More and more people are using the Internet to access health-related information. In 2003, an estimated 93 million Americans used the Internet for this purpose, a 27% increase over 2002 (1,2). People use the Internet because of several favorable features: the convenience of being able to search quickly for information at any time, unlimited access to inexpensive information resources, and user anonymity when searching for subject-sensitive health information (1,3-5).

Several studies have found that the health topics most frequently searched on the Internet include the following: 1) background information as well as symptoms and treatment of a disease or condition, 2) medications and prescription drugs, 3) diagnostic tools, 4) new or experimental treatments, 5) diet, 6) exercise, and 7) support groups (2,6-9). Although 86% of Internet users indicated concern about obtaining unreliable information, 52% of participants who used health sites thought that "almost all" or "most" health information found on the Internet was credible (1). In contrast, 84% of physicians rate their patients as only "fair or poor" at appraising the quality of Web-accessed information and believe that although accurate or relevant information can be beneficial, inaccurate or irrelevant information can be harmful to patients' health care and outcomes (10). Results from a systematic review, however, found a low risk of harm associated with using the Internet as a health information resource (11).

Because of the wide range in quality of health information available on the Internet, one major concern is where people search for health-related information on the Internet. In addition, having access to numerous inexpensive sources that may or may not be credible raises questions about whether people who search the Internet for health information also seek advice from their physicians when experiencing serious health problems (12). Using data from a recent mail survey of the U.S. population, we characterized the population that uses the Internet for accessing health information, and we evaluated use of various types of health-related Web sites. We also explored how often individuals in this population shared health-related information found on the Internet with their physicians. Because the Internet can be used as a valuable tool for both health promotion and disease prevention, we examined factors that would encourage using the Internet to search for health information.

Because 67% of all deaths are attributable to five chronic diseases (heart disease, cancer, stroke, chronic obstructive pulmonary disease, and diabetes) and because these diseases account for 75% of the nation's total health care cost (13), we examined whether using the Internet as a health information resource differed among individuals having one or more chronic diseases compared with those having no chronic disease.

Of the total 8432 eligible HealthStyles respondents in 2002 and 2003, 52% were women and 48% were men. The mean age was 48 years, and 22% were aged between 35 and 44 years. Most respondents were white (72%) and were married (59%); approximately one third had some college education (32%) and earned $60,000 or more per year (32%). Sixty-nine percent reported having good health or better, and 59% had one or more chronic diseases.

Thirty-nine percent of the women and 33% of the men used the Web as a health information resource (Table 1). On average, use of the Web for seeking health information was highest among respondents aged 25 to 34 years (43.5%), Asians or Pacific Islanders (49.2%), and those who were married (37.6%). Seeking health-related information on the Web increased with level of education (P < .001) and income (P < .001). Participants who reported their general health as being fair or poor sought online health information more often than participants who reported their health as being excellent, very good, or good.

We included 2966 (35%) respondents who reported seeking health information on the Internet in a subgroup analysis. Among these respondents, health information portals were the most frequently used Internet source for health information (Figure 1); 58% of respondents reported using health information portals, 31% reported using government agency sites, and 29% reported using nonprofit organization sites. When respondents were asked which features of the Internet encouraged them to search for online health-related information (Figure 2), 68% reported ease of finding the desired information as the primary factor. Other factors included clarity and ease of use of accessed information (60%), not having to pay or register to use the Web-based health site (53%), and trustworthiness of the information source (43%). Among respondents who reported accessing health information on the Web, 53% reported sharing the information with their physicians sometimes; 11%, usually; 3%, always; and 32%, never (data not shown).

Of the respondents who reported seeking health information on the Internet, 57% reported having one or more chronic diseases. Of these respondents, 37% reported having high blood pressure; 33%, high cholesterol; 33%, arthritis; 31%, depression; 18%, asthma; 16%, diabetes; 5%, osteoporosis; and 3%, cancer. Internet users who had one or more chronic conditions differed significantly by age (P < .001), race (P < .001), income (P = .009), education level (P < .001), marital status (P = .006), and health status (P = .02), but they did not differ significantly by sex (data not shown). Respondents with no chronic diseases differed significantly in their use of type of Web site from those who reported having one or more chronic diseases only in the use of health-related chat rooms (P = .002). Similarly, significant differences were found in the following features of the Internet that would encourage use of the Web more often to find health-related information: 1) easy to find the desired information (P < .001), 2) health sites have less advertising (P = .01), 3) Internet connection is faster (P = .01), and 4) health sites have information about people like me (P < .001) (data not shown).

After adjusting for sex, age, education, and income, results from multivariate logistic regression indicated that women were 1.36 (95% confidence interval [CI], 1.21–1.52) times as likely as men to use the Internet as a resource for health information (Table 2). People with at least a college degree were 2.68 (95% CI, 2.30–3.13) times as likely to seek health information online as people with a high school degree or less. In addition, respondents who earned $60,000 or more per year were twice as likely as respondents who earned less than $25,000 (odds ratio [OR] = 2.10; 95% CI, 1.78–2.48) to seek health information online. The odds of searching the Internet for health information decreased with increasing age; no significant effect of race on Internet use was found.

Internet users with one or more reported chronic disease were 1.30 (95% CI, 1.16–1.45) times as likely to seek health-related information as Internet users with no chronic disease. Furthermore, respondents with depression (OR = 1.47; 95% CI, 1.27–1.71) and high cholesterol (OR = 1.18; 95% CI, 1.02–1.37) were as likely to seek health-related information on the Web compared with those without these conditions. From regression analyses on number of chronic diseases, in contrast to the unadjusted analyses, respondents who reported having one chronic disease were 1.26 (95% CI, 1.11–1.44) times as likely to seek health information on the Web and respondents who reported having two or more chronic diseases were 1.35 (95% CI, 1.16–1.56) times as likely as respondents who reported having no chronic diseases.

Thirty-five percent of our study population used the Internet to search for health-related information. In contrast, results from the Health Information National Trends Survey (HINTS), the purpose of which is to identify the health information needs and practices among the American public (16), showed that 53% of respondents used the Internet to search for health or medical information for themselves and 48% searched for someone else (17). This difference may be explained by distinct differences in study design between the HealthStyles survey and HINTS.  

Internet use differed according to sex and age and was strongly associated with income and education. Women were more likely than men to seek health information, perhaps because women are more often considered caretakers and thus may be more concerned with their own health as well as the health of their family and friends. Although older people (aged 65 or older) are shown to be the fastest growing age group of online users (18), the activity of searching the Internet for health information decreased with age in our study. Our finding could possibly be attributed to the fact that older people are more likely than younger people to be new Internet users and thus often lack familiarity with computers and have limited knowledge of how to conduct and interpret Internet searches (18). One study found that 77% of Internet users who had 2 or 3 years of online experience searched for health information on the Web and that 59% of new Internet users made such searches (2). Similarly, previous use of health-related Web sites was found to be a significant predictor of interest in online health information (19). 

Although Internet access is strongly associated with income and education, differences in Internet use for health-related information cannot be solely attributed to lack of Internet access (8,18). In a study in which participants received free Internet access, only one fourth of them actually searched the Internet for health information (20). Consequently, differences in Internet use also may reflect other priorities, more compelling needs, and lack of interest in health information as well as education, income, and content barriers such as literacy, language, and cultural diversity (8,18,21,22). 

As previously stated, respondents with either one or two or more chronic conditions were more likely than respondents with no chronic conditions to search for online health-related information. This finding could be attributed to the possibility that people with chronic health conditions want to control and manage their illnesses. Several studies have shown that using the Internet to obtain health-related information improved people's understanding of a disease or condition, increased their knowledge of treatment options and outcomes, influenced their exercise or eating behavior, influenced their treatment decision making, and helped in coping with their illness (5,6,8,10,12,22,23). In addition, 85% of physicians interviewed reported that patients experienced benefits from using the Internet to search for health-related information (22). Many health care providers also believe that patients who are well educated about health are more involved in their own health care and adhere better to treatment regimens (5,10,23-26). A significant difference in the use of health-related chat rooms among people with one or more chronic diseases compared with those who have none could be because chat rooms not only allow patients to exchange information and provide emotional support but also promote disease management and adherence to medication (27).

Although using the Internet did not affect the number of visits or phone calls patients made to their physicians (6), the majority (53%) of our study sample sometimes shared information that they accessed on the Web with their doctors. Possible reasons why patients might hold back information included feeling uncomfortable discussing highly personal issues, having less time to talk with their physician, and fear of receiving negative feedback from their physician (5,22). Nonetheless, in a survey of doctors' experiences of patients using the Internet, 44% of the physicians reported that their patients experienced problems in using the information obtained from the Internet, and 8% reported that their patients experienced actual physical harm (22). The most commonly reported problems included obtaining misinformation about a condition and patient requests for new, unavailable treatments and alternate or experimental therapies (5,22). Hence, it is imperative that patients be encouraged to always share concerns, including health-related information that they access from the Internet, with their doctors to avoid adverse health outcomes.

In this study, the three types of Web sites (i.e., health information portals, government agencies, and nonprofit organizations) most frequently cited by Internet users for researching health-related information indicate that the public is accessing high-quality, credible online health information from trustworthy sources (28). Web sites from health-related government agencies, academic health science centers, and commercial health information publishers are likely to have high-quality information (25). Web sites ending in ".com" may have self-interests that compromise the quality or completeness of the information presented compared with Web sites ending in ".org," ".edu," or ".gov" (26).

The two most important factors that would encourage Internet users to use the Web more often as a health information resource are ease of finding the desired information and health information that is clear and easy to use. Hence, Web sites should present information clearly and should preferably be written at a seventh- or eighth-grade reading level so that all Internet users can use the information as a tool to improve their health. It is also important that 37% of respondents reported that having fewer advertisements would also encourage them to use a Web site. Advertisements are viewed as a red flag, indicating that the information might be biased toward marketing partners (28).

The findings reported in this study may be subject to several limitations. First, because HealthStyles is a mail survey, the sampling methods used generate a sample population in which minority and low-income households are underrepresented. Nonetheless, the HealthStyles survey is based on a national sample, and its data correlate well with surveillance data from the Behavior Risk Factor Surveillance System (14). Second, we relied on self-reported data, which might not provide complete or accurate findings on Internet use for obtaining health-related information.

Facilitating appropriate use of the Internet to locate and interpret high-quality health information is a worthy challenge for the future. Public health professionals and the medical community can play key roles in this endeavor by becoming more aware of the types of information available on the Internet, by critically appraising the available information, by rating the value and trustworthiness of the sources, and by helping to identify high-quality sites. In addition, health care providers should consider opportunities for using the Internet to communicate with their patients. Results from the HINTS survey indicate that 7.6% of respondents reported using e-mail to communicate with a physician or physician's office (17). Furthermore, studies on the use of e-mail and Web messaging as tools for communication between patients and their physicians have shown that these were easy to use, allowed flexibility in communicating nonurgent problems, improved continuity of care overall, and enhanced disease management (29,30).

In recognizing the increasingly important role that the Internet plays in our nation's health, both the public health and medical communities have a unique opportunity to use it as a valuable tool for health promotion and disease prevention. The Internet can help educate patients, enrich the patient–physician relationship, and influence patients' health care and health outcomes. We encourage further studies that evaluate physicians' reactions to patients searching for online health information, explore additional barriers that the public faces when searching for online health information, assess how successful people are in finding health information, and determine how people use the information (i.e., diagnosis, treatment, or prevention).