Community preparedness for emergency: a cross-sectional survey of residents in Heilongjiang of China.

OBJECTIVE: This article aims to identify factors that shape the knowledge, attitudes and behaviours of community residents in China's Heilongjiang province towards emergency preparedness. Findings of such a study may provide evidence to support the development of effective public risk communication strategies and education campaigns.
DESIGN: A cross-sectional household questionnaire survey was conducted in Heilongjiang province in 2014. A stratified cluster sampling strategy was employed to select study participants. The questionnaires were administered using face-to-face interviews. 2800 questionnaires were completed, among which 2686 (95.9%) were considered valid for data analyses. A multivariate logistic regression model was adopted to identify the extent to which the independent variables were associated with emergency preparedness.
RESULTS: Fewer than 5% respondents were well prepared for emergency. Over half (52%) of poorly prepared respondents did not know what to do in emergency; women (OR=1.691), higher household income (OR ranging from 1.666 to 2.117), previous experience with emergency (OR=1.552), higher levels of knowledge about emergency (OR=2.192), risk awareness (OR=1.531), self-efficacy (OR=1.796), as well as positive attitudes towards emergency preparedness (OR=2.265) were significant predictors for emergency preparedness. Neither educational attainment nor exposure to awareness-raising entered into the logic regression model as a significant predictor for emergency preparedness.
CONCLUSIONS: The level of emergency preparedness in Heilongjiang residents is very low, which is linked with poor knowledge and attitudes of the residents towards emergency preparedness. Future emergency awareness campaigns should be more focused and tailored to the needs of intended audience, taking into consideration of their usual source of information and knowledge in relation to emergency. Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/

The questions measuring awareness and preparatory behaviours for emergency were intentionally broad, covering four broad types of events, which may have contributed towards unclear or vague responses from some participants. Owing to limited resources, the survey was conducted only in Heilongjiang province. Despite these limitations, this study provides important information for understanding resident preparedness behaviours and evidence to support the development of effective public risk communication and education campaigns.

Introduction

The past 10 years have seen a 14% rise in affected populations and a 39% rise in deaths as a result of disasters, both natural and manmade, according to the World Disasters Report 2012.1 In 2014, 210 million people were affected by disasters across the globe. China is one of a number of countries particularly vulnerable to natural disasters. It was estimated that, during February 2015 alone, 238 000 people in China were affected by natural disasters and more than 2000 people needed emergency assistance. The direct economic loss reached ¥230 million.2

Effective responses to disasters depend not only on the strong leadership and coordinated actions of governmental agencies and professional bodies, but also on the efforts of local residents. Emergency preparedness and early self-help efforts of local residents are usually critical for disaster victim survival.3

Extensive studies have investigated the effectiveness of professional responses to emergency events4; however, there is paucity in the literature documenting individual and community preparedness for emergencies.5 In a 2009 national survey conducted by the US Federal Emergency Management Agency (FEMA), 57% of respondents reported having an emergency kit prepared within the home.6 The Council for Excellence in Government found that one-third (32%) of US residents took no steps to prepare for potential emergencies, with lack of awareness, beliefs and resources (time, knowledge, money) being cited as the main reasons.7 8 The level of individual emergency preparedness in China is even lower. A study of 2000 Beijing residents in 2009 revealed that 2.5% had prepared emergency kits,9 similar to the finding (2.2%) of another study conducted in Shanxi province.10

Low levels of public awareness and failure to prepare present challenges for emergency response management.3 This study aims to identify factors that may shape the knowledge, attitudes and behaviours of community residents in China's Heilongjiang province towards emergency preparedness. Findings of such a study may provide evidence to support the development of effective public risk communication and education campaigns.

Methods

Study population

A cross-sectional household questionnaire survey was conducted in Heilongjiang province between September and October 2014. Heilongjiang is located in northeast China, with a population of over 38 million. In 2014, the Gross domestic product per capita in Heilongjiang reached ¥39 226.4, ranking in the lowest range of all provinces.14

A stratified cluster (in terms of demographic and social economic diversity) sampling strategy was used to select study participants. Five major municipalities in Heilongjiang (Harbin, Qiqihar, Mudanjiang, Jiamusi, Daqing) were involved in the study. One urban district and one rural county of the five municipalities were randomly selected. Two communities/villages within each selected district/county were sampled. All households were identified as the participants for survey and each household identified one adult member to complete the questionnaire. The questionnaires were administered through face-to-face interviews, each of which took on average 20 min to complete. A total of 2800 questionnaires were completed, among which 2686 (95.9%) were valid for data analyses.

Data collection

Trained postgraduate students from the school of public health of Harbin Medical University collected data. Those students spoke the same dialect as those interviewed, were involved in the development of the questionnaire (eg, literature review and identification of relevant theories), conducted a pilot study in a convenience sample (n=110) and a random cluster sample (n=468) to test the validity and reliability of questionnaire,12 as well as participated in several research workshops prior to embarking on the full-scale survey. One experienced researcher supervised the data collection activities in each community.

Measurement

The questionnaire was developed based on an emergency preparedness model proposed by Jessica Enders.13 A total of 50 items were included in the questionnaire (excluding demographic and social economic characteristics), which yielded a Cronbach’s α of 0.847 and a Spearman-brown formula of 0.884, demonstrating acceptable internal reliability and validity.

Dependent variable

Emergency preparedness was the outcome measure (dependent variable); defined as the necessary actions for an emergency. In this study, we measured emergency preparedness by asking participants what items they had in an ideal emergency supplies checklist.14 The ‘ideal’ checklist consists of 20 commonly used items in case of an emergency defined by the public education ‘ready’ programme in the USA and the Manual for All Citizens to Prevent Disaster edited by Ministry of Civil Affairs of the People's Republic of China.8 15Those who possessed all or more than the five essential items (3-day supply of food and water, first aid kit, battery-powered radio, flashlight, and extra batteries)16 were categorised as well prepared.

Independent variable

There are several steps involved between individual receiving information and then acting based on the information received. Enders13 summarised six major factors that may determine/change the course of actions during the process.

Demographic and social economic characteristics of participants, such as gender, age, marital status, education background, residential location and household income.

Previous experience coping with an emergency.

Emergency cognition—knowledge in relation to three types of emergency events was tested in this study using 16 statements: natural disaster (eg, shelter location and self-saving principles following an earthquake), man-made disaster (eg, evacuation routes and methods in household fire), and public health emergency (eg, symptoms, route of transmission, and preventive measures of respiratory infectious diseases; food handling and poisoning rescue measures). Respondents were asked to choose one answer for each of the statements: 1=correct, 2=incorrect, 3=don't know. One point was awarded for each correct response.

Risk awareness—we asked participants to rate on a five-point Likert scale about the likelihood of occurrence of emergency events in relation to natural disaster (earthquake), accidental disaster (fire), public health (communicable disease) and social safety events (terrorism).

Attitudes towards emergency preparedness—respondents were asked to rate their attitudes and beliefs towards emergency preparedness. Five questions were asked for this purpose, each being assessed on a five-point Likert scale (strongly agree, agree, neutral, disagree, strongly disagree). The questions were: ‘I do not believe preparation will make any difference in emergency events’; ‘I am very interested in emergency-related information/news’; ‘I pay special attention to all kinds of warnings issued by governmental authorities’; ‘I often discuss with my friends and family members about emergency response’; ‘I have confidence on governments in responding to emergencies’.

Self-efficacy—respondents were asked to rate their own ability to respond to emergencies and mitigate risks on a five-point Likert scale (ranging from totally disagree to totally agree).

Those considered factors can find their foundations from a few theories. For example, theory of rational action (TRA),17 which believes that human behaviours are consciously influenced by individual attitudes; Bandera's self-efficacy theory, which addresses the confidence and ability of a person to adjust and control his/her own behaviours; Tran-theoretical model, which emphasises the importance of both internal and external factors for driving changes in human behaviour18 and KAP theory, which puts knowledge, attitudes, belief and action in sequence with each serving as a precondition for the latter one.19

Statistical analysis

The outcome (dependent) variable was treated as a nominal measure: ‘well prepared’ versus ‘poorly prepared’.

We calculated dimensional scores for knowledge, risk awareness, attitudes towards preparedness and self-efficacy, and then transformed them into a nominal measure: ‘above average’ versus ‘on/below average’.

χ2 Tests were performed to determine the differences in emergency preparedness across different levels of independent variables. A multivariate logistic regression model was adopted to identify the extent to which the independent variables were associated with outcome measures (preparedness). An enter regression based on the maximum likelihood estimation method was used, with an enter/exit criterion (α) of 0.05/0.10.

All statistical analyses were performed using SPSS V.19.0.

Results

Participant characteristics

Slightly less than half (48%) of respondents were in the age between 31 and 50 years. Women comprised 56% of the respondents. Most (59%) respondents lived in an urban community. The majority (79%) were married at the time of the survey. More than half of respondents completed senior high school level of education. A household monthly income of between ¥2000 and ¥4999 accounted for 53% of all respondents (table 1).

Table 1

Characteristics of respondents and emergency preparedness outcomes (n=2686)

Variables	Respondentsn (%)	Well-preparedn (%)	χ2	p Value
Gender			4.846	0.028
Male	1177 (43.9)	46 (3.9)
Female	1509 (56.1)	87 (5.8)
Age			6.979	0.031
18–30	717 (26.7)	30 (4.2)
31–50	1281 (47.7)	56 (4.4)
51+	688 (25.6)	47 (6.8)
Residential location			4.302	0.038
Urban	1586 (59.1)	90 (5.7)
Rural	1100 (40.9)	43 (3.9)
Marital status			0.244	0.621
Married	2115 (78.7)	107 (5.1)
Others	571 (21.3)	26 (4.6)
Educational attainment			6.260	0.044
≤Junior high school	1260 (46.9)	49 (3.9)
Senior high school	644 (24.0)	41 (6.4)
University	782 (29.1)	43 (5.5)
Household monthly income (¥)			9.526	0.009
0–1999	853 (31.7)	28 (3.3)
2000–4999	1409 (52.5)	75 (5.3)
5000+	424 (15.8)	30 (7.1)
Living with a child under 18 years			0.585	0.444
No	1258 (46.84)	58 (4.6)
Yes	1428 (53.16)	75 (5.3)
Previous experience with emergency			10.910	0.001
No	2106 (78.4)	89 (4.2)
Yes	580 (21.6)	44 (7.6)
Exposure to awareness raising during the past year			4.236	0.040
No	2056 (76.5)	92 (4.5)
Yes	630 (23.5)	41 (6.5)
Knowledge about emergency			25.583	0.000
On/below average	1301 (48.4)	36 (2.8)
Above average	1385 (51.6)	97 (7.0)
Risk awareness			7.646	0.006
On/below average	1444 (53.8)	56 (3.9)
Above average	1242 (46.2)	77 (6.2)
Attitudes towards preparedness			33.895	0.000
On/below average	1307 (48.7)	32 (2.4)
Above average	1379 (51.3)	101 (7.3)
Self-efficacy			18.926	0.000
On/below average	1652 (61.5)	58 (3.5)
Above average	1034 (38.5)	75 (7.3)

Less than 5% (133/2686) of respondents were rated as ‘well-prepared’, that is possessing all five essential emergency items, while more than 9.5% of respondents did not possess any emergency items at all. Emergency preparedness was found to be associated with gender (χ2=4.846, p=0.028), age (χ2=6.979, p=0.031), residential location (χ2=4.302, p=0.038), education (χ2=6.260, p=0.044), household income (χ2=9.526, p=0.009), past experience with emergency (χ2=10.910, p=0.001), exposure to awareness-raising (χ2=4.236, p=0.040), as well as knowledge (χ2=25.583, p<0.001), risk awareness (χ2=7.646, p=0.006), attitudes (χ2=33.895, p<0.001) and self-efficacy (χ2=18.926, p<0.001). Women, older people, urban residents and those who completed senior high schools and had a higher household income were more likely to be rated as well prepared. The respondents who had previous experience of emergencies and exposure to awareness-raising had a higher percentage of well prepared than those who did not. Higher levels of knowledge, risk awareness, self-efficacy and positive attitudes towards emergency preparedness were positively correlated with preparedness outcomes. Marital status and living with a child (under 18 years of age) were not found to be associated with emergency preparedness (table 1).

Logistic regression analysis on emergency preparedness

In the multivariate logistic regression model, 11 independent variables were included which had statistical significance in χ2 test (p<0.05). Seven independent variables were identified as significant predictors for emergency preparedness. Women (OR=1.691) and those who had a higher household income (OR ranging from 1.666 to 2.117) were more likely to be rated well prepared. Previous experience with emergency (OR=1.552), higher levels of knowledge about emergency (OR=2.192), risk awareness (OR=1.531), self-efficacy (OR=1.796), as well as positive attitudes towards emergency preparedness (OR=2.265) were also significant predictors for emergency preparedness; however, neither educational attainments nor exposure to awareness-raising featured in the logic regression model as a significant predictor for emergency preparedness (table 2).

Table 2

Logistic regression analysis on emergency preparedness

Variables	Walds	p Value	OR	95% CI
Gender
Female	7.495	0.006	1.691	1.161	2.462
Male (reference)
Age
18–30 (reference)
30–50	0.093	0.760	0.930	0.582	1.484
51+	3.492	0.062	1.626	0.977	2.707
Residential location
Rural	0.002	0.964	1.011	0.623	1.641
Urban (reference)
Education
≤Junior high school (reference)
Senior high school	2.327	0.127	1.471	0.896	2.415
University	0.584	0.445	1.233	0.720	2.112
Household monthly income (¥)
0–1999 (reference)
2000–4999	4.152	0.042	1.666	1.020	2.722
5000+	6.031	0.014	2.117	1.163	3.851
Previous experience with emergency
Yes	4.929	0.026	1.552	1.053	2.286
No (reference)
Exposure to awareness-raising in last year
Yes	0.847	0.358	0.827	0.552	1.239
No (reference)
Knowledge about emergency
Above average	13.978	0.000	2.192	1.453	3.308
On/below average (reference)
Risk awareness
Above average	5.285	0.022	1.531	1.065	2.202
On/below average (reference)
Emergency attitude
Above average	14.196	0.000	2.265	1.480	3.465
On/below average (reference)
Self-efficacy
Above average	9.500	0.002	1.796	1.238	2.606
On/below average (reference)
Constants	54.929	0.000	0.003

Reasons of poor preparation for emergency

Lack of awareness and knowledge about emergency was cited by the respondents as the most important reasons of poor preparation for emergency (figure 1). More than half (51.7%) of poorly prepared respondents did not know what to do, while over 30% simply disregarded emergency preparation as an issue that affected them. More than 20% of respondents preferred to leave emergency preparedness in the hands of professional workers (22.7%), blame the government and media for limited information disclosure (21.8%), or/and maintain optimistic perspective (23.9%), hoping that any disaster will not befall them. Interestingly, <15% of respondents attributed their poor preparedness to limited access to resources, such as time, money and personal ability.

Figure 1

Reasons of poor preparation for emergency.

Further data exploration (table 3) revealed that mass media was the major contributor to emergency knowledge, with more than 77.7% respondents being exposed to it; closely followed by personal communications (33.9%). By contrast, educational programmes were an information source for only 15.8% (through community emergency education programmes) to 22.1% (through school education) of respondents. Most respondents (90%) had never participated in emergency response exercises.

Table 3

Sources of emergency knowledge (N=2686)

Sources of knowledge	Number of respondents	Per cent
Mass media	2087	77.7
Personal communication	911	33.9
School education	593	22.1
Community education programme	424	15.8
Emergency response exercise	279	10.4

Discussion

The level of emergency preparedness in Heilongjiang residents is very low: fewer than 5% of respondents reported having essential emergency supplies in hands; and 9.5% possessed no items in the emergency supplies checklist (20 specific items). Such a low level of emergency preparedness has also been found in studies undertaken elsewhere in China.20 21 Despite differences in data collection methods, it is evident that the US populations have a much higher level of emergency preparedness compared with Chinese populations. FEMA reported that 57% of US citizens have emergency supplies prepared and ready to be used solely in the case of a disaster.6

As found in the literature, poor knowledge and attitudes towards emergency preparedness are the major factors that have contributed to the low level of community preparedness outcomes22 significant associations between preparedness outcomes and individual knowledge and attitudes have been demonstrated in the χ2 tests, the multivariate regression model, and the subjective perception survey in this study. More than half of the poorly prepared respondents simply did not know what to do.

Knowledge and attitudes towards emergency preparedness were found to have the greatest OR (>2) in the multivariate regression model on preparedness outcome. Indeed, knowledge and attitudes are regarded as a key driver of human behaviours in several behavioural theories.23

Although financial capability remains as a significant factor influencing emergency preparedness, the study participants perceived limited resources and ability to prepare as a contributor for a very small proportion (<15%) of poor preparedness outcomes. Nevertheless, there is an increased need for help in emergency for the households with low income.24 25

Effective education and exercise can help community residents recognise and identify hazards to improve their understanding about how to respond.26 27 This study showed that mass media is perhaps the most important channel for people to obtain information about emergency. It is important to note that educational attainments and exposure to awareness-raising were not found to be associated with emergency preparedness. Obviously, educational campaigns need to be carefully designed and tailored to the needs of their audiences.

An effective communication and education programme should target the most important audience in a way that is most appropriate to the audience. Women, for example, may play a critical role in community emergency preparedness.28 29 This study likewise found that women are more likely to be prepared for emergencies than men. The finding is somehow different from that of the FEMA study,6 in which US men were reported having a higher level of emergency preparedness. Such inconsistency may have actually reflected cultural differences: in China, women are usually responsible for shopping and housing arrangements, and thus determine what items to be stored at home.

Risk awareness should also be a key element in emergency education programmes. Several studies confirmed that individuals with low risk awareness are less likely to respond to warnings and undertake preparatory actions compared to those with higher risk awareness.30–32 Risk awareness plays an important role at the early stage of the continuum from information attainment to behavioural changes.33 It encourages people to seek information. Gisela Wachinger proposed a heuristic framework to understand the relationship between risk perceptions and individual behaviours.34 Three factors may mediate the relationship: experience and motivation, trust and responsibility, and ability (economic and personal condition). We found that past experience with emergencies is associated with improved preparedness. It is understandable that those who have experienced with emergency events may have increased awareness of the importance of preparedness.35 In general, they tend to be more proactive in obtaining knowledge and information and taking preparatory actions36–41; however, those who have not experienced personal damage during a hazardous event are less likely to believe that a future event may bring harm to them, which may jeopardize their preparatory efforts.42 A survey conducted in Gansu reported that past experience with disastrous events was associated with risk perceptions, but the perceptions failed to translate into preparatory actions.43 Some researchers argue that there is not necessarily a direct link between awareness, risk perception and desired preparation or behavioural response.27 Perceived coping ability (self-efficacy) may mediate the link between risk awareness and emergency preparedness outcomes. External conditions may also alter the course of actions.44 Emergency response exercises could help build the link between risk awareness and preparatory actions.13

Limitation

This study has several limitations. The questions measuring awareness and preparatory behaviours for emergency were intentionally made broad, covering four types of events (natural disaster, public health emergency, human-made accidence, and social security event). The broad notion and non-specific events may cause unclear or vague responses from some participants, especially in those with low levels of education. However, the questionnaires were administered through face-to-face interviews and all of the interviewers had been made aware of this potential problem to minimise such a bias. Owing to limited resources, the survey was conducted only in Heilongjiang province. In this article, we examined potential factors that are associated with preparedness outcomes; further studies are needed to explore the mechanisms of the links between those factors and preparedness outcomes.

Conclusion

The level of emergency preparedness in Heilongjiang residents is very low (<5%), which is linked with a lack of knowledge and resident attitudes towards emergency preparedness. Future education and communication campaigns should be more focused and tailored to the needs of intended audience, taking into consideration of their usual sources of information and knowledge in relation to emergency preparedness. A variety of strategies should be taken to increase community risk awareness, including emergency training and exercise; mobilisation of volunteers; emergency day event for publicising purpose; and mass media (web, TV, radio, etc) campaigns. Women may become the major audience because they are responsible for housing arrangements and are more likely to become well prepared than men. It is also important to emphasize the translation process from increased awareness into preparatory actions, possibly through emergency response exercises. Currently, only 10% of Heilongjiang residents participated in emergency response exercises, falling far short of expectations.