Abstract

Using the Internet as a mode for health promotion is appealing. There are important methodological considerations to the approach, but there are also important reasons why people will and will not participate in Internet interventions. This is a report on data from 4601 people who completed an online survey of sexual risk behavior in 2000. Most indicated they would visit a website for STD/HIV prevention information (61%), but fewer would open an E-mail (45%) or chat (30%) about the topic. Top reasons for rejecting website, E-mail and chat room education about STD/HIV are given. Logistic regression results showed men who have sex with men (MSM) and persons with a history of testing for STD are consistently more likely to endorse STD/HIV prevention through chat rooms (MSM 1.8, STD testers 1.3), E-mail (MSM 1.6, STD testers 1.2) and websites (MSM 1.8, STD testers 1.2). The data demonstrate the Internet may facilitate health promotion among MSM who may not be reached in a publicly funded STD prevention setting. The Internet may also act as a good adjunct to STD information obtained in clinic settings among those who seek STD testing.

Introduction

The Internet is becoming an attractive medium for health promotion and delivery of behavior change intervention (Glasgow et al., 1999; Sharp, 1999; Winker et al., 2000). The future for such intervention appears bright, with advantages such as cost-effectiveness, convenience and `mass media' ability to reach many persons with relevant, tailored information (Gustafson et al., 1999). There are, however, some important considerations for Internet intervention, including barriers we face in implementation. This paper discusses barriers to Internet intervention strategies specific to STDs, including HIV prevention from the perspective of Internet users who participated in an online survey of risk behaviors related to acquisition of and transmission of STD/HIV.

Barriers to STD/HIV Internet interventions are multiple; they include uncertainty regarding efficacy, questions about quality of information available online, privacy, access to information and methodological concerns. Such barriers are relevant and useful to examine for any type of online health intervention.

Do Internet interventions work?

Evidence shows that people use the Internet to solicit sexual partners, increasing their risk for STD infection (McFarlane et al., 2000). Outbreaks of syphilis have been traced back to partners met on an online chat room (Klausner et al., 2000). To address this new risk for STD infection, researchers are contemplating online behavioral interventions. However, the efficacy of such interventions is unknown. While computer-based (e.g. Internet, CD-ROM, hand-held devices) behavior change interventions have shown promise with regard to short-term change in health behaviors, little is currently known about their efficacy with regard to maintenance of behavior change over the long run (Brug et al., 1999). In one study of HIV-positive individuals, participants who accessed a specific, tailored, computerized program for 6 months maintained favorable health behaviors 2 months after use of the program had been discontinued. Those who were only able to access the program for 3 months did not maintain favorable health behaviors (Gustafson et al., 1999).

Quality of information available online

Efficacy of computer-based interventions notwithstanding, issues regarding participants' trust in the validity of information provided through the Internet in particular may present barriers. A poll by Louis Harris & Associates indicates that the Internet was used as a healthcare information resource by approximately 60 million US adults (Gustafson et al., 1999). On the other hand, it is difficult to determine if information can be trusted and the potential for misinformation is great. A recent presentation demonstrated that persons who access STD care at an inner city clinic who use the Internet want to find STD prevention information online (Rietmeijer et al., 2001). In a study that focused on locating sex education information on the Internet, out of nearly 6 million web pages identified by key word searches, only 41 pages provided relevant, educational information and many of these contained subjective or incomplete content, or reinforced `myths'. The other pages consisted of advertisements, personal home pages and organizational position statements (Smith et al., 2000).

Privacy concerns among Internet users

If users can trust information obtained online, can they be sure the information they provide is protected? Users of Internet sites providing medical information indicate that personal privacy is their number one concern (Winker et al., 2000). To address this issue, use of the Internet for health research and health interventions must be accompanied by strong protection of confidentiality to foster the trust of person who can most benefit from online programs. Steps to help ensure online confidentiality include limiting and/or encrypting email contact and advising participants of security limitations (Childress and Asamen, 1998).

Access to the Internet

Internet interventions appeal in part because they may reach large numbers of people with relatively less effort than face-to-face interventions. However, ongoing evidence of a socioeconomic `digital divide' implies that Internet-based interventions might not reach the populations at highest risk for disease. Access to the Internet is currently estimated at 130 million US households (A. C. Nielson Survey, 2000) and there is some evidence that the Internet is becoming increasingly available to groups that previously had limited access (Lake, 2000; Mandl et al., 2000; Witte et al., 2000). However, while some have maintained that the digital divide is shrinking, Internet sites indicate an under-representation of individuals with less than a high-school education, non-whites and households whose income is less than $30000 (US Department of Commerce, 2000).

Methodological challenges

While the advantages of Internet-based applications are numerous, particularly with regard to cost-effectiveness, numbers of potential users and data collection capabilities (Gustafson et al., 1999; Witte et al., 2000), there are methodological cautions that may present barriers to designing interventions. Online interventions may take a number of formats including: (1) passive information systems offered on websites, (2) interactive approaches such as bulletin boards and chat rooms, (3) applications that allow the individual to complete a self-assessment and receive tailored feedback or (4) some combination of the three. Passive information systems may allow limited interaction, i.e. users may be able to click to get more information, or to link with other sites, but do not interact with any live individual. Examples of passive formats include position statements, on-line journal information and organizational web pages, such as Mayo Clinic's Health Oasis (www.mayohealth.org), or government sponsored sites, such as The National Women's Health Information Center (www.4woman.gov).

The interactive approaches to online health interventions, i.e. bulletin boards and chat rooms offer two-way communication, peer support or `expert' resources. There are clear methodological issues, however, with these intervention strategies, including the inability to verify self-report data, determine eligibility, ascertain if participants are representative of the general population, determine the characteristics of individuals who choose not to participate and control loss to follow-up (Soetikno et al., 1997; Binik et al., 1999; Cho and LaRose, 1999; Treadwell et al., 1999). In addition to the methodological concerns, little is known about personal motivation to access specific Internet features such as websites, chat rooms and E-mail. Interventions designed to utilize these features should be informed about the intrinsic barriers faced by potential users.

To help surmount the aforementioned methodological concerns, some interventions extend contact with the clinical provider by offering support or information as a supplement to a clinic visit (McTavish et al., 1995; Glasgow et al., 1999; Gustafson et al., 1999). While these are important to study, the focus of this paper is to consider those online interventions that do not have a purpose that is adjunct or complimentary to specific clinic encounters and to consider the challenges we face in implementing such interventions, particularly from the consumer viewpoint. The Internet has potential to reach a good number of people who would not otherwise seek care or be targeted for health related intervention. We report on findings from a survey of 5474 persons about their sexual risk behavior that was conducted between 3 April 2000 and 3 August 2000 (18 weeks), and focus specifically on what they told us about their willingness to participate in Internet-based interventions that involved a website, chat room discussions, or reading and/or responding to E-mail messages.

Method

We developed a 68-item self-administered anonymous survey with the main objective of documenting Internet behaviors related to seeking sexual partners online, and STD/HIV-related risk behaviors with Internet and non-Internet partners. As part of the study, we also recorded demographic data and asked people to tell us if they would be willing to participate in an Internet intervention that involved visiting a website, or one that involved posting/responding to messages on a bulletin board, or participate in a chat room, or an intervention that involved opening and responding to E-mail messages. The study was approved for persons aged 18 and older by the Colorado Multiple Institutional Review Board (COMIRB) and by Institutional Review Boards at the Centers for Disease Control and Prevention and the AMC Cancer Research Center. Persons logging onto www.SexQuiz.org during the study period saw a `home page' of information about the study, with links to other websites offering STD/HIV prevention information, local testing information and frequently asked questions (FAQs) about SexQuiz. Those interested in taking the SexQuiz could click on a button to take them to the next page, which contained the study informed consent. After reading the consent, those who wished to continue with the study clicked on a button saying, `I agree' and began to self-administer the 68-item survey.

Measures

The survey included questions on demographics, i.e. age, gender, employment, insurance and education. In order to assess general health behaviors, we asked people to tell us if they were smokers, if they exercised three times a week or more and if they ate five servings of fruit/vegetables daily. We asked people to tell us if they had ever had sex with a person they first met online (termed Internet partners) and if they ever had sex with a person they did not first meet online (non-Internet partners). We asked respondents to indicate the gender of their Internet and non-Internet partners. Those men who indicated that any partners were male or male and female were classified as men who have sex with men (MSM). After responding to all the STD/HIV-related risk assessment questions we asked respondents to offer preferences for receiving STD/HIV information electronically. Specifically we asked, `Would you visit a website with information about STD/ HIV?', `Would you participate in a chat room discussion or a bulletin board list about STD/HIV?' and `Would you open an E-mail message from someone you knew that was about STD/HIV?'. For those who answered negatively to any of these questions, we asked them to tell us why they would not participate in the particular activity.

Recruitment and enrolment

The survey was posted online at the website www.SexQuiz.org from 3 April through 3 August 2000. In an earlier phase of the study staff identified chat rooms and bulletin board sites where people exhibited behaviors suggesting partner solicitation (Bull and McFarlane 2000). Staff returned to these locations to recruit persons for the online survey. They posted information about and links to the survey in chat rooms, on bulletin boards, and through multiple STD/HIV prevention list serves and websites. Staff made an effort to identify potential recruits from rooms where ethnic/ minority participation was likely (e.g. rooms titled `Ethnic Ebony Love', `Ebony and Ivory', `Latin Lover', `BlkM4M'). Of 71 chat rooms used for online recruitment over the course of the study, staff visited minority-oriented sites 11 times or 15%. Once in a chat room, staff would send out private messages (that other chatters would not see) to participants to encourage their enrollment in SexQuiz. They could occasionally infer ethnicity or race of participants by their `handles' or `profiles', i.e. information chatters offer about themselves (e.g. 19/bl/orlando, Latino in LA, etc.); staff sent 37 private messages in the 11 racial/ethnic-oriented rooms, of 430 sent in total (9%). However, there is no consistent way to ascertain racial/ethnic information of chatters online; non-white persons may participate in chat rooms without a racial/ethnic title and they may or may not offer clues to race/ethnicity in profiles or handles. In addition to staff recruitment that occurred throughout the study, media attention about the survey resulted in high recruitment—we received 61% of all responses (3503 of 5474) in 1 week coinciding with national media attention paid to the research.

Persons enrolled were offered a `risk score' after the survey as a small incentive for completion. This was a crude assessment developed algorithmically from the risk assessment that weighted various risk behaviors (e.g. sex without a condom, number of partners, being high or drunk during sex more than 50% of the time) and placed people into `limited', `moderate' and `higher' risk categories, with explanations of what each risk behavior was, why it was risky and how risk could be reduced. Upon receiving their score, persons were redirected to the page containing links to other STD/HIV-related websites, including testing resources, for further information.

By 3 August 2000, there were 5474 surveys completed. Of these, 64 persons who stated their age was under 18 or over 87 (i.e. age 116) completed the survey, and were ineligible because of age. An additional 561 persons living outside North America (i.e. the US, Canada and Mexico) and 248 persons with irreconcilable inconsistencies in survey responses were removed from data analysis. The final sample reported on here includes 4601 persons. There were 231 persons who read the consent form and then clicked on a button labeled `I do not agree to continue'. Most common reasons offered for refusal to participate included lack of relevancy (persons who never used the Internet to find a sex partner—although this was not a criteria for eligibility) 38%, invasion of privacy (19%) and no time (11%). We cannot assess the number of persons who accessed the site and left before clicking on the `consent' or `do not consent' buttons.

Data analysis

As mentioned, persons who responded negatively to questions about willingness to participate in Internet-based interventions were offered an opportunity to describe why they would not choose any given approach (e.g. chat, E-mail, website). These questions were open ended and we coded each response by placing it into one of several categories that will be described in detail in the results. There were two people involved in coding responses; samples of 100 responses were selected randomly for each open-ended question and the second person coding re-coded these responses. The inter-rater reliability for these sample coding segments was 0.91.

We tabulated data to describe the sample and their preferences for chat rooms, E-mail and websites related to HIV and STD prevention. We also examined frequencies of reasons people offered as to why they would not use a particular Internet feature for a STD/HIV prevention intervention. We examined χ2 tests for significance using the SAS program to compare demographic and behavioral characteristics with preferences for chat rooms-, E-mail- and website-related STD/HIV prevention interventions. In addition, we constructed multivariate logistic regression models to control for relationships between these variables and chat room, E-mail and website preferences. Because one independent variable in these models, MSM, excludes 29% of the sample (women), we ran models with and without MSM to verify consistency of findings. The magnitude and significance of results for each model was not altered, so we retained MSM in the models. In addition, to eliminate concerns regarding multicollinearity, we assessed the correlation between males and MSM, both independent variables in the logistic regression models, and found values below 0.30, suggesting multicollinearity would not be present in the models.

Results

Table I shows the demographic characteristics of the sample. Respondents are predominately male (71%), white (81%) and well educated (51% with a college or post-graduate degree). The age distribution of the sample shows relatively even proportions aged 18–35, with slightly lower proportions aged 36–50 and few over 50 years. Most (85%) respondents are insured and most (88%) are employed.

The top reasons people gave for not using chat rooms, E-mail or visiting websites devoted to STD/HIV prevention are shown in Table II. Less than a third of respondents (29.5%) indicated they would participate in a chat room discussion regarding STD/HIV prevention. Those indicating they would not chat about STD/HIV say they do not feel they need STD/HIV information from chat (40%). Examples of responses coded in this category included `I'm much too smart to need some Internet idiot's help on STD prevention' and `I'm HIV-positive and I already know the risks'. Others say they get their STD/HIV information elsewhere (38%) and examples coded in this category include `I go elsewhere, this is not helpful' and `I don't need to communicate in both directions—I do my own STD research'. Fourteen percent indicate chat rooms are suspect—not a good source for reliable information—and 9% say chat rooms are not anonymous. Finally, a small percentage (6%) indicate they simply do not like chat, with responses like `I've become bored with chat', `chat has a lot of hostile straight people' and `chat rooms attract some very strange people'.

A higher proportion of people (45%) indicate they would open an E-mail message about STD/HIV prevention. For those who would not, top reasons for declining this option include not needing STD/HIV information via E-mail and opening E-mail only from known persons (both 39%) followed by disdain for junk or `spam' mail (36%), saying `spam is a poor way of reaching people and I won't honor it', `If it is sent to a related list (e.g. listserv or E-group) then yes, blanket forward, no' and `I get too much E-mail period'. Many (21%) indicate they fear misuse of their E-mail [e.g. America Online has temporarily restricted accounts where hackers steal E-mail identities and use them to spam others]. Finally, a small group (14%) indicated they feared receiving a virus over E-mail.

The largest proportion of people was willing to visit an STD/HIV-related website (60.6%). Reasons given for not visiting a site are similar to those for not chatting, i.e. that people do not need STD/HIV information (62%) and they get it elsewhere (39%). Far fewer (6%) indicate they do not know which sites to access (`where would I get this information?') and that they know everything about STD/HIV (`I already have this information', `I've been monogamous for 5 years and know how to take care of myself').

Table III presents findings from bivariate associations between selected demographic and behavioral characteristics and willingness to chat, E-mail or visit a website related to STD/HIV prevention. The data show that while differences exist, the relative risk is low, suggesting the magnitude of the difference is not great. Data appear consistent for each type of activity, i.e. that younger people, non-smokers, exercisers, men, those with a job, MSM, insured, those with no college degree and those who have been tested for STD appear to be more willing to chat, open E-mail and visit websites.

The results from a logistic regression controlling for these effects are shown in Table IV. These data indicate that two variables consistently stand out as predictors of willingness to chat, E-mail and visit website, i.e. testing for STD (those testing are 1.28 times more likely to chat, 1.23 times more likely to E-mail and 1.17 times more likely to visit the web) and being MSM (who are 1.79 times more likely to chat, 1.64 times more likely to E-mail and 1.84 times more likely to visit the web). Education is inversely related to willingness to chat (those with a college degree or more are less likely to chat), but a predictor of visits to a website (1.10 times more likely to visit a website). Being white is a deterrent to all three activities, with whites less likely to chat, E-mail or visit websites. Age, employment and behavioral factors such as exercise, smoking and fruit/ vegetable consumption did not retain significance when included in multivariate analyses.

Discussion

Our findings are useful for anyone considering using the Internet for intervention. We have demonstrated that we can recruit large numbers of people to an online survey; while 65% of the responses coincided with national media attention, there were 1971 additional respondents, the majority of whom logged on prior to any media coverage of the work. While we have shown that we can recruit people to an online survey, we have also uncovered important information from them on what approaches to health information and intervention are most appealing, and what approaches may not work.

We have also demonstrated that we can ask people sensitive, private questions online and they will answer them. There will always be concern for the validity of responses, particularly in situations such as this survey where participants are anonymous. On the other hand, data entered anonymously may be less likely to be corrupted or less subject to reporting error. The data that we gathered does have consistent similarity to data on Internet and STD/HIV risk collected in a more controlled setting, i.e. an inner-city STD clinic (McFarlane et al., 2000).

We have focused on understanding support for and barriers to three avenues for reaching people online: through chat rooms, E-mail and websites. This sample most strongly supported using the Internet to access STD/HIV prevention information through a website (60.6%), although almost half (45%) favored E-mail and almost a third (29.5%) favored chat participation. This illustrates preferences for information exchange in this particular sample. By closely examining reasons for rejecting an activity, we can develop strategies for how to improve chances of acceptability of chat, E-mail and website visit options. The biggest barrier to getting endorsement for any activity is a sense that people `don't need information'. While it is a commonly held notion (supported historically through the Health Belief Model) (Rosenstock, 1974) that people need to perceive themselves to be at risk before being open to a risk reduction message, it is challenging to consider ways to do that via the Internet. For STD/HIV prevention, perhaps increasing evidence about the Internet as an STD/HIV risk environment (McFarlane et al., 2000; Bull and McFarlane, 2001) may help. However, other viable approaches to consider may be using additional, alternative strategies, capitalizing on the `teachable moment', that can simultaneously build on preference for website information. For STD/HIV prevention, this may involve partnership with websites where partner solicitation is common to include links to STD/HIV prevention websites. For non-STD/HIV areas, similar strategies to send messages when people may be thinking about a risk behavior, e.g. posting links to anti-tobacco websites on sites targeting young girls may be an effective way to engender interest.

People may increase acceptance of E-mailed messages about risk behavior or health promotion if program planners can avoid the perception that their message is `spam'. This finding is supported by evidence from the San Francisco Health Department, whose experience in tracking partners of persons with known syphilis infection was substantially augmented by a personal endorsement of the infected partner when E-mail messages were sent (Klausner et al., 2000). Personalizing E-mails, getting permission and endorsement to use list-serves, and finding ways to make the information as credible as possible are all approaches that might improve acceptability to the E-mail health promotion approach.

Fewer persons accept chat rooms as a viable, credible source for STD/HIV information, but strategies that can address perceptions that chat offers misinformation and that confidentiality will be honored may improve chat room intervention acceptability. The San Francisco Health Department facilitated their use of chat rooms to encourage cooperation in the partner notification process during a syphilis outbreak by partnering with a web-based Gay-advocacy group that gave credibility to their effort (Klausner et al., 2000).

Our data on predictors for acceptability of any one of these strategies to reach people on the Internet show that persons with some level of health-seeking behavior, i.e. those having tested for STD, might be more reachable with interventions. This may be a self-selecting group and underscores the difficulty of reaching those who do not perceive themselves to be at risk. However, this does suggest promise for clinics that want to use the Internet as a source for health promotion that can serve as an adjunct to reinforce education received at a clinic visit.

Our data also show that MSM are more likely than non-MSM to participate in these activities. We can use this finding to our advantage. MSM are a group susceptible to STD in part because of Internet sex partner solicitation (McFarlane et al., 2000; Bull and McFarlane, 2001). Unlike other health promotion programs that attract the `worried well' and do not effectively reach those at high risk, these data suggest the Internet may be of high appeal to MSM and may prove an ideal medium to promote risk reduction among this group, particularly because the Internet may be able to reach this group more effectively than a more traditional venue for STD/HIV prevention messages, e.g. a publicly funded health clinic.

The findings cannot be extrapolated to a larger population, but they do offer some useful ideas of how to use the Internet for interventions. All interventions should endeavor to demonstrate rigid privacy protection and to offer high quality information. When relying on E-mail for intervention, programs can use listserves and known contacts for sending out messages so they are not perceived as `spamming' recipients, and can partner with a credible online source to sponsor chat room discussions. The Internet may be a useful tool as an adjunct to information obtained in clinic settings. Finally, for MSM who appear willing to read E-mail and participate in chat, the Internet may offer a way to reach those who may not be accessing clinics with important health promotion messages.

Table I.

Demographic characteristics of sample (n = 4601)

Characteristic n Percent 
aData missing from 121 cases (3%). 
bData missing from 165 cases (4%). 
Male 3266 71 
Race/ethnicity 
    white 3737 81 
    black 211 05 
    Hispanic 188 04 
    mixed 164 04 
    Asian 139 03 
    Native-American 55 01 
    other 34 01 
    won't answer 66 01 
Education 
    eighth grade or less 00 
    some high school 71 02 
    high school graduate 502 11 
    some college/technical school 1663 37 
    college graduate 1430 32 
    post-graduate degree 854 19 
Age 
    18–25 1469 32 
    26–35 1640 36 
    36–50 1226 27 
    51+ 266 05 
Insureda 3936 85 
Employedb 4045 88 
Characteristic n Percent 
aData missing from 121 cases (3%). 
bData missing from 165 cases (4%). 
Male 3266 71 
Race/ethnicity 
    white 3737 81 
    black 211 05 
    Hispanic 188 04 
    mixed 164 04 
    Asian 139 03 
    Native-American 55 01 
    other 34 01 
    won't answer 66 01 
Education 
    eighth grade or less 00 
    some high school 71 02 
    high school graduate 502 11 
    some college/technical school 1663 37 
    college graduate 1430 32 
    post-graduate degree 854 19 
Age 
    18–25 1469 32 
    26–35 1640 36 
    36–50 1226 27 
    51+ 266 05 
Insureda 3936 85 
Employedb 4045 88 
Table II.

Top five barriers cited to use of Internet features for STD/HIV prevention interventions

Chat rooms Indicating barrier (%) (n = 2467)b E-mail Indicating barrier (%) (n = 2467)b Websites Indicating barrier (%) (n = 1782)c 
aMissing 98 cases. 
bMissing 114 cases. 
cMissing 78 cases. 
Do not need STD/HIV information from chat 40 Do not need STD/HIV information via E-mail 39 Do not need STD/HIV information 62 
Get STD/HIV information elsewhere 38 I only open E-mail from people I know 39 I get STD/HIV information elsewhere 39 
High potential for mis-information 14 I get too much junk E-mail—I wouldn't open it 36 Do not know which sites to access 
Chat rooms are notanonymous/confidentiality may be compromised I fear misuse of my E-mail address 21 I already know everything about STD/HIV 
Do not like chat I fear receiving a virus 14 It is too hard to access sites 
Chat rooms Indicating barrier (%) (n = 2467)b E-mail Indicating barrier (%) (n = 2467)b Websites Indicating barrier (%) (n = 1782)c 
aMissing 98 cases. 
bMissing 114 cases. 
cMissing 78 cases. 
Do not need STD/HIV information from chat 40 Do not need STD/HIV information via E-mail 39 Do not need STD/HIV information 62 
Get STD/HIV information elsewhere 38 I only open E-mail from people I know 39 I get STD/HIV information elsewhere 39 
High potential for mis-information 14 I get too much junk E-mail—I wouldn't open it 36 Do not know which sites to access 
Chat rooms are notanonymous/confidentiality may be compromised I fear misuse of my E-mail address 21 I already know everything about STD/HIV 
Do not like chat I fear receiving a virus 14 It is too hard to access sites 
Table III.

Bivariate associations between demographic and behavioral characteristics and preference for chat rooms, E-mail and website STD/HIV prevention intervention

Characteristic Endorsing chat rooms (%) (total 29.5%) P value, RR Endorsing E-mail (%) (total 45%) P value, RR Endorsing website (%) (total 60.6%) P value, RR 
Smoker 9.72 0.01, 0.95 14.22 0.01, 0.97 17.86 0.01, 0.88 
Non-Smoker 19.75  30.74  42.72  
Exerciser 17.98 0.01, 1.12 27.57 0.01, 1.17 36.79 0.01, 1.18 
Non-Exerciser 11.48  17.46  23.82  
Fruit/veg daily ≥ 5 9.20 0.01, 0.94 13.94 0.01, 0.93 18.25 0.01, 0.93 
Fruit/veg daily < 5 20.31  31.11  42.39  
Employed 26.81 0.01, 1.08 40.90 0.01, 1.02 55.74 0.01, 1.42 
Not employed .46  4.02  4.68  
Age ≤ 25 19.21 0.01, 1.05 30.18 0.01, 1.02 40.86 0.01, 1.01 
Age 26+ 10.26  14.84  19.74  
White 23.16 0.01,0.77 35.66 0.01,0.76 48.44 0.01, 0.78 
Non-white 6.31  9.36  12.16  
Male 20.39 0.01, 1.07 31.05 0.01, 1.12 41.96 0.01, 1.14 
Female 9.08  13.97  18.64  
Education ≥ college 14.03 0.01, 0.08 23.17 0.01, 1.08 31.88 0.01, 1.25 
Education < college 15.77  21.82  28.66  
Insured 25.63 0.01, 1.01 39.90 0.01, 1.11 53.89 0.01,1.19 
Not Insured 3.80  5.23  6.91  
Man who has sex with men 19.12 0.01, 1.67 28.64 0.01, 1.75 37.56 0.01,1.85 
Man who doesn't have sex with men 10.35  16.38  23.04  
Tested for STD 18.38 0.01, 1.46 27.32 0.01, 1.41 35.94 0.01, 1.42 
Never tested for STD 11.06  17.82  24.69  
Characteristic Endorsing chat rooms (%) (total 29.5%) P value, RR Endorsing E-mail (%) (total 45%) P value, RR Endorsing website (%) (total 60.6%) P value, RR 
Smoker 9.72 0.01, 0.95 14.22 0.01, 0.97 17.86 0.01, 0.88 
Non-Smoker 19.75  30.74  42.72  
Exerciser 17.98 0.01, 1.12 27.57 0.01, 1.17 36.79 0.01, 1.18 
Non-Exerciser 11.48  17.46  23.82  
Fruit/veg daily ≥ 5 9.20 0.01, 0.94 13.94 0.01, 0.93 18.25 0.01, 0.93 
Fruit/veg daily < 5 20.31  31.11  42.39  
Employed 26.81 0.01, 1.08 40.90 0.01, 1.02 55.74 0.01, 1.42 
Not employed .46  4.02  4.68  
Age ≤ 25 19.21 0.01, 1.05 30.18 0.01, 1.02 40.86 0.01, 1.01 
Age 26+ 10.26  14.84  19.74  
White 23.16 0.01,0.77 35.66 0.01,0.76 48.44 0.01, 0.78 
Non-white 6.31  9.36  12.16  
Male 20.39 0.01, 1.07 31.05 0.01, 1.12 41.96 0.01, 1.14 
Female 9.08  13.97  18.64  
Education ≥ college 14.03 0.01, 0.08 23.17 0.01, 1.08 31.88 0.01, 1.25 
Education < college 15.77  21.82  28.66  
Insured 25.63 0.01, 1.01 39.90 0.01, 1.11 53.89 0.01,1.19 
Not Insured 3.80  5.23  6.91  
Man who has sex with men 19.12 0.01, 1.67 28.64 0.01, 1.75 37.56 0.01,1.85 
Man who doesn't have sex with men 10.35  16.38  23.04  
Tested for STD 18.38 0.01, 1.46 27.32 0.01, 1.41 35.94 0.01, 1.42 
Never tested for STD 11.06  17.82  24.69  
Table IV.

Logistic regression: predictors of willingness to chat, E-mail and visit a website for STD/HIV prevention

Characteristic Chat E-mail Website 
 P value Odds ratio (CI) P value Odds ratio (CI) P value Odds ratio (CI) 
Smoker 0.32 NS 0.39 NS 0.07 NS 
Exercise 0.16 NS 0.11 NS 0.01 1.17 (1.03, 1.33) 
Five fruit/veg daily 0.03 1.17 (1.01, 1.37) 0.03 1.16 (1.00, 1.33) 0.35 NS 
Employed 0.40 NS 0.48 NS 0.00 1.44 (1.15, 1.79) 
Age 0.19 NS 0.38 NS 0.57 NS 
Test for STD 0.00 1.28 (1.11, 1.48) 0.00 1.23 (1.08,1.40) 0.01 1.17 (1.03, 1.33) 
Male 0.35 NS 0.08 NS 0.22 NS 
MSM 0.00 1.79(1.53, 2.1) 0.00 1.64 (1.44,1.87) 0.00 1.84 (1.61, 2.09) 
White 0.00 0.75 0.63, 0.89) 0.00 0.71 (0.61, 0.84) 0.00 0.73 (0.62,0.86) 
Insured 0.54 NS 0.45 NS 0.42 NS 
Education 0.00 0.90 (0.84, 0.97) 0.33 NS 0.00 1.10 (1.03, 1.18) 
Characteristic Chat E-mail Website 
 P value Odds ratio (CI) P value Odds ratio (CI) P value Odds ratio (CI) 
Smoker 0.32 NS 0.39 NS 0.07 NS 
Exercise 0.16 NS 0.11 NS 0.01 1.17 (1.03, 1.33) 
Five fruit/veg daily 0.03 1.17 (1.01, 1.37) 0.03 1.16 (1.00, 1.33) 0.35 NS 
Employed 0.40 NS 0.48 NS 0.00 1.44 (1.15, 1.79) 
Age 0.19 NS 0.38 NS 0.57 NS 
Test for STD 0.00 1.28 (1.11, 1.48) 0.00 1.23 (1.08,1.40) 0.01 1.17 (1.03, 1.33) 
Male 0.35 NS 0.08 NS 0.22 NS 
MSM 0.00 1.79(1.53, 2.1) 0.00 1.64 (1.44,1.87) 0.00 1.84 (1.61, 2.09) 
White 0.00 0.75 0.63, 0.89) 0.00 0.71 (0.61, 0.84) 0.00 0.73 (0.62,0.86) 
Insured 0.54 NS 0.45 NS 0.42 NS 
Education 0.00 0.90 (0.84, 0.97) 0.33 NS 0.00 1.10 (1.03, 1.18) 

The authors would like to gratefully acknowledge Patrick Piper for his assistance in coding over 5000 open-ended responses to the survey. This research was supported through a Cooperative Agreement with the Centers for Disease Control and Prevention (R18/CCR815954-02).

References

A. C. Nielson Survey (2000) A. C. Nielsen survey finds nearly two-thirds of U.S. population age 12 or older are online. Available: http://www.acnielson.com/news/american/us/2000/2000508.htm
Binik, Y. M., Mah, K. and Kiesler, S. (
1999
) Ethical issues in conducting sex research on the Internet.
Journal of Sexual Research
 ,
36
,
82
–90.
Brug, J., Campbell, M. and Van Assema, P. (
1999
) The application and impact of computer-generated personalized nutrition education: a review of the literature.
Patient Education and Counseling
 ,
36
,
145
–156.
Bull, S. S. and McFarlane, M. (
2000
) Soliciting sex on the Internet: What are the risks for sexually transmitted diseases and HIV?
Sexually Transmitted Diseases
 ,
27
,
545
–550.
Bull, S. S. and McFarlane, M. (
2001
) HIV and sexually transmitted infection risk behaviors among men seeking sex with men on-line.
American Journal of Public Health
 ,
91
,
988
–989.
Childress, C. A. and Asamen, J. K. (
1998
) The emerging relationship of psychology and the Internet: proposed guidelines for conducting Internet intervention research.
Ethics & Behavior
 ,
8
,
19
–35.
Cho, H. and LaRose, R. (
1999
) Privacy issues in Internet surveys.
Social Science Computer Review
 ,
17
,
421
–434.
Glasgow, R. E., McKay, G., Boles, S. M. and Vogt, T. M. (
1999
) Interactive computer technology, behavioral science, and family practice.
The Journal of Family Practice
 ,
48
,
464
–470.
Gustafson, D. H., Hawkins, R., Boberg, E., Pingree, S., Serlin, R. E., Graziano, F. and Chan, C. L. (
1999
) Impact of a patient-centered, computer-based health information/support system.
American Journal of Preventive Medicine
 ,
16
,
1
–9.
Klausner, J. D., Wolf, W., Fischer-Ponce, L., Zolt, I. and Katz, M. H. (
2000
) Tracing a syphilis outbreak through cyberspace.
Journal of the American Medical Association
 ,
284
,
447
–449.
Lake, D. (2000) Access up, divide shrinks. The Standard, 20 August.
Mandl, K. D., Feit, S., Pena, B. M. G. and Kohane, I. S. (
2000
) Growth and determinants of access in patient E-mail and Internet use.
Archives of Pediatric Adolescent Medicine
 ,
154
,
508
–511.
McFarlane, M., Bull, S. S. and Rietmeijer, C. A. (
2000
) The Internet as a newly emerging risk environment for sexually transmitted diseases.
Journal of the American Medical Association
 ,
284
,
443
–446.
McTavish, F. M., Gustafson, D. H., Owens, B. H., Hawkins, R. P., Pingree, S., Wise, M., Taylor, J. O. and Apantaku, F. M. (
1995
) CHESS (Comprehensive Health Enhancement Support System): an interactive computer system for women with breast cancer piloted with an underserved population.
Journal of Ambulatory Care Management
 ,
18
,
35
–41.
Rietmeijer, C. A., Bull, S. S., Landrigan, J., Douglas, J. M. and McFarlane, M. (2001) The Internet: opportunities for sexually transmitted infection (STI) risk, STI prevention, or both? Presented to the Conference of the International Society for STD Research, Berlin.
Rosenstock, I. M. (
1974
) Historical origins of the health belief model.
Health Education Monographs
 ,
2
,
328
–335.
Sharp, J. W. (
1999
) The Internet: changing the way cancer survivors obtain information.
Cancer Practice
 ,
7
,
266
–269.
Soetikno, R. M., Mrad, R., Pao, V. and Lenert, L. A. (
1997
) Quality-of-life research on the Internet: feasibility and potential biases in patients with ulcerative colitis.
Journal of the American Medical Informatics Association
 ,
4
,
426
–435.
Treadwell, J. R., Soetikno, R. M. and Lenert, L. A. (
1999
) Feasibility of quality-of-life research on the Internet: a follow-up study.
Quality of Life Research
 ,
8
,
743
–747.
US Department of Commerce (2000) Falling through the Net: Toward the Digital Divide. US Department of Commerce, Washington, DC.
Winker, M. A., Flanagin, A., Chi-Lum, B., White, J., Andrews, K., Kennett, R. L., DeAngelis, C. D. and Musacchio, R. A. (
2000
) Guidelines for medical and health information sites on the Internet: principles governing AMA Websites.
Journal of the American Medical Association
 ,
283
,
1600
–1601.
Witte, J. C., Amoroso, L. M. and Howard, P. E. N. (
2000
) Research methodology: method and representation in Internet-based survey tools—mobility, community and cultural identity in Survey 2000.
Social Science Computer Review
 ,
18
,
179
–195.

Comments

0 Comments