Web-based Treatment for Rural Women with Alcohol Problems: Preliminary Findings

Abstract

Alcohol abuse among women in rural areas is a concern, particularly since treatment is not readily accessible, and women are highly susceptible to the ill affects of alcohol misuse. The Internet may offer a treatment delivery alternative for women who cannot easily take part in gender-focused programs that are located in urban centers. The purpose of this randomized study was to evaluate a 90-day web-based treatment program for women in rural areas of Missouri with problem drinking habits. The online treatment program consisted of reference and decision-making modules, synchronous and asynchronous communication features, and the option to interact privately with the researcher. Preliminary findings relating to participant demographics, website use, and satisfaction are reported in this article. The enrolled participants (n = 46) averaged 50 (SD = 11) years of age, were relatively well educated and, to a large extent, financially secure. Based on their website usage, participants appeared to prefer reading communiqués over actively posting messages to the asynchronous bulletin board. In general, participants indicated satisfaction with the program, and 83% noted that they would recommend it to a friend.

It is estimated that 19% of women in the United States have a history of alcohol abuse or alcoholism sometime during their lifetime, and few of them receive treatment.¹ Barriers to professional care are greatest in rural areas and include limited financial resources, a scarcity of gender-focused services, privacy issues, and stigma associated with alcohol abuse among women. Lay-led support groups that may exist in rural areas are likely to be populated by men, and child-care services are usually not available.²

The lack of treatment options is highly detrimental to women since they are particularly susceptible to the harmful effects of alcohol misuse. Even when a woman consumes less and for a shorter period of time than a man, she is at greater risk of developing alcohol-related health problems.^3,4,5 These problems include heart, liver, and brain disease; and cancers of several body systems, including the digestive track, breast, head, and neck. Women who misuse alcohol also have a greater chance of becoming victims of violence, sexual assault, and injury due to accidents. Among women of childbearing age, fetal alcohol syndrome is a concern,⁵ whereas, older women are particularly susceptible to medication interactions and osteoporosis.⁶

It has been suggested that the World Wide Web may offer a treatment delivery option for women who cannot easily access more traditional services.^7,8,9 Approximately 66% of American women go online. Of that number, 74% are thought to use the Internet to look for health and medical information, and 66% get support for a medical issue or personal problem. Nine percent have searched for information related to alcohol or drug abuse.¹⁰ Among those who currently use online alcohol treatment sites, 67–72% of users appear to be women.^7,8

Based on findings from the first known randomized-controlled study (N = 261) of an Internet-based intervention to promote low-risk drinking among the general public, researchers concluded that the web may be well-suited to meet the needs of adult women. At 6-month follow-up, 17.2% of individuals who received an online intervention were drinking within normative limits versus 5.4% in the control group. These results reflect a high 1:1 (woman to man) participation rate versus a 4:1 rate among problem drinkers in the general population.¹¹

In light of these findings and the treatment barriers confronted by women with alcohol problems in rural areas, a 90-day web-based self-help treatment program was developed and evaluated. Preliminary results from this pilot investigation relating to participant demographics and use and evaluation of the treatment website are reported in this article.

METHODS

Recruitment and Sample Selection

For purposes of this study, rural was defined as counties that are classified as 25 to 100% rural by the United States Census Bureau (USCB).¹² This resulted in the exclusion of 10 out of 114 counties and 1 independent city within Missouri. Due to state licensing regulations, which prevent inter-state nursing practice via the Internet, only Missouri residents were eligible to participate. This investigation received Institutional Review Board (IRB) approval.

Recruitment efforts spanned from January 2006 until February 2007. During that time, two high-intensity newspaper and radio advertising campaigns were conducted. Recruitment was also carried out using unpaid television interviews and announcements plus information posted to community websites and closed circuit television monitors at university outreach centers. E-mail messages were sent to public health agencies, nursing organizations, rural extension sites (educational and health-related), and university faculty and staff located in rural counties. Hardcopy flyers were also mailed to public health agencies and churches in rural communities with traditionally Black congregations. University students were not targeted for enrollment since many undergraduate students cannot legally purchase alcohol, and this was not strictly an abstinence based program. In addition, the treatment materials were developed with an older demographic in mind.

Interested individuals contacted the researcher by telephone for information about the study. During this conversation, the investigation was fully explained and callers' questions were answered. Individuals who remained interested in the study were then asked to respond to a brief screening questionnaire to determine their potential eligibility to participate. Women who were 18 years of age or older and from rural areas of Missouri qualified to enroll providing they had maintained a drinking pattern of more than 3 drinks per day or 7 drinks per week for at least a year. These limits are based on criteria set for women by the National Institute on Alcohol Abuse and Alcoholism.¹³

Key exclusion criteria included (a) current participation in an in- or out-patient treatment program; (b) use of street drugs or prescription anti-anxiety or hypnotic medications, habit-forming stimulant medications, or pain relief medications in doses higher than recommended by a health care provider; (c) use of prescription medication to treat an alcohol problem; (d) a history of severe alcohol withdrawal symptoms (e.g., hallucinations, convulsions); (e) mandates to participate in alcohol treatment due to legal infractions; or (f) current suicidal ideation. Women who did not have access to a computer with Internet capabilities or who were unfamiliar with how to use this type of equipment were excluded.

Women who qualified to participate and who continued to express interest in the program during the initial phone contact were randomized to standard or web-based treatment. A no-treatment control or wait list group was ruled out due to the potential for serious health-related consequences such as automobile accidents. For the same reason, a rolling enrollment process was used so that motivated individuals could start the treatment program as soon as their signed consent form was returned to the researcher. All participants had access to the treatment materials for 90 days regardless of when they enrolled.

In adherence with IRB guidelines, all qualified individuals were sent two hardcopy consent forms (1 to retain, and 1 to sign and return) via ground mail. Upon return receipt of their signed consent forms, web-based participants received a letter that included instructions for accessing the secure treatment website using an alias and a personalized password. Only individuals who were officially enrolled in the treatment program and had been issued an alias and a personalized password could access the treatment website. Individuals who did not access the website within a few days after receiving their letters were sent an e-mail message encouraging them to do so.

With the exception of one assessment instrument that was difficult to use within the software selected for this project, online participants completed all baseline assessments on the treatment website. The remaining assessment instrument was mailed and returned with their consent form. Online participants could only access the treatment portion of the website after they had submitted all baseline data. Women in the standard-care group completed baseline assessment instruments in hard-copy format and returned them in pre-addressed postage-paid envelopes along with their consent forms. Individuals who did not return hardcopy materials within two weeks were sent duplicates. This process is illustrated in Figure 1.

Figure 1.

Sample Selection Process.

Treatment Components

WebCT (Blackboard, Washington, DC) software was used to deliver the online treatment components, which were developed based on principles of the Harm Reduction Model,¹⁴ self-efficacy theory,¹⁵ the Transtheoretical Model of Change,¹⁶ motivational interviewing,¹⁷ and cognitive behavioral skills therapy.^18,19 The treatment program consisted of 8 reference modules and 15 decision-making modules, which are listed in Figures 2 and 3. The modules ranged from 1 to 3 pages in length. Except for a single module that provided a list of references, the treatment materials did not contain links to web pages outside of the treatment program.

Figure 2.

Reference and Decision Making Modules: Hits per Page.

Figure 3.

Reference and Decision-making Modules: Average Time per Hit.

The reference modules were comprised of didactic content relating to stages of change, differences between alcoholism and alcohol abuse, alcohol withdrawal, health problems associated with alcohol problems, fetal alcohol syndrome, safe drinking guidelines, simultaneous use of alcohol and medications, and additional reference sources. The decision-making modules were comprised of problem solving exercises and activities. They included content pertaining to the pros and cons of change; moderate drinking versus abstinence; goal setting; strategies for cutting back, turning down a drink, rewarding oneself, and balancing life. Topics such as management of slipups, craving, triggers, bad moods, and sleep problems; daily alcohol use monitoring; filling time; and involving others in change were also addressed in the decision-making modules.

Web-based treatment components were developed with the limitations of rural-based Internet connections in mind.²⁰ Due to bandwidth restrictions and relatively slow Internet connections, file sizes were kept small (55 to 1,560 kilobytes). Colorful clip art was used to enhance the appearance of documents, which were mounted in Rich Text Format.

A highly-prescriptive dose-based approach to treatment is antithetical to the underlying principles of the Harm Reduction Framework¹⁴ and is inconsistent with the progressive-regressive-progressive nature of resolving alcohol problems within the context of the Transtheoretical Model of Change.¹⁶ In keeping with these perspectives, participants had access to all of the modules throughout the 90-day treatment period, and they were encouraged via synchronous bulletin board messages to access them on an as-needed basis. Women in the standard care group were mailed the same modules in a 48-page spiral-bound treatment booklet along with instructions for contacting the researcher by telephone regarding questions.

Web-based participants also had access to an asynchronous bulletin board and a synchronous chat feature, both of which could be used to interact with the researcher and the other participants. The intent of these communication features was to provide a supportive community for change. In addition, participants could send private messages to the researcher and other enrollees. Upon accessing the full treatment website, participants were immediately invited to introduce themselves using the asynchronous bulletin board. Participants were also encouraged to use the synchronous chat feature once they had become familiar with the web environment.

Researcher-designed satisfaction instruments (web and standard-care versions) were mailed 90 days after participants accessed the treatment website or they received their materials by mail. The Participant Satisfaction Questionnaire (PSQ) was based on results from a preliminary investigation⁷ and Zeithaml and colleagues²¹ empirically-identified evaluative dimensions of service quality. Face validity was established.

Each version of the PSQ (standard-care and web) consisted of items that participants ranked using a 5-point Likert scale. Responses ranged from strongly disagree (1) to strongly agree (5). For example, web-based participants were asked to record their level of agreement with the following statement: The treatment website was easy to use. Conversely, women in the standard care group were asked to respond to a corresponding statement: The program materials were easy to use. Each version of the instrument also contained open-ended statements such as the following. Please share any ideas you have about how the treatment program can be improved.

Web-based participants were not asked to complete these instruments online due to limited use of the website near the end of the treatment period. If follow-up assessments were not returned within two weeks, a second mailing was forwarded, and an e-mail reminder was sent. All participants received $10 when they returned their satisfaction questionnaires.

Data Analysis

Simple descriptive statistics were used to analyze demographic and participant satisfaction data. The Wilcoxon Rank Sum test was used to compare treatment program satisfaction between groups. Since the standard care and web versions of the PSQ were not identical, questionnaire items were categorically grouped, and the means were compared. Participants offered very few relevant responses to open-ended requests for evaluative feedback, and it was impractical to systematically draw substantive inferences from these data.

Messages posted to the asynchronous bulletin board were electronically compiled and placed into a data analysis matrix (i.e., table). Messages sent by the researcher and the study participants were separated so that they could be analyzed independently. Each message was carefully read, coded, and categorized based on the type of communication it represented (see Table 1).

Table 1.

Communication Categories

Messages from Researcher
Categories	n	%
Positive reinforcement	22	32.8
Encouragement	12	17.9
Reflection	10	14.9
Suggestion	8	11.9
Probe	4	6.0
Share information	3	4.5
Expression of hope	3	4.5
Instructions	2	3.0
Response to procedural matter	1	1.5
Welcome	1	1.5
Introduction	1	1.5

Messages from Participants
Categories	n	%
Progress report	12	11.4
Introduction	11	10.5
Goal setting	11	10.5
Drinking triggers	11	10.5
Solicit interaction	9	8.6
Barriers to change	8	7.6
Assets	7	6.7
Self-assessment	6	5.7
Identification with others	5	4.8
Strategies to cut back/abstain	5	4.8
Thanks	4	3.8
Emotional venting	3	2.9
Anonymity important	3	2.9
Website anxiety	2	1.9
Procedural matters	2	1.9
Expression of hope	2	1.9
Empathizing	1	1.0
Advice to others	1	1.0
War stories	1	1.0
Problem using website	1	1.0

Results

Demographics

Recruitment efforts resulted in 88 telephone inquiries. Newspaper advertising accounted for 74% (n = 65) and radio advertising was responsible for 17% (n = 15) of the telephone calls. The remaining inquiries 9% (n = 8) stemmed from a variety of recruitment strategies such as e-mail and closed-circuit television announcements. Of the 88 callers, 21 (24%) could not be randomized to a treatment group. Eight did not reside in a rural county, 5 did not currently have access to the web, and 1 declined to participate due to concerns about anonymity/confidentiality. Three women had a history of severe alcohol withdrawal symptoms, and one felt suicidal. These individuals could not be safely treated within the study protocol. Three additional individuals did not meet the threshold for problem drinking that was established for this investigation (more than 3 drinks per day or 7 drinks per week¹³).

Sixty-seven women were randomized to the web-based care (n = 34) or standard care (n = 33). Twenty-four individuals within the standard care group returned their signed consent forms and hardcopy baseline assessment instruments, and they were forwarded bound treatment materials via surface mail. Twenty-six web-based participants returned their signed consent forms, and 20 completed all online assessment instruments and gained full access to the treatment website. Twenty-nine individuals (13 web; 16 standard care) returned 3-month follow-up data.

As a result of random assignment, the web-based and standard care groups were not significantly different in terms of demographics. The average age of the women (n = 46) who completed the baseline assessment instrument was 50 (SD = 11). All of the women were White and of non Hispanic or Latino descent. Ninety-eight percent (n = 45) had a high school diploma or general equivalency degree; 46% (n = 21) had a bachelor's degree or above. Seventy-six percent (n = 35) were employed full- or part-time. Annual household income ranged from $20,000 or less to more than $125,000. Forty-eight percent (n = 22) came from households in which $50,000 or more was earned per year.

Website Use

WebCT software is designed to monitor participants' website activity. In addition, the researcher monitored the treatment website on a daily basis. Participants were apprised that their web activities would be monitored in the consent materials.

Everyone who accessed the program materials was considered to have received treatment, since participants could print the materials from the website. In addition, some women logged on infrequently, but their visits were relatively lengthy.

Participants had access to the treatment website for 3 months, and they remained active anywhere from 1 to 88 days. The average was 40 days. Individual use of the website ranged from a low of 5 to 185 hits per individual (median = 71). This includes the number of times a single person visited the treatment site's homepage, baseline assessment instruments, and reference or decision-making modules. The average number of hits per person was 78.

Use of the 23 reference and decision-making modules included in this treatment program is illustrated in Figures 2 and 3. Of the 10 modules visited most often, 9 primarily offered information and/or assisted individuals to self-assess their situations. With diminishing frequency, participants viewed modules that emphasized taking action; such as goal setting, cutting back, and managing drinking triggers.

Although the module dealing with balance in one's life was viewed the least (i.e., 11 hits), it rated highest in terms of average time per hit (~30 minutes). The module relating to useful resources rated second in terms of average time per hit (~10 minutes), however, this probably reflects the fact that it contained links to resources outside of the treatment website, which could be viewed without closing the original treatment module. Other modules that individuals viewed for more than 3 minutes on average offered a mix of reference and self-assessment information along with some action-oriented content.

WebCT software allows participants to forward private messages to the researcher and each other, but the system does not track private interactions among enrollees. Participants were not encouraged to communicate with each other privately, and results from the follow-up satisfaction survey suggest that they did not. There is also no evidence to suggest that participants traded personal e-mail addresses, telephone numbers, or ground mail addresses.

The researcher received 13 private messages. Nine came from a single individual, and 6 dealt with computer problems resulting in large part from the use of an old operating system. This individual also initiated a brief interaction about a newspaper article relating to moderation of drinking habits. Two other individuals sought clarification regarding use of the website. One person requested information about local Alcoholics Anonymous meetings, and another asked if it was possible to extend her treatment period. Among the standard care participants, the researcher only received a few telephone calls, which routinely related to study mechanics (e.g., questions about the consent form).

The researcher encouraged the organization of real-time chat groups, but participants did not appear interested in committing to formalized discussion times, topics, or groups. Instead, they preferred the informal 24/7 availability of the asynchronous bulletin board. On a few occasions, women invited others to interact using the synchronous chat feature, but specific times to interact were not established.

Bulletin Board Messages

Sixteen participants read and/or sent messages to the asynchronous bulletin board. The total number of messages posted was 69. Of that number, 43% (n = 30) were sent by the researcher. The number of messages posted to the asynchronous bulletin board per participant ranged from 0 to 7 with an average of 2 (median = 1). The number of messages read ranged from 0 to 61 with an average of 25 (median = 13). Analysis of these messages resulted in the categories that are listed in Table 1. Categories pertaining to the researcher's and participants' messages are displayed separately and in descending order of occurrence.

Researcher

The researcher interacted with participants on the bulletin board using communication techniques such as positive reinforcement, encouragement, reflection, suggestions, and probes. Positive reinforcement was the most frequently used communication strategy. The primary purpose of this technique was to acknowledge success and to promote ongoing engagement in the program. It was used to praise participants for doing things such as posting a message to the bulletin board, sharing information, demonstrating insight into situations, taking steps in the right direction, or achieving a small goal.

Encouragement was the second most frequently used communication technique. The researcher employed it to initiate action such as reviewing treatment modules, posting to the bulletin board, setting drinking goals, and problem solving. Reflection was also used to help individuals focus on important information. For instance, the researcher pointed out perceptive insights and personal assets that might assist the women to resolve their alcohol problems.

Direct probes were used on occasion to promote problem solving, however, this communication technique was employed sparingly to avoid alienating respondents and to encourage them to move at their own pace. Suggestions were offered, but participants were routinely placed in the role of decision maker. To insure that participants had an informed basis for making decisions, suggestions were frequently linked with information or reflective inferences.

Participants

Participants used the bulletin board primarily to introduce themselves, provide progress reports, and discuss drinking goals and triggers. The women also solicited interaction from other participants within the treatment program, explored personal assets and barriers to change, and performed self-assessments. Less frequently, the women discussed active attempts to cut back on their drinking such as walking, alternating a non-alcoholic drink with one containing alcohol, and posting messages to the treatment website. Thanks were expressed when others provided helpful feedback.

Communiqués were generally here-and-now focused, and the bulletin board was sometimes used for emotional venting. Incidental messages were posted to share hope, empathy, advice, and war stories; but these were more the exception than the rule. On occasion, the women interacted directly with one another, but messages were usually posted without a particular recipient in mind. Infrequently, participants communicated about procedural matters, the importance of anonymity, and anxiety about using the website. Only one person addressed problems with the technology.

Treatment Satisfaction

The majority of web-based participants who completed the researcher-designed follow-up satisfaction questionnaire found the treatment website attractive (67%; n = 8) and easy to use (75%; n = 9). They also perceived that the website promoted communication (66%; n = 8), and 50% (n = 6) felt that the asynchronous bulletin board was a valuable resource. In contrast, 67% (n = 8) had no opinion about the synchronous chat function, and only 33% (n = 4) felt that being able to communicate privately with the researcher was a valuable component of the program. Only one person had concerns about the privacy of her website activities, and 83% (n = 10) indicated that they would recommend the program to a friend. Sixty-seven percent (n = 8) indicated that it was important for them to be able to print the website modules for use at a later time.

Although 75% (n = 9) of the web-based participants felt that the modules containing reference information were difficult to understand, 67% (n = 8) thought they contained valuable information. Sixty-seven percent (n = 8) of the women also perceived that the decision-making exercises were helpful, and the goal setting module was endorsed by 75% (n = 9) of the respondents. Seventy-five percent (n = 9) of individuals found the self-assessment modules helpful. There were no significant differences between the standard-care and web-based groups in terms of treatment program satisfaction.

Discussion

Based on the average age of the randomized respondents it is inferred that this online treatment program appealed to an older group of individuals who may not normally access traditional treatment programs. The fact that 76% were employed full-or part-time also indicates that there may be a need for treatment programs that accommodate the busy schedules of working women.

Despite the fact that recruitment advertising for this investigation targeted rural residents, 9% of the telephone inquiries came from women who did not reside in rural locations. Accordingly, it appears that face-to-face treatment barriers for women with alcohol problems are not limited to rural areas, and targeting participants across urban and rural locations is appropriate.

According to the USCB,¹² approximately 36% of Missouri households earn more than $50,000 per year. This compares with 48% of the households in this study. In terms of education, approximately 82% of White non-Hispanic/Latino women in Missouri have a high school diploma or GED, whereas, 98% of the women in this study had a high school diploma or general equivalency degree. Despite growth in rural Internet access in recent years,²⁰ it appears that individuals in higher socio-economic brackets are still more likely to have a web connection.

As the number of online health care resources increases, individuals who are not linked to the web may find themselves at a significant disadvantage in terms of treatment options and access to essential information. On a more optimistic note, it appears that once individuals have access to the Internet, they do not have a great deal of trouble learning how to use it. Only one participant appeared to struggle with basic computer skills, and even that person was able to function within the online environment with encouragement.

Twenty-three individuals were lost to treatment because they did not return their signed consent forms and baseline assessment instruments. To avoid attrition in the future, it is recommended that respondents complete and submit informed consent and baseline data online. Individuals in the web group could then gain immediate access to the treatment website, and those in the standard care group could receive an electronic message indicating that their treatment materials were in the mail.

Within the web group, 35% (n = 7) of enrollees were lost to 3-month follow-up. Although this rate of attrition is less than ideal, it is consistent with results from other online treatment programs. For example, Riper et al. reported a dropout rate of 42.1% at 6-month follow-up.¹¹ In general, high attrition rates are common to self-help type online services, since non-use and discontinuation of treatment are very easy.²² This limitation may be offset by the vast number of people that can presumably be reached with website interventions.

Estimates of the amount of time participants spent viewing modules are limited by the fact that they could have stepped away from their computers while a webpage was open. In addition, 67% of the women found it important to be able to print materials from the treatment website. It is possible that individuals were thinking ahead to a time when they would no longer have access to the online information. In real-world circumstances, it is surmised that web-based materials may be intermittently accessed over time, and having a stable web presence is important. Also, making web pages printer friendly is recommended.

Participants' use of the asynchronous bulletin board was limited. These findings are consistent with those generated from a study of the AlcoholHelpCenter.net online support group. Of 155 participants, only 50 (32%) posted a total of 474 messages over the course of 283 days. Support staff were responsible for an additional 200 messages. Based on these results, the researchers concluded that without staff participation, the support group was not self-sustaining. The researchers surmise that enrollees may have been primarily interested in reading messages (i.e., lurking) rather than actively posting.²³

Another explanation is that participants primarily logged on to use informational resources or self-assessment tools, and they were not interested in the website communication features.²³ This may explain why relatively few messages were posted by the participants in this study since they responded to an advertisement about a web-based treatment option versus a traditional face-to-face treatment program in which interpersonal interaction is the norm. In addition, the fact that participants entered and exited the program at different times may have discouraged the development of cohesive relationships.

Given the limited amount of active posting that occurred on the treatment website, it is interesting to note that 50% of the follow-up respondents felt that the asynchronous bulletin board was a valuable resource. Based on this finding, it is inferred that passive participation (i.e., lurking) may be therapeutic. Moreover, lurking may have the same therapeutic effect as listening to group members' stories in face-to-face settings. The latter is speculative, however, and there is no substantive evidence to support the idea that online groups can serve as a substitute for traditional group interaction.²³

Another factor that may play a role in limited bulletin board activity relates to the Pew organization's report that the proportion of women Internet users who participate in online chats and discussion groups dropped from 28% in 2000 to 17% in 2005. This decline is attributed to an increased awareness of and sensitivity to worrisome behavior on interactive sites.¹⁰ In the current study, there also may have been some concerns about anonymity and confidentiality, since the importance of anonymity was noted three times within participants' posts, and one individual declined to enroll due to privacy issues.

Although participants' active use of the asynchronous bulletin board was limited, the messages posted were all appropriate to the mission of the treatment program. The messages were also largely parallel to the kinds identified by other researchers who have analyzed communiqués sent to professionally-moderated alcohol treatment websites.^8,23 The fact that participants demonstrated very limited interest in communicating privately with the researcher probably relates to her ongoing availability on the treatment site bulletin board.

Given the number of enrollees, the time demands involved in maintaining the treatment website and reading and responding to communiqués was very reasonable. In comparison, the mechanics of managing the research project (recruiting participants, mailing enrollment and follow-materials, complying with IRB regulations, etc.) were more time consuming. This finding bodes well for the financial feasibility of online treatment.

The fact that 75% of the web participants found the modules difficult to understand was surprising, since everyone in the web group had at least a high school diploma or a general equivalency degree, and the average Flesch-Kincaid grade level of the modules was 6.45 (SD = 1.3). This finding may be explained by the fact that several of the reference modules contained medical terminology. In addition, many of the decision-making modules provided participants with basic information and then asked them to problem solve. Some individuals may have found the latter challenging. Another possibility is that participants may have linked to web pages outside of the treatment site that were difficult to understand since they were not vetted for readability.

Summary

Despite the fact that there were no significant differences between the standard care and web-based groups in terms of treatment program satisfaction, an online delivery system has the potential to cost-effectively reach individuals who would not normally have access to alcohol treatment services. Thus, this treatment delivery method is worthy of further investigation. In particular, researchers are encouraged to evaluate the potential therapeutic value of self-assessment aids, reference materials, and lurking.