1.1 The multidimensional nature of 12-step experience

Conceptually, 12-step activity is a multidimensional phenomenon, involving not only attendance at meetings, but also active participation during meetings (e.g., talking/sharing), and social interaction with a sponsor and/or other fellowship members before, after, and outside of formal meeting contexts. Active work on the 12-steps, particularly those that require more tangible action (e.g., writing a personal inventory in step 4, and discussing it in step 5), are further 12-step specific activities presumed critical to recovery (AA, 2001). Given the entirely voluntary aspect of 12-step fellowships’ emphasis which state the only membership requirement is a personal “desire to stop drinking/using”, an important contextual element is the degree to which individuals have been mandated by the judicial system to attend. This important dimension may affect many other indices of activity and affiliation.

Based on previous empirical and theoretical literature, we identify seven key dimensions of 12-step participation (Table 1). The first five dimensions (meeting attendance, meeting participation, fellowship involvement, step work, and mandated attendance) involve objective behaviors and events. There is empirical support linking meeting attendance with abstinence (Chi, Kaskutas, Sterling, Campbell, & Weisner, 2009; Kelly, Brown, Abrantes, Kahler, & Myers, 2008; Kelly, Dow, Yeterian, & Kahler, 2010; McKay, et al., 1994; Tonigan, Connors, & Miller, 2003), and with improvements in psychosocial functioning (Tonigan, Toscova, & Miller, 1996). Very little is known about the specific types of meetings individuals attend (e.g., speaker meetings, literature-focused meetings, meetings for beginners, women, young persons, or gay persons), preventing more detailed examination of the effects of meeting types on individuals’ fellowship engagement, retention, and derived benefit. Some meetings tend to be larger and do not provide the opportunity for during-meeting participation (e.g., speaker meetings) whereas others (e.g., literature-focused meetings) tend to be smaller and more intimate providing much greater opportunity for active participation and social engagement (Kelly, Humphreys, & Youngson, 2004). While established members may participate in a variety of meeting formats, predominant attendance by a new member at one or another type of meeting could either accelerate or diminish social engagement and integration.

Behaviors reflecting a more active level of engagement in the program itself, such as having a sponsor, working the 12-steps, and reading 12-step literature, have been linked with positive outcomes (Chi, et al., 2009; McKellar, Stewart, & Humphreys, 2003; Montgomery, Miller, & Tonigan, 1995; Weiss, et al., 2005). Similarly, despite the common 12-step fellowship recommendation to become involved in group service work, a principle shown to have therapeutic benefit (Pagano, Friend, Tonigan, & Stout, 2004), little has been documented regarding the extent to which attendees obtain service positions (e.g., help set up meetings, or make coffee) or sponsor or help others. These aspects of “service” are captured in the “meeting participation” and “fellowship involvement” dimensions in Table 1.

Absent from existing measures of 12-step participation are considerations of the extent to which group attendance has been mandated by the courts or legal authorities. This is important as studies suggest poorer outcomes for this subgroup (Kownacki & Shadish, 1999). The criminal justice system plays a large role in influencing entry into specialized addiction treatment programs (SAMHSA, 2009), with studies demonstrating variable associations with retention and active involvement in treatment (Joe, Simpson, & Broome, 1999; Stevens, et al., 2005; Wild, Roberts, & Cooper, 2002). Given these associations, further clarification of the extent and impact of criminal justice system influence on 12-step group effectiveness is needed.

The remaining two dimensions listed in Table 1 (affiliation and safety) tap subjective aspects of the overall experience of 12-step participation. Perceptions of the value and effectiveness of participation may help to clarify past and present attendance behaviors, and may predict continued attendance or future re-engagement following a period of non-attendance. Perceptions that meetings are unsafe for particular sub-populations (e.g., women, young people) (Kelly, Dow, Yeterian, & Myers, in press) or for specific reasons, such as potential exposure to substances, may be expected to influence referral patterns by treatment professionals and individual attendance behaviors. However, empirical study of these subjective aspects of 12-step participation, and their links to behavioral aspects of involvement and downstream outcomes, is particularly lacking.

1.2 The need for comprehensive measurement

Measurement of the 12-step construct has been largely limited to meeting attendance, or to scales that capture a single or small number of the above dimensions in limited detail (e.g., McKay, et al., 1994; Snow, et al., 1994; Tonigan, Conners, et al., 1996). Those more comprehensive in content (Morgenstern, et al., 1996) have nonetheless lacked an ability to differentiate between common fellowship types, such as AA, NA, and CA. Although simple summary measures are analytically appealing, they arguably fail to yield the richness of data required to answer many key research questions pertaining to 12-step organizations and recovery.

In the absence of comprehensive coverage of 12-step experiences, questions pertaining to the inter-relationships between dimensions, and the incremental predictive utility of distinct behaviors and perceptions for outcomes, remain unanswered. In addition, collapsing information across fellowship types (e.g., AA, NA, CA and others) prohibits the study of important nuances with potentially critical importance for determining the effectiveness of matching substance use profiles to fellowship focus (e.g., an individual with primary opiate dependence attending AA instead of NA, or a primary alcohol dependent individual attending NA instead of AA). Overall, the bulk of the literature evaluating the effectiveness of 12-step groups is focused on AA, to the exclusion of other fellowship types (Kelly, 2003). Finer distinctions of participation and engagement are needed to enhance our understanding of the effectiveness of 12-step fellowships as recovery resources and to enhance the effectiveness of clinical referrals. In the increasingly cost-constricting environment of specialized addiction treatment, clearer empirical specification of the acceptability and effectiveness of these free community resources, across different subgroups of individuals in need, is timely and highly relevant.

2.2 Procedures

Recruitment and assessment procedures were similar in the two studies. For the adolescent sample, potential participants and their parents (if the adolescent was <18) were informed about the study during the clinical assessment at treatment intake. Study staff contacted interested participants and their parents to provide a brief overview of the study and schedule an appointment for an interview. Consent was obtained from participants aged 18 and older and from the parents of younger participants (who assented to participate). Study interviews were conducted by trained research staff, on average within 10.6 days (SD=12.4) of admission. Participants were reimbursed $50 for their time. Of the 160 adolescents who were invited to participate in the study, 95% (n=152) agreed to be contacted by study staff and, of these, 83.6% (n=127) were enrolled. Reasons for non-participation included being unable to schedule an interview during the first month of treatment (48%), deciding not to attend treatment (24%), parental non-consent (20%), participant non-consent (4%), and transportation difficulties (4%).

A small reliability substudy was conducted at one of the follow-up time points within the adolescent study, in which 11 participants completed two study interviews approximately 1 week apart, providing data for a preliminary consideration of test-retest reliability. The main study and the reliability substudy were approved by the Partners Health Care System Internal Review Board and all participants signed informed consent documents.

For the young adult sample, potential participants were approached by study staff in person shortly after admission. All admitted individuals aged 21–24 were approached, as well as every second individual aged 18–20, to ensure representation of the older age group given the predominance of those aged 18–20 at the treatment center. A small number of potential participants left treatment before recruitment could take place (n=6) or were not approached by staff for recruitment (n=14). Of the 384 clients who were invited to participate, 78.9% (n=303) were enrolled. Reasons for non-participation included not wanting to participate in follow-up interviews (44%), not being interested in the study (31%), wanting to focus on treatment (14%), and legal issues (2%). Following enrollment, an additional 17 participants withdrew prior to the assessment. Trained study staff conducted the interviews, which were completed either in person or by telephone. Participants were reimbursed $30 for their time. The study was conducted in accordance with the Institutional Review Board at Schulmann Associates IRB, an independent review board, and all participants signed informed consent documents.

2.3 Data collection

As part of the larger assessment battery for the study, short structured interviews were conducted to provide in-depth information on 12-step participation across the seven dimensions and types of fellowships (AA, NA, CA, and “other”). The interviews were conducted by research personnel, and took approximately 3–8 minutes to complete depending on respondent history of 12-step attendance. The questions, cross-tabulated by dimension, are summarized in Table 2; the full measure can be found in the Appendix. Items used a variety of response formats and specified either a lifetime or 3-month time frame for recall, depending on the nature of the dimension. In addition to those listed in Table 2, four open-ended items assessed likes and dislikes of 12-step groups and any reasons for non-attendance and dropout. Cutting across the identified dimensions of participation, these were designed to elicit additional information to contextualize respondent experiences and to capture barriers to participation and reasons for drop-out, furthering meaningful interpretation of the quantitative data.

Responses to the two items assessing lifetime and 3-month attendance provided an anchor for the remaining content of the measure for each respondent. Specifically, those reporting any attendance in the past 3 months were asked to provide additional details on fellowship involvement, participation in meetings, and progress in step work during this time. The items assessing step work specifically targeted the five steps that require more distinct tangible action on the part of respondents (i.e., steps 4, 5, 8, 9, and 11). Those reporting any lifetime attendance at meetings, regardless of their past 3-month pre-treatment attendance, were asked items on the total number of steps they completed at least once, their affiliation and sense of safety, and to identify, in open-ended format, the most and least helpful aspects of 12-step groups. Those who reported having attended meetings in the past, but not in the past 3 months, were also asked to indicate, in an open-ended format, why they had stopped attending meetings. Those with no lifetime attendance were similarly asked to provide reasons for non-attendance. Finally, in recognition of the fact that individuals may choose not to fulfill court mandates to attend meetings, all respondents, regardless of the lifetime or recent attendance, were asked to report on mandated attendance. With the exception of the open-ended items and those assessing mandated participation, all questions collected information separately by fellowship type.

3.2 Recent levels of participation, involvement and step work

Table 4 summarizes recent levels of fellowship involvement, meeting participation, and progress in step work in the subset who reported attending each fellowship type in the 3 months prior to treatment entry. Only a small number of adolescents considered themselves to be a member of a fellowship at this point in time. Higher proportions of young adults endorsed this item, with 44%–52% identifying as members of AA or NA. Approximately one-third of young adults who were attending AA reported having a sponsor, and almost all of these individuals had contact with their sponsor outside of meetings. Fewer of the adolescents and young adults attending other fellowships reported a sponsor. Sponsoring other individuals was particularly rare in these samples. In contrast, contacts with other group members were more commonly endorsed, particularly among the young adults.

Participation in meetings, defined in terms of sharing or talking in a meeting, was endorsed by approximately three-quarters of young adults and 40% of adolescents who had attended each type of fellowship in the past 3 months. A minority reported helping to set up or run meetings during this period.

As with fellowship involvement, step work was limited in these samples. Among those with recent meeting attendance, between 59% and 71% of respondents across samples and fellowship types reported that they had not completed any of the 12 steps, and only a handful of participants in either sample reporting having completed any of the steps requiring specific actions (e.g., writing an inventory, making direct amends). That said, a majority of young adults who had attended groups during the past 3 months (57%–74%) indicated that they had completed Step 11, which involves praying or meditating outside of meetings.

3.3 Court mandated attendance

Similar proportions of respondents in the two samples had been mandated by the courts to attend 12-step groups, including 10% of adolescents (n=13) and 11% of young adults (n=33). Among the adolescents, respondents reported being mandated to attend a median total of 6 meetings (ranging from 2–90) or a total of 2 meetings per week (ranging from 1–7), for a median period of 3 months (ranging from 0.5–24). Among the young adults, respondents reported being mandated to attend a median total of 5.5 meetings (ranging from 1–25) or a total of 2 meetings per week (ranging from 1–14), for a median period of 2 months (ranging from 1–36). Respondents varied in the degree to which they attended these meetings as required: 8% of the adolescents and 15% of young adults attended no meetings, while 75% of adolescents and 47% of young adults attended fully as required, with the remainder of clients falling between these extremes.

3.4 Affiliation and safety

For the subset of respondents who had ever attended each of the fellowship types, Table 5 summarizes overall experiences of affiliation and safety. Respondents varied in terms of the degree to which they perceived 12-step groups to be important and helpful to their recovery efforts. Mean ratings among adolescents for AA/NA were 5 or lower on the 10-point scale, indicating low average perceived importance or helpfulness. Young adults reported a higher degree affiliation with AA/NA, on average, but with lower ratings for CA and “other” fellowship types. Average ratings of the level of enjoyment of meetings followed a similar pattern.

Threats to safety, including incidents of intimidation, threats, or sexual harassment, were rare. These exceeded 10% percent for NA attendance among adolescents, although the absolute number of affected individuals (3) was low. Overall, respondents in both age groups reported high levels of perceived safety at meetings, with average ratings all exceeding 7.5 out of 10.

3.5 Open-ended responses

Of those who had never attended any 12-step meetings, 98% of adolescents (72/73) and 91% of young adults (95/105) provided reasons in the open-ended question addressing non-attendance. Of those with prior 12-step experience, 96% of adolescents (52/54) and 72% of young adults (141/197) provided information on what they found most helpful, while 96% of adolescents (52/54) and 57% of young adults (112/197) provided information on what they liked least. Among those with lifetime, but not recent, attendance, 82% of adolescents (14/17) and 63% of young adults (55/87) provided information on why they had stopped attending meetings.

This study was supported by the National Institute of Alcohol Abuse and Alcoholism (NIAAA) grant numbers R21AA018185-01 and R01AA015526.