An Evidence-Based Approach

Introduction: Evaluation of IHC

Consumers & IHC Evaluation

Developers & IHC Evaluation

Policy Issues Relevant to IHC

Health Care Providers, Purchasers & IHC

SciPICH Final Report

Wired for Health and Well-Being: The Emergence of Interactive Health Communication

Editors: Thomas R. Eng, David H. Gustafson

Suggested Citation: Science Panel on Interactive Communication and Health. Wired for Health and Well-Being: the Emergence of Interactive Health Communication.  Washington, DC: US Department of Health and Human Services, US Government Printing Office, April 1999.

Download in PDF format:  [Entire Document] [References]

Chapter III.

Underlying Evidence and Science of IHC

An evidence-based approach to IHC application design is grounded on developing applications by taking into account the best available evidence from research and generally accepted theories and concepts of behavior change and decisionmaking (Jadad, 1998a). Stakeholders, especially health professionals and purchasers of IHC applications, should be familiar with the social science concepts commonly employed in application design. Social science theories, models, and evidence from research provide guidance about important design considerations, such as the characteristics of individuals, ways in which people process information, and likely consequences of behavioral change strategies. Understanding these theories, models, and evidence from research is also helpful in critically appraising the true value of an IHC application. For example, without knowledge of these concepts, the use of technologies that appeal to the senses may distract someone from focusing on application content or on the evidence-based methods employed. Utilizing the latest technology is not sufficient when the content or approach is inappropriate. In this section, the Panel briefly reviews major psychological theories and models frequently used by developers in selecting appropriate content, media, and methods.

No unified theory exists to provide direction in IHC application design and development. Rather, a variety of social science theories and models effectively describe how people think, reason, act, and make choices. These theories and models can help illuminate the processes related to health-related behavior change and decisionmaking. They include: the theory of reasoned action (Ajzen and Fishbein, 1980; Ajzen, 1991); theories of learning (Bandura, 1986); group decisionmaking (Janis and Mann, 1977); transtheoretical stages of change (Prochaska et al., 1992; Prochaska et al., 1994); decision analysis (Weinstein et al., 1988; Mulley, 1989); and other theories and models (Rosenstock, 1974; Nisbett and Ross, 1980; Petty and Cacioppo, 1986; Locke and Latham, 1990; Strecher and Rosenstock, 1998).

Social science models that describe cognitive and behavioral concepts, and the relationships between these concepts, provide important orientation to designers of IHC applications. Listed below are selected psychosocial concepts with illustrative examples of IHC applications that may be of particular relevance to developers.

1. Outcome expectations associated with the behavior in question—Outcomes (both positive and negative) one expects as a result of engaging in a particular behavior.

Theories: (Rosenstock, 1974; Ajzen and Fishbein, 1980; Bandura, 1986; Weinstein, 1988; Ajzen, 1991; Prochaska et al., 1992; Strecher and Rosenstock, 1998)

IHC Applications: (Velicer et al., 1993; Campbell et al., 1994; Skinner et al., 1994; Strecher et al., 1994; Brug et al., 1996; Dijkstra et al., in press)

2. Self-efficacy expectations—Confidence in one’s ability to engage in a particular behavior.

Theories: (Bandura, 1986; Strecher et al., 1986)

IHC Applications: (Campbell et al., 1994; Dijkstra et al., in press)

3. Goal setting—Setting goals for change.

Theories: (Locke and Latham, 1990; Strecher et al., 1995)

IHC Applications: (Strecher et al., 1995)

4. High-risk situations—Situations that trigger a particular behavior.

Theories: (Shiffman, 1996)

IHC Applications: (Shiffman et al., 1997)

5. Attributions for previous failures—Interpretations one makes for the causes of previous failures in changing a particular behavior.

Theories: (Foersterling, 1986; Weiner, 1986)

IHC Applications: (Strecher et al., 1994)

6. Stage of change—Degrees of motivation and current experience in changing a particular behavior, ranging from precontemplation, contemplation, preparation, and action to maintenance.

Theories: (Weinstein, 1988; Prochaska et al., 1992)

IHC Applications: (Velicer et al., 1993; Campbell et al., 1994; Skinner et al., 1994; Strecher et al., 1994; Brug et al., 1996; Dijkstra et al., in press)

7. Prescriptive decision theory—Use of explicit quantitative estimates of probabilities of good and bad outcomes, and the utilities of those outcomes from the decisionmaker’s perspective to inform a decision; and descriptive decision theory—Modeling decisions in the face of uncertainty in an attempt to predict actual behavior (e.g., prospect theory).

Theories: (Mulley, 1989)

IHC Applications: (Brennan, Moore et al., 1995; Gustafson, Hawkins et al., 1999)

Several psychosocial concepts that are particularly important in IHC application development are empowerment, self-efficacy, and motivation. Empowerment can be generally defined as the process that enables people to exert control over their lives and their destinies (Peterson and Stunkard, 1989; Feste and Anderson, 1995). It is closely related to health outcomes in that powerlessness has been shown to be a broad-based risk factor for disease. Studies demonstrate that people who feel "in control" in a health situation have better outcomes than those who feel "powerless" (Israel and Sherman, 1990; Anderson et al., 1995). Empowerment can be enhanced by online support groups that allow patients to feel "connected" to others with a similar health condition (Gustafson et al., 1992; Pingree et al., 1993; Gustafson, Hawkins, Boberg, Bricker, Pingree et al., 1994). Interactive self-assessment tools can also help in this regard by helping individuals focus on central issues. Similarly, self-efficacy is a person’s level of confidence that he or she can perform a specific task or health behavior in the future (Bandura, 1977; Holman and Lorig, 1987; Lorig et al., 1989). Clinical studies show that self-efficacy is most predictive of improvements in patients’ functional status (O’Leary, 1985; Cunningham et al., 1991). Perceived self-efficacy has been shown to play a significant role in smoking cessation relapse rates, pain management, control of eating and weight, success of recovery from myocardial infarction, and adherence to preventive health programs (Strecher et al., 1986; Mullen et al., 1987; O’Leary et al., 1988; Allen et al., 1990; Maibach et al., 1991).

Motivation is a major factor in explaining the effectiveness of any instructional event, especially those that are voluntary and dependent upon intrinsic (as opposed to extrinsic) motivation as many IHC contexts are. Attribution theory states that the degree to which people attribute their own successes or failures to ability, effort, task difficulty, or luck differentially predicts whether, to what degree, and what kinds of subsequent learning opportunities they will voluntarily seek (Dweck and Leggett, 1988). Differences in attribution (e.g., "I achieved because I made the effort" versus "I achieved because the test was easy") explain why some people feel in control of their learning whereas others feel helpless in learning. For example, if when searching online to learn about treatment options, a person experiences difficulty, and he or she usually attributes success or failure to blind luck, the person may feel the task is just too difficult and/or his or her luck has run out. As a result, the person may give up and not seek online help again. On the other hand, another person may experience the same difficulty, but because the person attributes his or her success to effort, this individual persists and continues to use online resources. One of the shortcomings with attribution theory and other motivation theories is that little is known about whether these are "trait" variables or "state" variables. Someone whose motivation might be quite low in a classroom context might be much more motivated by online instruction, indicating that their motivation is a "state" variable. On the other hand, some people are not motivated by any learning opportunity, indicating a "trait" variable. Attribution theory is related to the "confidence" factor in Keller’s ARCS model, which is designed to help developers be more attentive to the motivation aspects of their instruction (Keller and Suzuki, 1988).

Use of the above theories and models as the bases for IHC application development may also contribute to their further development. Interactive media may allow researchers to collect better and different kinds of data on behavior change processes—data that could lead to more refined or comprehensive theories and models.

Almost all IHC applications seek to change individual behavior. They may lead to better health status, healthy lifestyles, or more appropriate uses of health services. In all cases, positive change is the goal. As described above, many behavior change theories have been developed and tested and all of them involve the use of one or more of the following concepts that are relevant for developers.

1. Motivation for change. Users need to believe that they cannot continue with their current behaviors. In some cases, this is a forgone conclusion because the user’s life is suddenly out of control due to an illness or injury. In other cases, the user’s motivation for change needs to be increased. This is the case for primary prevention efforts such as smoking prevention with teenagers. As mentioned above, a developer’s understanding of an intended user’s motivation for change is an important determinant of application effectiveness.
2. Superior alternative. Users need to believe that the proposed behavior change will improve their situations. Hence, a woman with breast cancer needs to believe that regular arm exercises after surgery will enhance her arm mobility. Developers need to be very clear about what changes they hope to achieve and design programs that help users to believe that such changes will address their pressing and long-term needs.
3. Social support. People facing significant life transitions or health problems, or those adopting complex behavior changes, often experience stress. When they do, they need the support of others who care about them and have experienced similar problems. Emotional support may help them overcome setbacks and renew their commitment to change. Developers need to build in mechanisms for emotional support to help overcome stress.
4. Skills and self-efficacy. Changes often require new skills or the application of old skills in new settings. In both cases, users need to not only know these skills but also have the confidence that they are capable of implementing them in different settings. Developers may need to build into their program opportunities to learn and practice such skills.
5. Plan. Change is difficult. The simpler the change, the easier it will be to adopt. A well-thought-out, easy-to-implement, and well-documented plan should be the centerpiece of change efforts. Developers hoping to effect change among users should include simple and easy-to-use implementation plans.
6. Pilot tests. Most changes fail the first time—and often several times—when they are implemented before becoming an effective part of a user’s life. It is essential that the user learn from the failures and continue working toward change. Developers should expect this and set up mechanisms to allow the user to fail safely and to learn in the process.
7. Monitoring and feedback. An important part of learning from failure is to have a monitoring mechanism that allows users to track their behaviors and the impact of their behaviors and to give feedback, not only to themselves but also to developers. Developers should build into their applications effective feedback mechanisms that will allow them to learn from users and to assist users to learn from their experiences.

Evidence from research on health communication interventions, technology-based approaches to communication, and other discrete elements should be taken into account by developers, sponsors, evaluators, and users of IHC. Some health communication interventions have been shown to be efficacious (Robinson, 1989; Campbell et al., 1994; Gustafson, Hawkins, Boberg, Bricker, Pingree et al., 1994; Strecher et al., 1994; Balas et al., 1996, 1997; Krishna et al., 1997; Shiffman et al., 1997), but research on the effectiveness of computer-based approaches is limited. In fact, most applications have not been evaluated for effectiveness. Only a small number of studies have examined the effectiveness of IHC applications in improving health status indicators (AHCPR, 1997). Of these, only a few were randomized controlled trials (Robinson, 1989; Brennan et al., 1995; Chewning, 1996; Barry et al., 1997; Morgan et al., 1997; Brennan, 1998; Gustafson, Hawkins, et al. 1999).

The potential effectiveness of IHC also is suggested by research on related discrete elements. Studies show that access to health information can enable patients to be more active participants in their care and lead to better medical outcomes (Greenfield et al., 1985; Brody et al., 1989). Patients report that they want to be informed about their medical condition (Korsch, 1984; Mahler and Kulik, 1990). The process of sharing information enhances the doctor-patient relationship. In addition, research on the effectiveness of various formats and types of media for conveying health information generally indicates that video and slides are more effective than books and audiotapes in educating consumers (Alterman and Braughman, 1991; Funnell et al., 1992; Gillipie and Ellis, 1993; Consoli et al., 1995).

The following sections examine the potential impact of IHC on satisfaction and relationship with providers, health care practice patterns, personal lifestyles, and utilization of health services.

Impact on Satisfaction and Relationship With Providers

Users are generally satisfied with IHC applications (Hassett et al., 1992). This is not surprising, because dissatisfied users are unlikely to utilize them and therefore may not be represented in studies of satisfaction. A better picture of satisfaction with IHC emerges when level of use is examined. Average use of IHC applications is high, especially for electronic support groups. For example, one study reported that, during one-year of study, caregivers of persons with Alzheimer’s disease used electronic support groups twice per week for an average of 13 minutes (Brennan et al., 1995). Similarly, another study reported that cocaine-using pregnant women used electronic services over a 7-month period an average of 3.2 times per week (Alemi, Stephens, Javalghi et al., 1996). Extensive use of IHC applications may indicate user satisfaction.

The impact of IHC on overall satisfaction with the health care system is not well understood. Some data suggest that when patients have access to both online and face-to-face counseling, they prefer online counseling. For example, in an unpublished study of recovering patients who had access to both online and outpatient substance abuse treatment, 30 percent presented for outpatient treatment and 87 percent accessed online treatment (Mahboeba Mosavel, TelePractice, Inc., personal communication, September, 1998). A randomized study of postpartum mothers showed that they were eight times more likely to use electronic support groups than face-to-face groups (Alemi, Mosavel et al., 1996). Another study showed that women with breast cancer preferred online counseling and support groups to face-to-face interactions (Gustafson et al., 1992).

The above studies suggest that use of—and, by inference, satisfaction with—face-to-face interactions may decrease when electronic-mediated options are available. One study, however, reported that IHC applications could improve patients’ confidence in their physician (Gustafson, Hawkins et al., 1999). These seemingly contradictory findings may be dependent on the extent of integration of IHC with face-to-face services. When electronic-mediated and face-to-face visits are closely integrated (e.g., both interactions are with the same clinician), then IHC may increase satisfaction with face-to-face services. When online and face-to-face encounters are not fully integrated, then online services may reduce satisfaction with face-to-face visits. In some cases, however, online encounters may strengthen trust in regular health care providers if online encounters reaffirm the advice of such providers.

Integration of online and face-to-face services is also related to providers’ attitudes towards online services. Health care providers’ satisfaction with IHC is not well documented. One of the few studies available surveyed 325 members of the American Association of Diabetes Educators about their preferences for different methods of education including books, videotapes, computer-based programs, and audiotapes (Funnell et al., 1992). Providers were least enthusiastic about computer-based applications, but this finding may reflect the quality of early IHC applications. Providers’ negative reaction to IHC is surprising in light of the findings that patients generally prefer IHC to other forms of health communication (Alemi and Higley, 1995), and evidence that IHC applications can be effective in changing patient behaviors.

It is possible that many providers have not been exposed to high-quality applications, and their attitudes may change once they use them and become more involved in discussions about IHC. Some providers’ negative attitudes toward IHC, however, may be a function of the difficulties they face in integrating these technologies into their practices (Alemi, 1998). Effective implementation of IHC applications requires not only substitution of, or integration with, educational books and pamphlets but also changes in the way clinicians interact with patients. The very nature of clinical visits changes when information can be tailored to the patient’s condition; when some components of care, such as education, can be completed before or after the visit; or when follow-up care can be accomplished without an office visit.

The impact of IHC on patient behavior may be an indirect result of changing provider practice patterns. Patients may interact with IHC applications but the results of these interactions are shared with providers who may change their advice to the patient. Good examples of such applications include shared medical decisionmaking and informed consent applications. Studies show that multimedia applications can be used to assess patient preferences (Barry et al., 1995; Jimison et al., 1998), but limited data are available on the effectiveness of shared decisionmaking applications in changing practice patterns.

Another way practice patterns might be affected by IHC applications is through computerized history taking. Studies show that people may be more likely to be truthful to a computer than to a clinician (Erdman et al., 1985). One study found that patients donating blood were more likely to report their HIV-related risk factors to a computer than to a clinician (Locke et al., 1992). These studies suggest that IHC may solicit more accurate information that ultimately changes clinical decisions and courses of treatment.

IHC also may change practice patterns by improving the efficiency of clinical visits. In one study, patients were interviewed by a computer before their visits. Findings were put into patient medical records and made available to attending clinicians (Lloren, 1998). Clinicians were not only satisfied with this service, but also thought that it had changed their practices in a positive way. Independent verification showed that these clinicians were detecting 15 percent more alcoholics than the clinic’s average detection rate.

Several studies show that mass media can effect behavior change among communities. For example, one study showed that 26 hours of mass media promotion of healthy behaviors led to a 16 percent reduction in cardiovascular risks across a community (Farquhar et al., 1990). Although mass media are not interactive, to the extent that online communications are evolving into mass media, they may be effective in bringing about widespread behavior change. With the advent of "push" technology, multimedia and video, online applications are becoming increasingly similar to established mass media, such as broadcast television.

IHC can change health behaviors, but not all applications have been successful in bringing about such change (Fitzgerald and Mulford, 1985; Alterman and Braughman, 1991; Brennan et al., 1995; Consoli et al., 1995; Balas et al., 1997). Furthermore, the impact of IHC on health behavior is not always sustained. No single health education/communication intervention has been shown to be effective in changing behavior over a time period; there is no "magic bullet." For example, one study found that computer-based instruction improved smoking quit-rates for 6 months but not for 18 months post-baseline (Lando et al., 1997). It is possible that some IHC applications have not had a lasting impact on behavior change because information alone is not sufficient for behavior change. Health education has been shown to be more effective when combined with other supporting interventions. For IHC, this reinforces the need to combine information with interactive role-playing and peer support through electronic bulletin boards. IHC applications that have integrated health information, role-playing, and support groups, have been successful in bringing about behavior change (Alemi et al., 1989; Gustafson et al., 1992).

Health education also can be made more effective by tailoring the information to key issues and patient characteristics. For example, smokers who received a letter tailored to their circumstances were more likely to quit than those who received a general message (Strecher et al., 1994). Similar results were obtained for people trying to reduce their fat intake (Campbell et al., 1994; Watkins et al., 1994).

Given that people face multiple sources of information, computerized health education is more likely to be effective when the effort is frequent and sustained over time; however, the effectiveness of brief interventions may eventually diminish. Growing evidence suggests that there must be a minimal level of interaction before the IHC can have a measurable impact on behavior. In one study, for example, no beneficial impact was measured unless people had used the system at least three times per week over a 7-month period (Alemi, Stephens, Javalghi et al., 1996). Benefits were observed for patients with even higher use patterns in another study (Taylor and Gustafson, 1998). Although both of these studies involved self-selected users, they suggest that impact of IHC may be more pronounced when patients use the service at least three times per week. Another study, however, found that it was not the amount but the type of use that most affected outcomes (Smaglik et al., 1998). Patients who had a clear purpose for using an IHC application benefited more than people who used it more often but did so in an undirected manner, and more than those who only used the application’s support group functions.

It is possible that there is a positive "dose-response" effect with certain types of IHC applications. However, one study has raised the possibility that greater use of the Internet may lead to declines in face-to-face communications with family members and increases in depression and loneliness (Kraut et al., 1998). Randomized studies are needed to clarify the relationship between exposure to IHC and positive behavior change and potential related side effects.

IHC may have a positive impact on utilization of resources either by reducing use of unnecessary services (e.g., use of emergency rooms for non-urgent problems) or by increasing use of cost-effective services (e.g., immunizations). Health education can reduce unnecessary health care visits (Fries and McShane, 1998), and there is evidence that IHC applications can have a similar impact. One study compared randomly assigned groups of university students who did or did not receive computerized health education (Robinson, 1989). The group that was exposed to the intervention had a 22.5 percent lower rate of medical visits than the group that did not receive the application.

Of particular interest is a randomized control study involving 204 HIV-infected patients (Gustafson, Hawkins et al., 1999). The experimental group was provided access to an IHC application with multiple functions, including online peer support. Computer-mediated social support was the most frequently used function of the application. Investigators found that patients with access to the application, as compared to control patients, were more likely to report higher quality-of-life in several dimensions, including social support and cognitive functioning. They also had shorter time-per-visit to ambulatory care, and less frequent and shorter hospitalizations than the control group. The experimental group had lower total health care costs than the control group.

In another study, voice-based electronic support groups were compared to face-to-face support groups (Alemi, Mosavel et al., 1996). Over time, the groups that met online were eight times more likely to meet than other groups. Furthermore, subjects who used the electronic bulletin boards were less likely to visit a health care provider. Reduced visits did not impact on health status. This study suggests that electronic support groups may help reduce inappropriate use of health services.

The above studies support the importance of IHC in bringing about behavior change and show that use of electronic support groups can lead to substantial reductions in use of services and cost-of-care in certain groups.³ Because no studies have examined the cost-effectiveness of IHC applications in large populations, however, the cost-effectiveness of employing IHC versus traditional health communication media in the general population is unknown.

Other examples of how IHC applications may influence utilization of health services are applications that focus on reminding patients (Austin et al., 1994). With some exceptions (Austin et al., 1994), when computers call to remind parents to visit a clinic, on-time immunization and vaccination rates improve (Brimberry, 1988; Linkins et al., 1994; Dini et al., 1995; Alemi et al., 1996; Lieu et al., 1998). These data suggest that IHC applications can affect health outcomes by encouraging patient compliance with scheduled visits.

Use of IHC applications, however, does not always lead to lower utilization of services. The impact of these applications on utilization and cost-of-care may depend on the effectiveness of the application and the message being communicated (Bass et al., 1998). If the content encourages a visit, then more visits are likely. If messages are neutral (as in electronic support groups) or discourage visits (as in messages encouraging self-care) then IHC applications may reduce visits.

A comprehensive review of the scientific literature on consumer health informatics related to patient decisionmaking concluded that, because of the relative paucity of studies in this area and their varying methodological quality, it was not appropriate to draw solid conclusions about the effectiveness of these applications in reducing costs, improving health outcomes, or in regards to other important measures (AHCPR, 1997).⁴ The authors identified four priority areas for research:

• Assess the effects of informatics tools on a full range of outcomes
• Identify factors that influence use of informatics tools
• Assess effects of informatics tools on patient-clinician communication
• Assess the cost-effectiveness of different types of patient informatics tools

After examining the methodologies employed by researchers in this area, the authors also proposed the following lessons for future research in this area (AHCPR, 1997):

• Describe the nature of the implementation and use of informatics tools
• Develop clear hypotheses about measurements
• Provide adequate length of follow-up
• Incorporate a no-treatment or minimal-treatment control group
• Provide adequate sample size to assess effects among key subgroups

In addition to the above guidelines, the Panel emphasizes the need to select a truly representative sample of the population to study and avoid use of self-selected participants. Research efforts in this area should also incorporate strategies to avoid important sources of bias identified in biomedical research (Jadad, 1998b; Jadad and Rennie, 1998).

Valid, reliable, and sensitive measures of IHC effectiveness are limited. Existing measures of effectiveness may not be specific enough to detect some program effects. There are many outcome measures with proven reliability and validity (e.g., the SF-36 or the FACT cancer quality-of-life scale) (Ware et al., 1994; Brady et al., 1997; McQuellon et al., 1997), but these may not be appropriate for evaluating all IHC applications. Outcome measurement scales often combine several statistically related but conceptually different elements and an application’s effect on one variable may be diluted by its lack of effect on another. For example, a scale from an instrument of proven validity and reliability examines the physician-patient relationship by inquiring about patients’ confidence in their physician and their perceived availability of the physician. An application intended to improve confidence but not physician availability may show no or less impact because of a dilution effect occurring with the use of this scale. Developers, therefore, need to precisely define the objectives of their applications before selecting outcome measures. These measures would ideally measure only those effects of interest. An expanded discussion of other challenges to evaluation of IHC applications is presented in the next chapter.

Several user-related factors may determine developers’ selection of application content, interfaces, media, and approaches (Jimision et al., 1999). In designing effective applications, awareness of individual characteristics, preferences, and other individual factors are critical. The explicit involvement of members of the target audience in application design is often essential to a successful product. Developers are frequently challenged to implement specifications that can both meet individual needs and accommodate a wide variety of users.

Individual characteristics, such as age, gender, disability, race and ethnicity, cultural factors, and socioeconomic status, may influence health-information-seeking behavior, and can account for differences in the amount and type of health-related information and support that individuals seek. Some people do not seek much information or support, and others who do may encounter serious barriers to the use of IHC applications (Eng et al., 1998). Willingness to use health information technology may also be an important consideration in designing IHC applications. Individual characteristics and preferences can be accommodated by "tailoring" content and interfaces. Ensuring accessibility of an application’s interface is essential to reach users with physical disabilities (WWW Consortium, 1998). Tailored information has been found to be more effective in providing consumer information (Mullen et al., 1985; Skinner et al., 1993, 1994; Strecher et al., 1994; Brennan et al., 1998) and is preferred by patients (Jimison, Fagan et al., 1992). Some users, however, may resist over-customized applications that are too narrowly focused.

The concept of individual preferences is important for IHC applications that focus on health decisionmaking (Mulley, 1989; Barry et al., 1988, 1995). Although patients need information about the quality-of-life associated with the medical outcomes of possible decisions, reliable assessment of individual preferences and risk attitudes for clinical outcomes are probably weak links in clinical decisionmaking. Recent efforts to explore the use of computers in communication about health outcomes and in assessment of patient preferences for various health outcomes have started to address these issues (Jimison and Henrion, 1992; Goldstein et al., 1994; Nease, 1994). Information on patient preferences is important for tailoring information to patients and for providing decision support (Jimison, 1997). In addition to differences in preferences for health outcomes, patients differ in the degree to which they choose to be involved in decisionmaking. Age (younger more than older persons), gender (women more than men), and education level (higher-educated more than less-educated persons) are generally strong predictors of desire to be involved in medical decisions. In addition, there is a greater desire to be involved in decisions in health areas that generally require less medical expertise, such as a knee injury, than those that are more complex, such as cancer (Thompson et al., 1993).

An individual’s reading ability impacts on application design. The problem of low literacy skills is widespread in the United States (Holt et al., 1990; Reid et al., 1995; Baker et al., 1996), and about one of every five adults reads at or below the fifth grade level. Only about half of people examined comprehended written health education materials and average reading levels were well below what is needed to understand standard health brochures (Morgan, 1993; Davis and Mayeaux, 1994; Feldman and Quinlivan, 1994; Baker et al., 1996). Additionally, an analysis of medical information on Web sites showed that, on average, materials were written at a 10^th grade reading level, which is not comprehensible to the majority of people (Graber et al., 1999). Lack of health literacy may be an acute problem among the elderly (Gazmararian et al., 1999).

A person who has completed a certain grade level should not be assumed to be able to read at that level. Generally, health materials should be written at least three grade levels lower than the average educational level of the target population (Jubelirer and Linton, 1994). There is a danger, however, that excessively simplifying materials may reduce the value of the program to more educated users. Interactive media can help in this situation because they can be used to accommodate a range of users with varying levels of health and technology literacy. Text characteristics and organization and clarity also impact on comprehension and retention of material (Reid et al., 1995). To address shortcomings in reading literacy, multimedia techniques can be used to facilitate comprehension. Information can be conveyed through video, audio, and graphics, in lieu of text. Additionally, presenting material in multiple languages would increase comprehension for non-native English speakers.

Point-of-Access

The ideal point of access for many of the functions of IHC (as discussed in Chapter II) is the home because this allows the user to access the application at any time of day, in privacy and comfort. However, some applications can function effectively in more public settings, such as schools and worksites. For example, shared decisionmaking applications, because of their typical one-time-use nature, may function effectively through these and other access sites, including libraries and health professional offices. Disease coping and behavior change applications that offer both information and emotional support, however, need to be immediately accessible at any time. Hence, to be used effectively, they need to be available in the home and/or portable. In addition, access to the Internet at work and home expands the availability of online employer-sponsored wellness programs that traditionally were only available at worksites.

Public access points need to be selected with a thorough understanding of the target audiences and will depend on the type of application and the relationship between intended users and the setting. For instance, many underserved populations harbor a distrust of certain institutions that might otherwise be appropriate candidates for delivering IHC applications. If government-sponsored sites, such as clinics and public buildings, are not trusted by these populations, then alternate settings, such as community centers and places of worship, may need to be considered.

The capability and performance characteristics of the hardware and software and communication pathways used by target audiences to access IHC applications are important considerations for application design. The functions and content of the application should be matched to the level of technology available to typical users. For example, integrating extensive graphics or full motion video into an application that is intended for groups of users who do not typically have computer graphics accelerators or large bandwidth access is counterproductive. The quality of a user’s experience when accessing an interactive application via a slow dial-in modem versus a much faster T1/T3 connection is so different as to almost render them as different programs.

³ Other forms of communication, such as telephone-based interventions, also can reduce cost-of-care. In a very successful trial, investigators replaced face-to-face, follow-up visits with three scheduled telephone calls (Wasson et al., 1992). Over a 2-year period, estimated total expenditures for “telephone care” were 28 percent less per patient compared to the usual care patients.
⁴ The authors of this report used the term “informatics tools” to refer to tools that “describe and present information regarding screening or treatment alternatives in order to help patients in making decisions about alternatives.” The authors examined both interactive computer-based tools and noninteractive tools such as brochures.

Return to Table of Contents

Comments: SciPICH@nhic.org   Updated: 05/01/08