Lesli's Random and Non-So-Random Musings

Article Review
"Sillence, E., et. al. (2007). How do patients evaluate and make use of online health information? Social Science and Medicine, 64 (2007). 1853-1862.

Critique:
This small-scale longitudinal study of 15 women in the UK describes study results on the evaluation of trustworthiness and usefulness of online health information websites. The study subjects, ages 41-60, were to search the Internet for information on menopause and hormone replacement therapy (HRT).

Of the 15 women in the study, all of whom were at or near menopause, each had at least a high school education with half completing at least some college. Half of the women were currently working in business or administrative jobs and half self-classified as "housewife" or "retired." All of the women used the Internet on at least a weekly basis and were familiar with basic search techniques. The women were recruited through advertisements and were all interested in finding out more information on menopause and HRT. This study took place from August 2003 to April 2004, in N.E. England (UK).

The study design consisted of three phases. The first, the observation phase, consisted of the observation of online search behaviors. There were four two-hour sessions at an Internet cafe. Weeks one and four of this phase consisted of the women performing independent searches using their own criteria for acceptance or rejection. All websites visited, regardless of time spent, were to be logged by each woman. Independent computerized logging was also performed. Weeks two and three consisted of the women using directed searching designed with specific trust design elements (based on e-commerce trustworthiness elements). The second phase, was designed to understand Internet use and the integration of information across sources over time. Phase two consisted of each woman keeping a diary or log for six months from the end of phase one. In the diary, they were to log each visit to an online health site as well as to describe other, off-line health-search behaviors (friends, physicians, periodicals, etc.). In the final phase, phase three, the researchers conducted telephone interviews one month after the end of phase two, to gather patient reflections on the value of Internet-based health information.

Phases one and three used back-up methods for gathering search and reflective information. In phase one, computer logging software was utilized for comparison purposes with the information actually logged by each subject. Discrepancies were discovered, with fewer websites logged by the subjects than by the software, although instructions were explicit for each subject to log every site visited regardless of rejection or acceptance. In phase three, telephone interviews were recorded and transcribed by the researchers.

In phase one it was discovered that the subjects thought content (83%) was more important than design (17%) in the selected or trusted sites, while design (94%) was deemed more important than content (6%) for the rejected or mistrusted sites. Additionally, the subjects deemed most trustworthy those sites by "reputable organizations" but did not trust "corporate" sites. Pharmaceutical sites were not deemed trustworthy by the subjects. Also considered trustworthy, with good information, were those that presented the information according to the subject's search query and those sites the subject could socially identify with -- "written by people like me for people like me." One common complaint of the subjects had to do with risk comprehension. The subjects felt they were "overloaded with numbers" and "alarmed by some of the bolder risk statements." Phase two had the least compliance by the subjects. Three women did not turn in diaries at all and two more returned empty diaries stating that they either had no time or no need to look at any websites. This is the one phase that was not monitored in any other fashion by the researchers. Of those that completed the diaries, entries were generally brief with few new sites being visited. The diaries did prove useful to the researchers regarding the integration of material from different online and offline sources. In the final phase of telephone interviews, the interviews were recorded and transcribed by researchers. Then two researchers read and coded the transcripts using "constant comparison" to gather emerging themes. Overall, researchers in this study found that:

• Few of the women made repeat site visits -- they got what they wanted and moved on
• The Internet played an important role in early decision making (what to do about HRT)
• The Internet allowed the women to "act like scientists" by creating hypotheses and doing their own research about the way menopause affected them personally
• Online advice did not usually lead to any "fundamental changes" in behavior but did reinforce decisions already made
• Subjects developed more self-confidence in their own decision-making abilities.

While a very small study, the results were consistent with other current similar studies. The ubiquitious nature of the Internet and the growing online presence of health-related sites will likely result in more people of all ages using the Internet for health information that they may or may not then share with their physicians or other healthcare providers. Sensitive, formerly taboo health subjects such as sexuality, STDs, and menopause are now being openly and frankly discussed on numerous blogs and discussion groups. Mainstream health-related web sites like WebMD and AARP also include information on these issues.