Emergency risk communication and sensemaking during smoke events: A survey of practitioners

Emergency risk communication (ERC) for smoke emitted from major fires continues to challenge governments. During these events, practitioners (including scientific, communication, and emergency response government staff) are tasked with quickly making sense of the public health risks and the communication options available. Practitioners’ sensemaking—the process of creating meaning from information about an unfolding emergency—is key to effective ERC. This article identifies the factors that ERC practitioners consider the most important to their sensemaking for smoke events. A survey of practitioners (n = 86) was conducted to elicit their views on the level of importance of 22 different factors (individual, organizational, and contextual) on their sensemaking. The results indicate that the majority of the factors tested are very important to practitioners. This finding likely reflects the multidimensional nature of emergency smoke events and provides evidence as to why practitioners are challenged when trying to make sense of emergency situations. Despite multiple factors being considered very important to practitioners, the time‐limited nature of emergencies means that practitioners will inevitability be forced to prioritize in their sensemaking efforts. Our results also provide insight into practitioners’ prioritization of different information sources. Specifically, practitioners prioritize their own knowledge and the knowledge of other practitioners. The two most important factors were information from other incident management stakeholders and the practitioners’ past experience. Other information, including community‐based and academic knowledge, appear to be of lower priority for practitioners. Based on the study results, recommendations for practice and future research are discussed.


INTRODUCTION
Smoke emitted from fires, such as bushfires 1 and waste facility fires, presents a significant and increasing threat to human health (Bowman et al., 2017;Reisen et al., 2015). Smoke events require urgent action from governments to protect public health and emergency risk communication (ERC) is an essential component of the protection strategies (World Health Organization [WHO], 2018). During smoke events, ment practitioners, including scientific and communication advisors and emergency managers. Emergency smoke events are complex situations characterized by high levels of uncertainty and an urgency to act. These events require practitioners to urgently make sense of the situation, including the risk to public health and the communication options that are available and required to protect the public (Thomas et al., 2021). Sensemaking-described as an iterative process of creating meaning (Weick, 1988;Weick et al., 2005)-has been established as a key process underpinning the practice of ERC (Boholm et al., 2012;Curnin & Owen, 2013;Taarup-Esbensen, 2019;Thomas et al., 2021).
The observed breakdown of sensemaking during emergencies (Mills & Weatherbee, 2006;Weick, 1993Weick, , 2010) may help explain poor ERC outcomes. As practitioners attempt to make sense of the situation at hand, they are simultaneously required to communicate with the affected public. Due to the time-limited nature of emergencies, practitioners will inevitably be forced to prioritize in their sensemaking efforts. Practitioners' prioritization of the information they perceive and interpret, and the actions they take are pivotal to the implementation of ERC. Practitioners' sensemaking, and the factors that practitioners consider the most important to their sensemaking, therefore, play a critical role in determining the effectiveness of ERC during smoke events.
There is a vital need to understand the most important factors to practitioners' sensemaking in this setting. For example, there is a need to understand what information and personal attributes practitioners rely on when sensemaking, and the contextual factors that influence their sensemaking. The existing academic literature has begun to highlight numerous factors that can potentially influence practitioners' ERCrelated sensemaking (e.g., Boholm, 2019aBoholm, , 2019bBoholm et al., 2012;Taarup-Esbensen, 2019). These factors are identified in Section 2.1 and relate to the practitioners themselves (individual factors), the institutional/organizational and collaborative nature of this work (institutional/organizational factors), and the specific public health event (contextual factors). However, to date, the available academic research that examines sensemaking in risk communication is primarily qualitative in nature and considers a small number of practitioners. As a result, there is a need to confirm the importance of the factors identified in these studies and to test their generalizability for a broader group of practitioners.
The aim of the current research is to understand the importance of different individual, institutional/organizational and contextual factors for ERC practitioners' sensemaking. It also investigates whether the level of importance placed on the different factors is influenced by practitioners' characteristics, including their professional area of expertise or their level of past experiences working in the field. To this end, an online survey was undertaken to elicit the views of practitioners working for Australian government organizations in roles that involve ERC for smoke events. Understanding the level of importance of the various factors can aid academic researchers and government organizations in develop-ing interventions to better support ERC practitioners in their sensemaking endeavors.

EMERGENCY RISK COMMUNICATION AND SENSEMAKING PRACTICE FOR SMOKE EVENTS
ERC is described as the "intervention performed … to enable everyone at risk to take informed decisions to protect themselves, their families and communities against threats to their survival, health and well-being." (WHO, 2018, p. vii) The protection of public health during environmental health incidents is primarily the responsibility of governments (Glik, 2007;WHO, 2018). This study examines ERC practitioners from government organizations because they are most commonly the primary issuers of emergency and warning messages. For these practitioners, smoke events (e.g., bushfires, structural fires, or industrial/waste facility fires) are a challenge as they represent an unanticipated and visible public health threat that attracts attention and comment from citizens and news media (Greven et al., 2018;Parliament of Victoria, 2019;Teague et al., 2014).
Sensemaking is an established concept in the field of emergency response (Maitlis & Sonenshein, 2010;Weick, 1988;Weick & Sutcliffe, 2015). It is an iterative and adaptive process that accommodates the inherent complexity and uncertainty of emergency situations (Weick, 1988;Weick et al., 2005). Sensemaking is triggered by disruptive events (such as a smoke event), and involves perceiving cues, interpreting cues, and taking action (Weick, 1988;Weick et al., 2005). The iterative nature of sensemaking allows practitioners to process evolving multidimensional situations and shifting priorities over time, which informs their perception of the risk as each new situation is presented (Slovic, 2015). Weick (1988) proposed that in a crisis practitioners' sensemaking is influenced by three aspects: capacity, commitment, and expectation. A practitioner's capacity to recognize a situation will increase for situation where they feel capable to act, their commitment to a view or action can influence the level of attention they give to certain cues and the meaning they create will be influenced by their expectation of the situation. As discussed in Section 2.1, since Weick's (1988) formative research, numerous studies have expanded the understanding of the variety of factors influencing sensemaking.
To study sensemaking practice for ERC, the research draws on the theoretical understanding of the practice of risk communication, including Chess et al. (1991) and Boholm (2019aBoholm ( , 2019b, and the body of literature exploring sensemaking as a key process of risk-related practice, including Boholm et al. (2012), Taarup-Esbensen (2019), and Thomas et al. (2021). The study has been designed considering Rickard's (2019) two functions of risk communication, and primarily adopts the pragmatic function, while also considering the constitutive function.

Factors important to ERC practitioners' sensemaking
The existing academic literature identifies numerous factors as potentially important to ERC practitioners' sensemaking. Several identified factors relate to the practitioners themselves (referred to as individual factors). Emergency response teams comprise various specialized individuals coming together to make sense of the situation (Bigley & Roberts, 2001;Maitlis & Christianson, 2014). The existing literature highlights that the individual practitioners' knowledge base is an important factor influencing their sensemaking (Boholm, 2019b;Boholm et al., 2012). Practitioners draw on their academic and professional development training and their past lived experiences of smoke events or other public health events (Boholm, 2019a(Boholm, , 2019bBoholm et al., 2012;Taarup-Esbensen, 2019;Thomas et al., 2021). Our recent research studying practitioners' sensemaking for smoke events in Victoria, Australia showed that the lived experience of working in these settings can be stressful and exhausting (Thomas et al., 2021). Practitioners are required to make sense of the situation in tight timeframes and deal with high levels of uncertainty (Weick, 1988). When trying to make sense of a situation, practitioners can face competing demands, such as having to address several questions at once from different teams within the organization or from different stakeholders (Olsen et al., 2014).
Several identified factors relate to the organizational and collaborative setting of this work (referred to as institutional/organizational factors). Government practitioners sit within government organizations, meaning that sensemaking is undertaken within this institutional/organizational setting. Institutional mechanisms, including plans and procedures, can both guide and focus practitioners' sensemaking (Bigley & Roberts, 2001;Boholm et al., 2012;Maitlis & Christianson, 2014;Savoia et al., 2017). For emergency smoke events, this can include emergency management plans and plans specific to communicating during smoke events. Limited and inadequate human resources place pressure on the practitioners who are working (Olsen et al., 2014;Oltra & Sala, 2015). Collaboration of practitioners (both within teams and across organizations) enables collective sensemaking and shared meaning about an incident (Boholm et al., 2012;Taarup-Esbensen, 2019). During these collective meaning creation and diffusion events, practitioners draw on information from their colleagues and share their own understandings (Boholm et al., 2012;Taarup-Esbensen, 2019;Weick et al., 2005). Shared past lived experiences help reduce confusion and establish a shared understanding about the incident (Boholm et al., 2012). Personal relationships are also developed through working together on previous events, and these relationships aid collective sensemaking (Thomas et al., 2021). Institutional mechanisms (such as plans and procedures) can also support collaboration (Savoia et al., 2017).
Several identified factors relate to the context of the sensemaking, which in this case is an emergency smoke event (referred to as event-context factors). Sensemaking is situational, in that it is influenced by the circumstances surrounding it (Boholm et al., 2012;Taarup-Esbensen, 2019). For emergency smoke events, practitioners' sensemaking occurs within the context of emergency response and focuses on the specific smoke event at hand. Practitioners draw on various cues and information to make sense of the event (Curnin & Owen, 2013). For example, practitioners consider environmental monitoring data and scientific literature (Boholm, 2019b;Boholm et al., 2012). Practitioners also draw on information from the community (Savoia et al., 2017;Seeger et al., 2018), information in the news media and information from other incident management stakeholders about the event (Bigley & Roberts, 2001;Olsen et al., 2014;Savoia et al., 2017). To target communication and assess what available communication options are required, practitioners also draw on information about the affected communities (Savoia et al., 2017).
As shown, the extant literature highlights numerous factors as potentially important for ERC-related sensemaking. Figure 1 presents the factors identified for each of the three categories identified: individual, institutional/organizational, and event-context factors. In considering the complexity of emergencies and the limited time practitioners have to make sense of smoke events, a key question is: which factors do practitioners prioritize when sensemaking? Our recent study highlights the key role of past lived experiences in facilitating sensemaking (Thomas et al., 2021). However, these findings are based on one study of 15 practitioners, and as such, the generalizability of these findings is unknown.
For effective ERC outcomes, academic literature recommends risk communication practitioners embrace two-way communication mechanisms to engage with affected communities and establish an understanding of their communication needs (e.g., Loroño-Leturiondo et al., 2018;Savoia et al., 2017). If practitioners are prioritizing recommended practices for ERC, then we would expect information from and about the community to be two factors that are prioritized. However, the frequent criticisms of risk communication suggests it is currently not meeting community needs (Parliament of Victoria, 2019; Teague et al., 2014), and potentially indicates information from and about the community is currently not prioritized by practitioners. Understanding the factors practitioners are prioritizing may help us understand why ERC is met with continued criticism from the public. Therefore, the study addresses the primary research questions of: • RQ1: What factors do emergency risk communication practitioners view as the most important for their sensemaking during smoke events?

Practitioners' past experience
Past lived experiences are identified as a key facilitator of sensemaking (Thomas et al., 2021). We propose that the perceived importance of past experience is likely F I G U R E 1 Factors potentially important to ERC practitioners' sensemaking (presented in alphabetical order within the categories) related to the degree of experiences a practitioner has accumulated. Therefore, this study investigates the following questions: • RQ2. How does a practitioners' level of past experience influence the perceived importance they place on: a. their individual past lived experiences? b. their shared past lived experiences?

Practitioners' professional areas of expertise
Many practitioners are involved in ERC for smoke events and these practitioners have varying professional areas of expertise (Thomas et al., 2021). The practitioners involved have different responsibilities during smoke events based on their professional areas of expertise and related role in emergency response (Emergency Management Victoria, 2019). For example, environmental scientists (such as air quality scientists) advise on the conditions and behavior of the smoke plume; environmental health practitioners' (such as public health practitioners, toxicologist, and human health risk assessors) advise on the risk posed to public health from the smoke; communication and community engagement practitioners advise on how best to communicate with the affected community; and emergency and fire response practitioners advise on the emergency management.
It is considered likely that the most important factors for practitioners will differ depending on their professional areas of expertise. For example, environmental science practitioners may perceive environmental monitoring data as the most important factor; environmental health practitioners may perceive scientific literature as the most important factor; communication and engagement practitioners may perceive information about the affected community as the most important factor; and emergency and fire response practitioners may perceive emergency management procedures as the most important factor. Therefore, this study also investigates the relationship between practitioners' professional areas of expertise and the importance they place on the different factors: • RQ3. How does a practitioners' professional area of expertise influence the perceived importance they place on the different factors influencing their sensemaking?

METHOD
A questionnaire was designed to elicit practitioners' views on their perceived importance of the factors identified as potentially influencing their sensemaking, and an online survey was used for data collection. The survey was conducted from May 03 to June 21, 2021 using the Qualtrics 2 online platform. The sample comprises 86 practitioners with experience working in six states, two territories, and the federal level of government across Australia. The study was approved by the Monash University Human Research Ethics Committee (Project ID: 28113).

Target population
The target population was practitioners who are (or may have been) required to assist with emergency risk communication during a smoke event on behalf of governments in Australia. Practitioners self-identified as meeting the target population inclusion criteria.

Questionnaire design
The questionnaire is provided in Supporting Information Appendix A. The initial two questions required respondents to confirm they met the target population inclusion criteria.
Respondents were then provided with background information on the scope of the survey, including a description of sensemaking. The factors identified in Section 2 formed the basis for the questionnaire design (also see Table 2). For each of the 22 identified factors, participants were asked to rate how important this factor was to them when making sense of a smoke event. Response options were on a 5-point Likert scale; where 1 = not at all important and 5 = extremely important. The factors were presented in blocks, with one block for each of the three categories-individual, institutional/organizational, and event-context factors. To remove potential selection bias associated with the display order, the display order of the blocks and the factors within each block were randomized through Qualtrics. Demographic data were collected, including gender, age, and education. In addition, data were collected about respondents' relevant careers, including the location of relevant experience in Australia, primary professional area of expertise, years of experience in their primary professional area of expertise, and experience working during smoke events. Practitioners were asked to estimate the number of smoke events that they had been involved with. Smoke events were described in the survey as incidents where people are potentially exposed to smoke from a fire (e.g., bushfires, structural fires, or industrial/waste facility fires). Respondents were then asked to estimate how many of these smoke events they considered to have been crisis incidents. Crisis incidents were not defined in the survey, as the primary objective was to capture the respondents' subjective experiences and how many crisis events they perceived to have been involved in.

Recruitment
The survey was advertised using publicly accessible email addresses of the public health, environmental protection, and emergency management agencies in each state or territory and the federal level of government in Australia. Recruitment was supported by several professional bodies, including the Australian College of Toxicology and Risk Assessment, Australasian Environmental Law Enforcement and Regulators neTwork, Environmental Health Australia, and Clean Air Society of Australia and New Zealand. Recruitment support included advertising the survey in their monthly newsletter, emails to members, and information posted to online membership notice boards.

Survey respondents
The survey was accessed by 129 people: 86 participants completed the survey (67% of those who accessed the survey); 32 participants (25%) did not complete the survey due to not meeting the target population inclusion criteria; and 11 participants (9%) met the criteria but exited the survey before completing it. Table 1 summarizes the respondents' characteristics. Respondents reported being well educated, with 90% holding a postgraduate degree, master's degree, graduate diploma/certificate, or bachelor's degree. Publicly available demographic information for the target population (i.e., government ERC practitioners in Australia) is not available, however, this level of education is higher than the Australian population more broadly, with 62% of Australians (aged 15-74 years) having a certificate, diploma, or degree (Australian Bureau of Statistics, 2021). While the sample population comprises practitioners from across Australia, it is noted that the majority of respondents reported working in the states of Victoria (31%) and Western Australia (29%), and South Australia (3%) and Tasmania (3%) appear underrepresented. However, it is considered unlikely that this would have significantly influenced the results because smoke events are a common hazard across Australia and these events are managed in a similar way in each of the states and territories.

Data analysis
Descriptive statistics are used to identify the factors most important for sensemaking and address RQ1. Due to the ordinal nature of the data (Likert scale), frequencies (i.e., percentages of responses in each response level) and median values are presented. To rank factors on their level of importance to practitioners, a frequency percentage combining very important and extremely important was adopted. Inferential statistics were undertaken to investigate RQ2a, RQ2b, and RQ3, to evaluate differences in the level of importance placed on certain factors based on respondent characteristics. Statistical analysis was undertaken using SPSS version 27. Shapiro-Wilk tests were performed to check the data's normality. These tests indicated generally nonnormal data, and therefore, nonparametric tests were performed. A series of Kruskal-Wallis H omnibus tests were conducted. This H test is a one-way analysis of variance test for ranks. Distributions of the level of importance scores were assessed by visual inspection of a boxplot for each of the tests. Generally, distributions were not similar and, therefore, mean rank values were adopted for analysis. Where statistically significant results were identified in the omnibus test, post hoc pairwise comparisons were undertaken, using Dunn's (1964) proce-

Important factors for sensemaking (RQ1)
The first research question seeks to understand the factors ERC practitioners consider the most important for their sensemaking during smoke events. Table 2 presents the percentage of respondents that selected each of the importance levels and the median importance score for each factor (for the sample population as a whole, N = 86). The median importance score results show the majority of the factors (17 out of 22 factors) are considered to be very important to respondents. The remaining five factors are considered moderately important. Table 2 also ranks the factors based on the combined percentage of respondents indicating that a factor was very or extremely important. From this, several factors are identified as being more or less important than other factors.
Information from other incident management stakeholders ranked the highest, with 91% of respondents indicating this factor was very or extremely important to their sensemaking. The second highest ranking factor is practitioners' individual past experiences of working during smoke events, with 87% of respondents indicating this factor was very or extremely important. Practitioners consider having past personal experiences of working during smoke events as more important than having personal experiences working during other public health events more generally. Shared lived experiences of working together during smoke events are also viewed as important to practitioners, but were not ranked as high as having individual experiences. Information in the news media about the incident is the lowest ranking factor, with only 33% of respondents indicating this factor was very or extremely important. This suggests practitioners see less value in this source of information compared to other information they can gather about a smoke incident. The second lowest related factor is scientific literature, with 34% of respondents indicating very or extremely important. Linked to scientific literature, an individual's academic training also ranked low (35% of respondents indicating very or extremely important). As discussed in Section 4.2.3, the importance placed on scientific literature and academic training varied depending on practitioners' professional area of expertise.
The two factors relating to the community, being information about and information from the affected community, both received a median value that indicates these factors are very important. However, when considering the ranking of the factors, information about the affected community ranked 12th (69% of respondents indicating very or extremely important) and information from the affected community ranked 14th (64% of respondents indicating very or extremely important) out of 22 factors. This potentially suggested information from and information about affected communities is not being prioritized by practitioners. However, as discussed in Section 4.2.3, the importance of information from the community varied depending on practitioners' professional area of expertise.

Influence of practitioners' past experience (RQ2a and RQ2b)
Research questions RQ2a and RQ2b investigate the relationship between the importance respondents place on their individual or shared past experiences (working during smoke events or other public health events), and the actual degree of past experiences the respondents have. Actual past experience is measured by the respondents' experience working during past smoke events (see Section 4.2.1) and the respondents' years of professional experience (see Section 4.2.2). Two types of smoke events were tested, with respondents asked to estimate the number of smoke events that they had been involved in and to estimate the number of these events they perceived to be crisis incidents. Table 3 presents the Kruskal-Wallis H tests performed to assess differences in importance rankings based on practitioners' past experiences, and where statistically significant results were identified Table 4 presents the post hoc pairwise comparisons.

4.2.1
Past experience working during smoke events The importance of individual past experience (working during smoke events or other public health events) is not influenced by practitioners' actual levels of past experience working during smoke events or crisis smoke events. This result indicates that individual past experiences are considered important for sensemaking by all practitioners, irrespective of the actual level of past experiences they have had.
The importance of shared past experiences (working during smoke events or other past public health events) is influenced by practitioners' actual level of past experiences working during smoke events but is not influenced by their actual past experience working during crisis smoke events. For the importance placed on shared past experience working during smoke events (Test: T1.3, Table 4), the results indicate that practitioners who have worked during 31+ events perceive greater importance than practitioners who have worked during 3-30 past smoke events. For the importance placed on past shared experience working during other past public health events (Test: T1.4, Table 4), the results indicate that practitioners with experience working during 0-2 events perceive greater importance in shared experiences during other past public health events than practitioners with experience working during 3-30 events. A pattern appears to exist where shared experiences of past smoke events and other public health events are viewed as more important by practitioners with limited actual experiences (<3 events) and then again after experiencing many fires (31+ events), with the perceived importance decreasing for practitioners who have experience working during 3-30 events.

Years of professional experience
The results show that the perceived importance of individual or shared past experiences (working during smoke events or other past public health events) is not influenced by practitioners' years of professional experience.

Influence of professional area of expertise (RQ3)
Research question RQ3 investigates the relationship between the importance respondents place on each of the factors and the respondents' professional area of expertise. Table 5 present the Kruskal-Wallis H tests performed to assess differences in importance rankings based on practitioners' professional area of expertise, and where statistically significant results were identified Table 6 presents the post hoc pairwise comparisons. The results show the perceived importance of several of the factors is influenced by practitioners' professional area of expertise. The Kruskal-Wallis tests revealed statistically significant differences in the mean ranks of the professional area of expertise groups for the following factors: environmental monitoring data (T4.5), emergency response plans (T4.8), information about the affected community (T4.12), academic training (T4.20), and scientific literature (T4.21). However, for emergency response plans, despite a statistically significant result in the Kruskal-Wallis H test, the post hoc analysis revealed no statistically significant pairwise comparisons between any of the group combination.

Practitioner characteristics a (Independent variables) Level of importance of factors(Dependent variables) Test ID χ 2 df p
Cohen's f The results indicate that environmental health practitioners place greater importance on several factors (including environmental monitoring data, their academic training, and scientific literature) than other practitioner groups. The finding that environmental health practitioners perceive the greatest importance in academic training and scientific literature cor-relates with this group having the highest rate of higher education qualifications. For environmental health practitioners, 74% of respondents reported having a master's or postgraduate degree. This is compared with 57% of environmental science practitioners, 50% of communication and engagement practitioners, and 18% of emergency and fire response practitioners (refer to Supporting Information Appendix B).
The results also show that for information about the affected community, both communication and community engagement practitioners and environmental health practitioners perceive greater importance than emergency and fire response practitioners.

Implications of findings
The primary aim of this study was to identify the factors that practitioners view as the most important for their ERC-related sensemaking during smoke events. The results indicate that the majority of the factors tested are viewed as very important to practitioners. This finding is not surprising when considering the complex and multidimensional nature of emergency smoke events. These events comprise several aspects (i.e., a fire, a smoke plume, potential environmental, and human health impacts, and a need for multiagency collaboration) and the numerous factors that are important to a practitioners' ability to make sense of the situation likely mirrors the numerous dimensions of these events. The fact practitioners are reliant on numerous factors for their sensemaking may also explain why practitioners find sensemaking challenging in this setting. As an emergency event occurs in realtime, practitioners need to quickly make sense of the smoke event, assess appropriate response guidelines and processes, The inherent time-limited nature of emergencies means that practitioners are inevitability forced to prioritize in their sensemaking efforts. Our results provide insight into practitioners' prioritization of different information sources for their ERC-related sensemaking. Specifically, practitioners appear to prioritize the knowledge of practitioners. The two highest ranking factors, information from other incident management stakeholders and a practitioner's past experience, relate to the expert knowledge of other practitioners and the experiential knowledge of the practitioners themselves. Other types of information, such as community-based knowledge and academic knowledge, appear to be of lower prioritization for practitioners. Understanding practitioners' prioritization of different information sources enables us to appreciate the information that is most critical to practitioners when sensemaking and, subsequently, the information sources that should be leveraged to efficiently and effectively support practitioners. In addition, understanding the information sources that practitioners currently consider to be of lesser importance identifies opportunities for further research to aid more effective ERC-related sensemaking.
The most important factor to practitioners for making sense of smoke events is the information they receive from other incident management stakeholders. Information sharing and coordination is a key practice widely recommended for successful ERC (Boholm, 2019a;Savoia et al., 2017). Therefore, our results confirm the key role of collaborative practices in supporting successful ERC-related sensemaking. Effective collaboration between all individuals and organizations involved is key to ensure information is shared and coordinated during an incident.
The second highest ranking factor was practitioners' past lived experiences of smoke events. This finding supports our previous research that identifies the important role of past experiences in facilitating ERC-related sensemaking. The current results indicate that practitioners' past experiences of smoke events are viewed as more important than their past experiences working during other public health events more generally. This shows that practitioners consider it important to gain context-specific knowledge (Yanow, 2004), which is knowledge about the specific public health incident at hand. The current analysis also shows personal past lived experiences are viewed as important by practitioners of any level of experience (i.e., irrelevant of years of professional experience or the number of past experiences working during smoke events or crisis smoke events).
Shared past lived experiences of smoke events are also viewed as important to practitioners but were not ranked as highly as having individual experience. The value of shared experience working during smoke events appears to be influenced by the number of past smoke events practitioners have experienced; that is, shared experiences (of working during past smoke events and other public health events) appear to be viewed as more important by practitioners with limited actual experiences (<3 smoke events) and practitioners who have experienced many fires (31+ smoke events). The reason for this observation is not known, and requires further study to better understand this relationship.
Two-way communication with affected communities is widely recognized as recommended ERC practice (e.g., Loroño-Leturiondo et al., 2018;Savoia et al., 2017). The factors associated with two-way communication (i.e., information about and from the affected community) were viewed as very important to practitioners for their sensemaking. However, in the overall ranking of factors, these two factors were not prioritized in comparison to the other factors tested. This suggests that when forced to prioritize factors, practitioners may not be prioritizing community-based knowledge. The lower prioritization of community-based information might be a reflection of the primary ERC method adopted during smoke events. In the early stages of emergency smoke events, with the aim of reaching a large and diverse audience, ERC generally comprises generic emergency and warning messages (Fish et al., 2017). The use of generic messages potentially means practitioners do not see the need to prioritize community-based information.
For successful ERC, community-based knowledge is critical to enable the tailoring of communication messages and methods of delivery for the audience (Covello, 2003;Savoia et al., 2017). The lower prioritization of community-based knowledge may help to explain why the public has criticized ERC for smoke events in the past. This criticism has included communities feeling they were not listened to, and that the communication is inadequate and does not meet their needs (e.g., Parliament of Victoria, 2019; Teague et al., 2014). To improve ERC, prioritization of community-based knowledge is essential. Our analysis did show that communication and engagement practitioners, as well as environmental health practitioners, awarded greater importance to information about the affected community than other practitioner groups. Therefore, these results highlight a key role for com-munication and engagement practitioners and environmental health practitioners in advocating for information about the community being considered. However, for making sense of the situation and effective communication, there is also a need for practitioners to gather and prioritize information from the affected community.
The results indicate that practitioners see less importance in scientific literature and their academic training when making sense of a smoke event. This result is surprising considering the high level of higher education reported by the respondents. A potential explanation for the limited utility seen in scientific literature is the inaccessibility of this information (such as scientific journal articles) during a fast-paced emergency. However, for an individuals' academic training, this knowledge should be readily available to the practitioner. A gap between the theoretical understanding of risk communication and risk communication activities in practice has been observed in the field and is commonly termed the theory-practice gap (Boholm, 2019b;Chess et al., 1995;Höppner et al., 2012). If practitioners continue to see limited value in academic knowledge, the theory-practice gap is likely to persist. Decades of academic research provides theoretical recommendations for optimal risk communication (e.g., Covello, 2003;Heath & O'Hair, 2009;Lundgren & McMakin, 2018). Supplementing practitioners' knowledge with the available academic knowledge should lead to improved ERC practice and communication outcomes. Therefore, the results highlight a need for academic training and research to be tailored and accessible to better support practitioners' undertaking sensemaking for ERC.
The analysis did show a difference in the perceived importance of academic training and scientific literature between different professional areas of expertise. Environmental health practitioners reported higher importance of academic training and scientific literature. These results correlate with a higher level of education (master's and postdoctoral degrees) for environmental health practitioners. Research has highlighted the core role of scientific and academic knowledge for risk-related professions (Boholm, 2019b;Boholm et al., 2012). Environmental health practitioners are responsible for making sense of the risk to human health, and the results are likely reflective of a high reliance on academic knowledge and data for this assessment. These results suggest that academic training and scientific literature are most useful to sensemaking for specific aspects of a smoke incident (such as human health risk assessment), when compared to the knowledge required to make sense of other aspects (such as emergency response).

Limitations
The limitations associated with the survey methodology are acknowledged. Due to the survey recruitment methods, it was not possible to derive a specific response rate. In addition, survey methodology has inherent limitations associated with self-reporting, including recall bias and response bias, such as respondents being more likely to agree with socially accepted factors (such as the importance of procedures) and tend to undervalue factors relating to emotions. Limitations associated with the need to simplify a complex issue, such as an emergency smoke event, when compiling a list of factors associated with the issue are also acknowledged. While our list of factors is based on a close reading of the available literature and the research team's knowledge in the field, the list is not exhaustive, and other factors may exist. For example, in addition to environmental monitoring data, smoke plume modeling (which is a key source of information for understanding the behavior of the smoke plume) could have been included as a factor. Nevertheless, we believe our list of factors balances an appropriate number of factors to represent the complexity of emergency smoke events, with the importance of presenting these factors in a way that allows a wide range of practitioners to complete the survey in a reasonable time period. Despite the limitations identified, we believe the survey results provide meaningful insights and contribute to the wider understanding of ERC practice.

Recommendations for practice and future research
The value practitioners see in past lived experience emphasizes the critical need to capture and share lessons learned from working during past smoke events with practitioners who are yet to experience working in this setting. One suggestion is that practitioners with no past experience gain experience by shadowing more experienced practitioners during a smoke event. Of the respondents, 37% reported having 21+ years of professional experience, 40% reported having 10-20 years of experience, and only 23% reported less than 10 years of experience (see Table 1). If the sample population is representative of the wider practitioner population, a significant proportion of this cohort may be nearing retirement. This possibly highlights the essential need for succession planning in the field to ensure that the lessons learnt are not forgotten. Operation-based preparedness activities (such as drills and field exercises) can also provide an opportunity to share lessons learned (Skryabina et al., 2017). In addition, these collaborative activities have the added benefit of bringing the two factors identified as most important to practitioners' sensemaking-being individual past experience and the knowledge of other practitioners'-together in a practice setting before an actual event occurs.
The complex nature of emergency smoke events means that practitioners will often face an overwhelming amount of information and an urgency to act. Our study emphasizes the importance of a diverse and adequately resourced team of practitioners, who are able to each prioritize different aspects of a complex smoke event. The key role of environmental health practitioners and communication and engagement practitioners is identified to ensure community-based and academic knowledge receives attention. Further research is recommended to understand why in practice community-based knowledge is currently not being prioritized. Research should aim to understand how the use of information from and about the community can be facilitated in the early stages of an emergency smoke event. Similarly, the limited importance of scientific literature and academic training requires reflection. It is suggested that academic ERC researchers consider how research findings can be made more accessible to practitioners to enable evidence-based sensemaking during emergency smoke events.

A C K N O W L E D G M E N T S
This research was completed as part of a PhD undertaken at Monash University, supported by the Australian Government Research Training Program and the Environment Protection Authority Victoria. We would like to thank the survey participants for their time and the anonymous reviewers for their constructive comments.
Open access publishing facilitated by Monash University, as part of the Wiley -Monash University agreement via the Council of Australian University Librarians.

D ATA AVA I L A B I L I T Y S TAT E M E N T
De-identified survey data is available online in the Supplementary Information section of the online article.