of solid fuel use in a variety of contexts, the complexities of gendered household decision-making, and food-related cultures and traditions is therefore necessary.
In undertaking these ICS intervention studies, common methodologies to explore perceptions have included questionnaires, interviews and focus groups. While these have clear value in exploring perceptions of solid fuel use, these methods do not necessarily reflect participant's lived experiences which can be vital to uncover the more 'hidden' factors such as culture and tradition. Our scoping review has highlighted the importance of the senses in shaping people's perceptions of solid fuel use, whether that be smell, taste or vision. There is therefore potential here to innovate in terms of the methods being used to explore people's perceptions of solid fuel use and to bring the participant and their solid fuel experiences closer together. We know from ongoing work by some of the authors in projects in Malawi and Kenya that a range of participatory methodologies (mapping, body mapping, walking interviews, video diaries, photovoice) and more creative forms of community engagement -created and practiced routinely in regions such as Latin America and Africa-such as theatre for development, storytelling and drawing, may not only fill this methodological gap, but also capture more accurately people's lived experiences of solid fuel use in the home.
The findings show that every day in LMICs solid fuel users are making household decisions that affect their HAP-related health, e.g. choosing to ventilate their homes or not. This combined with the fact that participants consistently focused on the short-term (e.g. coughing) rather than long-term (e.g. lung cancer) adverse health consequences of exposure to HAP highlights that more awareness-raising interventions that provide participants with the evidence base needed to make betterinformed decisions that affect the health and wellbeing of the household are required. While this alone may not change household use of solid fuels, because of other more pressing issues such as cost and availability, it will provide people with the full evidence base to feed into decision-making and enable people to advocate for change more widely.
This review offers, for the first time, an amalgamation of LMIC solid fuel users' perceptions, focused specifically on solid fuel use and collection. The review has demonstrated the richness and potential of exploring perceptions, and yet perception of solid fuel use has not well explored. This review highlights the importance of understanding and considering perceptions of solid fuels from a social ecological perspective (i.e., in terms of health, family and community life, home, space, place and roles, cooking and cultural practices, environment and practice and policy development) in order to reduce exposure to HAP, be this through (infra)structural or behavioral interventions. Of particular importance, future interventions need to take into account complex gendered and community dynamics to make 'connections' that are currently lacking in order to overcome the various disconnects identified throughout this review. It is anticipated that such a participatory approach from a social ecological perspective will ensure that any intervention or alternative fuel source is available and affordable to the poorest communities, as well as seen as acceptable in the local and cultural context.

Funding
This work was supported by internal funding from the University of Stirling Crucible program which went towards the development of the scoping review.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.