Expanding the Reach and Impact of
|Table of Contents|
|Executive Summary (Stand-Alone)|
|Preface: A Vision of e-Health Benefits for All|
|Chapter 1. Introduction|
|Chapter 2. Mapping Diversity to Understand Users’ Requirements for e-Health Tools|
|Chapter 3. Assessing the Evidence Base for e-Health Tools for Diverse Users|
|Chapter 4. Strategic Factors in Realizing the Potential of e-Health|
|Chapter 5. Partnerships for Meaningful Access|
|Appendix 1. Environmental Scan of 40 e-HealthTools|
|Appendix 2. Project Interviewees, Experts Consulted, and Reviewers|
|Appendix 3. Chapter 3 Literature Review Summary|
|Appendix 4. A Comparison of Internet Use and Health Status of Populations That Experience Health Disparities|
Chapter 3. Assessing the Evidence for e-Health Tools for Diverse Users (Part 1)
This chapter summarizes and analyzes recent research literature on e-health tools to clarify what about e-health tools for diverse users is working well and where more and different research is needed. Critics argue that over-reliance on e-health tools can increase disparities rather than reduce or eliminate them; therefore, it is vital to identify when e-health tools can help to narrow gaps. The Institute of Medicine (IOM) report, Speaking of Health, proposes that several factors are relevant for assessing e-health for diverse populations: access, availability, appropriateness, acceptability, and applicability of content (2002). This chapter uses these concepts, referred to as the “Five A’s,” to organize key research findings and discuss their implications for tool design, use, dissemination, and impact. The review suggests that design and dissemination factors are closely connected to and likely to affect the impact of the tools according to a variety of outcome measures.
Previous reviews also have looked at the evidence base for e-health but have not focused as closely on design, use, and dissemination issues as the present review (Eng, 2001; IOM, 2002; Neuhauser and Kreps, 2003; HHS, 1999). These other reviews point not only to the great promise of e-health tools, but also to the need to moderate enthusiasm by recognizing factors that can limit the tools’ potential. Numerous individual examples of research-based tools usually produce the desired effects. To date, however, no systematic body of knowledge or theoretical frameworks explain what processes or contextual factors produce and mediate these effects or what the effects would be for different kinds of e-health tools used by different audiences (Neuhauser and Kreps, 2003). Given that some population groups experience a disproportionate amount of disease and overall poor health, it is critical to use the research enterprise to understand if and how e-health tools might be designed and deployed to reduce rather than exacerbate disparities and improve individual and population health.
Methodology and Rationale for Review
This review selected research studies using experimental design, as well as relevant review articles, that either were meta-analyses or summaries of experimentally based research studies. After the initial round of article selection, the inclusion criteria were made less stringent to increase the breadth of coverage in certain areas. For example, no randomized controlled trials were found for healthcare tools because they are relatively new in the e-health arena. Therefore, studies were included that surveyed user satisfaction and ease of use to provide some insight into these tools. Similarly, in the area of online communities and health information, studies using content analysis provided important findings relative to the potential utility of these tools for different subpopulations; these were also included. Only studies published in peer-reviewed journals were considered. The intent was to identify those studies that used scientific methods and had already been reviewed by the field and found to be significant enough for publication.
Although this approach differs from the most rigorous evidence reviews, such as those conducted by the Cochran Collaboration or sponsored by the Agency for Healthcare Research and Quality, the purpose of the present review is not to differentiate research based on methodological rigor. The intent is to highlight the presence or absence of solid research on key elements affecting e-health use and dissemination. The recent Cochrane Collaboration review, “Interactive Health Communication Applications for People With Chronic Diseases,” should be consulted for an example of a rigorous review of the science and conclusions about the effects of e-health tools on persons with chronic diseases (Murray, Burns, See Tai, et al., 2006).
The literature search used the overarching purpose categories to identify studies for inclusion: health information, behavior change/prevention, online communities, healthcare tools, decision support tools, disease management, and health self-management. Research studies for these categories were identified through the use of the following databases: PubMed, Medscape, Medline, PsycINFO, CINAHL, and the Social Sciences Citation Index. The searches covered the time period from January 1999 to September 2004 to identify recent literature. The CRISP (Computer Retrieval of Information on Scientific Projects) database maintained by the National Institutes of Health and covering federally funded biomedical research projects was searched twice approximately 6 months apart in 2004 to identify new research either just being concluded or in progress; the same search terms were used as above. Review of the reference lists and suggestions from an expert panel and expert interviews also identified articles.
Critical information was extracted from each article and summarized into a matrix table. The matrix, presented in Appendix 3, contains data on the study’s author, research design, sample, health topic area, locus of use, technology, tool description, study overview, measures, and outcomes. The table is subdivided by study design. The first section includes the studies using randomized controlled designs. The table then moves through quasi-experimental designs, single-group studies, and content analyses. Within each research design subsection, studies are arranged alphabetically by author. Each study has been assigned a unique identifying number to allow easy location of that study in the table. Each citation in this chapter includes a table reference number (TR#). To return to the text from the table, the chapter section in which the study is cited is indicated in brackets after the citation in the table.
Overview of e-Health Tools in Studies Reviewed
Most of the e-health tools in the studies reviewed below are multicomponent interventions designed to impact many aspects of personal health self-management, including prevention, behavior change, decisionmaking, and chronic disease management (see Chapter 1). This review found that although e-health tools have been developed for a wide variety of health topics and purposes, some topics and purposes appear to have greater representation in the research literature. Areas with the largest numbers of tools are nutrition education, weight management, tobacco cessation, and cancer and diabetes prevention and management. Although most of the tools in these studies are designed for adults, some target children and adolescents. Some tools, such as those for behavior change, are grounded in a theoretical framework. Others, such as healthcare tools, are emerging in response to market and policy demands and do not yet have much of a scientific basis to suggest that they will have their intended effect.
Each tool contains health information specific to its intended purpose. This information can be general, targeted to a specific user group, or tailored to an individual user. In addition to information, other features might include interactive games and simulations, video clips, chat rooms, message boards, e-mail to and from healthcare providers, self-assessments, decisionmaking tools, disease management tools, and links to other sites. Tools designed for a similar purpose do not always contain the same components.
Several studies in the review do address the effectiveness of specific components of the computer-based intervention (Baranowski, Baranowski, Cullen, et al., 2003, TR#39; Feil, Noel, Lichtenstein, et al., 2003, TR#10; Napolitano, Fotheringham, Tate, et al., 2003, TR#23; Neighbors, Larimer, and Lewis, 2004, TR#24; Tate, Wing, and Winett, 2001, TR#34). Tate and colleagues used two different e-mail approaches in their study (Tate, Jackvony, and Wing, 2003, TR#33). Both the control group and the intervention group received access to a weight-loss Web site and weekly e-mail reminders to submit their weight; the intervention group also received individual e-counseling from a weight-loss counselor. The researchers found that, compared to the control group without the individualized counseling, the intervention group doubled the percentage of initial body weight lost.
Neighbors and colleagues studied the unique impact of personalized normative feedback alone on drinking behavior in college students and found changes in misperceptions about drinking norms and on drinking behaviors (2004, TR#24). Studies from the D-Net (diabetes) projects indicated that participants using interventions with a support component improved in perceptions of support and actually had higher login rates than the other intervention groups and the controls (Barrera, Glasgow, McKay, et al., 2002, TR#2; Glasgow, Boles, McKay, et al., 2003, TR#13). Studies of CHESS (Comprehensive Health Enhancement Support System), an Internet-based program to help patients cope with cancer and other diseases, have found that use of the component parts of the system vary by a number of demographic factors, including race and income (Gustafson, Hawkins, Pingree, et al., 2001, TR#15; McTavish, Pingree, Hawkins, et al., 2003, TR#88). These types of studies are an important beginning to help clarify what about e-health tools for diverse user groups is working and what is not.
The majority of the tools reported in the research studies were Internet-based interventions that could be accessed from personal computers. Some studies used CD-ROMs to deliver the intervention. Other delivery mechanisms used in these studies included a telephone-linked communications system (Delichatsios, Friedman, Glanz, et al., 2001, TR#9; Pinto, Friedman, Marcus, et al., 2002; TR#27), videophones (Ryan, Kobb, and Hilsen, 2003, TR#73), computers in freestanding kiosks in community settings (Anderson, Winett, Wojcik, et al., 2001, TR#1; Radvan, Wiggers, and Hazell, 2004, TR#70; Valdez, Banerjee, Ackerson, et al., 2002, TR#35), a fingerprint reader (Sciamanna and Clark, 2003, TR#31), and home telehealth units (Finkelstein, O’Connor, and Friedman, 2001, TR#11; Kaufman, Starren, Patel, et al., 2003, TR#63; Ryan et al., 2003, TR#73).
In their reports of findings, researchers do not often discuss their rationale for choosing a specific delivery method. The intended locus of use and the amount of graphics are current factors that appear to influence the decision. For example, Napolitano et al. (2003, TR#23) and Lenert and Cher (1999, TR#65) report that they delivered their interventions via the Internet to reach a potentially wide audience of users who could access the intervention from any location. Proudfoot, Goldberg, Mann, et al. used a CD-ROM-based program with video vignettes, which was designed for delivery in a clinical setting (2003, TR#28). Because it is possible to convert content on compact discs (CDs) for use on the Internet and vice versa, the distinction between formats will likely become less relevant. At the present time, when graphics-heavy CDs are moved onto the Internet, there may be lengthy download times that can affect usability and satisfaction, particularly for those using older computers or slow Internet connections (Baranowski et al., 2003, TR#39). If broadband costs decline and more users opt for high-speed access, connection speed may become less of a problem, but not necessarily, given the size of the access gaps described in Chapter 2.
Synthesis of Findings From Research Studies of e-Health Tools
Issues of access underlie all studies of consumer e-health tools. This brief section focuses on the impact of disparities in access on the validity of findings reported in the literature. (See Chapter 2 for a general discussion of access issues.) The most important issue relates to the external validity of the research. Findings from this review indicate that many studies included only participants who have computers, thereby excluding those who lack computers or Internet access. A few studies recruited participants directly from Internet Web sites, making it less likely that people without regular access would be considered for the sample. The access criterion for study participation affects the generalizability of the findings for other population groups or the population at large. Because people without computers also tend to have less education, lower incomes, and poorer health, the bias in the current literature must be recognized, and the need for ongoing and future research to include diverse populations is critical.
Access for all population groups is an issue. A few studies, particularly in the area of online communities, have provided participants with computers and expected no computer experience from their participants (Gustafson et al., 2001, TR#15; McTavish et al., 2003, TR#88). These studies are encouraging in that the researchers found that user technology support was not difficult and, ultimately, users were able to use the technology to give and receive support in the online communities. Providing computers for public use can be another avenue for increasing access; however, Radvan et al. found that one reason people did not use a community-placed computer-kiosk for health information was that they did not feel comfortable using the kiosk in public (2004, TR#70).
In a study of older adults, Kaufman et al. found that use of the computer and mouse was very difficult for elderly participants with diabetes who had limited computer experience (2003, TR#63). For this age group, more attention may need to be paid to choosing technology that is suitable to the users’ needs. For example, Ryan et al. in the Community Care Coordination Service of the U.S. Department of Veterans Affairs (VA) used a unique approach in which they matched technology to users based on their clinical need and ability, rather than on the availability of a specific kind of technology (2003, TR#73). Their matching process was based on the patient’s education, vision, manual dexterity, willingness to use technology, and adherence to medical regimen. Using this approach, they were able to demonstrate improved clinical outcomes in a group of veterans with chronic illnesses.
Davis found that only 19 percent of 500 Web sites representing common illnesses or conditions were accessible for users with visual impairments who used automated screen readers (2002, TR#54). He also notes that almost 65 percent of the Web sites that failed the accessibility test had just a single type of fixable problem. Davis further points out that the best way to make sure a Web site is accessible is to do so from the beginning by following established guidelines, such as those described in Chapter 2 and Appendix 1.
In sum, there appears to be a bias in the literature toward studying those persons who have easy Internet access, can use readily available technologies without adaptation, and do not need much if any technical support. Identifying ways to include currently excluded or understudied groups in future research is critical to creating an evidence base of results that can be generalized as well as specified for select user groups.
In addition to technology access, people must also have available the information and tools they want and need—that is, meaningful access. Because the Internet seems to be an “always on,” universally available channel, there is often the assumption that posting something on the Internet automatically increases information availability. Developing a Web site that contains relevant information is not enough, however, if people cannot locate the site. The studies discussed below suggest that research on information-seeking behaviors is still needed to understand how well different groups can locate health information and tools. (See Chapter 2 for additional information on health information-seeking issues.)
One approach to assessing availability is to go directly to the target audience to conduct a needs assessment. For example, Rozmovits and Ziebland conducted focus groups and interviews with people who had breast or prostate cancer (2004, TR#72). They found that cancer patients had information needs that changed during the course of their illness, and they were not always able to find the information they wanted. Similarly, Goldsmith, Silverman, and Safran found through formative research that parents of children with cancer reported a primary need for help with medication management (2002, TR#60).
Understanding the strategies that people use to locate information is key. Eysenbach and Kohler observed study participants as they tried to locate answers to specific researcher-generated health questions using the Internet (2002, TR#58). They found that although all 16 participants used search engines as starting points and somewhat suboptimal search strategies, they were able to find answers to the questions. However, the researchers did not provide an analysis of the accuracy of the answers or ascertain whether the participants were satisfied with the information they found.
The Pew Internet & American Life Project’s 2005 report on search engine use found that 84 percent of Internet users have used search engines, 92 percent of those who use search engines are confident about their searching ability, and 87 percent report successful search experiences most of the time (Fallows, 2005, TR#59). Some user groups, however, have special challenges related to information-seeking. Zarcadoolas, Blanco, Boyer, et al. examined the navigation skills of adults with low literacy and identified several factors that affect availability for this group (2002, TR#81). These include spelling problems that interfere with searching, difficulty entering Web addresses, and difficulty using navigational tools such as graphic links, back arrows, and scrolling.
Users can have access to technology and the skills to locate information and tools, but still encounter issues related to appropriateness. Appropriateness refers to the fit between the user and the tool. In an attempt to assess appropriateness, researchers have conducted studies on cultural relevance, users’ perceptions of the credibility of content, content analyses focused on information quality and readability, and the use of tailoring.
Few of the reviewed studies specifically examined cultural relevance or recruited samples based on racial and ethnic characteristics. Most of the studies did include members of the target audience segmented by age (e.g., college students) or by health or disease status (e.g., women with breast cancer, people at risk for heart disease). Only a few studies conducted research with members of specific ethnic groups to assess cultural relevance (e.g., Campbell, Honess-Morreale, Farrell, et al., 1999, TR#4; Duncan TE, Duncan SC, Beauchamp, et al., 2000, TR#41; Jantz, Anderson, and Gould, 2002, TR#45); Zimmerman, Akerelrea, Buller, et al., 2003, TR#82).
Users’ Perceptions of the Credibility of Content
Measuring users’ perceptions of the credibility of available information is another means to assess appropriateness. Rozmovits and Ziebland found that study participants were aware of the credibility issues surrounding health information on the Internet, and reported that they often compared information from several different sources before taking it as fact (2004, TR#72). These users preferred information about cancer treatment from noncommercial sites and specifically from institutions with good reputations, such as universities or medical centers.
Eysenbach and Kohler found that users identified many criteria for establishing credibility, such as the source of the information, a professional layout, understandable and professional writing, and citation of scientific evidence (2002, TR#58). Similar to Rozmovits and Ziebland’s findings (2004, TR#72), a few users felt that it is easier to assess information quality on the Internet because they could cross-check information on different sites. When they were actually observed searching for information, none of the participants checked the source of the information and fewer than 25 percent could even tell the broad category of the site they used (e.g., university, Government agency, business).
Barnes, Penrod, and Neiger found a similar disconnect between what users reported as important factors to consider when establishing credibility and actual behavior in assessing Web site quality (2003, TR#46). Walther, Wang, and Loh found an interaction effect of advertisements on user perception of credibility (2004, TR#36). The presence of advertisements on sites with .org domains made the site appear less credible than ads on sites with .com or .edu domains.
Physicians or other healthcare providers could serve as intermediaries to direct patients to appropriate Internet content. The study by D’Alessandro, Kreiter, Kinzer, et al. had physicians provide information prescriptions to patients that contained relevant Internet sites for health information (2004, TR#8). One-third of participants used these prescriptions and were then more likely to state that they would use them again and had already recommended them to others.
Researchers also assess appropriateness, particularly of publicly available Web sites, by conducting content analyses of the information and performing readability analyses. The overall goal is to measure information quality. Inconsistent findings are reported related to Web site quality. For example, a study by Madan, Frantzides, and Pesce (2003, TR#87) on laparoscopic bariatric surgery and a study by Fahey and Weinberg (2003, TR#85) on LASIK (laser-assisted in situ keratomileusis) eye surgery found that the information on the Web in both of these areas was poor and unreliable. One study on diabetes sites found that information quality varied widely (Seidman, Steinwachs, and Rubin, 2003, TR#91). Oermann, Lowery, and Thornley reported that better quality content was found on Web sites sponsored by a university, professional organization, medical center, or Government agency (2003, TR#90). Only the study by Cheh, Ribisl, Wildenmuth, et al. on smoking cessation Web sites found that a majority of the information was accurate (2003, TR#83).
Evers, Prochaska, Prochaska, et al. examined the quality of Internet programs designed to help users change behavior in seven key areas: tobacco use, physical activity, alcohol, diet, diabetes, depression, and pediatric asthma (2003, TR#84). Of the 273 sites examined, only 42 (15 percent) met four of the five minimum criteria determined to have the potential to change behavior. These 42 sites then underwent a full review. All included self-assessments and some form of contact. Only 12 percent included individually tailored feedback, and none included information about evaluation for effectiveness, which was a key recommendation of the 1999 Science Panel on Interactive Communication and Health.
Content readability is usually assessed using readability formulas that provide grade-level assessments. Birru, Monaco, Lonelyss, et al. (2004, TR#48), Kusec, Brborovic, and Schillinger (2003, TR#64), and Oermann et al. (2003, TR#90) found that the average reading levels of the sites they examined was at a 10th-grade level. Birru et al. found some methodological difficulties assessing respondents’ comprehension of information on the Internet (2004, TR#48). For example, some respondents could correctly answer interviewers’ questions on the content by reciting directly from the Web site. However, when prompted, respondents could not put the answers in their own words. This finding is not surprising because readability analyses do not provide much insight into users’ understanding of the content and their capacity to apply the information to specific circumstances. (See Chapter 2 for additional discussion of health literacy issues.)
Eysenbach and Kohler conducted a systematic review of studies that assessed the quality of health information on the Internet (2002, TR#58). Differences in study methodology and quality criteria were used in the reviewed studies, a fact that could explain differences in study results and conclusions. For example, they found that many studies assessed completeness of information; however, this approach generally did not take into account the context or stated purpose of the site or links provided to additional information. They point out that the Internet is not the only type of media delivering information of inconsistent quality, and thus must be considered against the “background of imperfect consumer health information in other media” (p. 2697). One strategy they recommend includes improving the user’s ability to locate credible sites and to filter out inadequate ones.
As Chapter 2 indicates, tailoring is thought to be one of the most promising methods to improve the appropriateness of content for users because tailoring simulates an individualized assessment and response. Several tools in the behavior change area evaluated tailored information and feedback using randomized controlled trials (Bernhardt, 2001, TR#3; Campbell et al., 1999, TR#4; Oenema and Brug, 2003, TR#25; Oenema et al., 2001, TR#26). All these trials involved tools tailored to the user’s stage of readiness to change. Other tailoring variables included knowledge, dietary intake and habits, awareness of dietary intake as compared with published guidelines, and perceived overweight. These studies all showed positive effects for the tailored information as compared to the control conditions.
In general, the study findings that address appropriateness indicate that users may find it difficult to connect with tools that fit their interests and needs. The success of tailoring suggests the need for much greater attention to the design and testing of elements that make tools a better fit in terms of cultural relevance, consistency, comprehensiveness, and understandability for diverse users.
Acceptability refers to whether people find the tools satisfactory. Satisfaction is typically one criterion that is applied to the evaluation of commercial tools. The fact that millions of people are actively seeking health information online and the phenomenal increase in Internet use speak to a high initial level of acceptability. Researchers and tool developers have focused on usability studies to gauge and improve acceptability, recognizing it as a necessary condition for the ultimate success of e-health tools. Examining use over time can provide an additional measure of acceptability in that it makes it possible to gauge ongoing satisfaction with or usability of programs based on whether people continue to use them.
Ease of Use
Studies of e-health tools designed for a variety of purposes generally found that users report they are easy to use, although some studies found that this was not always the case. Block, Miller, Harnack, et al. reported that 97 percent of users found a nutrition education program easy to use (2000, TR#49). Feil et al. reported that 63 percent of users rated their smoking cessation Web site “easy” or “very easy” to use (2003, TR#10). Some users commented that the smoking cessation site used in the study by Lenert and Cher was complex and difficult to navigate (1999, TR#65). Oenema et al. found that those who had less familiarity with computers also found their tailored program more difficult to use (2001, TR#26).
People using e-health tools designed to allow access to medical records and/or to provide a means to communicate electronically with their healthcare providers were able to use these tools. Participants were able to master the complex login procedures required for privacy and to use the systems effectively; however, these users tended to be more educated, have personal computers, and be covered by a private health insurer (Cimino, Li, Mendonca, et al., 2000, TR#51; Hassol, Walker, Kidder, et al., 2004, TR#62; Masys, Baker, Butros, et al., 2002, TR#68). Sciamanna and Clark examined the acceptability of a fingerprint reader as an alternative means to authenticate users in a medical clinic, thus eliminating the need for complex login procedures (2003, TR#31). Those who used the fingerprint reader did not appear to under-report information and had fewer concerns about the reader than did those who did not use the reader.
More difficulties were found when the study populations were chronically ill, elderly patients with little or no computer experience. Caregivers of patients with dementia generally found the telephone-linked support system easy to use, but a small percentage of users had difficulty reading the screen or hearing the messages (Czaja and Rubert, 2002, TR#53). Kaufman et al. found that the use of the computer mouse for a diabetes home telemedicine system was exceedingly difficult for some of their elderly participants (2003, TR#63). Furthermore, all of the novice users experienced difficulty in developing a coherent mental model of the system and were frustrated by their inability to navigate screen transitions.
McKay, Glasglow, Feil, et al. found that the diabetes self-management component of their Web site, which guided participants in tracking blood glucose levels throughout the day, was not used often (2002, TR#21). They concluded that the tool might have been too complex for participants to use regularly. The VA program by Ryan et al. that matched technology to user ability found that patients were highly satisfied with the technology and 95 percent of users rated their technology “easy to use,” indicating that with careful selection of technology, these types of problems can be solved (2003, TR#73).
Self-reported satisfaction levels have been high for tools across a wide range of purposes. People showed high levels of receptivity to e-health tools to aid decisionmaking for the treatment of benign prostatic hypertrophy (Lenert and Cher, 1999, TR#65), genetic testing for breast cancer (Green, Peterson, Baker, et al., 2004, TR#14), and contraceptive use (Chewning, Mosena, Wilson, et al., 1999, TR#6).
Healthcare tool users were also very satisfied. Liederman and Morefield found that 78 percent of their sample of RelayHealth users rated Web messaging “better” or “much better” than calling their doctor, and they reported that electronic communication improved access to their practitioner (2003, TR#67). Tang, Black, Buchanan, et al. found that patients using the PAMFOnline system (Palo Alto Medical Foundation) rated online messaging highly, even though a subscription fee was associated with this function (2003, TR#76). The researchers also found that the majority of users identified getting lab results as the most important benefit of having access to their medical records (2003, TR#76). Hassol et al. surveyed members of the Geisinger Health System who were “early adopters” of the MyChart application (2004, TR#62). They reported that patients saw online communication as especially useful for general medical questions or prescription renewals.
Constraints of the technology at times affected satisfaction. Liederman and Morefield found that satisfaction with Web-based messaging correlated with response time (2003, TR#67). Those who felt they received a timely response to their messages were “very satisfied” (74 percent) with the system; likewise, those who reported a slow response from the clinic were dissatisfied (6 percent). Patients used the telephone when the electronic system was not in place yet, when they wanted quicker responses, or when it was easier to explain the problem orally than in writing.
Others liked using e-health tools as an adjunct to medical care in physicians’ offices or clinics. Wilkie, Huang, Berry, et al. found that patients liked using computerized assessments to help assess their levels of pain and fatigue (2001, TR#78; Wilkie, Judge, Berry, et al., 2003, TR#79). Patients reported that the tool gave them the ability to describe their pain more specifically, enabling better discussions with their physicians.
In addition, surveys conducted with people who use online health communities show that they identify many advantages of online community use. For example, groups are available 24 hours a day, 7 days a week (Han and Belcher, 2001, TR#61; Shaw, McTavish, Hawkins, et al., 2000, TR#74). They do not have to be concerned about their appearance (Shaw et al., 2000, TR#74) or other issues related to attending face-to-face groups (Shaw et al., 2000, TR#74; Czaja and Rubert, 2002, TR#53). They perceive equalized participation among group members due to anonymity (Colvin, Chenoweth, Bold, et al., 2004, TR#52) and the lack of social context cues, such as dress or appearance (Shaw et al., 2000, TR#74).
Other advantages are that people also can exchange information (Finn, 1999, TR#86; Mendelson, 2003, TR#89); share personal feelings (Shaw et al., 2000, TR#74), support, and coping strategies (Mendelson, 2003, TR#89); feel less alone (Reeves, 2000, TR#71; Shaw et al., 2000, TR#74) and less depressed (Lieberman, Golant, Giese-Davis, et al., 2003, TR#66); help others (Reeves, 2000, TR#71); and gain feelings of empowerment (Finn, 1999, TR#86; Reeves, 2000, TR#71). Preece, Nennecke, and Andrews found that people who posted to online communities had a greater sense of belonging and satisfaction than people who visited the communities but did not post (2004, TR#69).
Online community users do report some disadvantages, such as the time commitment needed to review large volumes of postings (Han and Belcher, 2001, TR#61; Shaw et al., 2000, TR#74), a lack of physical contact or proximity to other group members (Colvin et al., 2004, TR#52; Han and Belcher, 2001, TR#61), dealing with “noise” or off-topic postings, and the generation of negative emotions because they were exposed to others’ losses or problems (Han and Belcher, 2001, TR#61). Technical problems, such as difficulty with posting, can also be a disadvantage (Colvin et al., 2004, TR#52; Lieberman et al., 2003, TR#66).
Users were generally satisfied with tools designed to help them adopt healthier behaviors. For example, Lenert and Cher reported that 94 percent of the users of their smoking cessation site felt the site had helped their quit effort (1999, TR#65). In a tailored nutrition program, 79 percent of users reported that the program was helpful and most would use it again (Campbell et al., 1999, TR#4). About 90 percent of users of a nutrition education program reported that they had learned something new and would recommend the program to others (Block et al., 2000, TR#49). In a study by Woodruff, Edward, Conway, et al., 95 percent of teens would recommend the smoking cessation site to other teen smokers (2001, TR#80). McKay, King, Eakin, et al. found that the users in the intervention group were more satisfied with an intervention designed to increase levels of physical activity than were users in the computer-based information-only control group (2001, TR#22).
Only one reviewed study reported participants’ negative feelings about an Internet group (Harvey-Berino, Pintauro, and Gold, et al., 2002, TR#16). The researchers found that people preferred in-person groups for weight-loss maintenance rather than Internet groups; however, all of these participants had previously attended in-person weight-loss groups.
In contrast, McKay et al. found that nearly 60 percent of patients with diabetes in primary care practices were willing to participate in a computer-based diabetes management intervention (2002, TR#21). They believe this reflects a substantially higher percentage than would be willing and able to attend traditional educational programs.
Most surveys of satisfaction examine the tools as a whole. The study by Weis, Stamm, Smith, et al. of users of a site for persons with multiple sclerosis examined satisfaction with components of the site (2003, TR#77). They found that, in general, users preferred the information functions to the support functions of this site. Users who used both functions gave the site the highest overall ratings. Women rated the information function higher than did men; adults with children rated all functions higher than did those without children; and younger users rated the support functions higher than older users did. Escoffery, McCormick, Bateman, et al. also found that participants who used their smoking cessation site preferred the informational components to the “ask the expert” and message board features (2004, TR#57).
|Back to Top||Chapter 3 (Part 2) >|