A national telephone survey (Online Health Search 2006) of 2,928 adults was conducted between August 1 and August 31, 2006. Interviews were conducted by Princeton Survey Research Associates International, an independent survey research firm with>20 years of experience in designing and conducting interviews globally on a wide range of topics. The survey was conducted under the direction of the Pew Internet and American Life Project, a nonprofit initiative of the Pew Research Center that is funded by the Pew Charitable Trusts to examine the social impact of the Internet.
Data collected by telephone can systematically underrepresent certain groups in a population. Data were therefore weighed based on Census 2005 parameters (i.e., sex, age, education, race, Hispanic origin, population density, and region) to ensure that sample characteristics were consistent with those of the population of U.S. adults living in households with telephones. All analyses for this article refer to weighted data. Weights were derived using the Deming algorithm, an iterative technique that simultaneously matched the distribution of all weighting variables to their population targets. Participants were 1,992 adults (68% of the full sample) who indicated that they use the Internet “at least occasionally.” Their mean age was 43.3 years (median=42; standard deviation=16.9) and 51.7% were women. Racial/ethnic status was 74.9% White, non-Hispanic; 8.9% Black, non-Hispanic; 9.8% Hispanic, and 6.3% other race. Eighty-three percent of participants were employed or retired. Total family income levels were 8.3% under $20,000, 27.0% between $20,000 and $49,999, 16.4% between $50,000 and $74,999, 12.9% between $75,000 and $99,999, and $16.0% over $100,000, with 19.4% missing data. More than one-third (36.1%) obtained a college degree, 27.1% reported some college, 31.4% had a high-school degree only, and 5.4% did not complete high school. Community type (i.e., rural vs. nonrural) used Census Bureau criteria whereby respondents were categorized as “rural” if they reside in a non–metropolitan statistical area (MSA) county. Nonrural participants included “suburban” adults (i.e., living in a portion of an MSA county that is not in a central city) and “urban” adults (i.e., living within a central city of an MSA). Rural adults represented 15.5% of the sample.
The full telephone interview (available online at www.pewInternet.org
) assessed demographics, frequency of Internet use, how users spend their time on the Internet, and access and use of health-related information online. Questions measuring access and use of health information online were categorized as follows: (1) types of health information accessed, (2) use of health-related information, and (3) reactions to health-related information.
Participants were included in this analysis if they answered affirmatively when asked “Do you use the Internet, at least occasionally?”
Types of Health-Related Information Accessed
Sixteen items assessed adults' history of searching online for health information. Participants were asked whether they had “ever looked online” for information about a specific disease or medical problem; a certain medical treatment or procedure; experimental treatments or medicines; alternative treatments or medicines; diet, nutrition, vitamins, or nutritional supplements; exercise or fitness; prescription or over-the-counter drugs; immunizations or vaccinations; how to quit smoking; problems with drugs or alcohol; depression, anxiety, stress, or mental health issues; environmental health hazards; sexual health; a particular doctor or hospital; health insurance; and dental health.
Use of Health-Related Information Online
Online health information seekers (n=1,594) were asked how they used the information they obtained. First, they were asked “Did you later talk with a doctor or other healthcare professional about the information you found online, or didn't you happen to do this?” Second, “Did the health information you found in your last search online have a major impact on your own healthcare or the way you care for someone else, a minor impact, or no impact at all?” Those indicating that the information had some impact (n=854) were asked six questions about how the information affected their own healthcare routine or the way they cared for someone else. These items, asked in random order, were structured as follows: “Did the information you found online: (a) affect a decision about how to treat an illness or condition? (b) change your overall approach to maintaining your health or the health of someone you help take care of? (c) change the way you cope with a chronic condition or manage pain? (d) affect a decision about whether to see a doctor? (e) lead you to ask a doctor new questions, or to get a second opinion from another doctor? and (f) change the way you think about diet, exercise, or stress management?”
Reactions to the Information
The final two questions asked about perceptions of harm or help: “Finally, have you or has anyone you know been seriously harmed by following medical advice or health information you found on the Internet?” and “Have you or has anyone you know been significantly helped by following medical advice or health information you found on the Internet.”
Random-digit-dial methodology was used to recruit adults living in U.S. households with telephones. New pieces of sample were released daily and were kept in the field for at least 5 days. A minimum of 10 attempts were made to complete an interview before a case was dropped. Call attempts occurred on different days and at different times to maximize likelihood of contact. Once contact was made, interviewers asked to speak with the youngest adult man currently in the home. If an adult man was not available, interviewers asked to speak to the youngest adult woman. This procedure has been found to yield samples that closely approximate the population with regard to age and gender.27
Of the residential phone numbers in the sample, 75.7% were contacted by an interviewer; 38.3% agreed to participate in the survey, which is typical for a national telephone survey.28
The eligibility rate for households among those agreeing to participate was 86.1%. The completion rate for interviews was 93.7%.