Understanding the barriers to smoke alarm installation and maintenance in rural Alberta
Development of a questionnaire based on the Theory of Planned Behaviour
Mahendra Wijayasinghe, PhD
10-credit project for the degree of Master in Safety Promotion
Supervisor
Celine Farley, PhD.
Stockholm, June 2006
ABSTRACT
Background: Fire statistics for Alberta and the US indicatea highrisk of fires, fire deaths and fire injuries in rural area homes. A major deficiency in home fire safety, especially in rural areas, is a lack of both the installation and maintenance of smoke alarms. The three maintenance functions that ensure smoke alarms are in working condition to alert occupants in a fire emergency are monthly tests, annual replacement of batteries and avoidance of the tendency to disable smoke alarms that are too sensitive or sound false alarms.
The Theory of Planned Behaviour (TPB) has been widely appliedto the explanation of health behaviours and to inform behaviour change interventions. The TPB integrated within the PRECEDE-PROCEED Model was therefore selected to understand the barriers to smoke alarm installation and maintenance in rural Alberta.
Main aim and objectives:This study sought to understand the barriers to smoke alarm installation and maintenance in the Town of Wainwright, a rural Alberta community, by administering a preliminary questionnaire based on the TPB integrated in to the PRECEDE-PROCEED model.
Methods: Two questionnaires, for homes with and without smoke alarms, were developed to measurerespondents’ demographic and housing characteristics, smoke alarm status, risk perception, fire drill frequency, reasons for not performing fire safe behaviours and the sources of fire prevention and safety information. In addition, the questionnaires carried a set of items to measure TPB constructs, enabling factors and reinforcing factors. The questionnaires were administered to a randomly selected group of 36 homes in the Town of Wainwright, with the assistance of the local fire department.
Results and discussion: Analysis of survey responses from homes that had at least one smoke alarm (N=35) indicated that the internal consistency was high for TPB constructs – attitude (0.78), behavioural beliefs (0.76), and motivation to comply (0.91) as measured by Cronbach’s alpha. Positive fire safety behaviours among the respondents included 94% of households having smoke alarms in working condition, 70% replacing batteries once a year and only nine percent reacting to false alarms by removing batteries. The perception of a low fire risk by 70% of the respondents may be related to the lack of fire drills in 78% of the homes and lack of testing smoke alarms at the recommended monthly intervals (only nine percent). Other possible reasons for the low perception of fire risk the related behaviours and challenges to interventions are discussed.
Conclusions: Results from the survey, though preliminary, provides some understanding of the barriers to smoke alarm maintenance. The results point to increased fire department involvement and changes in fire prevention messaging to improve smoke alarm maintenance behaviours. Future research will focus on an elicitation study to transform the preliminary questionnaires for large-scale application.
Key words: Smoke alarms, rural communities, Alberta, home-fires, Theory of Planned Behaviour.
CONTENTS Page
ABSTRACT2
BACKGROUND4
Community differences in fire risk, deaths and injuries4
Smoke alarms4
Smoke alarm statistics in Alberta and other jurisdictions4
Research studies on smoke alarm giveaway programs5
Using theory to guide intervention6
The Theory of Planned Behaviour6
The PRECEDE-PROCEED Model 7
MAIN AIM AND OBJECTIVES 7
METHODS 7
Development of preliminary questionnaires 7
Data analysis 7
Selection of rural community and administration of the questionnaires 8
Ethical considerations 8
RESULTS 9
DISCUSSION10
CONCLUSIONS12
ACKNOWLEDGEMENTS13
APPENDICES14
Appendix A: References14
Appendix B: Smoke Alarms18
Appendix C: Theory of Planned Behaviour & the PRECEDE-PROCEED Model19
Appendix D: Preliminary Questionnaire for Homes with Smoke Alarms22
Appendix E: Cronbach’s Alpha Values for Theory of Planned Behaviour Constructs32
Appendix F: Town of Wainwright34
Understanding the barriers to smoke alarm installation and maintenance
in rural Alberta
Development of a questionnaire based on the Theory of Planned Behaviour
BACKGROUND
The 2001 Annual Report of Fire Losses in Canada, published by the Council of Canadian Fire Marshals and Fire Commissioners (CCFM/FC), a national organization dedicated to fire protection in Canada, reported a total of 55, 323 fires associated with 338 fire deaths; 2,310 fire injuries; and a total property loss of $1.421 billion Canadian. The "Residential Property” classification accounted for the largest number of fires. In 2001, there were 21,494 fires in this category, 39% of the Canadian total. These incidents resulted in 273 deaths or approximately 73% of the nation's fire fatalities. The monetary loss amounted to $632 million, which is 45% of the total loss.
Most people who die in fires in Canada and in Alberta are victims of home fire incidents (CCFM/FC, 2004; Alberta Fire Commissioner’s Office, 2004). Homes defined as one/two family dwellings, apartments and manufactured homes accounted for approximately 26% of all fires, 65% of all fire deaths and 60% of all fire injuries in Alberta in 2003. Seniors (65+ years), young adults and children under five years of age are at high risk of fire related injuries and deaths in Alberta (Alberta Fire Commissioner’s Office, 2004). Home fires in the U.S., during 2004, caused 82% of all civilian fire deaths and approximately 80% of all fire injuries (Karter, 2005). In London, UK, 86% of fire deaths occurred in homes during 1996-2000 (Holborn, 2001).
Community differences in fire risk, deaths and injuries
A comparison of rural and urban fire losses during 1992-1996 in Alberta revealed that the risks of fire incidence, fire death and property losses are 1.2, 2.4 and 3.4 times higher in rural compared to urban municipalities. (Wijayasinghe, 1997). Fire death rate expressed as number of deaths per 100,000 population in Aboriginal communities (6.1) is ~ 4 times higher than in rural municipalities (1.7) and ~ 9 times higher than in urban centres (0.7). Fire rate (number of fires per 1,000 population) is ~1.5 times higher in rural communities (2.8) compared to both urban and Aboriginal communities (Wijayasinghe, 2005). In the U.S., per capita fire death rates in rural communities (under 2,500 population) are more than twice the rates found in communities with more than 500,000 (Ahrens, 2003). The major concern lies with the very limited fire protection resources typically available to a rural community. Rural areas are inherently spread out and mostly protected by small, volunteer fire departments with limited resources to attack specific fire problems. Response times of fire departments is also longer than in urban centres (Wijayasinghe, 1997). A strong statistical correlation between the rural poverty level and rural fire death rates has been established. Being poor, means a reduced ability to respond to fire safety initiatives. Heads of poor households are less likely to have high school education, so they may not grasp the full import of public fire safety education messages, if they do grasp the messages, they are more likely to lack the discretionary income to obtain smoke detectors, or safe heating systems, or code compliant electrical service, or fire safe security measures, or anything else that involves buying a larger measure of fire safety (National Fire Protection Association, 2002).
Smoke alarms
Functioning smoke alarms serve to provide early warning of a fire incident, and allow occupants time to evacuate the dwelling. Appendix B provides background information on smoke alarms and show the human intervention steps required to gain the full benefit from them in a home environment. However, human safety during a fire emergency requires not only an early warning from smoke alarms but also successful evacuation of occupants, who must hear and correctly interpret the alarm signal, and then be able to exit the dwelling. The latter step is critical in surviving a fire emergency and requires householders to have planned and practiced a fire escape plan (Fire Protection Handbook, 2003).
Smoke alarm statistics in Alberta and other jurisdictions
Smoke alarms are required by law in Alberta (Alberta Building Code, 1997; Alberta Fire Code, 1997), and their use has been promoted by the fire service educational programs since the late seventies. The first legislated requirement for smoke alarms in Alberta homes was introduced in 1977. The decline in fire death rate in the province has been partly attributed to this legislation (Alberta Fire News, 2005). According to Statistics Canada (1998), in 1997, 96.2% of all households in Alberta were installed with at least one smoke alarm, located inside the dwelling. In U.S., the use of smoke alarms increased rapidly during 1975 to 1984, and slowly thereafter, and evidence from actual fire stories and fire statistics showed the life-saving effectiveness of these alarms. The smoke alarm usage in the U.K. has lagged behind that of the U.S. In 2002, surveys have shown 96% and 81% of households in the U.S. and the U.K. had smoke alarms installed (Hall, 2005).
Fire statistics for Alberta homes, that experienced fire incidents, indicate that in spite of the legislation and promotional activities by the fire service, 31% of urban homes were not installed with smoke alarms. In rural and First Nation/Metis communities this figure rose to 60 and 62%, respectively (Alberta Fire Commissioner’s Office, 2005). Similarly, in 2002, smoke alarms were not present in 57% of reported home fires in the U.K. and in 38% of reported home fires in the U.S. (Hall, 2005). Thus, it appears that homes where people have reported fires are less likely to have smoke alarms. According to Ahrens (2004), two factors may explain this disproportionate share of home fires occurring in U.S. homes with no smoke alarms. First, there are, proportionally, so many more smoke alarms in homes in general than in homes with reported fires, and households that have fires tend, for a variety of reasons, to be the kind of households that would be less likely to buy or own smoke alarms. Second, smoke alarms discover some fires so early that the occupants can control the fires without involving or notifying the fire department. The author elaborated further that the households that still do not have smoke alarms are more risk prone but in ways that do not correlate strongly or neatly with the socioeconomic characteristics – poverty, race, age, etc. – that usually correlate with the risk of having a reported fire or having a fatal fire.
A comparison of urban, rural and First Nation/Metis communities in Alberta revealed that only 39%, 25% and 7%, respectively, of the installed smoke alarms activated during fires (Alberta Fire Commissioner’s Office, 2005). About 25% of the smoke alarms in homes that experienced fires in the U.S. did not function (Ahrens, 2004). Similarly, in 2002 in the U.K., 27% of the homes with smoke alarms present failed to operate (Hall, 2005). According to this report, smoke alarms failed to operate in 43% of fires where the units were battery-powered but only 12% of fires where the units were connected to main power. The majority (53%) of U.K. smoke alarms reported in home fires were connected to main power. The major reason for non-functioning smoke alarms appears to be due to dead, disconnected or missing batteries. People are most likely to remove or disconnect batteries because of nuisance activations (Ahrens, 2004).
Although there are many studies on smoke alarms and related home fire risks (DiGuiseppi, 2000; CDC, 1998, Istre, 2001; Shaw, 1988), only a few systematic documented studies on actual field application are available. These have taken the form of large-scale smoke alarm giveaway programs, the subject of the next topic.
Research studies on smoke alarm giveaway programs
Two groups of researchers have addressed the problem of increasing smoke alarm use in materially deprived communities in Oklahoma City, U.S. and in inner London, U.K. The Oklahoma City study was a non-randomized controlled trail in which ~10,000 free smoke alarms were distributed in targeted neighbourhoods along with written educational pamphlets and home based follow up to test whether the alarms were functioning correctly. During the six years following the project, the residential fire-related injury rate decreased 81% in the target population but only 7% in the rest of Oklahoma City. This dramatic decline in fire related injuries in the target area was largely attributed to the free smoke alarm distribution as well as to educational efforts promoting awareness about residential fire prevention (Mallonee, 1996).
The smoke alarm give away program, in two deprived multi-ethnic inner London boroughs, was conducted as a randomized controlled trial, where ~20,000 smoke alarms, with batteries, fittings, and fire safety brochures (in English and other local languages) were distributed door to door in randomly selected wards by a coalition of agencies. One year later, postcards reminding that the battery should be changed were sent. Control wards received no intervention (DiGuiseppi, 1999). The results indicated that the programme did not reduce injuries related to fire, admissions to hospital and deaths, or fires attended by the fire brigade, mostly because few alarms had been installed or were maintained. It was concluded that giving away smoke alarms may be a waste of resources and of little benefit unless alarm installation and maintenance are assured (DiGuiseppi, 2002). Pless (2004) concluded that the inner London study, although modeled after the apparently effective Oklahoma study (Mallonee, 1996) in which an 80% reduction in fire related injuries was reported, failed to replicate these results probably due to a regression towards the mean. Another report from the same study focused on determining the prevalence of and predictors for installed, functioning smoke alarms in public housing found that few properties in this multi-ethnic, materially deprived urban area had any installed,functioning smoke alarms, despite a high risk of residential firesand fire related injuries in such areas. Alarmsmost commonly failed because they lacked batteries (72%). Having an installed, functioning alarm was most stronglyassociated with living in a house versus an apartment, having two resident adultsversus one and recognising stillsfrom a Home Office television smoke alarm campaign. Furthermore, the authors concluded that smoke alarms are associated with a significantlyreduced risk of death in residential fires, and are more protectivein households with young children. It was concluded that effective methods toincrease the prevalence of installed and functioning alarms mustbeidentified (DiGuiseppi, 1999).
In an effort to identify the barriers to the continued use of smoke alarms, a qualitative research study was conducted on a sample of adults and children from the randomizedcontrolled trial of free smoke alarm installation in the inner city London boroughs (Roberts, 2004). This study found that the main barrier to smoke alarm use was the distresscaused by false alarms. Although trial participants consideredthemselves to be at high risk for fires and would recommendsmoke alarms to others, respondents’ reports on the distresscaused by false alarms suggest that people balance immediateand longer-term risks to their health and wellbeing when theydisable alarms.This study identified some of the reasons for thelow level of functioning smoke alarms, and problems experiencedwith alarms.
Using theory as a guide to intervention
Fire deaths and injuries represent a significant public health issue in Alberta. As discussed above, the fire risk to rural Albertans is particularly high. In rural Alberta homes that experienced fire, the percent installation of smoke alarms was half that observed in similar homes in urban centres. Activation of installed alarms was 1.5 times less in rural compared to urban homes. Therefore, a smoke alarm intervention program aimed at rural Albertans can be considered beneficial.
Although many health promotion interventions are developed and implemented without any specific reference to theory, there is much evidence in health promotion literature that an intervention theory can provide the guidance necessary to develop an effective intervention. Theory can: help practitioners understand the nature of the problem being addressed; explain the needs and motivations of target populations; lead practitioners on how to change health status, health-related behaviours and their determinants; and inform the methods and measures used to monitor the problem and the program (Nutbeam, 2004).
The Theory of Planned Behaviour (TPB) (Ajzen, 1991) was selected as the best choice to guide a smoke alarm intervention program in rural Alberta, based on the criteria proposed by Glanz (2002). The TPB has been applied in diverse settings to a variety of health behaviours, including exercise behaviour, smoking and drug use, HIV/STD prevention behaviours, mammography use, clinical preventive services, and oral hygiene behaviours (Glanz, 2002). Other interventions based on the TPB directed at individuals who do not currently intend to perform the desired target behavior: to get individuals who have no intention to do so to perform testicular self-examinations (Brubaker & Fowler, 1990), to ride the bus instead of driving the car (Bamberg, Ajzen, & Schmidt, 2003), to wear a bicycle safety helmet (Quine, Rutter, & Arnold, 2001), to increase regular condom use (Fishbein et al., 2001), and so forth.
Theory of Planned Behaviour (adapted from Glanz, Rimer & Lewis, 2002 and Ajzen I, 1991)
A brief description of the constructs in the TPB and their measurement are provided in Appendix C.
The PRECEDE-PROCEED (Green and Kreuter 1999, Glanz et al., 2002) model has been used widely by health promotion practitioners. This planning model follows a step-wise sequence including planning, implementation and evaluation stages. Different theories can be made use of at each of these stages to guide and inform practitioners (Nutbeam, 2004). The acronym PRECEDE stands for Predisposing, Reinforcing, and Enabling Constructs in Educational/Environmental Diagnosis and Evaluation. PROCEED stands for Policy, Regulatory, and Organizational Constructs in Educational and Environmental Development. The constructs in the Theory of Planned Behaviour fit within the Predisposing Factors in the model.