ORIGINAL ARTICLE

On the Road With Injury Prevention—An Analysis of the Efficacyof a Mobile Injury Prevention Exhibit

Mariann Manno, MD, Allison Rook, MPS, EdM, Amanda Yano-Litwin, BS, Louise Maranda, MVZ, MSc, PhD,Andrew Burr, DO, and Michael Hirsh, MD

Background: To assess the effectiveness of a mobile injury preventionvehicle (mobile safety street [MSS]) with a hands-on curriculum on instruc-tion and retention of safety knowledge compared with traditional classroomsafety curriculum among grade 5 elementary school children.Methods: Grade 5 students (n � 1,692) were asked to participate in thestudy as either the intervention group (MSS experience) or the comparisongroup (traditional classroom safety curriculum). Each student in the inter-vention group was asked to complete a series of three surveys. The firstsurvey was given before the MSS visit (Fall 2009), the second immediatelyfollowing the MSS visit (Fall 2009), and a third given 6 months after theMSS visit (Spring 2010) to measure knowledge retention. Students in thecomparison group were asked to complete two surveys. The first survey wasgiven at the same time as the intervention group (Fall 2009) and the secondwas given after the completion of the traditional classroom safety curriculum(Spring 2010).Results: Students scored on average 5.67 of 10 (5.56–5.80) before anysafety instruction was given. After MSS instruction, mean scores showed asignificant increase to 7.43 of 10 (7.16–7.71). Such increase was stillmeasurable 6 months after the intervention 7.34 (7.04–7.66). The compari-son group saw a significant increase in their mean scores 6.48 (6.10–6.89),but the increase was much smaller than the intervention group.Conclusions: Community-based injury prevention programs are essential toreducing preventable injury and deaths from trauma. This study demonstratesthat a hands-on program is more effective than traditional methods forproviding safety knowledge.Key Words: Safety, Education, Interactive.

(J Trauma. 2011;71: S505–S510)

Unintentional injury remains the leading cause of deathamong children older than 1 year, adolescents, and

young adults.1 The cause of injury varies with age. Amongschool-aged children, motor vehicle crashes, drowning, and

burns account for most fatal injuries; falls and bicycle andpedestrian crashes are responsible for most nonfatal injuries.2In addition, children who live in poverty are at greater risk forinjury and live in environments where they are simultane-ously exposed to multiple safety hazards.3 There is evidencethat risky behaviors increase with age in school-aged childrenand that children who have a history of injuries are morelikely to report risky behaviors.4

Schools are a constant and important part of children’slives and an obvious venue for teaching safety behaviors.School-based interactive injury prevention programs havereported positive results in improved knowledge among stu-dents in areas of bike and pedestrian safety, falls, playgroundsafety, fire safety, and poisoning when compared with tradi-tional methods of instruction.5–7 This is consistent with liter-ature that has shown shorter periods of time spent in moreactive learner-to-learner interactions are more productivethan longer periods of more passive teacher-driven sessions.8

Successful injury prevention education has occurredthrough multifaceted approaches that include hands-on learn-ing, visual aids, creative writing, and verbal interaction.Inquiry-based instruction allows children to develop criticalreasoning skills, superior to declarative methods where chil-dren are simply provided with the correct answer.9 Onemobile safety center reported that their mobile nature offeredthem the ability to provide urban families greater access topersonalized, low-cost injury prevention education.10

Our mobile safety street (MSS) is an injury preventionprogram that seeks to incorporate proven effective strategies toenhance the learning of safety behaviors of fifth-grade students.It is a mobile exhibit that travels to the schools, maximizingefficiency in the students’ school day and creating a familiarenvironment in which students learn. The MSS curriculum hasbeen developed to parallel the safety and health curriculum thatis required by the Department of Education in Massachusetts. Ituses a hands-on, learner-centered method of instruction bestsuited to the learning style of grade school boys and girls. Thepurpose of this project is to test the effectiveness of the MSScurriculum and compare it with a traditional classroom safetycurriculum among grade 5 elementary school children.

MATERIALS AND METHODS

Curriculum DesignA comprehensive elementary school injury prevention

curriculum was developed and piloted by the Injury Free

Submitted for publication January 29, 2011.Accepted for publication September 28, 2011.Copyright © 2011 by Lippincott Williams & WilkinsFrom the Division of Pediatric Emergency Medicine/Department of Pediatrics

(M.M.), Division of Pediatric Surgery/Department of Surgery (A.R., A.B.,M.H.), Division of Clinical Research/Department of Pediatrics (L.M.),UMassMemorial Children’s Medical Center, Worcester, Massachusetts.

Supported by Wal-Mart, Ford Motor Company, Department of Public Health ofthe City of Worcester, Hoche Scofield Foundation.

Presented at Forging New Frontiers: Making Communities Safe for Children &Their Families Injury Free Coalition for Kids 15th Annual Conference,November 12–14, 2010, Chicago, Illinois.

Address for reprints: Mariann Manno, MD, 55 Lake Avenue North, Worcester,MA 01655; email: mannom@gmail.com.

DOI: 10.1097/TA.0b013e31823a49bc

The Journal of TRAUMA® Injury, Infection, and Critical Care • Volume 71, Number 5, November Supplement 2, 2011 S505

Coalition for Kids (IFCK) in Worcester, MA. The MSS curric-ulum was adapted from the standard classroom health curricu-lum used by Worcester Public Schools, the Michigan Model forComprehensive School Health Education. The same safety in-formation was taught in two different ways—at MSS and in theclassroom. This study was designed to compare the effective-ness of the hands-on MSS curriculum in improving and retain-ing safety knowledge with what is achieved through the standardclassroom health education. The MSS curriculum uses applied

learning techniques including problem-solving, role-playing,and hands-on activities. MSS is a 40-foot mobile safety unit thatincludes four separate learning stations within the school bus andsafety trailer (Fig. 1). MSS travels to each school and is set upon school property lending to the realistic nature of thestudents’ experience. Instruction at each station includes10-minute lessons that focus on specific pedestrian, home,and bus passenger safety using visual simulations andhands-on activities (Fig. 2).

Figure 1. Learning stations at MSS. (A) Urban pedestrian safety, (B) suburban pedestrian safety, (C) home safety, and (D)bus safety.

Manno et al. The Journal of TRAUMA® Injury, Infection, and Critical Care • Volume 71, Number 5, November Supplement 2, 2011

© 2011 Lippincott Williams & WilkinsS506

Survey DesignWe developed a 14-question survey that focused on

important safety behaviors emphasized in both the school-based and MSS curricula. The evaluation tool from theMichigan Model is available in open-ended question formatonly. We adapted this tool and created multiple choice ques-tions. For example, the Michigan Model evaluation regardingcalling 911 asks the student to describe how to make anemergency phone call. A correct student response would beto dial 911, state his or her name and address, and tell thenature of the emergency. We adapted this on the survey asmultiple choice question 5 (Fig. 3).

Study DesignThis study followed a randomized block design. The 33

elementary schools in the Worcester Public School systemwere placed in homogenous blocks based on socioeconomicand demographic data reported by each school. The percent-age of enrolled students who are eligible for governmentsubsidized lunch was the primary criterion and served as aproxy for household income. Using this criterion, the 33elementary schools were matched into 16 pairs, excluding 1school because the reported demographic and socioeconomicdata were widely divergent from the other schools. A dis-crepancy of �10% in the criterion was determined as an

acceptable variance. Only one matched pair exceeded the10% variance in the category of percentage of studentseligible for government subsided lunch (10.5% variance).One of the two schools in each of the 16 matched pairingswas randomly assigned to participate in the MSS educationalcurriculum.

Intervention and Comparison GroupsAll grade 5 students (n � 1,692) were asked to partic-

ipate in the study as either the intervention group (whoreceived the MSS experience) or the comparison group (whoreceived traditional classroom safety education). Before re-ceiving MSS, each child in the intervention group took awritten safety knowledge survey. During the next healthclass, each student participated in the MSS 30-minute to40-minute curriculum. Students completed a written safetyknowledge survey immediately after participating in MSS toassess changes in their safety knowledge. Six months later,students completed a follow-up survey to evaluate retentionof safety knowledge following their MSS experience. Stu-dents in the comparison group completed a baseline survey atthe same time as the intervention group. They participatedin their regularly scheduled health class and safety curricu-lum. They took a follow-up survey after the completion of thetraditional classroom safety education curriculum. After the

Figure 2. Safety knowledge and behaviors covered at MSS.

The Journal of TRAUMA® Injury, Infection, and Critical Care • Volume 71, Number 5, November Supplement 2, 2011 Efficacy of a Mobile InjuryPrevention Exhibit

© 2011 Lippincott Williams & Wilkins S507

data collection period was completed, the comparison groupschools received the MSS educational curriculum so studentsattending comparison group schools were not deprived ofparticipating in the MSS program.

Data AnalysisDescriptive statistics were calculated to establish a

baseline evaluation of intervention and comparison groups.Variables include gender, socioeconomic status as deter-mined by eligibility for free and reduced lunch, and age foreach of the intervention and comparison groups. Also re-corded were the total number of students who completed eachsurvey and their mean score, with its associated 95% confi-dence interval.

A two-way analysis of variance (ANOVA) was thenperformed to compare means (grade of 10) within groups,adjusting for gender and age. Log transformation was per-formed to account for the lack of normality for the measuredoutcome. Success rates for each question were comparedseparately for boys and girls and between tests using a �2

analysis. Ninety-five percent confidence intervals were con-structed around differences between means and percentages.

RESULTSInspection of baseline values of gender, age, and so-

cioeconomic status showed that the intervention and compar-

ison groups were comparable, with the presence or absence ofthe MSS intervention as the only significant difference be-tween groups (Table 1).

Table 2 reports the number of students present and theirmean test scores within each group studied. Test scores arethe average number of right answers of 10 questions adjustedfor age, exponentiated from results obtained with one-wayANOVA done on log-transformed data. As can be seen,children scored on average 5.67 of 10 (5.56–5.80) before anysafety instruction was given. Once the MSS program wasperformed, mean scores showed a significant increase to 7.43of 10 (7.16–7.71). Such increase was still measurable 6months after the intervention. Children who were submittedto the regular school curriculum also saw a significant in-crease in their mean scores, but the increase was muchsmaller than for the MSS group.

In Table 3, mean test scores are stratified over genderand adjusted for age. Means are exponentiated from resultsobtained with a two-way ANOVA done on log-transformeddata. Again, results show that the increase in test scores islarger for the MSS intervention than for the regular schoolcurriculum and that boys seem to perform better than girlswith hands-on experiences.

Individual question analysis indicates that questions 1,5, 11, 13, and 14 show no measurable change in responses

Figure 3. MSS survey.

Manno et al. The Journal of TRAUMA® Injury, Infection, and Critical Care • Volume 71, Number 5, November Supplement 2, 2011

© 2011 Lippincott Williams & WilkinsS508

with and without intervention or before and after interven-tion. Girls’ answers to question 4 show a significant increasein correct responses between baseline and intervention (per-cent change � 36.2%, p � 0.001), a slight decrease betweenintervention and the testing done 6 months later (percentchange � 4.0%, p � 0.098), and all results are significantlyhigher than for the comparison school curriculum program(percent change � 16.4% between baseline and school pro-gram, p � 0.004; 21.2% between the intervention and theschool curriculum, p � 0.001; and 15.8% between the 6months postintervention evaluation and the school curricu-lum, p � 0.002). Boys’ answers to question 4 followed thesame pattern. Answers to questions 2, 3, and 12 also show thesame trend for both boys and girls.

Girls’ responses to question 3 show no change incorrect responses between baseline and intervention (percentchange � 5.3%, p � 0.157), a slight increase between

intervention and the testing done 6 months later (percentchange � 3.4%, p � 0.465), but all results remain signifi-cantly higher than for the nonintervention school curriculumprogram (percent change � 11.4% between baseline andschool program, p � 0.041; 16.7% between the interventionand the school curriculum, p � 0.007; and 20.1% between the6 months postintervention evaluation and the school curric-ulum, p � 0.001).

Boys’ responses to question 7 are similar to the girls’responses to question 3 stated above, except for the percentchange between baseline and intervention which seems to benonexistent (percent change � 0.6%, p � 0.866).

DISCUSSIONInjury is the leading cause of death for children in the

United States. Unintentional childhood injuries dispropor-tionally affect families challenged by poverty.1 The goal ofthis work was to examine the effect of an interactive methodof instruction on a mobile safety vehicle in teaching safetybehaviors to fifth-grade students and compare these resultswith a traditional health education curriculum. Our findingssupport previous studies which show that the use of a hands-on, multifaceted curriculum, like the mobile safety vehicle,had a positive effect on the retention of injury preventionknowledge within the community.8,11,12

Mean test scores illustrated that the MSS approachmade an immediate and delayed posttraining improvement inthe intervention group over the comparison group. Thesefindings could be attributed to the fact that the MSS curric-ulum uses a more interactive, hands-on approach to the safetycurriculum resulting in more knowledge gained and betterretention of that knowledge by the MSS students.

Although girls scored higher than boys in all threesurveys done with the intervention group, boys showed agreater improvement in scores than girls. This could beexplained by the fact that boys in our intervention cohort age(mean, 10.09 years) have been shown to benefit from learningactivities that require some physical activity, such as thosethey experienced during MSS.13

The practical realities of partnering with a large publicschool system presented several challenges. These includedthe following: (1) relying on public school health educatorsoutside of our injury prevention group; (2) instruction timeavailable to MSS in the course of the school day; (3) teacher

TABLE 1. Baseline Comparison of Studied Groups

VariableIntervention Group

n�1415†Comparison Group

n�155 P*

Gender (% males) 50.5 44.5 0.10

Age of males

(yr, mean, � 95% CI) 10.09 �10.00–10.18� 10.26 �10.18–10.34�

Age of females

(yr, mean, � 95% CI) 10.48 �10.37–10.59� 10.52 �10.38–10.66�

Social economic status, % eligible for government subsidy 60.8 65.7 0.023

CI, confidence interval.* From �2 test to compare 2 proportions.† Thirty-four students from the intervention group and one student from the non-intervention group did not indicate their gender.

TABLE 2. Test Scores Adjusted for Age

Test Group N Mean Test Score*

At baseline 1354 5.67 �5.56–5.80�

After MSS intervention 425 7.43 �7.16–7.71�

6 months after MSSintervention

336 7.34 �7.04–7.66�

School curriculum(comparison)

154 6.48 �6.10–6.89�

* Mean � 95% Confidence Interval.

TABLE 3. Comparison of Test Results (the Increasein Score)

Test Group Comparisons Males Females

Between baseline and MSSintervention

1.28 �1.21–1.36� 1.30 �1.23–1.37�

Between baseline and 6 mo afterMSS intervention

1.22 �1.13–1.31� 1.26 �1.19–1.34�

Between baseline and schoolcurriculum

1.04 �0.95–1.16� 0.86 �0.80–0.93�

Between MSS intervention andschool curriculum(comparison)

1.23 �1.10–1.34� 1.12 �1.02–1.22�

* Mean difference � 95% Confidence Interval.

The Journal of TRAUMA® Injury, Infection, and Critical Care • Volume 71, Number 5, November Supplement 2, 2011 Efficacy of a Mobile InjuryPrevention Exhibit

© 2011 Lippincott Williams & Wilkins S509

and student reassignment and student absenteeism; and (4)traveling to public schools. Health educators taught both thetraditional classroom curriculum and assisted in teaching inMSS. This may have altered the delivery of the school-based curriculum and altered scores in the comparisongroup. Detailed in-services regarding survey administra-tion were conducted by the IFCK Worcester staff for allhealth teachers before the initiation of the study. However,administration of surveys was not observed by IFCKWorcester staff, so individual characteristics of teacherscould have impacted the method of administration of somesurveys and the survey results.

MSS functioned on the schools’ regular schedules; thisimposed limitations on instructional time. Educators did notalways have time to address all topics completely. Questions6, 8, 9, and 10 are related to topics that were not consistentlytaught. We eliminated these questions from our analysis andanalyzed the remaining 10 questions. Finally, teacher andstudent reassignments occur throughout the school year. Sub-stitute teachers assigned midyear to a class may not haveparticipated in in-service training about MSS and surveyadministration. Similarly, it is unknown how many studentsswitched between comparison and intervention groups whenthey were transferred from one school to another throughoutthe course of the study. In addition, outside factors such asillness or student absenteeism could not guarantee that eachstudent participated in each phase of the intervention orcomparison group.

MSS is an outdoor exhibit that is limited by weatherconditions, restricting its availability to the early fall andspring months. MSS was not able to visit 5 (of 16) schools inthe intervention group because there was no safe place forMSS to park and set up (two schools) and cancellationsbecause of inclement weather (three schools). This contrib-uted to the skew in socioeconomic status that was observedwhen we analyzed this characteristic of the intervention andcomparison groups (60.8% vs. 65.7%).

CONCLUSIONMSS is an innovative, interactive, and effective ap-

proach to teaching safety behaviors. The positive resultsseen in this study suggest that mobile safety units such asMSS are an important way to impart knowledge aboutsafety behaviors to grade school students living in low-income urban environments where children are at thegreatest risk of injury.

REFERENCES1. Available at: http://www.cdc.gov/Injury/publications/FactBook/Introduction-

2006-a.pdf.2. Schnitzer PG. Prevention of unintentional childhood injuries. Am Fam

Physi. 2006;74:1864–1869.3. Chaudhari VP, Srivastava RK, Moitra M, Desai VK. Risk of domestic

accidents among under five children. Internet J Fam Pract. 2009;7:1–22.4. Morrongiello BA, Cusimano M, Orr E, et al. School-age children’s

safety attitudes, cognitions, knowledge, and injury experiences: how dothese relate to their safety practices? Inj Prev. 2008;14:176–179.

5. Kendrick D, Groom L, Stewart J, Watson M, Mulvaney C, Casterton R.“Risk Watch”: cluster randomized controlled trial evaluating an injuryprevention program. Inj Prev. 2007;13:93–98.

6. Greene A, Barnett P, Crossen J, Sexton G, Ruzicka P, Neuwelt E.Evaluation of the THINK FIRST for KIDS injury prevention curriculumfor primary students. Inj Prev. 2002;8:257–258.

7. Hotz G, Kennedy A, Lutfi K, Cohn SM. Preventing pediatric pedestrianinjuries. J Trauma. 2009;66:1492–1499.

8. Haidet P, Morgan RO, O’Malleey K, Moran BJ, Richards BF. Acontrolled trial of active versus passive learning strategies in a largegroup setting. Adv Health Sci Edu. 2004;9:15–27.

9. Lawson AE. What can developmental theory contribute to elementaryscience instruction. J Elemen Science Edu. 2008;20:1–14.

10. Gielen AC, Mcdonald E, Frattaroli S, et al; CARES (Children ARESafe) Mobile Safety Center Partnership. If you build it, will theycome? Using a mobile safety centre to disseminate safety informa-tion and products to low-income urban families. Inj Prev. 2009;15:95–99.

11. Towner E, Dowswell T. Community-based childhood injury preventioninterventions: what works? Health Promot Internat. 2002;17:273–284.

12. Azeredo R, Stephens-Stidham S. Design and implementation of injuryprevention curricula for elementary schools: lessons learned. Inj Prev.2003;9:274–278.

13. Bonomo V. Gender matters in elementary education: research-basedstrategies to meet the distinctive learning needs of boys and girls. EducHoriz. 2010;88:257–264.

Manno et al. The Journal of TRAUMA® Injury, Infection, and Critical Care • Volume 71, Number 5, November Supplement 2, 2011

© 2011 Lippincott Williams & WilkinsS510