The Internet Diabetes Self-Management Workshop for American Indians and Alaska Natives

Abstract

Type 2 diabetes disproportionately affects American Indians and Alaska Natives (AI/ANs). In the larger population, patient self-management has become an increasing focus of the health care system to help reduce the impact of diabetes. However, little is known about patient self-management programs designed for AI/ANs. This study reports on the feasibility of implementing the Stanford Internet Diabetes Self-Management Workshop within the AI/AN population using a participatory research approach. This is a continuation of self-management studies to assist in meeting the needs of both patients and the health care system for health services that are effective (evidence based), efficient, and culturally appropriate. To our knowledge, this is the first study examining the effectiveness of an Internet-based diabetes patient self-management program among AI/ANs. This article reports on a pilot for a larger randomized study that is ongoing.

Type 2 diabetes disproportionately affects Indigenous people, referred to in this article as American Indians and Alaska Natives (AI/ANs). Indian Health Service data found that between 1994 and 2002 the age-adjusted prevalence of diabetes increased 33.2% among AI/AN adults, from 11.5% to 15.3%, and 73.7% among those aged 20 to 34 years, from 1.8% to 3.1% (Acton, Burrows, Geiss, & Thompson, 2003). The Strong Heart Study, a longitudinal study examining cardiovascular disease and associated risk factors among American Indians, also found that the prevalence of diabetes increased from 6% to 12% during a 4-year period, from 1989–1991 to 1993–1995 (Welty et al., 2002). Among the communities participating in the REACH 2010 initiative funded by the Centers for Disease Control and Prevention, American Indians have a higher prevalence of diabetes than any other ethnic group (Galloway, 2005).

Patient self-management has become an increasing focus of the health care system to help reduce the impact of diabetes; however, little is known about the efficacy of diabetes self-management programs for AI/ANs. Although some clinics within the Indian health care service setting offer diabetes self-management programs (Indian Health Service, 2007; www.ihs.gov), the outcomes of these programs are not discussed within the scientific literature. The literature to date includes community-based interventions focusing on increasing physical activity and healthy eating (Gilliland, Azen, Perez, & Carter, 2002; Griffin, Gilliland, Perez, Helitzer, & Carter, 1999; Leonard & Wilson, 1986; Narayan et al., 1998) and awareness campaigns about diabetes are discussed within the literature (Roubideaux et al., 2000).

The potential for Internet-based patient self-management programs has rapidly expanded as more than two thirds of the U.S. population report being online (Fallows, 2005). Although Internet usage varies by income, education, employment status, age, and race, there is evidence that this digital divide is quickly narrowing (Heuertz, Gordon, Moore, & Gordon, 2002). A 2002 Bill and Melinda Gates Foundation report stated that 69% of AI/ANs report using the Internet (Heuertz et al., 2002), compared with 73% of Whites, 61% of Blacks, and 76% of English-speaking Hispanics (Pew Internet and American Life Project, 2006). Internet-based diabetes self-management programs have been found to be accepted and effective approaches for changing behaviors and improving health outcomes, with high participation and satisfaction rates and participant reports that they are easy to use (Glasgow, Boles, McKay, Feil, & Barrera, 2003; Glasgow & Toobert, 2000; McCoy, Couch, Duncan, & Lynch, 2005; McKay, Feil, Glasgow, & Brown, 1998).

This article examines the feasibility of implementing the Stanford Internet Diabetes Self-Management workshop (IDSMW) within the AI/AN population using a participatory research approach. To our knowledge, this is the first study examining the effectiveness of an Internet-based diabetes patient self-management program among AI/ANs. We report on a pilot for a larger randomized study that is ongoing.

METHOD

The aims of the feasibility study were to (a) conduct a process evaluation of IDSMW use (e.g., number of logins, completion of assignments, parts of workshop used); (b) assess AI/AN demographic and patient characteristics associated with accessing, using, and completing the IDSMW; (c) assess the feasibility of recruiting AI/ANs via the Internet into the workshop; and (d) assess the cultural appropriateness and acceptability of the IDSMW with AI/ANs.

Description of the Intervention

The Stanford IDSMW is a 6-week workshop offered for 20 to 25 persons. It uses two trained peer moderators. Participants are asked to log in at least three times each week for a total period of about 2 hr and to participate in the weekly activities. These include reading the online content for the week via the Learning Center. Topics include American Diabetes Association–required content that discusses causes of diabetes, principles of problem solving, healthy eating, menu planning, label reading, exercise, preventing and treating hyper and hypoglycemia, preventing foot and eye complications, managing sick days, medications, and dealing with negative emotions such as stress, anxiety, and depression, dealing with family relationships, and improving patient–provider communications.

There are four areas in each workshop: the Learning Center, the Discussion Center, My Tools, and Help. The Learning Center offers the didactic content of the workshop. Parts of the learning center are tailored to the individual participant by links with their baseline questionnaire, for example, feedback on the meaning of their hemoglobin A1c (HbA1c) and how to start or build on an exercise workshop. Each week, new content is made available. Past weeks’ content also remains available.

The Discussion Center is interactive and includes four directed bulletin boards centered on action planning, problem solving, difficult emotions, and celebrations. Each week, participants posted a new action plan, that is, something they planned to do, on the action planning board. These action plans were prompted by the Learning Center. Participants also received weekly prompts to post to one or more of the other bulletin boards. In addition, participants wishing to communicate with each other individually are able to do so with an anonymous internal post office message center. My Tools contains journals, glucose and exercise-monitoring logs, medication records, and food-planning tools. The Help section contains features that enable participants to take a tutorial on workshop use, learn computer skills, such as how to scroll and use pop-up windows, and contact study staff. Using a menu on each Web page, participants can move between areas as they wish.

The IDSMW is built for use by almost anyone who has access to the Internet and has a seventh to eighth grade reading level. Because many people log in via dial-up modems, video clips and extensive graphics are not used. All participants in the IDSMW are mailed a copy of Living a Healthy Life With Chronic Conditions (Lorig et al., 2000). This reference book was written based on needs assessments of persons with chronic conditions.

Theoretical Framework

The IDSMW is based on the self-efficacy theory developed by Albert Bandura (1986). The workshop integrates self-efficacy theory through the following ways:

Skills mastery.

Participants are asked to form a weekly action plan and try new behaviors such as menu planning. Each session includes a place for feedback on progress and discussion of problems. Other skills mastery techniques include exercise, practice in food selection, eating regular meals, and/or controlling portion size and quality and quantity of carbohydrates.

Modeling.

To enhance similarity between participants and moderators, it was ensured that at least one of the two peer moderators for each IDSMW had diabetes. For the AI/AN pilot, one of the two peer moderators was American Indian. The workshop offers structured opportunities for participants to support each other, with problem solving via bulletin boards. Thus, participants model for each other and, by serving as models, enhance their self-efficacy. Finally, class members are asked to check in with each other via the internal post office system between class meetings.

Interpreting symptoms.

Patients’ adaptations to diabetes are influenced by their beliefs about diabetes. Thus, if they believe that their fatigue is due to the disease process, they will rest. When it is explained that fatigue can also be due to reconditioning, poor nutrition, stress, or depression, participants have a rationale for trying new behaviors to manage fatigue. As each symptom or problem is discussed, the multiple possible causes are identified and a set of management techniques suggested. This allows participants to choose techniques that fit within their belief system. One can think of this as encouraging the participants to self-tailor the workshop.

Persuasion.

Participants are encouraged to share their action plans and other course activities with family and friends in order to gain a supportive environment for change. In addition, moderators urge participants to do a little more than they are doing now. Finally, the bulletin boards and the post office assist group members to interact with and support each other.

Workshop Recruitment and Sample Size

Two IDSMW pilot tests were conducted, one with AI/ANs (n = 27) and one with non-AI/ANs where participants were not specific to any one ethnic group (n = 27; total n = 54). Recruitment was conducted entirely via the Internet. Separate Web sites were used for recruitment, one for AI/ANs and one for non-AI/ANs. The AI/AN Web site included pictures and quotes from AI/ANs who had taken and helped to develop the community-based self-management course, and information about the workshop. The non-AI/AN Web site was an informational page with the Stanford Patient Education Research Center logo and information about the workshop. Interested AI/ANs were directed to the AI/AN Web site, whereas all others were directed to the study Web site. Both sites fed into the same database.

To direct individuals to the Web sites, a short listserver post was created for both AI/ANs and non-AI/ANs. To conduct targeted recruitment for AI/ANs, the AI/AN announcement was sent out through a listserver of approximately 150 people that included eight AI/AN organizations (four health clinics, two social service agencies, and one arts organization). These organizations were roughly equally distributed between urban and reservation settings. The same process was conducted for non-AI/ANs. The post was sent out to a diabetes health and wellness list server of nonprofit health agencies, a list server of e-newsletters, and Yahoo diabetes user groups.

Interested individuals linked to the Stanford Web sites and left their name, phone number, and e-mail address and revealed where they heard about the workshop. They were then sent an e-mail invitation with a link and login code to view the complete study description and online consent. The study was approved by the Stanford Institutional Review Board. If the individual consented to participate by clicking approve, they were given a link, a user name, and a password. These allowed them to access the baseline questionnaire. On completing the baseline questionnaire, the individual was mailed an HbA1c test kit, which required a finger prick and three drops of blood. The blood was returned for processing to a central lab (Biosafe Laboratories; http://www.ebiosafe.com/). When results were received by Stanford, the individual was enrolled in the workshop.

Within 2 days of posting on the AI/AN listserver, 53 people left contact information. All were invited to participate and 30 people (57%) completed the baseline questionnaire. Twenty-seven participated in the online workshop. For the non-AI/AN pilot, 65 people visited the Web site. To accommodate the 25-person-per-class limit, only the first 50 were sent the invitation to participate. Of those 50, 30 people (60%) completed the baseline questionnaire and all 30 people participated. Both pilots represent non-matched convenience samples. The recruitment and retention rates for both pilots were comparable to the other Internet-based workshops that the Stanford Patient Education Research Center offers (Lorig, Ritter, Laurent, & Plant, 2006).

Sources of Data

There were five sources of data: (a) baseline questionnaire administered, (b) a home blood test measuring HbA1c, (c) a qualitative analysis of the bulletin board posts for the AI/AN pilot, (d) an Internet adapted focus group with the AI/AN participants after workshop completion, and (e) a process evaluation to assess similarities and differences between AI/AN and non-AI/AN pilot participants.

Measures

All self-reported measures are listed in Table 1. Demographic variables included age, gender, ethnicity, and education. AI/ANs also reported their tribal affiliation and tribal enrollment status. Health care use questions included site for accessing health care services, insurance status, number of visits to physicians, number of emergency department visits, hospitalizations, and nights in the hospital.

TABLE 1.

Demographic Variables and Baseline Mean Scores for Native and Non–Native American Participants

Variable	Native American (n = 27)	Non–Native American (n = 27)	p (difference)
Demographic variables
Age (range 27–65)	42.9	54.2	.001
Percentage female, %	87.1	85.9	.83
Years of education (range 10 to 22)	15.6	17.1	.03
Internet usage
Comfort level using the Internet (1–10) ↑	9.20 (1.61)	9.44 (1.05)	.519
Internet use frequency within the past 6 months (1–4), 1 = less use	1.34 (0.669)	1.18 (0.557)	.338
Visits to health-related Web site(s) within the past 6 months (1–4), 1 = less visits	3.06 (1.22)	2.88 (0.933)	.540
Health care use
Physician visits (past 6 months)	4.22 (4.55)	3.40 (4.43)	.508
Emergency visits (past 6 months)	0.444 (0.640)	0.629 (1.64)	.587
Health indicators
Health distress (0–5) ↓	2.08 (1.46)	1.76 (0.863)	.008
Self-reported global health (0–5) ↓	2.89 (0.900)	2.77 (0.847)	.613
Symptoms of high glucose level (0–6) ↓	2.42 (1.77)	1.92 (1.57)	.272
Symptoms of low glucose level (0–6) ↓	2.60 (2.24)	1.77 (1.25)	.096
Activity limitation (0–4) ↓	1.35 (1.22)	1.38 (1.11)	.933
Fatigue (0–10) ↓	5.58 (3.19)	5.25 (2.47)	.668
Pain (0–10) ↓	3.62 (3.09)	4.07 (3.12)	.588
Depression
Depression (0–3) ↑	0.992 (0.735)	0.559 (0.469)	.011
Health behaviors
Aerobic exercise (min/week) ↑	100 (99.1)	93 (101.7)	.788
Stretching and strength exercise (min/week) ↑	87 (115.3)	55 (59.1)	.205
Communication with physician (0–5) ↑	2.75 (1.26)	3.16 (1.13)	.219
Stress management practice (times/week) ↑	2.44 (2.30)	2.81 (2.49)	.573
Self-efficacy (1–10) ↑	6.44 (2.24)	6.90 (1.61)	.381
Patient activation (0–100) ↑	71.3 (18.4)	62.6 (12.5)	.067
Metabolic control
Weight, lb	208 (55.7)	172 (46.9)	.016
Hemoglobin A1c	6.79 (1.33)	6.86 (1.63)	.904

Note: The range and direction are given with each variable, where applicable. An upward arrow indicates a higher value is desirable, a downward arrow that a lower value is desirable. Standard deviations are included in parentheses following each mean. The p values are from t tests comparing the two groups.

The following health indicators were measured. Health-related quality of life was measured by asking patients to rate, using Likert-type scales developed for the Chronic Disease Self-Management workshop, the four physical symptoms of fatigue, pain, hypoglycemia, and hyperglycemia (Dixon & Bird, 1981; Piette, 1999, 2000; Piette, McPhee, Weinberger, Mah, & Kraemer, 1999; Scott & Huskisson, 1976); activity limitation (Lorig et al., 1996); and the impact of the disease on the patients’ hobbies and recreational activities, household chores, errands and shopping, and social activities with friends (Stewart, Hays, & Ware, 1992). The participant was asked to rate his or her health using the National Health Survey (Idler & Angel, 1990; Roberts, Mason, & Nelson, 1980; Ware, Nelson, Sherbourne, & Stewart, 1992; Wolinsky & Johnson, 1992). Depression was measured by the Patient Health Questionnaire–9 (PHQ-9; Spitzer, Kroenke, & Williams, 1999). HbA1c was measured with a mailed home test kit developed and analyzed by BioSafe Laboratories.

Four health behaviors were measured using instruments developed for the Chronic Disease Self-Management study and the Spanish Diabetes Self-Management Study (Lorig, Chastain, Shoor, & Holman, 1989; Lorig et al., 1996): aerobic exercise, stretching and strength exercise, communication with health care providers, and practicing stress management.

Perceived self-efficacy was measured using an eight item scale for self-efficacy developed for the Chronic Disease Self-Management workshop; problems identified in a series of focus groups with diabetes educators, physicians, and nurse practitioners; and the key behaviors taught in the IDSMW (Bandura, 1992; Maddux, 1995; Schwarzer, 1992). When tested with 100 Spanish speaking people with diabetes, the scale had a Cronbach’s alpha of .85.

Critical in the management of chronic conditions are activated patients, defined as patients who have the skills, knowledge, and motivation to participate as effective members of the care team (Hibbard, Mahoney, & Tusler, 2004). The Patient Activation Measure (PAM) is an assessment tool that measures the broad range of elements involved in patient activation, including knowledge, skills, beliefs, and behaviors that a patient needs to manage a chronic illness (Hibbard et al., 2004). The PAM was administered to measure patient activation. Also included were questions about frequency of Internet use, use of health-related sites, and comfort with the Internet.

Qualitative Data

A content analysis was conducted of all the Native posts on the four study bulletin boards. The units of analysis (Graneheim & Lundman, 2004) were the participants’ posts.

Internet-Adapted Focus Group

Following completion of the workshop, an online focus group was conducted via a 1-week listserver with the AI/AN participants. Using a semistructured interview guide, participants were asked a new question every day and then asked to respond to the question and each other. Questions asked about the usefulness, cultural appropriateness, and acceptability of the workshop. Probes and prompts were used to stimulate further explanation or depth for the topics. The goal of the online focus group was to reveal the perspectives of the group participants and facilitate the discovery of new ideas and insights regarding AI/AN diabetes self-management and the use of the online workshop in facilitating self-management.

Analyses

Process evaluation.

The process evaluation examined the number and time of participant logins, completion of assignments, and parts of the workshop used or not used. To make comparisons between the AI/AN and non-AI/AN pilot groups, t tests were used. Characteristics associated with completion (defined as logging in at least once a week for 4 of the 6 weeks) and non-completion and cultural relevance of the workshop were examined. In addition, we explored the possible differences in demographic characteristics of the urban AI/AN pilot participants and those who reside in rural or tribal reservation areas.

Qualitative analysis.

Grounded theory was chosen in guiding the qualitative analysis of the bulletin board posts (Addison, 1999). Grounded theory is an inductive, theory discovery methodology that allows the researcher to develop a theoretical account of the general features of a topic while simultaneously grounding the account in empirical observations or data. Although the theory of self-efficacy underlies the IDSMW, how the program might operate among AI/ANs and general concepts of how diabetes is conceptualized and experienced within this population was an area we wanted to examine. Grounded theory was chosen in guiding the qualitative analysis of the bulletin board posts as it offered an opportunity to facilitate discussion of general concepts and features of the IDSMP in a less structured way while firmly basing or grounding the findings in the data collected. For the content analysis, each bulletin board post was considered a unit of analysis (Graneheim & Lundman, 2004; Mertens, 1998). The posts were comments and statements in the form of sentences or sometimes paragraphs. The posts were chosen as the units of analyses because each post was large enough to be considered a whole and for the context in which the post was written to be kept in consideration when looking at the meanings of each post (Graneheim & Lundman, 2004; Mertens, 1998). We next coded or labeled these units of analyses according to their meanings to categorize the data (Mertens, 1998). A category referred to a descriptive level of content and included subcategories (Mertens, 1998). The subcategories were then sorted and abstracted.

The content of the posts was broken into “source” posts (or person making the post) and “reference” posts (number of posts within a given category). Thus there was a possibility of 27 source posts as there were 27 people in the study. QSR NVivo version 7 (QSR International; http://www.qsrinternational.com/products_nvivo.aspx) was used to assist with the coding process.

Internet-adapted focus group.

For the Internet focus group, all of the comments and responses posted were collected and summarized. These summarized responses were fed back to the whole group online for their feedback and interpretation, to corroborate the overall findings and summary of the investigators.

RESULTS

AI/AN Participants Compared to Non-AI/AN Participants

Table 1 shows baseline findings for participants in both the AI/AN and non-AI/AN IDSMW pilots. AI/AN participants were slightly younger and had slightly lower educational levels compared with non-AI/AN pilot participants. Statistically significant differences between the groups were found in weight, symptoms of low glucose levels, and depression. AI/AN participants experienced a higher mean weight (208 lb mean weight among AI/ANs and 172 lb mean weight among non-AI/ANs). Although HbA1c levels were roughly equivalent between groups (AI/AN 6.79 vs. non-AI/AN 6.86), AI/ANs experienced more symptoms of low blood glucose levels (hypoglycemia; 2.60) than non-AI/AN participants (1.77). AI/ANs also experienced significantly more depression than non-AI/AN participants. Of the 27 AI/AN pilot participants, 4 participated in the workshop sporadically and 23 participated regularly. The participation rate for the AI/AN pilot was slightly above the rates of the non-AI/AN group.

In comparison to non-AI/AN pilot participants, the AI/AN pilot participants were more likely to log in to the workshop during daytime hours, typically 8 a.m. to 6 p.m., and less likely to log in during evening and weekend hours. This may be because more AI/AN participants had access to high-speed Internet from an office or library setting and dial-up at home.

Although weekly assignments for the IDSMW were voluntary, each participant was encouraged to create a weekly action plan and was asked to report on his or her plan at the end of each week. During the 6-week workshop, 1 of the 27 participants never made an action plan, 15 of the 27 participants made action plans every week, 12 of the 27 participants did not complete an action plan during one of the weeks, and 2 of those 12 did not complete an action plan for more than two of the weeks.

For the AI/AN pilot (n = 27), all were enrolled members of federally recognized tribes. A total of 18 tribes were represented. When asked for primary source of care, 13 participants reported Indian Health Service (IHS), 3 reported tribal health clinic, and 11 reported private hospital or doctor; 6 participants reported having no health insurance other than IHS. Of the participants, 60% reported they resided on or near tribal or reservation lands. All Indian Health Service areas were represented in the pilot.

Qualitative Content Analysis

There were a total of 354 bulletin board posts: 200 to problem solving, 83 to difficult emotions, 52 to action planning, and 19 to the celebrations. The posts on each of the bulletin boards were subcategorized according to themes. For each theme, the number of posts per person and the number of posts per theme were examined. These are listed within parentheses. The Problem Solving board was categorized into the following themes: family (16/37), weight management (15/19), eating habits (12/18), motivation (10/39), physical activity (10/26), glucose monitoring (9/20), time (8/12), and medical comorbidities (6/11).

The Difficult Emotions board included posts from 17 of the 27 participants (17 source posts) and included 68 reference posts. This bulletin board was not further subcategorized because all posts in the Difficult Emotions bulletin board fell into the subject category of depression.

The Celebrations board included posts from 9 of the participants (9 source posts) and consisted of congratulatory posts to one another for meeting workshop-related goals.

The Action Planning bulletin board had posts from 26 participants. The posts on the Action Planning bulletin board in order of categories with the most posts to least posts are physical activity (62), diet (41), glucose monitoring (19), goal setting (9), relaxation (2), and lastly confidence (1).

The Action Planning bulletin board is the place where participants shared their weekly goals, called action plans. During the 6-week workshop, 1 of the 27 participants never made an action plan, 15 of the 27 participants made action plans every week, 12 of the 27 participants did not complete an action plan during one of the weeks, and 2 of those 12 did not complete an action plan for more than two of the weeks. These action plan completion rates were consistent with the non-AI/AN pilot.

People not making action plans had the option of clicking “I did not make an action plan this week” and giving a reason. Each of the 12 people who did not make an action plan at least one of the weeks listed reasons. The most common reason for not making an action plan was lack of time (n = 6), followed by illness for that week (n = 3), family obligations (n = 2), and lastly depression (n = 1).

Internet-Adapted Focus Group

The Internet focus group examined the appropriateness, usefulness, and cultural acceptability of the workshop. The participants reported the blood glucose, fitness, and action planning tracking tools as being the most helpful parts of the workshop. They also reported enjoying the interaction with other individuals via the bulletin boards. Several participants commented on the circular model of the curriculum, the wheel, which they liked and felt resembled AI/AN concepts of health and well-being. Several people commented that the course provided them with information and links to information that supplemented the one-on-one 15-min counseling sessions they got with their diabetes educator. One person posted, “I get confused and overwhelmed when I visit the nurse. Having the online workshop helped me go through diabetes material at my pace and I learned much more about managing my diabetes.”

Three participants reported feeling rushed through the material and said that they needed more time to go through the material and read others’ comments on the bulletin boards. One participant felt there was too much reading involved and that learning how to navigate through the workshop took too much time.

The topic of cultural acceptability provoked the most response. Most of the participants liked the circle model (a wagon wheel). One person stated, “The wheel was about the only thing that seemed culturally specific. I didn’t find anything culturally inappropriate, though.” Most people discussed how this workshop was the first time they had been able to interact with other AI/AN people. They said that this interaction with other AI/ANs was what made the class culturally relevant. Below are comments from two participants:

• Participant 1: The thing that made it culturally specific was the fact that it was a group of Natives discussing these issues together. The non-linear wheel model was definitely culturally specific, but my identification with the cultural aspect of the group came from the participants more than the model.

• Participant 2: I don’t think the program was or was not culturally appropriate. I didn’t see much, save the participation of other Natives, that made this unique to Native peoples. I think the strongest, most culturally competent aspect of the program was the input, comments, suggestions, interactions, communication, ideas, etc. of the amazing Native participants—that is from where the true cultural appropriateness came, in my opinion.

A discussion ensued within the focus group about the interaction among AI/ANs and if the workshop should include both AI/ANs and non-AI/AN persons. Although most participants reported that having all-AI/AN workshops would be preferable, two participants reported that mixed classes that included both AI/AN and non-AI/AN persons would not affect their participation. Common sentiments among most of the group members were expressed in the following excerpts from the posts:

• Participant 1: There is a connection with other Natives that many non-Natives may not be aware of. Many times, Native people can relate to the same things whereas non-Natives cannot.

• Participant 2: I agree, Natives know Natives, and are sensitive to the needs of Natives. I am on a reservation and life is different. There is no culture, it is a way of life, it is how we live, so I think from Native to Native we understand this way of life.

• Participant 3: We understand each other. American Indians take in the whole situation, not just diabetes and the support, knowing makes it comfortable, at least for me. I think my participation would be different if it was a mixed class. I would have to be culturally competent and watch my words before speaking out.

• Participant 4: I totally agree. When I’m in a group of Natives I “let down my hair,” so to speak.

When asked about the feasibility of implementing the workshop in the AI/AN population, participants reported that this workshop would be particularly attractive to younger and middle-aged people as a result of its convenience and flexibility.

• Participant 1: I am part of an academic Native community, and this group would use the Internet. The big obstacle I found with the program was how much time it took to really get what I needed. Time is such a commodity that that would limit participation from other Natives in my community.

• Participant 2: It would be good with those that work for the government and those Natives who actually have a job and can afford a computer. For most of the people on my reservation, and half of them are diabetic, they don’t have access to the Internet. It would be more beneficial having a class or face to face meeting to come up with some results, where they can actually see, feel, hear and touch results from the group.

• Participant 3: I think that a lot of Natives have access to computers, and do indeed use computers frequently, at least in the urban setting where I live. I am inclined to think that the blanket statement that many Natives do not have access to computers is specific to particular living arrangements. I think one has to be careful when asking questions such as this, as it is common practice, albeit very incorrect, to group all Natives together and ignore the unique ness of not only different tribes, but relocated–urban vs. reservation Natives. I also agree very strongly with the comment that this does take a lot of time; in that sense, I don’t think it is feasible for a lot of people to dedicate such an amount of time to this, unless they find it to be very useful/helpful.

DISCUSSION

The IDSMW pilot suggests that AI/AN people are able to access and utilize this workshop in both the rural and tribal or reservation setting. The workshop was found to be culturally appropriate among different tribal and geographic communities and may be a mechanism to support diabetes education programs delivered in the clinic setting.

The IDSMW was built so that it could be used by almost anyone who had access to the Internet, including those with slower modems. A majority of the AI/AN participants used computers at their work settings to log in to the workshop during the daytime hours as although many of them had Internet access at home, they reported not having high-speed Internet at home but only dial-up. However, many of the participants did log in to the workshop from home, particularly on the weekends, and reported that the simple design of the workshop made it easier to log in with a slower modem.

The mean age of AI/AN IDSMW participants was 42 years, which is slightly less than that of AI/AN participants in diabetes self-management workshops in community settings. This suggests that middle-aged AI/ANs living in both the urban and reservation settings may benefit from an Internet intervention that allows them to receive health information during times more convenient for them and their schedules. Furthermore, participants reported having limited time with their health care providers and reported that the Internet offered much needed information to supplement their health education and in a time and pace convenient for them. Lastly, participants reported that the social support they received from the intervention was important and a valuable part of the workshop. As most AI/ANs participating in the workshop were rural or reservation-based, this suggests that for communities separated by distance, such as AI/AN tribes on reservations, the Internet may be an important source of health information and support for those working individuals with families, limited time, and living in geographically distant settings.

As with other diabetes programs described within the literature, using a culturally acceptable tool for introducing the workshop to AI/AN participants and for recruitment into the workshop appears to be critical. In the case of the IDSMW, the Web site created with pictures and stories of AI/AN participants from the community-based workshop introduced and marketed the workshop to AI/AN people. Although both AI/AN and non-AI/AN participants received the same intervention, the AI/AN Web site introducing the study presented the workshop in a culturally relevant context, advertising the experience of other AI/ANs and relaying the message that the workshop had something to offer AI/ANs. Just as having the knowledge of where to find and recruit AI/ANs in a community-based setting is needed for successful workshop recruitment, knowing where AI/ANs can be found online was similarly needed. A strong working knowledge of where AI/ANs congregate online as well as high-traffic Web site and listservers to reach AI/AN people was essential.

The qualitative analysis showed that AI/ANs found the workshop to be culturally acceptable because it was an all-AI/AN group. The participants reported having other AI/ANs understand the struggle of living on reservations or in urban communities and living with diabetes gave them a sense of a shared experience, which made the workshop culturally appropriate. Although symbols within the workshop were not AI/AN specific, that was not necessary, as the dialogue that took place online was AI/AN specific and diabetes specific. Furthermore, although the categories of the bulletin board exist to direct participants to topic areas, participants in the workshop wrote about various topic areas within each category. This suggests that as in a real-world setting, participants do not change location to discuss certain topics but rather discuss them where they are and where it is convenient. The posts revealed that family obligations, as well as lack of time, contributed significantly to not meeting diabetes self-management goals. Finally, depression was a significant issue for a majority of the participants, more than half of whom had posted a depression-related post. Participants posted that it was difficult to discuss their depression within their families and communities and that the online workshop provided them the opportunity to discuss depression openly.

The IDSMW integrates the theory of self-efficacy through skills mastery, modeling, interpretation of symptoms, and persuasion. No significant differences were found between AI/AN and non-AI/AN participants in workshop login, workshop participation rates, individual action planning, or group feedback using the bulletin boards. This suggests that the theory of self-efficacy that underlies the IDSMW, as well as the scales used in measuring workshop outcomes, operate similarly within the AI/AN population as within other populations. Furthermore, no significant differences were found in the return rate for the home blood test kit even though many AI/ANs resided in rural or reservation areas.

There were several limitations to our study. The focus of our study was on examining the feasibility of implementing the IDSMW with AI/ANs with Internet access in comparison with non-AI/ANs. As with all Stanford Internet-based patient self-management studies, only people with access to the Internet were included. Although we acknowledge that important differences may exist between AI/ANs who have Internet access and those who do not, we did not look at those differences and this may be considered a limitation of our study. Furthermore, AI/AN people were recruited into our study via a listserver that included urban and reservation Indian Health Care System clinics, urban and reservation social service agencies, and an AI/AN arts organization. Although a majority of AI/ANs in reservation settings use the Indian Health Care Service clinics, evidence suggests that a significant proportion of urban AI/ANs may not have access to the limited urban AI/AN health and social service agencies (Urban Indian Health Institute, 2004). Therefore, our sample included more reservation AI/ANs than urban AI/ANs. Urban AI/ANs continue to be difficult to identify and reach, even though they make up the majority (65%) of the AI/AN population in the United States (Urban Indian Health Institute, 2004). Thus, a future study may attempt a more representative sample of urban and reservation AI/AN populations. A larger trial is now ongoing examining the workshop within the AI/AN population nationally. Recruitment efforts are under way in both urban and reservation settings and a randomized wait-list control design is being used.

CONCLUSIONS

The IDSMW appeared to offer additional forms of information and support for cross-learning among diverse AI/AN people, despite differences in age, geographic location, and tribal affiliation. Several pilot participants volunteered to be peer leaders for future online workshops offered for Native people and have helped to recruit for the larger trial now being conducted.

The workshop is highly adaptable primarily because of the peer-led mechanism of delivery, which makes it one of few workshops that can be translated from research into practice successfully with little adaptation. This has been found with AI/ANs using the workshop in a community-based setting and now via the Internet with the IDSMW. The participatory approaches used to recruit AI/ANs, particularly the AI/AN-specific Web site and the personal invitation to participate from the AI/AN researcher involved in the project, facilitated the implementation of this workshop successfully via the Internet. It is unlikely that the workshop would have identified and recruited participants without the AI/AN-specific Web site, sharing the stories of AI/ANs who had participated in the workshop within the community setting, and the all-AI/AN classes and peer leaders that made the workshop culturally acceptable.

The success of the tailored recruitment effort using AI/AN-specific listservers suggests that similar ethnic specific outreach may be an effective approach for recruiting Latinos, African Americans, and other racial or ethnic minority groups for online workshops. With the increase in use of the Internet for health-related information and the rising tide of diabetes, particularly within racial or ethnic minority populations, the findings from this workshop suggest that it is feasible to implement an Internet-delivered disease self-management workshop within the AI/AN population. Furthermore, Internet delivered workshops may be useful supplements for individual patient health care for AI/ANs and other racial or ethnic groups disproportionately affected by health disparities.