This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Narimatsu, H. Articles by Nakata, Y. Search for Related Content PubMed PubMed Citation Articles by Narimatsu, H. Articles by Nakata, Y. Social Bookmarking                            What's this?

Detailed Analysis of Visitors to Cancer-Related Web Sites

Division of the Strategic Outcome Research Program for Cancer Control, Ministry of Health, Labour and Welfare Commission, Japan Cancer Society; Division of Exploratory Research, Institute of Medical Science, University of Tokyo, Tokyo, Japan

Tomoko Matsumura, Tomohiro Morita, Yukiko Kishi, Koichiro Yuji, Masahiro Kami

Division of Exploratory Research, Institute of Medical Science, University of Tokyo, Tokyo, Japan

Tsunehiko Komatsu

Third Department of Internal Medicine, Teikyo University School of Medicine, Tokyo, Japan

Yuji Tanaka

Division of Exploratory Research, Institute of Medical Science, University of Tokyo, Tokyo, Japan

Tomohiro Sawa, Yoshinori Nakata

Medical Information and System Research Center, Teikyo University School of Medicine, Tokyo, Japan

To the Editor:

Web sites are a valuable source of information for cancer patients.¹ Patients are seeking information necessary for their own treatment, as well as general cancer information. To satisfy such needs of cancer patients, it is necessary to build Web sites that are conducive to patients’ individual needs, as well as to have organic linkage between a wide variety of sites. Although achieving this end requires sufficient study of the characteristics of cancer-related Web site users, there is little research on the topic, leaving an unclear picture of the actual state of cancer-related Web site users. Therefore, in this study, we conducted an access analysis of cancer-related Web sites to shed light on the characteristics of their visitors, which is information necessary for improving the user friendliness of such Web sites.

Using Keyword Advice Tool (Overture KK, Tokyo, Japan),² we first selected 96 keywords pertaining to cancer that have been used in more than 3,000 searches per month on Yahoo! as of September 2006. Next, we used the 96 selected keywords to conduct Yahoo! searches,³ and then selected 2,000 Web sites that came up in these searches. We then used Keyword Advice Tool to obtain the number of searches performed with each keyword and ranked the Web sites proportionate to the number of searches. Then we computed a ranking score by giving the nth-ranking keyword of the converted ranking a 1/n value (eg, the first-ranking site gets 1,000 points, the second-ranking site half of that, and so on). We also assigned a hit frequency score for the frequency with which each Web site appeared in searches with each of the 96 keywords. We then computed the final score with the product of the ranking score and the hit frequency score and extracted the 100 highest scoring sites as the subject of this study. Blogs in the present study also included homepages on patients’ personal experiences fighting cancer.

There are two main methods for conducting access analysis; these are analyzing the Web server logs and obtaining access logs through JavaScript tags embedded in each page of a Web site.⁴ Both methods require the site author to collect the log data. It is possible to obtain the Uniform Resource Locator (URL) of the pages visited, the page viewed before visiting the site, the Internet Protocol (IP) address of the visitors, and the time of visit. In this study, we requested the following information from the selected 100 Web sites. We requested a summary of aggregated results from the Web sites already compiling access data on their own. For the Web sites that are not compiling access data on their own but that can obtain a server log, we obtained the access logs from the site authors and then compiled the data ourselves. For all of the other Web sites, we embedded tags to collect data for the purposes of this study and compiled the data.

Of the sites, 25 Web sites complied with our request and consented to participate in the study; characteristics of these sites are listed in Table 1. Each site operator agreed to participate in the study on the condition of anonymity. The number of visitors to cancer center sites was overwhelmingly higher than the number of visitors to cancer patients’ blogs.

View larger version (19K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 1. Average page views of hospital site by day of the week.

View larger version (26K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 2. Average weekly access to homepages on fighting cancer. The value on the y-axis is number of accesses to the site every 3 hours.

View larger version (29K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 3. Distribution of visitor familiarity and defection. The size of the spheres indicates the amount of aggregated data.

We selected Web sites in a wide range of categories for this study. In addition to hospital and pharmaceutical company Web sites, we also targeted a large number of homepage sites on patients’ personal experiences fighting cancer. We first screened for sites that are influential among users. Cancer patients’ homepage sites constituted 9% of the influential sites initially selected. Previous studies, however, have not focused on these cancer patients’ homepage sites.^1,5-8 Blogging on one's experiences with cancer has enabled the flow of information among patients that goes beyond time and space. Homepages provide a means for communication among patients and their families that is more convenient and costs less than traditional face-to-face patient organizations. It is possible that these sites provide information that is not provided by medical care providers but that is useful to patients. It is likely that patients’ blog sites will become an important category of Web sites in the future.

This study showed that visitors’ access patterns vary among different types of Web sites. Many people visited hospital-type sites on weekday afternoons, whereas few visited these sites on nonworking days. In contrast, there was hardly any variation between days of the week in visits to cancer patients’ blogs, which peaked at night (Fig 1). This fact demonstrates that the background or the status of use varies between users of hospital-type sites and homepage sites. Although we cannot draw any definitive conclusions as a result of insufficiently detailed data, we can infer from the fact that the peak in visits to hospital-type sites coincided with hospital consultation hours that many of these people use these sites as they prepare for hospital visits. At the same time, we can also deduce that there is a tendency for people to visit cancer patients’ blogs during their spare time. This may reflect the fact that people use cancer patients’ blogs not for one-way transmission of information but as a tool for communication among patients and their family members.

In this study, the rate of visitors who use MSN⁹ to reach cancer-related Web sites was generally low compared with the Japanese national average. Moreover, cancer patients’ homepage visitors tended to go through Yahoo! and Google¹⁰ more often than MSN compared with visitors to other categories of Web sites. In general, people who frequently access the Internet use Google, whereas those who access it less frequently use Yahoo!, and those who access the Internet even less frequently use MSN, which comes bundled in many computers’ initial setup configuration. Considering this, we can deduce that visitors to cancer-related Web sites and, in particular, to cancer patients’ homepages are highly literate with information technology and frequently access the Internet.

This study demonstrated that the repeat rate of visitors varies depending on the attributes of particular Web sites. The repeat rate of visitors to breast cancer–related homepages was extremely high compared with the average repeat rate of information service–type sites, which is approximately 25% to 30% (T. Nobue, personal communication, 2006). We observed the same trend from the results of visitor familiarity as well. This indicates that there are many avid fans of breast cancer–related homepages, which represents a significant departure from information service–type sites such as those of cancer centers. Even among cancer-related homepages, the repeat rate for leukemia-related sites was extremely low. There are a number of possible reasons for this. First, compared with breast cancer, there are many different subtypes of leukemia, with varying symptoms and duration. Therefore, visitors may more often find that the leukemia-related Web site they visited was not describing the exact subtype of leukemia they intended to look up. Second, the survival rate is higher and the duration of illness is longer for breast cancer than for leukemia. Third, because information technology is less diffuse among elderly individuals, it is possible that there are more visitors at the Web sites related to the types of cancer that occur at an earlier age, such as breast cancer.

To our knowledge, this study was the first to shed light on the characteristics of cancer-related Web site visits. However, there are a number of issues that need to be considered. First, because the number of sites from which we obtained data is limited, we cannot generalize for all cancer-related Web sites. More large-scale studies with a wider scope of target sites will be needed in the future. Second, this study demonstrated that homepage sites on people's personal experiences fighting disease are forums for the communication of information among patients. It is possible for patients and their families to obtain information through these sites that they cannot get from medical care providers. With few previous studies on these sites, more research is needed on the role of these sites in improving patient literacy, as well as the limitations of these homepage sites. Finally, this study showed that the background and behavior of cancer Web site visitors differ among different types of Web sites. This suggests that visitors have diverse needs. Cancer-related Web sites need to be designed from this perspective to make them easy to use and beneficial to their users.

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The author(s) indicated no potential conflicts of interest.

AUTHOR CONTRIBUTIONS

Conception and design: Hiroto Narimatsu, Masahiro Kami, Tomohiro Sawa, Yoshinori Nakata

Collection and assembly of data: Hiroto Narimatsu, Tomoko Matsumura, Tomohiro Morita, Yukiko Kishi, Koichiro Yuji, Tsunehiko Komatsu, Yuji Tanaka

Data analysis and interpretation: Hiroto Narimatsu, Masahiro Kami, Tomohiro Sawa, Yoshinori Nakata

Manuscript writing: Hiroto Narimatsu, Masahiro Kami, Yoshinori Nakata

Final approval of manuscript: Hiroto Narimatsu, Tomoko Matsumura, Tomohiro Morita, Yukiko Kishi, Koichiro Yuji, Masahiro Kami, Tsunehiko Komatsu, Yuji Tanaka, Tomohiro Sawa, Yoshinori Nakata

ACKNOWLEDGMENTS

We are grateful to Toshiteru Nobue (Maikai Solutions, Tokyo, Japan) for his technical support. Supported by Grant from Health and Labor Sciences Research Grants, the third-term comprehensive control for cancer (H18-012 to Dr Nakata) from the Ministry of Health Welfare and Labor of the Japanese Government, Tokyo, Japan.

REFERENCES

1. Newnham GM, Burns WI, Snyder RD, et al: Information from the Internet: Attitudes of Australian oncology patients. Intern Med J 36:718-723, 2006[CrossRef][Medline]

2. Overture. http://inventory.jp.overture.com/d/searchinventory/suggestion/

3. Yahoo! Japan. Homepage. http://www.yahoo.co.jp/

4. Nikkei BP Consulting Inc: Access analysis. http://consult.nikkeibp.co.jp/consult/bizmark/feature.html

5. Alper BS: Usefulness of online medical information. Am Fam Physician 74:482:485, 2006

6. Berland GK, Elliott MN, Morales LS, et al: Health information on the internet: Accessibility, quality, and readability in English and Spanish. JAMA 285:2612-2621, 2001[Abstract/Free Full Text]

7. D’Alessandro DM, Kingsley P, Johnson-West J: The readability of pediatric patient education materials on the World Wide Web. Arch Pediatr Adolesc Med 155:807-812, 2001[Abstract/Free Full Text]

8. Morita T, Narimatsu H, Matsumura T, et al: A study of cancer information for cancer patients on the internet. Int J Clin Oncol 12:440-447, 2007[CrossRef][Medline]

9. MSN Japan: Homepage. http://jp.msn.com/

10. Google Japan: Homepage. http://www.google.co.jp/

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Narimatsu, H. Articles by Nakata, Y. Search for Related Content PubMed PubMed Citation Articles by Narimatsu, H. Articles by Nakata, Y. Social Bookmarking                            What's this?