Measuring the effectiveness of a community-sponsored DWI intervention for teens.

Abstract: This study measured teen alcohol consumption, attitudes, and behaviors towards DWI. The intervention high school received the pre- and post-survey, with a another high school as the control group. In the previous month, students reported that 50% drank alcohol, 37% drank heavily, 33% rode with a DWI driver, and 16% reported DWI. The intervention program had a favorable impact on attitudes but not behavior. Survey and focus group data suggested there would be no sustained behavioral change without combining the intervention with stronger DWI law enforcement, community support, and educational programs.

**********

Teen use of alcohol is a significant community health problem. Among the leading causes of unintentional morbidity and mortality for youth in the United States, the use of alcohol ranks as one of the six major categories of risk behaviors (Kann, et al., 2000; and Meeks, Heir, & Page, 1996). When adolescents combine alcohol consumption with driving a motor vehicle or riding with an intoxicated driver, the consequences are often tragic. The principal cause of death for teens and young adults is alcohol-related motor vehicle accidents (Meeks, Heit, & Page, 1996). Results from the Youth Risk Behavior Surveillance Survey (YRBSS) in the United States for 1999 indicate that "during the 30 days preceding the survey 33.1% of students nationwide had ridden one or more times with a driver who had been drinking alcohol" (Kann, et al., 2000, p. 7). Also, in the YRBSS, 13.1% of teens nationwide admitted to having driven a vehicle one or more times under the influence of alcohol during the previous month (Kann, et al., 2000).

Drug prevention programs embrace a broad range of philosophies, from "macrolevel environmental approaches, such as national policy strategies (e.g., incentives for states to raise the legal drinking age to 21) to microlevel programs at the school or individual level (e.g., school-based prevention curriculum)" (Komro, Perry, Veblen-Mortenson, Williams, & Roel, 1999, p. 202). To be most successful, it is believed that drug and alcohol prevention efforts should combine both the macro- and microlevel approaches (Flay & Petraitis, 1998). Educational intervention programs aimed at school-aged students are the most common kinds of preventive approach in current use (Yuen & Pardeck, 1998). School-based drug prevention programs can be divided into three categories: knowledge programs, affective programs, and social influences programs (Ringwalt, Greene, Ennett, & Iachan, 1994). The knowledge-based prevention programs when used alone do not appear to prevent substance abuse and have a limited impact on behavior change (Coombs & Ziedonis, 1995). Educational guidelines support the theory that students should have the opportunity to personalize and assimilate learning in order to understand its effect on their lives (Kolaya & Grimes-Smith, 1999). The "Every 15 Minutes" intervention program measured in this study combines knowledge, affective response, and social influence to prevent teen drinking and driving.

PURPOSE

The Greene County DWI (Driving While Intoxicated) Task Force wanted to determine the effectiveness of a teen DWI intervention program known as "Every 15 Minutes" before deciding to sponsor it for a third time in 2000. Therefore, the purpose of this study was to measure the pre- and post-intervention prevalence of alcohol consumption, student attitudes, and behaviors towards drinking alcohol and driving at Kickapoo High School in Springfield, MO. The Greene County DWI Task Force is a community-based, volunteer organization in Springfield, MO. Its members include representatives from the local school PTSAs (Parent, Teacher, and Student Associations), school district administration, the Mayor's office, Springfield Police Department, Greene County Sheriff's Department, Missouri Highway Patrol, and community volunteers who meet to implement activities designed to reduce drinking and driving in Greene County, Missouri. The task force first piloted the "Every 15 Minutes," a teen drinking and driving intervention program at Glendale High School in May 1998, and planned to run the intervention again in May of 1999 at Kickapoo High School in Springfield.

METHODS

SUBJECTS

Springfield, Missouri is a small urban community in southwest Missouri with a population of 140,000. Kickapoo High School (KHS) is one of the five public 4-year high schools. The intervention group at KHS consisted of 1650 students in grades nine through 12. Parkview High School (PHS) is a 4-year high school in Springfield with an enrollment of 1,400 students and served as the control group for this study. Parkview High School did not receive the "Every 15 Minutes" intervention program.

STUDY DESIGN

This study was a pre-test/post-test design with the intervention program pesented at KHS, and PHS used as a control group. The study was based to a large extent, on information provided by a CDC publication on measuring the effectiveness of intervention programs (Thompson & McClintock, 1998). Permission to conduct this study was granted from the school board of the Springfield Public School District. The pre-intervention survey was conducted at KHS in March 1999. The teen alcohol intervention, "Every 15 Minutes," was presented within 14 days, and the post-intervention survey was administered 30 days later. The control survey was conducted at PHS within 10 days of the post-survey at KHS. It should be noted that the KHS post-survey and PHS survey were administered after each high school's prom.

SURVEY INSTRUMENT

The pre-survey and control survey instrument consisted of the same 15 questions. The post-survey contained nine questions from the pre-survey. Survey questions included demographic information (age, and gender), previous school drug education programs, as well as attitudes and behaviors regarding drinking and driving. Seven of the survey questions were taken directly from the 1997 CDC Youth Risk Behavior Surveillance Survey (1999). This was done to both confirm the validity of the responses and to allow some generalization of the data to other high schools within the community (see Table 1). The seven questions from the CDC survey dealing with attitudes and behaviors toward drinking and driving were the same in the KHS pre-survey, KHS post-survey, and PHS control survey. Significance testing for the effect of the intervention is based on five of these questions and reported as significant only if the KHS pre-survey results are different than both the KHS post-survey and PHS control. A panel of physicians, health educators, teachers, and social workers developed the pre-survey questions not taken from the YRBSS and the additional post-survey items. A pilot study was conducted to improve the format and item clarity of the pre-survey/control survey questions. The students in the pilot study were similar to the participants in this investigation. Students' suggestions were used to re-word survey questions for comprehension and to allow additional responses. The post-survey contained 20 questions: nine that matched the control and pre-survey, three concerning the "Every 15 Minutes" docudrama and assembly, and eight questions regarding the effect of the intervention on friends and discussions about DWI with friends after the intervention.

PROCEDURE

The prevalence of alcohol consumption, student attitudes and behaviors towards drinking alcohol and driving were measured by self-report. In order to ensure the validity of self-reported data, methods recommended by Rouse, Kozel, and Richards (1985) were followed to gather data through anonymous inquiry, using brief, easily understood directions to complete the survey. The use of self-report provided estimates of the students' most current behaviors and attitudes.

Preparation for administration of the pre- and post-survey at KHS included parental notification, teacher education regarding the rationale for the surveys, and teacher instructions for administration details. In addition, pre- and post-survey instructions were given to students to ensure that their participation was strictly voluntary.

The surveys were answered on scan cards that could be read by a Response Technology scan card reader. Statistical analysis was done using Minitab version 12.2, and all statistical tests were performed using 0.05 as the minimum level of significance. Descriptive statistics were reported for questions dealing with prevalence of drinking behavior and program effectiveness. Comparison of teen attitudes and self-reported behavior from the pre- and post-intervention surveys, and the PHS control survey was completed using Chi-square analysis.

INTERVENTION

The "Every 15 Minutes" teen alcohol intervention program originated in Chico, California and consisted of a staged drunk-driving accident scene (docudrama) viewed by students on the first day, an overnight retreat for selected students, and an assembly for the whole school on the following morning. At the assembly the selected students share the emotional impact of the retreat with the student body. Operational details, video, and written materials describing the program and survey instruments are available from the author on request. Additionally, the Missouri Department of Transportation distributes the video produced by the Kickapoo High School Media department to interested schools. This video describes the "Every 15 Minutes" program and how to produce this drinking and driving intervention.

DATA ANALYSIS

The data was examined for difference in responses based on gender, age groups, and self-reported use of alcohol. Comparison questions that allowed four or more categories of responses were evaluated by combining classes of responses into favorable versus unfavorable to avoid finding statistically significant but unimportant difference by Chi-square tests. For example, the question "If a friend has been drinking and is going to drive ..." allowed four responses. Responses of "I would still ride in a motor vehicle with him/her" and "I would ride with my friend unless he/she was really drunk" were combined into the unfavorable category. Responses of "I would not ride in a motor vehicle with him/her" and "I would take the keys from him/her" were combined into a favorable category. Where analysis demonstrated statistically significant results, the proportional change is reported to allow assessment of the magnitude of the effect.

Participation in the KHS pre- and post-survey was 98% and 92% respectively of students attending school on the day the survey was administered. Participation at PHS was a 69% of students attending school during the first hour when this survey was administered. Although the survey was distributed to all PHS first hour classes, a number of teachers did not administer the surveys. There is no material reason to think that the students not surveyed were different than those surveyed. The fact that the PHS data was virtually the same as the KHS pre-survey suggests that using PHS as a control is still valid. Less than 1% of all the scan cards were unusable. Usable data was available for 1,386 students for the KHS pre-intervention survey, 739 students for the PHS case controls, and 1, 152 students for the KHS post-intervention survey.

After the intervention and post-survey were administered at KHS, two focus group discussions were conducted in June 1999. The purpose of these discussions was to evaluate the validity of the survey results, evaluate possible student interpretation of the responses, and to identify the most effective components of the "Every 15 Minutes" intervention program.

RESULTS

DEMOGRAPHICS OF RESPONDENTS

The students at both high schools were 9th through 12th grade. The age distribution by year was similar for both schools. Less than 4% were under age 15 and less than 2% were 19 years or older for both schools. The populations of both schools were 98% Caucasian. Kickapoo students were 47.1% male and 52.9% female. Parkview students were 44.5% male and 55.5% female.

PRE-INTERVENTION RESULTS

The baseline prevalence of drinking behavior was similar at both high schools. The KHS pre-survey and PHS control survey showed similar responses in all categories except that PHS students reported drinking at a slightly younger age. Other differences did not reach statistical significance in terms of prevalence of recent alcohol use, riding with students who have been drinking, or actual drinking and driving behavior. More than 90% of students both at KHS and PHS had previously participated in the D.A.R.E. program or similar substance abuse programs. The response to questions about attitudes and behaviors toward drinking and driving was the same at KHS and PHS, and was similar to the 1999 Centers for Disease Control Youth Risk Behavior Surveillance Survey (YRBSS) (CDC, 2000) for both U.S. and State of Missouri (Table 1).

Approximately one-half of the students at both high schools were actively drinking in the thirty days prior to the survey. One-third of the students at both high schools reported heavy drinking (five drinks or more in one evening) at least once in the prior thirty days of the survey. One-third of students had ridden with someone who had been drinking during the previous thirty days. One in six students reported drinking and driving within the thirty days preceding the survey (Table 1).

PRE-INTERVENTION GENDER DIFFERENCES

Gender differences were similar to that reported in the YRBSS nationally and for CDC data specific to Missouri. In the pre-survey, 51% of both genders reported drinking in the past 30 days. More males reported drinking before age 13 (25.5%) compared to females (17.4%), [X.sup.2] = 23.7, p < .000. Females and males were equally likely to have reported drinking in the prior 30 days (47.8% versus 49.5%). Females were less likely to have drunk heavily (5 or more drinks on one evening on at least one or more occasions), 34% versus males at 40%, ([X.sup.2] = 8.0, p<.005). Both were likely to have ridden in a car with someone drinking and driving in the prior 30 days (32.3% versus 32.8%). Females, however, were much less likely to have been drinking and driving in the prior 30 days 13.5%) versus 21.4%, ([X.sup.2] = 17.7,p < .000). This is similar to the 1999 YRBSS CDC data for Missouri. Females were more likely to respond that drinking and driving was dangerous 77.6% versus 72.7%, ([X.sup.2] = 13.5, p < .000) and were more likely to try to stop a friend from drinking and driving 93% versus 87.1%, ([X.sup.2] = 21.1, p < .000). This gender difference was mirrored in the post-intervention survey.

The data was also examined to determine if any patterns existed for those students ages 16-19 who reported frequent (more than three drinking episodes in the prior 12 months) versus infrequent drinkers (0-2 such episodes). Frequent drinkers were significantly more likely to have responded that "drinking and driving was dangerous but that they could handle it" 31% versus 1.7%, ([X.sup.2] = 215.1, p < .000). These points were also supported by student responses in the focus groups. Both frequent and infrequent drinkers reported similar exposure to the D.A.R.E. or other substance abuse programs (92.3% versus 91.6%).

POST-INTERVENTION RESULTS

Uncontrolled intervention effectiveness questions for the "Every 15 Minutes" program are displayed in Table 2. The proportion of teens responding that the program favorably affected them was typically three to one. This uncontrolled data suggests a favorable affective response of the students to the program, but the controlled data (Table 3) demonstrates much less impact. The intervention significantly affected students' attitude toward drinking and driving as measured by questions asking if the students believed that drinking and driving is dangerous and if they would try to stop a friend from drinking and driving. The change in responses was significant for the KHS pre-survey results compared to the post-survey and the PHS control survey (Table 3). However, the magnitude of the change was not large. The proportional change (80.1% to 84.1%) in favorable response to the question asking teens if they believe it is dangerous to drink and drive reflects a absolute change in attitude of 4%. Simi lady, the favorable proportional change (90.7% to 94.2%) for teens that would try to stop a friend from drinking and driving after the intervention demonstrates an absolute change of only 3.5%. The relatively small magnitude of proportional change is due to a high baseline favorable response in the pre-surveys. Questions dealing with teen behavior for drinking and driving do not show a statistically significant change in behavior. The KHS pre-survey data for behavior changes does not reach statistical significance in comparison to KHS post-intervention nor with the PHS control for the same time frame (Table 3).

POST-INTERVENTION AGE AND GENDER DIFFERENCES

Questions on each of the three surveys (KHS pre-and post-surveys, and PHS control survey) were divided into age sets and analyzed for attitudes towards drinking, and driving and behaviors for drinking and driving. The age sets chosen were ages 14 and 15 (non-driving), and ages 16 and above (driving age). As would be logically expected, the non-driving ages showed no improvement in the answers related to drinking and driving, since there were so few who actually reported drinking and driving. Analyzing the data for ages 18-19 showed that older students were less likely to view drinking and driving as dangerous: 73.8% versus 81.3% for ages 14-17 ([X.sup.2] = 10.4, p<. 001). The pattern of gender-specific differences in responses from the pre-survey was also found in the post-survey. Consistent with the pre-survey results, females reported significantly less drinking and driving behavior than males: 12.7% versus 21.1%, ([X.sup.2] = 16.3, p < .000).

FOCUS GROUP RESULTS

Two separate focus groups were conducted in June of 1999 consisting of eight students each selected from each of the four grade levels at KHS. The purpose of the focus groups was threefold: to evaluate the validity of the responses in the survey; to evaluate the potential interpretation problems with responses to questions; and determine what parts of the program were most effective. The focus group students uniformly indicated that their peers had completed their scan cards responses honestly. The students believed that the survey responses might actually underestimate the drinking behavior. Students indicated that DWI enforcement with exposure of students who had lost the privilege of driving for a year would do more to deter drinking and driving than any other program.

In particular, the focus group students noted that the survey data may have underestimated the actual drinking and driving behavior, as they felt that the students would interpret the question to mean drinking and driving only when obviously impaired. This point is important because the post-survey indicated that even after the intervention, 12.5% of the students at KHS indicated that they knew drinking and driving was dangerous, but that they could "handle it." Students believed that the intervention was a catalyst for initiating discussions about alcohol behavior, particularly in terms of behaviors towards drinking and driving. The students noted that this was not an open topic previously and that the intervention gave them permission to discuss this freely. Students thought this discussion was beneficial in setting the stage for allowing them to openly resist drinking and driving behaviors. In addition, the focus groups felt that intervention programs should be aimed at eighth graders when drinking attitudes and behaviors are often initiated. This recommendation makes sense in that 44.3% of KHS students and 54.2% of PHS students reported having their first drink prior to age 15.

CONCLUSIONS

The prevalence of drinking, and drinking and driving behaviors in this study is similar to that reported in the 1999 CDC YRBSS (CDC, 2000). The focus group data and comparison to the CDC YRBSS corroborates the survey data. The survey data and focus groups questions specifically dealing with the effectiveness of the intervention demonstrate positive changes in student attitudes toward drinking and driving. Analysis of the data does demonstrate a statistically significant improvement in attitudes towards drinking and driving, but the incremental change is not large due to a high baseline favorable response. The data does not show a measurable improvement in self-reported behavior toward drinking and driving. The survey and the focus group data suggest that the intervention would not make a substantial impact on behavior unless combined with additional school and community-based interventions.

The Greene County DWI Task Force used the data from this study to modify the 2000 presentation of "Every 15 Minutes" intervention, and to strengthen the community awareness of the need for rigorous law enforcement of DWI laws. The study results were reported in the local media along with statements by the city and county prosecutors reflecting their commitment for more stringent enforcement and penalties for drinking and driving. The study was also presented to the Missouri Governor's State Highway Safety Commission to reinforce the need for enforcing zero tolerance for teens, to reduce the legal blood alcohol concentration from 0.10 to 0.08, and to advocate community support for vigorous, visible DWI law enforcement.

The National Highway Traffic Safety Administration (Department of Transportation, 1997, p. 25) plan noted that "some of the most profound and dramatic advances in traffic safety have resulted from innovative programs that have resulted in significant media attention and public awareness" (Department of Transportation, 1997, p. 15). The consensus of the Greene County DWI Task Force was that the public discussion of the study results along with media attention for the "Every 15 Minutes" program did increase public awareness of the teen DWI problem.

Although these findings have direct implications for the primary prevention of students drinking alcohol and driving, the limitations of the study must be acknowledged. First, the data is based on self-reported behavior of high school students. The pre- and post- survey, however, closely follow the nationally administered 1999 YRBSS, which also examines student self-reported alcohol consumption and driving while intoxicated. A second limitation is the scope of the study. The intervention results are based on short-term data from one high school. Future research is recommended and should investigate the long-term effect of this type of intervention on students' drinking and driving behavior, the effects of targeting this intervention to younger students, and effects of a coordinated community/school effort to decrease DWI among high school students.

Table 1. Baseline Pre-Survey Results Compared To 1999 CDC YRBSS

Questions                                 US *   MO *    KHS    PHS

Drank alcohol (ever) other
   than a few sips (yes)                  81.0%  79.4%  68.4%  73.4%
Drank alcohol before age 13 (yes)         32.2%  33.5%  18.4%  26.4%
Drank alcohol in the prior 30 days (yes)  50.0%  49.9%  48.5%  48.7%
Heavy ([right arrow] 5 drinks at one
   time) in the past month                31.5%  32.0%  37.0%  37.3%
Ridden with a driver who had been drin-
   king & driving in past month (yes)     33.1%  35.1%  31.8%  34.0%
Been drinking and driving in past
   month (yes)                            13.1%  15.9%  16.7%  16.7%

* Comparative data for US and MO derived from CDC YRBSS 1999
Table 2. Post-Survey Results.

Did the program affect
     your attitude toward
     drinking & driving?    Yes (81.6%0

                                           Did not
                            More           Change          Less
                            Likely         Mind            Likely

After seeing the program,
     are you more or less
     likely to:
     Ride with a friend
      who has been
      drinking?             2.6%           22.3%           75.0%
     Drive after
      drinking?             3.6%           20.0%           76.4%
     Try to stop a friend
      from driving who
      has been drinking?    71.1%          16.3%           12.6%

                            Strongly            Not           Strongly
                            Disagree  Disagree  Sure   Agree   Agree

I think my  friends who
     saw the  program
     will be  less likely
     to  drive while
     intoxicated.           4.9%      7.3%      28.8%  42.6%  16.5%
The program has further
     convinced me  not
     to drink & drive.      5.1%      2.9%      19.1%  36.2%  36.7%
I had not really talked
     with my friends about
     their decisions to
     drink & drive before
     the program.           9.7%      21.0%     30.5%  29.0%  9.8%
It is more likely that I
     will talk with my
     friends about
     drinking &  driving
     as a result  of the
     program.               6.2%      9.0%      33.4%  36.0%  15.3%
Table 3. Pre and Post-Intervention Comparison
of Attitude and Behavior Questions.

                                KHS
                                             P        PHS       P
                             Pre   Post    value    control   value

Ridden with someone
    drinking & driving in
    the past month (yes).   31.8%  31.0%   .660      34.0%    .197
                                          X2=.194            X2=1.81
Been drinking & driving in
    past 30 days (yes).     16.7%  16.%    .913      16.7%    .944
                                          X2=.012            X2=.005
Believe it is dangerous to
    drink & drive (yes).    80.1%  84.1%  0.01 *     80.2%    .0311 *
                                          X2=6.66            X2=4.62
Would not ride with someone
    drinking & driving
    or would take keys
    (yes).                  67.3%  74.4%   .074      72.6%    .152
                                          X2=3.19            X2=2.05
Would try to stop a friend
    from drinking &
    driving (yes)           90.7%  94.2%  0.001 *    89.8%   <.001 *
                                          X2=10.77           X2=12.41

* Significant at less than or equal to 0.05

REFERENCES

CDC Youth Risk Behavior Surveillance System 1997. (1998). Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 47(SS-3), 1-89.

CDC Youth Risk Behavior Surveillance System 1999. (2000). Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 47(SS-3), 1-96.

Coombs, R. H, & Ziedonis, D. (Eds.). (1995). Handbook on drug abuse prevention: A comprehensive strategy to prevent the abuse of alcohol and other drugs. Needham Heights, MA: Allyn and Bacon.

Department of Transportation, National Highway Traffic Safety Administration, Traffic Safety Program Office of Research and Traffic Records. (1997). Strategic plan for behavioral research in traffic safety. NHSTA Docket # 97-062-No 1.

Flay, B. R., & Petraitis, J. (1994). The theory of triadic influence: A new theory of health behavior with implications for preventive interventions. Advances in Medical Sociology, 4,

19-44.

Kann, L., Kinchen, S. A., Williams, B. I., Ross, J. G., Lowry, R., Grunbaum, J. A., & Kolbe, L. J. (2000). Youth Risk Behavior Surveillance-United States, 1999. Morbidity and Mortality Weekly Report; 49(SS05), 1-6.

Kolaya, L., & Grimes-Smith, B. (1999). From experimenting to dependency in 43 seconds: Teaching junior high and high school students about the progression of alcoholism. Journal of Health Education, 30(3), 185, 189.

Komro, K. A., Perry, C. L., Veblen-Mortenson, S., Williams, C. L, & Roel, J. P. (1999). Peer leadership in school and community alcohol use prevention activities. Journal of Health Education, 30(4), 202-208.

Meeks, L., Heit, P., & Page, R. (1996). Comprehensive School Health Education, 2nd ed. (pp. 258-259). Blacklick, OH: Meeks Heit Publishing Company.

Ringwalt, C., Greene, J., Ennett, S., & Iachan, R. (1994). Past and future directions of the D.A.R.E. program: An evaluation review. Research Triangle Park, NC: Research Triangle Institute.

Rouse, B. A., Kozel, N. J., & Richards, L. G. (Eds.). (1985). Validation of self-report: The research record. NIDA Research Monograph No. 57, DHHS Publication NO. ADM 85-1402 (pp. 12-21). Rockville, MD: National Institute on Drug Abuse.

Thompson, N. J., & McClintock, H. O. (1998). Demonstrating you. r program's worth: A primer on evaluation for programs to prevent unintentional injury. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury Prevention and Control.

Yuen, F. K., & Pardeck, J. T. (1998). Effective strategies for preventing substance abuse among children and adolescents. Early Child Development and Care, 145, 119-131.

Alexander R. Hover, M.D., FACP, is with the Quality Resources Department at St. John's Regional Health Center at 1235 E. Cherokee, Springfield, Missouri, 65804, [e-mail]. Barbara A. Hover, B.S., is the Kickapoo High School PTSA President in Springfield, Missouri. Janice Clark Young; EdD, CHES is an Assistant Professor in the Department of Health, Physical Education and Recreation at Southwest Missouri State University in Springfield, Missouri. Address all correspondence to Dr. Hover.