Preliminary evaluation of a prescription opioid misuse prevention program among rural middle school students

Abstract

Adolescent opioid misuse, addiction, and overdose have emerged as national health crises. Nearly 17% of high school students have misused prescription opioids. The purpose of this study was to evaluate the reach and acceptability of a widely-used prescription opioid misuse prevention program, This Is (Not) About Drugs© (TINAD), and its preliminary efficacy at improving opioid misuse knowledge, opioid misuse attitudes, self-efficacy to avoid opioid misuse, and intentions to misuse opioids. Participants were 576 7^th grade students (M_age = 11.8; 51% boys; 39% Hispanic, 31% White, 20% Black) from a rural county in the southeastern U.S. All participants received the TINAD program and completed pretest and immediate posttest assessments. The program was school-based and implemented in collaboration with school teachers and administrators. Over 91% of all eligible students in the school district participated in the TINAD program. Most participants found the program acceptable—over 83% of students liked the program. Approximately 9% of participants reported prior misuse of prescription opioids. After participating in TINAD, students self-reported higher knowledge and self-efficacy as well as safer attitudes. However, there was no change in intentions to misuse opioids in the future. Effects of the program were consistent across gender, socioeconomic status, race/ethnicity, and previous opioid misuse. TINAD is acceptable and shows promise for improving opioid-related cognitions. However, more rigorous experimental and longitudinal research is needed to understand whether TINAD reduces opioid misuse over time. Given the limited research on adolescent opioid misuse prevention, this study lays the ground work for future randomized control trials.

Introduction

On average, 130 people die each day in America from an opioid overdose [1]. As the number of youth, ages 11–30, misusing prescription opioids has become alarmingly high [2], youth opioid misuse and addiction have emerged as national health crises. National statistics show 14% of high school students have taken prescription pain medication without a doctor’s prescription or differently than how a doctor advised [3]. Youth who misuse opioids are at greater risk for health consequences, such as addiction, overdose, and even death [4–6].

Youth, ages 12–17, are in a developmental stage that leaves them particularly susceptible to engaging in risky behaviors, such as opioid misuse. This is a time of significant neurobiological development and corresponding physiological vulnerability to substance use [7]; thus, misusing opioids can leave youth particularly at risk for addiction. Additionally, differences in adolescent cognition can lead to risky exploration of new experiences, especially in the presence of peers [7]. Youth who take risks and misuse prescription opioids are more likely to use other drugs, including heroin [8, 9]. To that end, efforts are underway across the United States to address the rise of opioid misuse among youth [10, 11].

One school-based opioid prevention program, This is (Not) About Drugs© (TINAD), has been delivered to more than 45,000 students across 22 states in the United States. TINAD was developed by the non-profit organization, Overdose Lifeline [11]. This 45-minute program is designed to be delivered in schools to 6^th through 12^th grade students. The program includes PowerPoint slides that describe the opioid crisis and addiction, as well as a brief video about the impact of opioids on the lives of youth.

To date, there has only been one formal evaluation of its efficacy, which focused on changes in knowledge regarding misuse of prescription opioids among high school students [12]. This was a quasi-experimental study, conducted in Indiana, that compared students that received the TINAD program to students in a control group. Results show that TINAD was efficacious at improving knowledge. However, health behavior change theories, including the Reasoned Action Model and Information-Motivation-Behavioral Skills model [13, 14], demonstrate that other cognitive variables such as attitudes, self-efficacy, and behavioral intentions are also important predictors of behavior change.

TINAD may improve these other important cognitive variables as well. The program takes a “peer to peer learning approach” [11]—it includes a video of adolescents sharing their personal, negative experiences with opioid misuse—and thus may cause changes in adolescents’ attitudes about opioids, as peers are influential to attitude formation among adolescents [15]. TINAD also shares overdose death rates and describes the link between prescription opioids and heroin to promote less permissive attitudes towards prescription opioid misuse. The program describes the increased risk of addiction for people who start misusing substances before the age of 15 to specifically target attitudes about opioid misuse among young people. In addition, TINAD emphasizes the importance of help-seeking and alternative ways to relieve stress, which may increase adolescents’ self-efficacy to avoid opioids. Together, improvements in knowledge, attitudes, and self-efficacy may lead to improvement in adolescents’ intentions to misuse opioids in the future [13]. Furthermore, related empirical work on substance use and health behavior theory supports the prediction that improvements in all of these cognitions may lead to decreased rates of actual opioid misuse behavior [16, 17]. However, the efficacy of TINAD at changing attitudes, self-efficacy, and behavioral intentions has yet to be evaluated.

Further, given that opioids are extremely addictive, prevention programs must prioritize avoidance of first misuse. Prevalence of opioid misuse among high school students increases with age, with 11% of freshman and 17% of seniors having ever misused prescription opioids [3]. For these reasons, prevention efforts must start earlier than high school if they seek to prevent the first misuse of these highly addictive substances. Although TINAD shows promise for use with high school students and there are general substance use prevention programs available in middle schools [18], there are no opioid prevention programs with demonstrated effectiveness that are available for delivery to early adolescents.

To address these gaps in knowledge and practice, the purpose of this study is to evaluate the acceptability, reach, and preliminary efficacy of TINAD at improving knowledge of opioid misuse, attitudes towards opioid misuse, self-efficacy to avoid opioid misuse, and intentions to misuse opioids among middle school students in the southeastern U.S. We used an effectiveness-implementation hybrid type 2 design with a pre-post test assessment. Type 2 hybrid designs are those that involve simultaneously testing a clinical intervention and an implementation protocol to facilitate the speed of research dissemination [19]. Our goal was to simultaneously investigate the TINAD program’s efficacy at improving cognitions about opioid misuse as well as its potential for actual implementation (e.g., reach; acceptability). This study was conducted in collaboration with school teachers and staff, as part of a larger community-driven, county-wide initiative to prevent opioid addiction, closely simulating actual delivery conditions and thus providing a more ecologically-valid estimate of program success [19]. While we recognize a pre-post test design is not the most rigorous form of evaluation, this design was chosen for practical reasons outlined by school officials who were motivated, in light of rising rates of overdose deaths in the county, to deliver the TINAD intervention to all students within the school year.

Methods

In late 2017, a rural county in the southeastern United States formed a taskforce to address the rise in opioid addiction and overdose in their community. County leaders and school-board officials decided to target early adolescents with school-based prevention efforts and enlisted the help of our research team. In Fall 2018, 7^th grade students from all public middle schools in the county were recruited to participate in TINAD. An effectiveness-implementation hybrid type 2 design was used to evaluate program acceptability, reach, and preliminary efficacy (with pre- and posttest data from students who participated in TINAD). We used passive parental consent and active, written student assent to consent students into the study. Flyers were distributed to students to take home to their parents in all schools. Students were instructed to return the form with their parent’s signature if their parents did not want them to participate. All 7^th grade students were eligible to participate (n = 753). Students who were absent on the day the intervention was provided to their class (n = 61) did not receive the program and did not complete surveys. Of the students who were present, six students had parents that refused to have them participate and eight students did not assent to participate. Because the program was adopted by the school district as part of the regular health curriculum, students who did not assent did participate in the program, but did not respond to the surveys. Students without parental consent did not participate in either. Of the remaining, 686 students took part in the program, and 619 students completed anonymous electronic surveys immediately before the intervention (pretest) and after the intervention (posttest). For the current study, students missing 50% or more of data on variables of interest (n = 43) were excluded, resulting in a final sample of 576 students.

Design and Procedures

Our research team coordinated with school principals to identify days and times for program administration and data collection. On average, the program was presented to three classes per week. On the day of the program, students whose parents did not give consent for them to participate (n = 6) were moved to an alternate classroom. Participants began by completing anonymous pretest measures on laptop computers or tablets in their regular classroom. Next, one of four members of our research team who had completed TINAD presenter training delivered the program to the class. In most cases the classroom teacher was present during program activities, but not actively involved in assisting students with surveys or the intervention. Class sizes ranged from 21–29 students (M = 26 students). Small incentives (e.g., pens, stress balls) were distributed to students during the program for active participation. After the program, students completed an anonymous posttest assessment on a laptop computer or tablet. All intervention activities and assessments occurred within a single class period (range = 70–90 minutes). At the end of the class, all participants were given a take-away bookmark with local and national resources related to mental health and substance use (e.g., the National Crisis Text Line). All procedures were approved by the university IRB.

Description of the Program

TINAD is a prescription opioid misuse prevention program that not only emphasizes the avoidance of drug use but also the importance of youths’ personal choice to be healthy. It is a multimedia presentation utilizing PowerPoint slides with embedded discussion questions and a professionally-made, peer-to-peer video demonstrating the real-world consequences of opioid misuse. The overall goal of the program is to raise awareness of prescription opioid misuse prevalence and risk among adolescents. TINAD activities are designed to fit within one school class period, making the program more feasible to implement in schools than longer programs.

Presenter Training

Everyone who delivers TINAD is required by Overdose Lifeline to complete all elements of the presenter training. The presenter training includes two brief online courses [20], as well as live online training with a representative from the Indiana-based team. This training, as well as a 120-page manual for delivering the program, cost $275 per person and took approximately four hours. In addition, there is an annual program delivery fee for counties that wish to adopt the TINAD program which ranges from $250-$2000 and is based on the county population [20].

Measures

Participant Characteristics

We collected data on age, race, gender, and free/reduced-price lunch status (a proxy for socioeconomic status). We also collected information about participants’ lifetime substance use using items adapted from the National Youth Risk Behavior Survey [21].

Reach

Program reach was assessed using recruitment rates. We calculated the percentage of students who actually participated in the program out of the total number of students who were eligible to participate.

Acceptability

At posttest, participants were asked to report on the acceptability of the program with four items. Items such as “I liked the program” were rated on a scale from 1 = Not at all to 4 = A lot (see Table 2).

Table 2.

Posttest Items Assessing Acceptability of the TINAD Program Among Seventh-Grade Students

				No. (%)
	n	M (SD)	range	Not at all	A little	Some	A lot
I liked the program	574	3.29 (.82)	1–4	20 (3.5)	75 (13.0)	197 (34.2)	282 (49.0)
I learned new things from the program	569	3.43 (.80)	1–4	20 (3.5)	54 (9.4)	156 (27.1)	339 (58.9)
The program kept my attention	570	3.30 (.88)	1–4	28 (4.9)	79 (13.7)	159 (27.6)	304 (52.8)
I will use information from this program in the future	572	3.36 (.87)	1–4	27 (4.7)	68 (11.8)	148 (25.7)	329 (57.1)

Note. Each item in the above table was preceded in the survey by the following prompt: “These questions are about the This is (Not) About Drugs program. Tell us your honest opinions.”

Knowledge

At pretest and posttest, participants answered four items assessing their knowledge of prescription opioid pills and addiction. Items were rated on a scale from 1 = Strongly Disagree to 5 = Strongly Agree (for items, see Table 3). Items were averaged to capture participant knowledge at both pretest and posttest (pretest Cronbach α = .72; posttest Cronbach α = .72). These items were modeled after items from the Meth Prevention Lesson by the Meth Project under a Creative Commons License [12, 22].

Table 3.

Changes from Pretest to Posttest in Prescription Opioid Misuse Knowledge, Attitudes, Self-efficacy, and Intentions

	Pretest			Posttest
	M	SD	range	M	SD	range	Z	p	d
Knowledge Mean Score	3.52	.83	1–5	4.19	.78	1–5	−14.13	<.001	1.68
Individual Knowledge Items
Prescription pills such as hydrocodone and oxycodone are the same type of drug as heroin.	3.03	1.08	1–5	4.05	1.10	1–5
Using prescription pain pills not prescribed to you is as risky as using heroin.	3.51	1.16	1–5	4.06	1.20	1–5
The younger someone starts drinking alcohol or using drugs increases the risk of addiction or alcoholism.	3.96	1.11	1–5	4.32	.96	1–5
Drinking alcohol, using marijuana or other drugs increases the likelihood that someone would use heroin.	3.57	1.17	1–5	4.14	.95	1–5
Attitudes Mean Score	4.08	.95	1–5	4.37	.84	1–5	−8.35	<.001	.77
Individual Attitude Items
Opioid addiction is severe.	3.89	1.17	1–5	4.28	1.00	1–5
Opioid addiction is serious.	4.29	.96	1–5	4.48	.87	1–5
Opioid addiction would negatively affect my life.	4.07	1.19	1–5	4.31	1.09	1–5
Self-efficacy Mean Score	3.90	.96	1–5	4.08	.89	1–5	−5.03	<.001	.45
Individual Self-Efficacy Items
I am sure I could avoid misusing opioid prescriptions.	3.99	1.11	1–5	4.17	1.03	1–5
I am sure I could limit my use of an opioid prescription after an accident or injury.	4.00	1.07	1–5	4.11	1.01	1–5
I am sure I could avoid prescription opioids all together.	3.72	1.18	1–5	3.92	1.11	1–5
Opioid Misuse Intentions	8.55	1.35	1–9	8.53	1.35	1–9	−0.04	.97	.003

Note. The same pattern of results emerged when analyses were run separately among adolescents who reported previous opioid misuse and among adolescents who did not report previous misuse. Knowledge: n = 484; Attitudes: n = 537; Self-efficacy: n = 519; Intentions: n = 555

Attitudes

At pretest and posttest, participants answered three items assessing their attitudes towards misusing prescription opioid pills. Items were adapted from the Perceived Severity and Susceptibility scale [23]. Items were rated on a scale from 1 = Strongly Disagree to 5 = Strongly Agree (for items, see Table 3). Items were averaged to capture participant attitudes at both pretest and posttest (pretest Cronbach α = .79; posttest Cronbach α = .80).

Self-Efficacy

At pretest and posttest, participants answered three items assessing their self-efficacy for avoiding prescription opioid misuse. Items were adapted from the Perceived Severity and Susceptibility scale [23]. Items were rated on a scale from 1 = Strongly Disagree to 5 = Strongly Agree (for items, see Table 3). Items were averaged to capture participant self-efficacy at both pretest and posttest (pretest Cronbach α = .82; posttest Cronbach α = .80).

Intentions

At pretest and posttest, participants answered one item assessing their intentions to misuse prescription opioids: “In the next year, how likely are you to use pain pills without a prescription (e.g., codeine, Vicodin, Percocet, OxyContin, hydrocodone)?” Participants rated the item from 1 = Extremely Likely to 9 = Extremely Unlikely.

Analysis Plan

We used descriptive statistics to summarize sociodemographic variables and lifetime substance use behaviors. To address our first aim, evaluating the implementation of TINAD, we calculated the percentage of students who actually participated in the program out of the total number of students who were eligible to participate. To evaluate program acceptability, we calculated the percent of participants who selected each of the answer options (i.e., “not at all,” “a little,” “some,” or “a lot”) for the 4-items evaluating student perceptions of the program (see Table 2 for items).

To address our second aim, testing preliminary program efficacy, Wilcoxon Signed Rank Tests were used to evaluate differences from pretest to posttest in knowledge, attitudes, self-efficacy, and intentions. Wilcoxon Signed Rank Tests were used due to the skewness of the data. Intentions to misuse prescription opioids were evaluated using a single item. For the other outcomes, participant scores on all items assessing the construct were averaged and a mean score was used in analyses. We calculated Cohen’s d as a measure of effect size, which can be interpreted as small (.20), medium (.50), or large (.80); [24]. Only cases with matched pretest and posttest data were used in analyses to determine initial efficacy of the program. Power analyses, using G-power, revealed that to detect a small effect (.20) with desired power level equal to .95 and probability level equal to .05, we would need a total sample size of 343 participants; thus, these analyses were well powered to detect effects [25].

We included supplemental analyses to determine if there were differences in program efficacy by participant characteristics. Independent sample t-tests were used to determine whether change scores from pretest to posttest evaluations differed by gender, SES, and previous experience misusing opioids. Analysis of variance (ANOVA) models were used to determine whether change scores from pretest to posttest evaluations differed by race/ethnicity. A change score was calculated by subtracting pretest scores on each outcome from posttest scores on each outcome such that higher scores indicated greater improvement. We made a Bonferroni correction and a p-value of 0.003 was used to determine if analyses were significant for Wilcoxon Signed Rank Tests, Independent samples t-tests, and ANOVA models. Throughout analyses, sample size was allowed to vary so that all available data could be used. For all analyses, IBM SPSS® version 24 statistical software was used.

RESULTS

The mean age for the 576 middle school students included in analyses was 11.8 years old (SD = 0.68). Additional sample characteristics can be found in Table 1.

Table 1.

Sample Characteristics at Pre-test Among Seventh-Grade Students

Characteristics	No. (%)
Sociodemographics
Race/ethnicity - White	180 (31.3)
Race/ethnicity - Black	116 (20.1)
Race/ethnicity - Hispanic	227 (39.4)
Gender - Boy	295 (51.2)
Gender - Girl	273 (47.4)
Gender - Transgender/other gender identity/prefer not to say	8 (1.4)
Received free/reduced price lunch	336 (58.3)
Substance Use Behaviors
Ever tried cigarette smoking	56 (9.7)
Ever used an e-cigarette or vape	139 (24.1)
Ever drank alcohol	135 (23.4)
Ever used marijuana	48 (8.3)
Ever misused prescription pain medicine	51 (8.9)

While all 7^th grade students in the county were eligible to participate in TINAD (n = 753), 91% (n = 686) actually participated in the intervention. Additionally, most participants found TINAD acceptable (Table 2). Over 83% of 7^th grade students liked the program “some” or “a lot” and 86% learned new things from the program. Over 80% of participants felt the program kept their attention (“some” or “a lot”) and 83% felt they would use information from the program in the future.

Immediately following the TINAD program, participants showed a significant increase in knowledge of prescription opioid misuse (Cohen’s d = 1.68), attitudes towards prescription opioid misuse (Cohen’s d = 0.77), and self-efficacy to avoid prescription opioid misuse (Cohen’s d = 0.45), compared to their scores at pretest (Table 3). Effect sizes ranged from moderate to large and generally reflect a half-point increase on the outcome measurement scale. For example, knowledge started at a mean of 3.52 and ended at 4.19—indicating, on average, participants improved from selecting “neutral” in response to items like “using prescription pain pills not prescribed to you is as risky as using heroin” to selecting “agree”. Average scores on attitudes increased from 4.08 to 4.47 and average scores on self-efficacy increased from 3.90 to 4.08—indicating that scores improved but, on average, the meaning of participant responses (e.g., “agree”) remained similar.

There were no significant differences from pretest to posttest in participants’ intentions to avoid prescription opioid misuse (Cohen’s d = 0.003). For this item, participant scores started at a mean rating of 8.55 and ended at 8.53—indicating that at both pretest and posttest, participants were “somewhat”/“extremely” unlikely to misuse prescription pain pills.

We conducted supplemental analyses to examine if intervention efficacy was similar across gender, free/reduced price lunch status, race, and previous experience misusing prescription opioids (Tables 4 and 5). We found that the mean change did not differ as a function of these characteristics.

Table 4.

Differences in the Efficacy of the Program by Gender, SES, and Previous Experience Misusing Opioids.

	Change in Knowledge			Change in Attitudes			Change in Self-efficacy			Change in Intentions
	t(df)	p	M (SD)	t(df)	p	M (SD)	t(df)	p	M (SD)	t(df)	p	M (SD)
Gender	−0.35 (461.05)	.73		0.88 (527)	.38		0.24 (509)	.82		−1.02 (545)	.31
Girl			.68 (.76)			.26 (.77)			.17 (.77)			.05 (1.39)
Boy			.65 (.94)			.33 (1.03)			.19 (.98)			−.08 (1.57)
Lunch	1.73 (473)	.09		−1.08 (521)	.28		−0.14 (504)	.89		0.71 (539)	.48
Free/reduced			.61 (.86)			.33 (.94)			.20 (.94)			−.05 (1.36)
Not free/reduced			.74 (.83)			.25 (.82)			.18 (.81)			.05 (1.66)
Experience misusing	−1.27 (479)	.21		−0.43 (532)	.66		−0.77 (514)	.44		−0.91 (51.68)	.37
Yes			.82 (.87)			.35 (.94)			.28 (.94)			.28 (2.54)
No			.65 (.86)			.29 (.91)			.17 (.88)			−.05 (1.32)

Note. In these models, pretest to posttest change scores were used for each outcome variable with higher scores indicating greater improvement from pretest to posttest.

In the total sample, means and SDs for change in knowledge, attitudes, self-efficacy, and intentions, are as follows: .67 (.85), .29 (.91), .18 (.88), −.01 (1.48). For gender 1 = “boy” and 2 = “girl.” For lunch 0 = “not free/reduced” and 1 = “free/reduced.” For experience misusing opioids 0 = “no” and 1 = “yes.” SES = socioeconomic status.

Table 5.

Differences in the Efficacy of the Program by Race

	M (SD)	SS	(df)	MS	F	p
Change in Knowledge by Race		3.98	(2)	1.99	2.77	.06
White	.81 (.80)
Black	.64 (.88)
Hispanic	.60 (.87)
Change in Attitudes by Race		1.63	(2)	0.81	0.94	.39
White	.21 (.67)
Black	.34 (1.13)
Hispanic	.32 (.99)
Change in Self-efficacy by Race		1.17	(2)	0.59	0.72	.49
White	.13 (.72)
Black	.27 (1.13)
Hispanic	.19 (.90)
Change in Intentions by Race		2.37	(2)	1.18	0.56	.57
White	.11 (1.19)
Black	−.06 (1.60)
Hispanic	.00 (1.56)

Note. In these models, pretest to posttest change scores were used for each outcome variable with higher scores indicating greater improvement from pretest to posttest

DISCUSSION

Adolescents in the U.S. are misusing opioids at alarmingly high rates [3]. Adolescents living in rural areas are especially at-risk as they have a 35% greater chance of misusing prescription opioids compared to youth living in urban areas [26]. These substances are highly addictive; thus, it is imperative that interventions target youth and decrease their likelihood of misusing opioids before they first have the opportunity to misuse. TINAD is a brief opioid misuse prevention program designed for middle and high school students. In this hybrid effectiveness-implementation study, we found that the program was feasible to deliver in classroom settings over a period of four months to most eligible students in one medium-size school district in the rural southeast. The program was also acceptable to these students. In addition, using a pretest-posttest design, TINAD showed preliminary efficacy at improving knowledge, attitudes, and self-efficacy. However, TINAD did not change students’ intentions to misuse prescription opioids.

One of the greatest strengths of TINAD seems to be its wide reach and acceptability. This program has already been delivered in many communities across the U.S. [11] and, in collaboration with school staff and teachers, our team was able to deliver the program to over 91% of 7^th grade students in a rural southeastern county. The program is designed to be delivered in school. On most days, adolescents spend the majority of their waking hours at school [27]; thus, programs that take place in this setting have the best opportunity to reach a large number of adolescents. In addition, this study shows, middle school students like the program—it keeps their attention and they believe they learn new things from it. TINAD may represent a wide-reaching and acceptable option for beginning to educate adolescents about the opioid crisis.

Health behavior theories insist that changing knowledge, attitudes, and self-efficacy is an important step in promoting healthier behavior [13, 14]—in this case, decreasing opioid misuse. In addition, theory suggests [13], changes in intentions are most proximal to actual behavior change. In this preliminary pre-post study, TINAD is shown to be, in part, efficacious at improving some important elements of adolescent cognition. However, as we found no evidence that TINAD was able to shift early adolescents’ intentions to avoid opioid misuse, the program may not be doing enough to alone, have an impact on adolescent opioid misuse behavior. Because participant scores were very high at pretest, it is possible that a ceiling effect may have impacted our results.

In addition, results were consistent across race, gender, SES, and previous experience misusing opioids. This demonstrates the program’s potential for reaching large, diverse audiences in many school environments. The consistency of our findings across subgroups is important, considering the opioid epidemic does not discriminate based on socioeconomic and demographic factors [28, 29]. Future research that replicates these findings in other parts of the country that have been most heavily impacted by opioid addiction would add to the body of evidence showing where and for whom TINAD is most appropriate.

Although TINAD was a brief intervention that could be delivered in just a single classroom session, preliminary results from this study are in line with others that suggest even brief programming can impact adolescent health cognitions and behavior [30], and some meta-analyses that show that health interventions for youth can be effective regardless of their duration [31–33]. Since TINAD is wide reaching, acceptable, and shows preliminary efficacy at promoting healthier cognitions regarding opioid misuse, as demonstrated by this study with middle school students and one other study with high school students [12], it seems fruitful to build on and improve the TINAD program so that in the future, we can expect adapted iterations to change adolescent opioid misuse intentions and behavior. For example, interventionists might consider a longer-term approach to opioid misuse prevention, in which adolescents are provided support and information throughout the transition from middle to high school—a time when they are more likely to succumb to peer pressure and take health risks [7, 34]. It may also be helpful to involve parents in opioid intervention efforts [35], as there is research to suggest parental communication and monitoring decrease adolescent risk taking and substance use behavior [36]. Finally, broader structural interventions that involve clinicians and pediatricians [37] and aim to reduce adolescents’ access to prescription opioids are necessary, as poison control centers receive nearly 12,000 reports of children and adolescents being exposed to opioids each year [38].

Limitations and Future Directions

Our findings highlight the acceptability of the program as well as its preliminary efficacy. However, there were some limitations to our methodology that highlight important directions for future research. First, this study should be viewed as preliminary given that it was conducted at a single time point as a pre-post assessment and was not a randomized trial that could account for potential confounds. This study was conducted in collaboration with school teachers and staff, as part of a larger community-driven, county-wide initiative to prevent opioid addiction. While we recognize a pre-post test design is not the most rigorous form of evaluation, this design was chosen for practical reasons in line with the main priority of school officials at the time: to deliver the TINAD intervention to all students within the school year. TINAD must be evaluated among middle school students using a randomized controlled trial with outcomes assessed over time, with multiple, long-term follow up time points, to firmly establish treatment efficacy.

Second, although this program was able to reach the vast majority of eligible students, there were 67 students who were eligible but did not receive the TINAD intervention. Many of these students could not attend the intervention because they were absent or removed from regularly scheduled classes on the date the program was delivered. In the future, researchers and community-partners will need to consider how many sessions are needed to reach all intended students. Opportunities for adolescents to participate after school or at alternative times (e.g., during lunch or elective classes) may be needed.

Third, for this project, the acceptability of TINAD was only evaluated among students, not teachers, staff, and administrators, who may be responsible for sustaining the program in the long-term. In the future, acceptability of TINAD should be evaluated among these key stakeholders. In addition, the training required to deliver the program may be considered time-consuming and costly in many counties. Of note, all teachers were offered the opportunity to participate in the TINAD presenter training free of cost; however, no teachers actually completed the training. Teachers are extremely busy and pulled in many directions; thus, future implementation research is needed to understand whether it is feasible for teachers to complete the necessary training and deliver the TINAD program and what incentives may be necessary. Relatedly, the overall capacity of schools to deliver TINAD (e.g., availability of resources, stakeholder buy-in) must be evaluated to better understand the potential for schools to independently continue to implement the program.

Finally, this program evaluation took place in one rural county in the southeastern United States; however, TINAD is being implemented across the country. Future research should examine the generalizability of the current results and determine how the program is working in other communities.

Conclusion

TINAD is an opioid misuse prevention program for middle and high school students that has been implemented with over 45,000 students to date [11]; however, prior to the current study, its implementation and efficacy among middle school students had not yet been evaluated. In this study, we found that TINAD was wide-reaching and acceptable. It also showed preliminary efficacy at improving some opioid-related cognitions among 7^th grade students in a rural county. Further efforts are needed to expand upon this 45-minute, in-class program so that it can be expected to change adolescent opioid misuse intentions and behavior. This is one of the only prescription opioid misuse prevention programs available for middle school students. Many interventions aimed at combatting the opioid crisis are focused on decreasing opioid use among those who are already struggling with addiction; however, primary interventions that prevent the first opioid misuse are needed to keep a new generation from falling victim to this public health crisis [18]. In 2017 alone, there were over 13,000 overdose deaths among young adults (ages 25–34); [39]. If effective programs for preventing opioid misuse among adolescents are developed and implemented, it is possible that the nation can prevent a new wave of young adults from becoming addicted to opioids and reduce overdose deaths across the United States.