The original recruitment goal for this study was 200 youths and 10 sites (five American Indian N-O-T sites and five brief-intervention sites). We were able to recruit 79 youths from three American Indian N-O-T schools and two brief-intervention schools. Youth enrollment was thus lower than expected — 39.5% of the initial youth target. Seventy-four of the 79 youths recruited met the selection criteria of current smoking. Approximately 3 months post-baseline, 53.7% (29/54) of N-O-T youths and 70.0% (14/20) of brief-intervention youths were present for follow-up data collection. Overall, 58.1% (43/74) of eligible study youths participated in the follow-up evaluation. Compared with follow-up rates in our other N-O-T studies, this is the lowest rate (19
Anecdotal feedback from youths, program facilitators, school personnel, and community members suggest numerous reasons for recruitment and retention challenges for the American Indian N-O-T program. The first of these relates to cultural factors. Eastern North Carolina is a tobacco-growing region and many of the American Indian families in the study communities have strong economic ties to tobacco. Also, tobacco has historically been used by American Indians for spiritual and medicinal purposes. These factors may create ambivalence among youth about participating in a tobacco-cessation program. Second, American Indians place a high value on family and community; American Indian N-O-T adaptation recommendations include information on the impact of secondhand smoke on family and community members. Although family values can support positive behavior change, they also can act as attitudinal and motivational barriers when many household members, including elders, smoke and have lenient attitudes toward smoking tobacco (20
). Third, research requirements may deter youth recruitment. For example, youths in the American Indian N-O-T program were not permitted to participate without IRB-approved parental consent forms. Youths may not have provided these forms for various reasons: 1) they may simply have forgotten them; 2) they may have been afraid that their parents would be angry if they knew that they smoked; or 3) formal signed assent and consent forms may have reminded youths of research exploitation of American Indians in the past, causing them to choose not to participate. Nonresearch situations for which parental consent is not required may provide a better understanding of youth willingness to join the American Indian N-O-T program. Fourth, privacy may be a concern for youth from small tribes or communities where a stigma is associated with participating in what might be considered a drug-prevention class. Admitting to an addiction problem or seeking outside help may not be acceptable in some communities. Fifth,
youths may perceive the school climate to be unsupportive of quit efforts when both teachers and youths are permitted to smoke on school grounds. (In North Carolina, the majority of schools are not tobacco-free.) Only 43 of 115 public school districts or units in North Carolina have adopted the state's optional 100% tobacco-free schools policy. No schools in this investigation were in a county with a tobacco-free policy.
The final recruitment and retention challenge relates to racial factors. In the multiracial counties in North Carolina where the study was conducted, the ratio of American Indian teachers to American Indian students is low. According to community members, lack of American Indian adult role models or contacts in the schools may hinder support and encouragement for American Indian smoking-cessation efforts. This was illustrated in our challenges to recruit American Indian facilitators. Four of eight facilitators were American Indian; four were white. Research supports the importance of involving American Indian people in tobacco-cessation efforts for American Indians (20
Although student recruitment numbers were low, the study was successful at recruiting the target population of American Indian youth smokers. Overall, 82.2% of youths were American Indian. In addition, half of the facilitators were American Indian. More male than female youths participated in the study. Most participants had been smoking for about 5 years. On average, the youths in this study were smoking about 10 cigarettes per day. Interestingly, the FTQ revealed that the youths in this sample had a low–medium dependence on nicotine despite being daily smokers.
Consistent with other N-O-T studies, the percentage of N-O-T males who quit smoking was twice the percentage of brief-intervention males who quit smoking. The difference was not statistically significant; lack of significance likely resulted from small sample size. The effect size was meaningful in terms of intervention impact. Importantly, the quit rates for the males in this study are equal to or higher than male quit rates in other N-O-T studies of homogenous populations of youth. For example, in a recent 5-year review of N-O-T findings, the mean 3-month end-of-program quit rate for males was between 15.1% (intention-to-treat group) and 20.1% (compliant group) (19
). Approximately 10% to 14% of brief-intervention males quit smoking in the current study. This rate is slightly higher than spontaneous or care-as-usual rates found in other studies of smoking cessation among teenagers (19
Two of our study findings were unusual compared with other N-O-T core studies. One was that fewer females than males joined the American Indian N-O-T program. In past N-O-T core studies with predominantly white youth, male recruitment has been more challenging than female recruitment (10
). Another unusual finding was that no females quit smoking, which has never occurred in a N-O-T study (19
). A 5-year review of N-O-T core studies showed an overall quit rate for females between 14.7% (intention-to-treat group) and 18.5% (compliant group)Two of our study findings were unusual compared with other N-O-T core studies. One was that fewer females than males joined the American Indian N-O-T program. In past N-O-T core studies with predominantly white youth, male recruitment has been more challenging than female recruitment (10,11). Another unusual finding was that no females quit smoking, which has never occurred in a N-O-T study (19). A 5-year review of N-O-T core studies showed an overall quit rate for females between 14.7% (intention-to-treat group) and 18.5% (compliant group). Our American Indian N-O-T pilot findings, along with input from CAB members and tribal representatives, suggest that 1) there may be unique aspects of the social and cultural context of smoking and smoking cessation among American Indian females that call for further study; 2) the American Indian N-O-T curriculum needs further adaptation to meet the needs of American Indian females; and 3) facilitator training needs to incorporate additional information on building relationships with females. Our study is not the first smoking-cessation study to find a low rate of success among American Indian females. King et al found that adult women in minority racial and ethnic populations appear to be less responsive to smoking-cessation programs than white women (22
). Some research suggests that American Indian females may be less likely than males or females from other racial and ethnic groups to acknowledge the negative health consequences of tobacco use (20
). It is important to emphasize, however, that 5 of 19 females in our study succeeded in reducing their smoking.
Among youths who did not quit smoking, a greater number of American Indian N-O-T females than brief-intervention females reduced their smoking. American Indian N-O-T youths reduced smoking by a greater amount than brief-intervention youths. Specifically, American Indian N-O-T youths cut back on their smoking by more than half during the week and more than 75% during the weekend.
Using a quasi-experimental design rather than random assignment may have threatened the validity of our results. We concluded that random assignment would be difficult in this early phase of program study because we had yet not formed relationships with the schools. We thus chose a quasi-experimental design. CAB members helped to guide this decision. By following community-based participatory strategies, we established a foundation of trust among schools and communities in recruiting sites and facilitating participation. CAB members believed this was a necessary step for promoting participation. We selected brief-intervention sites and American Indian N-O-T sites based on common characteristics. The two intervention groups had similar baseline characteristics (), so the threat to validity resulting from participant differences was reduced. Another limitation was the lack of biochemical validation of self-reported smoking status and lack of documentation of days of continuous abstinence from smoking. We were not able to collect these data because of unexpected time constraints. Previous N-O-T studies have found high agreement between self-reports and exhaled carbon-monoxide–validated quit rates (10
A final limitation of this study is lack of generalizability. This pilot American Indian N-O-T program was implemented among tribes in North Carolina only. Although tribal commonalities may exist across the United States, we cannot assume that a one-size-fits-all approach is appropriate. As we move forward in efficacy testing, tribal involvement from various regions of the United States is critical.
To our knowledge, this is the first study on a smoking-cessation program tailored for American Indian youths, and it is the first examination of the American Indian N-O-T program. Lessons learned will improve methods and strategies in subsequent efficacy trials of American Indian N-O-T and general cessation programming for American Indian youths. The general outcomes of the pilot study highlight four key findings: 1) the American Indian N-O-T program served as cessation aid for males and as a reduction aid for females; 2) study youths seemed ready to change their smoking behavior (i.e., more than half of all available youths reduced cigarette use from baseline); 3) recruitment barriers need to be studied and overcome for greater American Indian youth participation in cessation programs; and 4) curriculum adaptation must give greater attention to cultural and contextual issues, especially related to differences between sexes. The current pilot study is the first step toward understanding the usefulness, efficacy, and long-term sustainability of the American Indian N-O-T program. Future research will focus on youth recruitment, gender issues, additional curriculum modifications, and efficacy testing.