Increasing smoke alarm operability through theory-based health education: a randomised trial

While working smoke alarms halve the number of deaths in residential fires, many families do not keep the alarms running.
We tested the theory.
Increase alarm operability based on education.
Methods random multi-arm test, one arm was randomly selected every day for low
Income community in Maryland, USA.
Weapons of intervention :(1)
The module of health belief and society-
Provide the cognitive theory module of hand
Practice on installing alarm battery and using alarm mute button(2)Hands-
About Social Practice
Cognitive module supplemented by typical fire department education; (3)
Accept only the current norms of typical fire department education.
430-
Six families were recruited by church or knocking at the door. Follow-
The alarm operability was checked in 370 households (85%)1–3.
5 years after installation.
Primary result measurement: number of homes with work alarm, defined as alarm with work battery or hard battery
Number of wired and work alerts per home.
Regression of pre-intervention control of alarm status;
Demographics and beliefs about fire risk and effectiveness of alerts.
As a result, families in the comprehensive education and practice sector are more likely to issue effective smoke alerts in follow-up activitiesup (OR=2. 77, 95% CI 1. 09 to 7. 03)
The average is 0.
Add 32 work alerts per home (95% CI 0. 09 to 0. 56).
The work alert for each household rose by 16%.
Complete education and practice have similar effects (p=0.
97 of the two outcome measures).
Conclusion The installer can achieve greater smoke alarm operability without exceeding the installation time of the typical Fire Department. Hands-
Practice is the key.
After two years of installation, every three families are received
In practice, one has an additional work alert.
Clinical trial registration numbernumber NCT00139126.
Background while working smoke alarms halve the number of deaths in residential fires, many families do not keep the alarms running.
We tested the theory.
Increase alarm operability based on education.
Methods random multi-arm test, one arm was randomly selected every day for low
Income community in Maryland, USA.
Weapons of intervention :(1)
The module of health belief and society-
Provide the cognitive theory module of hand
Practice on installing alarm battery and using alarm mute button(2)Hands-
About Social Practice
Cognitive module supplemented by typical fire department education; (3)
Accept only the current norms of typical fire department education.
430-
Six families were recruited by church or knocking at the door. Follow-
The alarm operability was checked in 370 households (85%)1–3.
5 years after installation.
Primary result measurement: number of homes with work alarm, defined as alarm with work battery or hard battery
Number of wired and work alerts per home.
Regression of pre-intervention control of alarm status;
Demographics and beliefs about fire risk and effectiveness of alerts.
As a result, families in the comprehensive education and practice sector are more likely to issue effective smoke alerts in follow-up activitiesup (OR=2. 77, 95% CI 1. 09 to 7. 03)
The average is 0.
Add 32 work alerts per home (95% CI 0. 09 to 0. 56).
The work alert for each household rose by 16%.
Complete education and practice have similar effects (p=0.
97 of the two outcome measures).
Conclusion The installer can achieve greater smoke alarm operability without exceeding the installation time of the typical Fire Department. Hands-
Practice is the key.
After two years of installation, every three families are received
In practice, one has an additional work alert.
Clinical trial registration numbernumber NCT00139126.
Working smoke alarms reduce the number of deaths in residential fires by at least 48%.
However, it turned out to be difficult for the home to check and replace the battery to keep the alarm running properly.
Nationwide, 96% of American households report smoke alarms, but only 80% report work or actionable alarms --
Hard alarm-
Wired or with working battery.
5 This percentage is low-
Usually 15-50% of the revenue area where the smoke alarm does not work properly.
A project in Kim County, Washington, where smoke alarms were installed among owners
Found that 20% of low-income families do not work
Almost always deliberately.
6 months and 23% months after installation.
In rural Iowa, 27% of alarms do not work for 42 months after installation.
14 in the country, during the period of 2006-2003, 31% of the households involved in the fire by the fire department did not have an alarm, and 21% of the households (
30% with alarm)had only non-working alarms.
1 Using a database of fire investigations, we reviewed all 402 home structure fires investigated in Prince George County, MD, in 2001;
25% of households have work smoke alarms; 50% had non-
Work alarm and no alarm for 25%. A theory-
Long-term and effective basic education programmes
The term operability of the smoke alarm seems to be a logical component of the alarm installer, but none of them exist.
Fire departments and community groups that install alarms are almost always educating recipients about fire safety and alarm battery maintenance.
When we look at samples of these educational programs, almost none of them are theoretical --based.
Their effectiveness was largely not evaluated.
Indeed, a review of more than 55 home fire prevention interventions concluded that none of them were based on theory.
Although the door-to-
The door canvassing and installation scheme proved effective in increasing the number of homes where the alarm was installed, and 4, 16 no educational scheme was available to ensure that the smoke alarm continued to use the working battery.
Our goal is to develop and evaluate the results of the theory.
Basic education designed to increase the number of smoke alarms has been running.
Method Design and intervention this study is a parallel multi-arm trial that randomly assigns home with equal probability to two therapeutic arms and one comparative arm (
1: 1 distribution).
The weapon is: Complete theory
Drive education (Full Education)—
Education program based on health belief model combined with hand
Practice based on observation learning, reinforcement and simulation of social rules
Cognitive theory.
18. the education program is mainly aimed at individuals --
Horizontal factors that affect the use of smoke alarms.
These are the beliefs that the smoke alarm is effective in preventing fire damage, because there may be a fire in the home, a fire in the home can cause damage and be able to keep the smoke alarm at home.
These factors correspond to the perceived interests, perceived susceptibility, perceived severity, and the structure of the health belief model of the self. efficacy.
Observation learning based on society
Cognitive theory strengthens the vital self
Efficacy construction in health belief model. The social-
The cognitive module consists entirely of hands.
In the practice of testing the battery, installing the replacement battery, and using the flush button.
As part of the hands-on work, residents have installed batteries in our installed alarms.
We have also completed the standard fire safety checklist in the Fire Department facilities.
It requires home guides and guides to identify home fire hazards, test whether existing alarms are valid, and decide how many alarms are installed or replaced and where to hang.
The reason why we choose a healthy belief model is because
The development level behavioral change theory is important to explain which factors are important to encourage preventive health actions that many people have not taken, although these actions have proved effective.
Observe learning, strengthening and simulating social rules
Cognitive theory is included when they explain how a person learns behavior, and these factors reinforce the selfefficacy.
Online Supplement Appendix A provides A script for the exercise module.
The lead author offers the enclosed brochure and all other materials free of charge. Focused Hands-on Practice (Practice)—
Compression model limited to hand
Education Practice module of typical Fire Department (Described below)
Leave a brochure describing the use of the alarm and explain the importance of replacing the battery.
Comparison arm: Education in typical Fire Department (Current Norm)—
General Education in the fire department, including completing the fire safety checklist and leaving a \"baseball card\" with fire safety education information on the back \".
This information covers fire escape planning and encourages smoke alarm testing and battery replacement once a year.
We promote ourselves as a Family Safety Program.
A health educator and an alert installer form an access team.
The security agreement requires educators to stay at home while the installer is working. Two master-
Health educators who have received education in accordance with the book.
Including research data collection, home visits to all weapons with large differences averaged 45 minutes.
If residents keep the topic, education takes less time than installation.
Therefore, the duration of the visit is sensitive to the identity of the residents, not to the educational programme.
Table 1 summarizes the data collection forms and other activities related to access.
Knowledge, attitude and behavior (KAB)survey (
See Supplementary Appendix B online)
Including four general questions about fire risk, 10 questions about smoke alarms, and 7 questions (
No analysis.
Risk of falling
We designed the KAB issues related to the stages in the health belief model, focusing on barriers and facilitators, but not verifying them.
The shortlist has 6 questions about fire fighting experience and 8 questions about falling experience.
We included questions about falls and installing touch lights in the dark stairwell as we described the project as a home safety check.
If participants know that our research focuses on the smoke alarm function, their behavior may change.
It is important that this approach compares arm with more balanced security information.
The protocol is only available when there are mute buttons and 10-Year Battery (
Zero qualified)or were hard-wired (
Two qualified).
We created the English and Spanish versions of all the materials.
All forms, procedures and modifications have been approved by the Pacific Research and Evaluation Institute (PIRE\'s)IRB.
The trial did not cause any harm.
View this table: View the contents and order of smoke alarm home access in inline View pop-up table 1, as well as the difference between armsInstallers using the Kidde model 0916 ionised smoke alarm with the flush button.
The manufacturer\'s specifications indicate that when 6% of the battery power remains the same, the alarm starts to rattle.
Alarms include test buttons that provide a second test operability mode.
Install batteries marked with nail polish
Provide differentiated services to residents
From the battery we\'re still working on.
We buy and use Rayovac D1604 Heavy Duty 9-in bulk-volt carbon-
Zinc batteries valid for April 2010, but have been installed for longer periods of alkaline batteries in all subsequent alarmsup.
The main outcome measures are (1)
The number of homes with at least one working smoke alarm and (2)
The average number of smoke alarms working in each home.
Recruitment, random and scheduling trials were conducted in Maryland, USA.
Installation began on September 2005 and ended on May 2008.
Our initial power calculation was for 1000 installed homes.
Each arm is up, which will support detection at a 95% confidence level of 4-6% of subsequent operability changes
Assume that half of the comparison arm has an operational alarm.
Usually, most residents who are canvassed by smoke alarm installers do not open the door.
The aim is to learn how to increase alert maintenance for those who accept the installation, not how to increase access to the home.
Therefore, when a resident aged 18 or over agrees to install an alarm, the entry into our sample begins even if the education is rejected (three homes).
The door for initial recruitment-to-
Four low-income communities in suburban Prince George County, door work days and weekend day canvassing.
Using census data, we divide the community into consecutive target areas of about 60 residential units and then use computers
Generate random numbers to select a simple random sample of the target area to be canvas.
Before the canvassing started, the change of the county fire chief forced us to install without the help of the fire truck, which greatly reduced our ability to open the door.
Random exercises (arm 2)
Stopped in February 2007 to increase the likelihood that the other two arms would reach a sufficient sample size.
To inform this decision, in the absence of a review of arm allocation, we performed a mid-term analysis of the mean and variance of the number of alerts at the time of registration and found two-
The third of the comparison arm has a pre-intervention of working alarm, SD is 0.
Therefore, sufficient power after detecting the 10% differential intervention only needs to be installed 100-200 in each arm.
Due to the ongoing shortage of samples, in January 2008 we added two recruitment streams in the city of Baltimore: Marketing of church services and canvassing of neighborhoods, where there was a fire truck with flashing lights that helped to get low
Income families in the fire department\'s target block.
To simplify the tasks of health educators, we assign all the participants of the day to one arm and, when the team gathers, randomly select each day by rolling the mold. Using pilot-
Tested procedures and materials, staff knocking and writing to arrange follow up
Check alarm operability and ask questions about alarm testing, fire fighting experience and fall experience.
Although follow
Up is not officially blinded and is generally executed by the installer;
They don\'t know how to distribute weapons. Ideally follow-
It won\'t start until all the batteries we initially installed are dead.
Otherwise, we cannot distinguish between situations where residents maintain alerts that have not yet required a new battery, or where an alarm cannot operate when the initial battery is dead.
Because there\'s no data on carbon.
The zinc battery life in the smoke alarm, we hung 100 alarms in PIRE\'s office.
Their battery failure rate caused us to wait at least 500 days before trying to return home to check for operability.
However, we returned 4 months in advance of the last 40 families and installed the alarm becauseup (and the grant)
Ended on September 2009.
Statistical analysis, we compared the treatment results with the current normal family using ordinary least squares linear regression to analyze whether the number of work alerts increased significantly from baseline to subsequent
Up and logistic regression analysis whether there is at least one working alarm after
The rise was obvious.
All analyses were performed in stata v.
11 people set out from the University of Texas station Stata Corp.
Before starting the data analysis, we decided to include the data from the pilot test house, because after the pilot, neither the method nor the material changed. testing.
The virtual variables in regression test whether the test data including the pilot test affect the results (
It\'s not).
Because of the results of two theories
It turns out that weapons based on health education are extremely similar and the number of samples is moderate, and most analyses combine the two weapons.
To process the missing revenue data, we coded the missing revenue to 0 and added a 0-1 variable with a value of 1 for those who reject the income issue.
Online Supplementary Appendix C describes all other model covariates and shows covariates-
Adjust the regression separately for each arm.
Regression analysis includes pre-intervention alarm status (
Is there an operational alarm or number of operational alarms)
As a predictor, any bias associated with functional differences between weapons at the time of registration can be adjusted.
Analysis whether there is time to follow up the test
In the home where the battery we originally installed is still not working, the affected results.
We initially tested the models of all population variables and all KAB fire and smoke alarm project variables.
We exclude most of the variables from the final model that lack statistical significance to minimize cases that are excluded due to lack of data and avoid multiple collinearity.
The design and intervention of this study was a multi-arm parallel trial that randomly assigned families with equal probability to two therapeutic arms and one comparative arm (
1: 1 distribution).
The weapon is: Complete theory
Drive education (Full Education)—
Education program based on health belief model combined with hand
Practice based on observation learning, reinforcement and simulation of social rules
Cognitive theory.
18. the education program is mainly aimed at individuals --
Horizontal factors that affect the use of smoke alarms.
These are the beliefs that the smoke alarm is effective in preventing fire damage, because there may be a fire in the home, a fire in the home can cause damage and be able to keep the smoke alarm at home.
These factors correspond to the perceived interests, perceived susceptibility, perceived severity, and the structure of the health belief model of the self. efficacy.
Observation learning based on society
Cognitive theory strengthens the vital self
Efficacy construction in health belief model. The social-
The cognitive module consists entirely of hands.
In the practice of testing the battery, installing the replacement battery, and using the flush button.
As part of the hands-on work, residents have installed batteries in our installed alarms.
We have also completed the standard fire safety checklist in the Fire Department facilities.
It requires home guides and guides to identify home fire hazards, test whether existing alarms are valid, and decide how many alarms are installed or replaced and where to hang.
The reason why we choose a healthy belief model is because
The development level behavioral change theory is important to explain which factors are important to encourage preventive health actions that many people have not taken, although these actions have proved effective.
Observe learning, strengthening and simulating social rules
Cognitive theory is included when they explain how a person learns behavior, and these factors reinforce the selfefficacy.
Online Supplement Appendix A provides A script for the exercise module.
The lead author offers the enclosed brochure and all other materials free of charge. Focused Hands-on Practice (Practice)—
Compression model limited to hand
Education Practice module of typical Fire Department (Described below)
Leave a brochure describing the use of the alarm and explain the importance of replacing the battery.
Comparison arm: Education in typical Fire Department (Current Norm)—
General Education in the fire department, including completing the fire safety checklist and leaving a \"baseball card\" with fire safety education information on the back \".
This information covers fire escape planning and encourages smoke alarm testing and battery replacement once a year.
We promote ourselves as a Family Safety Program.
A health educator and an alert installer form an access team.
The security agreement requires educators to stay at home while the installer is working. Two master-
Health educators who have received education in accordance with the book.
Including research data collection, home visits to all weapons with large differences averaged 45 minutes.
If residents keep the topic, education takes less time than installation.
Therefore, the duration of the visit is sensitive to the identity of the residents, not to the educational programme.
Table 1 summarizes the data collection forms and other activities related to access.
Knowledge, attitude and behavior (KAB)survey (
See Supplementary Appendix B online)
Including four general questions about fire risk, 10 questions about smoke alarms, and 7 questions (
No analysis.
Risk of falling
We designed the KAB issues related to the stages in the health belief model, focusing on barriers and facilitators, but not verifying them.
The shortlist has 6 questions about fire fighting experience and 8 questions about falling experience.
We included questions about falls and installing touch lights in the dark stairwell as we described the project as a home safety check.
If participants know that our research focuses on the smoke alarm function, their behavior may change.
It is important that this approach compares arm with more balanced security information.
The protocol is only available when there are mute buttons and 10-Year Battery (
Zero qualified)or were hard-wired (
Two qualified).
We created the English and Spanish versions of all the materials.
All forms, procedures and modifications have been approved by the Pacific Research and Evaluation Institute (PIRE\'s)IRB.
The trial did not cause any harm.
View this table: View the contents and order of smoke alarm home access in inline View pop-up table 1, as well as the difference between armsInstallers using the Kidde model 0916 ionised smoke alarm with the flush button.
The manufacturer\'s specifications indicate that when 6% of the battery power remains the same, the alarm starts to rattle.
Alarms include test buttons that provide a second test operability mode.
Install batteries marked with nail polish
Provide differentiated services to residents
From the battery we\'re still working on.
We buy and use Rayovac D1604 Heavy Duty 9-in bulk-volt carbon-
Zinc batteries valid for April 2010, but have been installed for longer periods of alkaline batteries in all subsequent alarmsup.
The main outcome measures are (1)
The number of homes with at least one working smoke alarm and (2)
The average number of smoke alarms working in each home.
Recruitment, random and scheduling trials were conducted in Maryland, USA.
Installation began on September 2005 and ended on May 2008.
Our initial power calculation was for 1000 installed homes.
Each arm is up, which will support detection at a 95% confidence level of 4-6% of subsequent operability changes
Assume that half of the comparison arm has an operational alarm.
Usually, most residents who are canvassed by smoke alarm installers do not open the door.
The aim is to learn how to increase alert maintenance for those who accept the installation, not how to increase access to the home.
Therefore, when a resident aged 18 or over agrees to install an alarm, the entry into our sample begins even if the education is rejected (three homes).
The door for initial recruitment-to-
Four low-income communities in suburban Prince George County, door work days and weekend day canvassing.
Using census data, we divide the community into consecutive target areas of about 60 residential units and then use computers
Generate random numbers to select a simple random sample of the target area to be canvas.
Before the canvassing started, the change of the county fire chief forced us to install without the help of the fire truck, which greatly reduced our ability to open the door.
Random exercises (arm 2)
Stopped in February 2007 to increase the likelihood that the other two arms would reach a sufficient sample size.
To inform this decision, in the absence of a review of arm allocation, we performed a mid-term analysis of the mean and variance of the number of alerts at the time of registration and found two-
The third of the comparison arm has a pre-intervention of working alarm, SD is 0.
Therefore, sufficient power after detecting the 10% differential intervention only needs to be installed 100-200 in each arm.
Due to the ongoing shortage of samples, in January 2008 we added two recruitment streams in the city of Baltimore: Marketing of church services and canvassing of neighborhoods, where there was a fire truck with flashing lights that helped to get low
Income families in the fire department\'s target block.
To simplify the tasks of health educators, we assign all the participants of the day to one arm and, when the team gathers, randomly select each day by rolling the mold. Using pilot-
Tested procedures and materials, staff knocking and writing to arrange follow up
Check alarm operability and ask questions about alarm testing, fire fighting experience and fall experience.
Although follow
Up is not officially blinded and is generally executed by the installer;
They don\'t know how to distribute weapons. Ideally follow-
It won\'t start until all the batteries we initially installed are dead.
Otherwise, we cannot distinguish between situations where residents maintain alerts that have not yet required a new battery, or where an alarm cannot operate when the initial battery is dead.
Because there\'s no data on carbon.
The zinc battery life in the smoke alarm, we hung 100 alarms in PIRE\'s office.
Their battery failure rate caused us to wait at least 500 days before trying to return home to check for operability.
However, we returned 4 months in advance of the last 40 families and installed the alarm becauseup (and the grant)
Ended on September 2009.
Statistical analysis, we compared the treatment results with the current normal family using ordinary least squares linear regression to analyze whether the number of work alerts increased significantly from baseline to subsequent
Up and logistic regression analysis whether there is at least one working alarm after
The rise was obvious.
All analyses were performed in stata v.
11 people set out from the University of Texas station Stata Corp.
Before starting the data analysis, we decided to include the data from the pilot test house, because after the pilot, neither the method nor the material changed. testing.
The virtual variables in regression test whether the test data including the pilot test affect the results (
It\'s not).
Because of the results of two theories
It turns out that weapons based on health education are extremely similar and the number of samples is moderate, and most analyses combine the two weapons.
To process the missing revenue data, we coded the missing revenue to 0 and added a 0-1 variable with a value of 1 for those who reject the income issue.
Online Supplementary Appendix C describes all other model covariates and shows covariates-
Adjust the regression separately for each arm.
Regression analysis includes pre-intervention alarm status (
Is there an operational alarm or number of operational alarms)
As a predictor, any bias associated with functional differences between weapons at the time of registration can be adjusted.
Analysis whether there is time to follow up the test
In the home where the battery we originally installed is still not working, the affected results.
We initially tested the models of all population variables and all KAB fire and smoke alarm project variables.
We exclude most of the variables from the final model that lack statistical significance to minimize cases that are excluded due to lack of data and avoid multiple collinearity.
As a result, we installed smoke alarms in 436 homes.
We\'re done.
To 370 (
Installation of 85%).
Loss to follow-
Flow chart (figure 1)
Summarize the home where we installed the alarm and why it was lost-up. Loss to follow-
From 12 to 12. 8% to 17. 6%.
The most common cause of loss isup included: (1)
The house is empty behind. up or (2)
No one is open on five working days and/or weekends
Up attempt, residents did not respond to follow-Up eventsup letter. Time to follow-
Average of 24 months (table 2).
View this table: View the inline View pop-up table 2 alert status in the new tabDownload powerpoint figure 1 flow chart by arm and time period
Analysis based on cause and arm.
There are 370 follow-up
Up data, for the reasons described in figure 1, 46 were removed from the result analysis.
About 75% of the devices entered the analysis and the changes between the arms were small.
The most common reason to drop the box is that all the original batteries are still working (
This means that we follow up too early to tell if the residents will replace the battery and remain operational)
All work alarms are hard-wired alarms (
Usually in middle-class families)
, Or the structure has been converted to nonResidential use.
Almost all non
The battery of the work alarm is dead.
No one has used their alarm battery for other purposes.
It is easy for residents to admit that the battery was lost;
They accused the alarm of nuisance.
When we got home, in our sample, the turnover of the new occupants was very low, with only 16% of the 184 houses (57%)
Where we record this information
Of the households involved, 70% of the population is headed by women.
Of the respondents, 85% were black, 52% were over 60 years of age, and 68% were children who did not live with them (
See online supplement table C1).
Good sample-
Educated, 39% took some university courses, another 26% completed university studies, and 18% studied in graduate school.
Just over 20% Baltimore 8% people participated in the pilot test (
Counting from Figure 1).
Spanish materials were used in 13 educational conferences.
27% of families did not have a work alert when they entered school;
There is an average of 1 per family.
3 working smoke alarm (table 2).
There were fewer work alerts for internships and full-time education families compared to school hours, but the difference was not statistically significant.
Subsequent alarm status-
Follow-up intervention and effects
Before the adjustment of the covariates, the proportion of families without a work alarm was small, and the work alarm of each family in the fully educated family increased greatly, and was at the intermediate level in the practice family, at the lowest level in the current standard family.
Analysis included 10% follow-up
Ups where residents report alarm operability but do not allow re-alarm
Into their home (table 2).
These cases are equally common in weapons.
In sensitivity analysis, cases with self are not included
The operational impact of the report is minimal.
Alarm status pre-installed Linear Regression Control shows that there are a lot more alarms for hands-on or full-education homes using operating batteries compared to subsequent current normal familiesup (table 3).
The pre-education income group and KAB help explain the operability of the alarm.
In the regression of the control of KAB differences, the significance of the treatment effect is clearer.
The estimated increase adds a work alert on the follow-up average
Every three internship or all-teaching families (p=0. 01, 95% CI 0. 09 to 0. 56).
Practice adds an average of 0 work alerts per family. 37 (p=0. 027, 95% CI 0. 04 to 0. 70)
More than 0. 30 increase (p=0. 020, 95% CI 0. 05 to 0. 55)
Comprehensive education (
See Appendix C2 for online supplement).
Before adjusting the KAB reaction, the difference between the two arms increased little and the effect was statistically similar (p=0.
971, please refer to the online Supplementary Appendix table C4).
View this table: View inline View pop-up table 3 Follow-up work alert increase-
Pre-installation, logistic regression relative to alarm status (table 4)
Shows a significant increase in the probability of a treatment home having at least one work alert in follow-upup (OR=2. 77, p=0. 032, 95% CI 1. 09 to 7. 03). An OR of 2.
77 compared to the current standard recommendation 64% ((2. 77 to 1. 00)/2. 77)
The increase in the number of households with work alerts is due to an increase in education levels, and if the installers provide current standards, the rest will result. 19 At follow-
The return of upward, practical and fully educated families decreased by 26%
The adjustment probability of no work alarm is higher than the current normal home (9. 7% vs 13.
1%, percentage not tabulated).
A significant increase in total education (OR=2. 68, p=0. 051, 95% CI 0. 99 to 7. 24;
See Appendix C3 for online supplement).
The exercise arm experienced a similar increase, (
Small sample size)
It is slightly significant (OR=3. 00, p=0. 097, 95% CI 0. 82 to 11. 04).
Similarly, before adjusting the KAB response, the effects of the two weapons were statistically similar (p=0.
967, please refer to Table C5 of the online Supplementary Appendix).
View this table: View inline View popupTable4 added in the next home with at least one work alert-
Relative to the increase in the current NormAs Online Supplementary Appendix C report, changes in regression models tested their sensitivity to the choice of analysis.
The size and significance of the treatment effect have undergone minor changes.
Loss to follow-
Flow chart (figure 1)
Summarize the home where we installed the alarm and why it was lost-up. Loss to follow-
From 12 to 12. 8% to 17. 6%.
The most common cause of loss isup included: (1)
The house is empty behind. up or (2)
No one is open on five working days and/or weekends
Up attempt, residents did not respond to follow-Up eventsup letter. Time to follow-
Average of 24 months (table 2).
View this table: View the inline View pop-up table 2 alert status in the new tabDownload powerpoint figure 1 flow chart by arm and time period
Analysis based on cause and arm.
There are 370 follow-up
Up data, for the reasons described in figure 1, 46 were removed from the result analysis.
About 75% of the devices entered the analysis and the changes between the arms were small.
The most common reason to drop the box is that all the original batteries are still working (
This means that we follow up too early to tell if the residents will replace the battery and remain operational)
All work alarms are hard-wired alarms (
Usually in middle-class families)
, Or the structure has been converted to nonResidential use.
Almost all non
The battery of the work alarm is dead.
No one has used their alarm battery for other purposes.
It is easy for residents to admit that the battery was lost;
They accused the alarm of nuisance.
When we got home, in our sample, the turnover of the new occupants was very low, with only 16% of the 184 houses (57%)
Where we record this information
Of the households involved, 70% of the population is headed by women.
Of the respondents, 85% were black, 52% were over 60 years of age, and 68% were children who did not live with them (
See online supplement table C1).
Good sample-
Educated, 39% took some university courses, another 26% completed university studies, and 18% studied in graduate school.
Just over 20% Baltimore 8% people participated in the pilot test (
Counting from Figure 1).
Spanish materials were used in 13 educational conferences.
27% of families did not have a work alert when they entered school;
There is an average of 1 per family.
3 working smoke alarm (table 2).
There were fewer work alerts for internships and full-time education families compared to school hours, but the difference was not statistically significant.
Subsequent alarm status-
Follow-up intervention and effects
Before the adjustment of the covariates, the proportion of families without a work alarm was small, and the work alarm of each family in the fully educated family increased greatly, and was at the intermediate level in the practice family, at the lowest level in the current standard family.
Analysis included 10% follow-up
Ups where residents report alarm operability but do not allow re-alarm
Into their home (table 2).
These cases are equally common in weapons.
In sensitivity analysis, cases with self are not included
The operational impact of the report is minimal.
Alarm status pre-installed Linear Regression Control shows that there are a lot more alarms for hands-on or full-education homes using operating batteries compared to subsequent current normal familiesup (table 3).
The pre-education income group and KAB help explain the operability of the alarm.
In the regression of the control of KAB differences, the significance of the treatment effect is clearer.
The estimated increase adds a work alert on the follow-up average
Every three internship or all-teaching families (p=0. 01, 95% CI 0. 09 to 0. 56).
Practice adds an average of 0 work alerts per family. 37 (p=0. 027, 95% CI 0. 04 to 0. 70)
More than 0. 30 increase (p=0. 020, 95% CI 0. 05 to 0. 55)
Comprehensive education (
See Appendix C2 for online supplement).
Before adjusting the KAB reaction, the difference between the two arms increased little and the effect was statistically similar (p=0.
971, please refer to the online Supplementary Appendix table C4).
View this table: View inline View pop-up table 3 Follow-up work alert increase-
Pre-installation, logistic regression relative to alarm status (table 4)
Shows a significant increase in the probability of a treatment home having at least one work alert in follow-upup (OR=2. 77, p=0. 032, 95% CI 1. 09 to 7. 03). An OR of 2.
77 compared to the current standard recommendation 64% ((2. 77 to 1. 00)/2. 77)
The increase in the number of households with work alerts is due to an increase in education levels, and if the installers provide current standards, the rest will result. 19 At follow-
The return of upward, practical and fully educated families decreased by 26%
The adjustment probability of no work alarm is higher than the current normal home (9. 7% vs 13.
1%, percentage not tabulated).
A significant increase in total education (OR=2. 68, p=0. 051, 95% CI 0. 99 to 7. 24;
See Appendix C3 for online supplement).
The exercise arm experienced a similar increase, (
Small sample size)
It is slightly significant (OR=3. 00, p=0. 097, 95% CI 0. 82 to 11. 04).
Similarly, before adjusting the KAB response, the effects of the two weapons were statistically similar (p=0.
967, please refer to Table C5 of the online Supplementary Appendix).
View this table: View inline View popupTable4 added in the next home with at least one work alert-
Relative to the increase in the current NormAs Online Supplementary Appendix C report, changes in regression models tested their sensitivity to the choice of analysis.
The size and significance of the treatment effect have undergone minor changes.
This random trial shows that based on society-
Cognitive theory combining hands
In the practice of installing alarm batteries, regardless of whether there is a health belief model component or not, the operability of subsequent smoke alarms is greatly increasedup.
Adding practice to traditional fire department education has tripled the likelihood that a family will have a work alert after 2 years.
It also increases the number of work alerts per installed home by 17% (0. 32/1.
89 alerts per current standard home).
However, replacing education in the fire department with education built around the health belief model does not further improve the operability of the smoke alarm.
Educational achievements, including theory, have been rewarded.
At the 95% confidence level, the statistical significance based on practice is significant, and the coefficients in the statistical model are stable.
The education incorporated into practice takes roughly the same amount of time as the fire department or community group provides security information during alarm installation.
Alarm status the average of 24 months in the current standard comparison home after installation-12.
1% no work alarm-
Comparable to other programs in fire department education.
Families in rural Iowa have received similar alerts and batteries.
9% there is no functional alert on 12 months 20 and 11.
At 42 months, 6% people did not.
A study of owners in Seattle
Occupied households in low-income areas have installed a similar alert for each household; 22.
9% people did not work within 15 months.
5-13% high across five states
Risk home does not have a work alert within 6-12 months of installation, including 11% in Baltimore.
The study is not entirely generalized.
It basically reached a low.
Income of white and white families.
In a small number of Hispanic families, we often leave materials in Spanish, but there is very little education available --on practice.
It is therefore unclear whether this intervention will play a role in the concentrated Hispanic community.
We also tested the intervention without nearly any Asians.
Elderly white people or families with higher incomes.
This intervention may apply to smoke-Carbon monoxide combined alarms, but has nothing to do with those with higher cost lithium batteries that cannot be replaced.
The statistical power of recruitment issues is limited.
They also prevent installers from reaching some of the houses that fire department personnel can access.
This again limits the coverage of the findings.
However, it doesn\'t seem likely to be a problem because in Baltimore, there is no difference in the outcome of firefighters helping staff get into the home.
Another limitation is that we recorded the number of alarms working when arriving at each home, but did not record the number of jobs working when installation access was complete.
We counted the alarms we installed.
Wired alerts that are still running, but there is no record of the number of existing alerts we left in place and equipped with new batteries.
Therefore, it is not possible to analyze the percentage of work alerts that are still working two years later, and this result was measured in two studies published after we started the study.
13, 14 These studies also found that the alarm function varies from location to location, and the possibility of keeping the kitchen alarm function is small.
Because we did not record the alarm position in the follow-up, we were unable to adjust this deviationup.
If the question of KAB is asked to strengthen the normal education of the fire department to make it more effective, the estimated effect will be underestimated.
KAB response helps explain the operability of subsequent alarmsup (
And enrollment).
Therefore, it is puzzling that the comprehensive education aimed at changing KAB does not add more operability than typical education and hands-on operationson practice.
Unfortunately, our follow-up
Up interviews are usually not conducted with people we were educated at the time of installation.
So we can\'t ask those questions again below about our education KAB
Analysis of learning outcomes.
Conclusion in the case of existing educational time and training, despite its limitations, this study provides important evidence that the fire department and community groups can adopt the theory-
Based on health education.
The key seems to be the hand-on practice—
Test and replace the alarm battery using the flush button.
Installation programme staff have the ability to do this education with minimal training or compliance with scripts.
Although the study was not designed as an equivalent trial, we found that the complete model of health belief and
Practice and typical education produce similar results in statistics.
What is already known on this topic is that working smoke alarms can reduce the number of residential fire deaths by at least 48%.
However, it turned out to be difficult to keep the alarm running in the home.
Two years after installing the smoke alarm at low load
Family income, 1-
Fourth, the lack of working batteries.
Without increasing the installation time, the installer can increase the operability of the smoke alarm by handon practice.
After two years of installation, if residents have practiced installing batteries and hushing alarms, then there should be 16% alarms working and the probability of not having a working alarm should be reduced by 26%.
The study is not entirely generalized.
It basically reached a low.
Income of white and white families.
In a small number of Hispanic families, we often leave materials in Spanish, but there is very little education available --on practice.
It is therefore unclear whether this intervention will play a role in the concentrated Hispanic community.
We also tested the intervention without nearly any Asians.
Elderly white people or families with higher incomes.
This intervention may apply to smoke-Carbon monoxide combined alarms, but has nothing to do with those with higher cost lithium batteries that cannot be replaced.
The statistical power of recruitment issues is limited.
They also prevent installers from reaching some of the houses that fire department personnel can access.
This again limits the coverage of the findings.
However, it doesn\'t seem likely to be a problem because in Baltimore, there is no difference in the outcome of firefighters helping staff get into the home.
Another limitation is that we recorded the number of alarms working when arriving at each home, but did not record the number of jobs working when installation access was complete.
We counted the alarms we installed.
Wired alerts that are still running, but there is no record of the number of existing alerts we left in place and equipped with new batteries.
Therefore, it is not possible to analyze the percentage of work alerts that are still working two years later, and this result was measured in two studies published after we started the study.
13, 14 These studies also found that the alarm function varies from location to location, and the possibility of keeping the kitchen alarm function is small.
Because we did not record the alarm position in the follow-up, we were unable to adjust this deviationup.
If the question of KAB is asked to strengthen the normal education of the fire department to make it more effective, the estimated effect will be underestimated.
KAB response helps explain the operability of subsequent alarmsup (
And enrollment).
Therefore, it is puzzling that the comprehensive education aimed at changing KAB does not add more operability than typical education and hands-on operationson practice.
Unfortunately, our follow-up
Up interviews are usually not conducted with people we were educated at the time of installation.
So we can\'t ask those questions again below about our education KAB
Analysis of learning outcomes.
Conclusion in the case of existing educational time and training, despite its limitations, this study provides important evidence that the fire department and community groups can adopt the theory-
Based on health education.
The key seems to be the hand-on practice—
Test and replace the alarm battery using the flush button.
Installation programme staff have the ability to do this education with minimal training or compliance with scripts.
Although the study was not designed as an equivalent trial, we found that the complete model of health belief and
Practice and typical education produce similar results in statistics.
What is already known on this topic is that working smoke alarms can reduce the number of residential fire deaths by at least 48%.
However, it turned out to be difficult to keep the alarm running in the home.
Two years after installing the smoke alarm at low load
Family income, 1-
Fourth, the lack of working batteries.
Without increasing the installation time, the installer can increase the operability of the smoke alarm by handon practice.
After two years of installation, if residents have practiced installing batteries and hushing alarms, then there should be 16% alarms working and the probability of not having a working alarm should be reduced by 26%.
Thanks to all the authors for the help of Bishop Dennis mcden and Bishop Mary Allen Firth as well as the priests of the Baltimore general parish, Baltimore and Prince George County fire department for the recruitment of this project.
Almost all of the work on the ground was done by health educators Tariq Chapman and Tabasha McNeil, as well as the installers Richard Bowen and Jeni Kanadi.
Leicester Becker and the smaller Dolin branch and Dennis Nguyen played a key role in the design and implementation of the instrument and on-site work agreement for the project.
Reference
Smoke alarms in the United StatesS. home fires.
Ma kunxi: National Fire Protection Association, 2009.
Runyan CW, Bangdiwala SI, Linzer MA, etc.
Risk factors for fatal residential fires.
N. Engl J. Med 1992; 37:859–63.
OpenUrl swiistre GR, Mai Ma, Moore BJ, etc.
Prevention of casualties from house fires: assessment of community outcomes
Installation scheme based on smoke alarm. Inj Prev 2014; 20:97–102.
OpenUrlAbstract/free full Text cucta VM, Frattaroli S, Bergen, etc.
Assessment of community fire safety interventions in the United States: a review of current literature.
2006 Community Health; 31:176–97.
If, Kresnow M-of OpenUrlCrossRefPubMedWeb Science GmbH Ballesteros-J.
Popular US residential smoke alarms and fire escape plansS.
: Results of the second injury control and risk investigation (ICARIS-2).
Representative of Public Health 2007; 122:224–31.
The science of OpenUrlPubMedWeb poepoehlman J, Flanigan T, Brown D, etc.
Evaluation of smoke alarm installation and fire safety education effect (SAIFE)
Project: final report.
Triangle Research Park, North Carolina, 2010.
Mickalide A, Validzic.
Alarm maintenance in low smokeIncome families.
No. 1999 public health; 89:1584–85.
Science Jones AR, Thompson CJ, Davis MK\'s openurlcrosspubpubmedweb.
Ownership and installation of smoke alarms: Comparison of rural and suburban communities in Georgia.
2001 community health; 26:307–29.
The School of Science, AE, Mike K, et al. openurlcrosspubpubmedweb.
Differences in popularity of American smoke alarmsS.
Family: conclusions drawn from published case studies.
J. Safety Res 2011; 42:409–13.
Openurlcrosspubmedweb, Goss CW, Higgins JP, science center.
Intervention to promote ownership and function of smoke alarms.
2010 Cochrane Library: 10. 1002/14651858. CD002246.
Chen L, Kelun AC, McDonald\'s EM.
Self-efficacy
Family safety measures were reported. Inj Prev 2003; 9:73–5.
Thompson CJ, Jones AR, Davis MK, etc.
Can the smoke alarm still work normally after one year of installation? A follow-up of the Get-
Vigilant campaign
2004 community health; 29:171–81.
Scientific openurlcross pubpubmedweb mueller Miller BA, West man EA, Alter H, et al.
A randomized controlled trial of ionization and photoelectric smoke alarm functions. Inj Prev 2008; 14:80–6.
OpenUrlAbstract/free full Text skip Peek-
Asa C, Yang J, Hamann C, etc.
Smoke alarms and battery functions 42 months after installation: a random trial.
J Prev Med 2010 in the morning; 39:368–71.
Gilon AC, sleet DA, Dillon AC, Dillon AC.
Chapter 13, house fire and other accidental injuries.
In: Gielen AC, sleet DA, DiClemente, eds.
Prevention of injuries and violence.
Theory, method and application of behavioral science.
San Francisco, California: Josie-
Bass 2006: 274-96.
Jackson ML, Martin MW.
Committed to eliminating the number of deaths from residential fires: smoke alarm installation and fire safety education in the Centers for Disease Control and Prevention (SAIFE)program.
Rehabilitation of Burn Care 2005; 26:434–9.
Scientific openurlcross pubpubmedweb gigielen AC, McKenzie LB, McDonald\'s EM, etc.
Use of computer kiosks for child safety: results of a randomized controlled trial in urban children\'s emergency department.
Pediatric 2007120:330–9.
OpenUrlAbstract/free full Text ↵ Glanz K, twisted hole BK, Lewis, FM, eds.
Health Behavior and Health Education: Theory, Research and Practice. 3rd edn.
San Francisco, California: Josie-Bass, 2002.
Alimhaddix communication, Teutsch SM, schaeffpa, Dunet DORothenberg RB, and Korean RA.
Appendix J: attribution measures.
In: Haddix AC, Teutsch SM, Shaffer PA, Dunet DO, eds.
Preventive effects: guidelines for decision-making analysis and economic assessment.
New York: Oxford University Press, 1996: 193-202. ↵Yang J, Peek-
Asa C, member Jones, etc.
Smoke alarms classified by type and battery life in rural households: a randomized controlled trial.
J Prev Med 2008 in the morning; 35:20–4.
Scientific openurlcrosspubpubmedweb harvey P, Aitken M, Ryan G, et al.
Strategies to increase the use of smoke alarms at high levelsRisk households.
2004 community health; 29:375–85.
All authors contributed to the results and revisions of the manuscript and approved the final manuscript submitted.
TRM is the lead researcher, leading the work throughout the process, doing data cleaning and statistical analysis, and drafting the manuscript.
He is the guarantor.
GB and ACG lead the development of educational scripts and instruments.
GB guides the manuscript with CDC permission.
MFB developed the concept and basic design of this study.
Statistical analysis was performed by SB, along with leading data cleaning and connection of measuring instruments.
The MSS oversees a lot of field work.
Part of the funding for this study supports grant under-17/CCU323334 from the National Center for Injury Prevention and Control, the United States Department of Health and Human Services, the Centers for Disease Control and Prevention (CDC).
The findings and conclusions of the report are the findings and conclusions of the authors and do not necessarily represent the official position of the CDC.
There is no competitive interest.
Ethics approval IRB of the Pacific Research and Evaluation Institute.
Uncommissioned source and peer review;
External peer review.
Data sharing statement with open-
The end response of the deletion is available from the corresponding author.
There is no consent to data sharing from participants, but the data presented is anonymous and the risk of identification is minimal.