Poster
# 4
Poster |
|
6th Internet World Congress for Biomedical Sciences |
J. W. Swanson(1), D. J. Capobianco(2), D. W. Dodick(3), J. Parker(4)
(1)(2)(3)Mayo Clinic - Rochester. United States
(4)Axia Health - Calgary. Canada
|
| ||
[Health Informatics] |
[Neurology] |
With the exponential growth of the Internet, critical appraisal of health-related information has proved a challenge. However, several studies and related commentaries are now appearing in the medical literature in which the quality of health-related information found on the Internet has been evaluated. In general, there is growing skepticism within the global medical community with respect to the overall credibility and utility of the ever-expanding and unregulated pool of health-related information available on the Internet.
To evaluate the accuracy, ease of access and variability of cancer information retrieval, Bierman and colleagues (7) performed a systematic study of cancer information found on the Internet, limiting the analysis to information specific to Ewing sarcoma. A number of prevalent search engines were used in the collection of such web sites and each site was evaluated in terms of its relevance to the search topic, medical information, anecdotal information and degree of peer review.
Of the 371 unique web sites dealing with Ewing sarcoma, 80 were presented in a language other than English or were otherwise inaccessible, 47 contained information not relevant to Ewing sarcoma and 30 were only peripherally related to the search term. In total, these web sites represented 50% of all sites uncovered by the search engines used in this study.
A total of 165 web sites contained medical information related to Ewing sarcoma. Of these, 35% (57 of 165) contained information that was not peer reviewed and 6% (10 of 165) of these web sites contained clearly erroneous or misleading information. These findings led the authors to conclude that erroneous or misleading information freely available on the Internet may have devastating effects on vulnerable patients and families and drive a wedge between patients and physicians. Furthermore, increasingly physicians are placed in an often compromising position of evaluating the credibility of information a patient may have encountered, either intentionally or inadvertently, on the Internet.
Henson, in his commentary of this study (8), concluded "the Internet may, in some cases, actually lead to adverse medical care by spreading misleading or even fraudulent information. Information obtained from the Internet may conflict with recommendations provided by physicians, thus leading to confusion and uncertainty in the minds of patients" (p. 373). In the arena of cancer care, Henson’s view is shared by others (9,10,11).
Given the lack of quality control and sheer volume of material available on the Internet, both credibility and utility of health-related information on the Internet should be evaluated (12,13,14). To this end, several rating instruments are now available on the Internet that review and rate web sites containing health-related information.
Jadad and Gagliardi (15) identified 47 such rating instruments and evaluated both validity and reliability criteria of each. Such instruments commonly evaluate, or rate, health-related web sites using a variety of awards, quality ratings or ‘seals of approval’. Of the rating instruments identified, only 14 provided descriptions of criteria utilized to generate ratings, 5 provided instructions for their use and none provided information concerning either interobserver reliability or construct validity.
Given the incomplete development of current rating instruments, the authors question whether such instruments should exist at all, whether they measure what they claim to measure and whether such instruments lead to more harm than good by providing misleading recommendations and ratings of health-related web sites on the Internet.
McKinley and colleagues evaluated the quality of surgery-related information available on the Internet (16). Significant variation in both accuracy and completeness of surgery information poses significant limitations on its reliability and safety for both patients and health care professionals. Such inaccuracy may have the greatest implications for patients who, through an inability to objectively evaluate surgery-related information found on the Internet, may be encouraged to practice potentially dangerous health behavior or may be mislead by false expectations regarding treatment outcomes and options.
Similar results were reported by Davison (17) who reviewed Internet web sites that provide dietary recommendations and compared these with the "Canadian Guidelines for Healthy Eating and Nutrition Recommendations for Canadians". Of the 365 such web sites found, 167 contained dietary recommendations. Of these, 76 sites (45%) provided recommendations not consistent with these established guidelines and most included commercial advertisements for dietary supplements, herbal remedies, weight loss products and promotion of specific diets. The percentage of those sites which offered inconsistent dietary information varied with the key word search term and ranged from 29% using the key word "diet" to 57% using the key word "nutrition".
The quality and accuracy of medical information shared within an electronic bulletin board for discussion of painful arm and hand conditions has been evaluated (18). Such a service, also known as a discussion group, enables ongoing discussion and contributions from patients and health care professionals alike. A total of 1658 consecutive messages were evaluated using a number of criteria including professional training of those offering medical advice, consistency of medical advice with established clinical recommendations and guidelines and the nature of evidence cited to support medical claims.
Of the messages monitored, 56% (927) addressed a medical topic. Of these, 89% (825) provided medical information by persons without medical training while 5% (37) were provided by trained health care professionals. Approximately one third of all advice provided was classified by the authors as either anecdotal or unconventional and only 9% of advice offered by patients and 18% of advice offered by health care professionals cited a published source. These findings led the authors to conclude that medical information found within such bulletin boards may be unconventional, inappropriate and based on limited medical evidence.
The relevancy of patient education resources in urology, found through two established search engines on the Internet, was evaluated by Sacchetti and colleagues (19). Relevancy rates were calculated by dividing the number relevant sites found through these search engines by the total number of sites found. Relevant sites were determined on the basis of quality and substantiation of information, comprehensiveness and objectivity. Using the search engine "HotBot", relevancy rates ranged from 10% for ".com" domains to 24% for ".edu" and ".org" domains. Overall relevancy rates when using the "Yahoo" search engine was 20%. The HotBot search engine was marginally more robust with respect to retrieving web sites relevant to the search topic. The authors conclude that, in light of the exponential growth of patient education resources available on the Internet, the number of such sites which provide complete, nonbiased information represents a small percentage of the total.
Similar results were reported in a study of health information and interactivity concerning female urinary incontinence (20). A total of 75 web sites, from academic, health care professional and commercial organizations were evaluated on the basis of several criteria including authorship, source citation, currency and balance. While there was significant variability among academic, professional and commercial sites across these criteria, 54 sites (72%) did not provide authorship of content, 50 sites (67%) did not provide source citations, 32 sites (43%) gave no indication as to currency of content and 34 sites (45%) were deemed biased in favor of commercial products or services. These results point to significant challenges with respect to overall reliability of patient education materials found on the Internet.
The author also submitted an email request for advice from a fictitious patient to each of the 75 web sites in the study. While two thirds of these sites responded to the request and, while the author suggests that information retrieved in this manner provides advice and comfort to women with urinary incontinence, there is no evidence offered which suggests that patients are able to objectively appraise such information in light of issues of reliability and credibility uncovered by the results of this study.
Reliability of information for parents on the Internet, relevant to established guidelines for managing fever in children, has been evaluated (21). A total of 41 web sites were retrieved and evaluated according to minimum temperature cited for fever, optimal sites for accurate temperature measurement, pharmacological and physical management of fever and conditions which warrant examination by a physician. Only 4 web pages (10%) adhered closely to recommended guidelines, 9 web pages (21%) provided clear direction with respect to taking a rectal temperature and 10 web pages (24%) provided accurate dose and frequency information concerning administration of antipyretics.
In addition, several web sites advocated administration of either unproven (ibuprofen) or contraindicated medications (salicylates) in the management of fever in children. Others advocated inappropriate methods to reduce body temperature, suggesting strategies of sponging (cold or alcohol) which may raise the thermoregulatory set point or, in the case of alcohol sponging, lead to alcohol inhalation induced hypoglycemia and coma in children.
These findings led the authors to conclude that that information found on the Internet must not only be complete and accurate but should be tailored to meet the specific learning needs and skills of its intended audience, in this case parents, a view shared by others (22,23).
In his commentary concerning this study, Wyatt (24) concludes "although the web makes it absurdly easy to disseminate information, by allowing anonymous authors to conceal commercial or other conflicts of interest, it does not help readers discriminate between genuine insight and deliberate invention" (p.1879). The author calls for a more systematic method of evaluating and insuring accuracy and completeness of information found on all health-related web sites for fear of "drowning in a sea of poor quality information" (p.1880).
In a recent commentary concerning issues of quality assurance for health-related information found on the Internet (25), Mcleod concludes "there is growing concern that a substantial proportion of clinical information on the Internet may be inaccurate, erroneous, misleading, or fraudulent, and thereby pose a threat to public health. It is also a concern that given the ambit of the Internet, the potential audience for misleading or dangerous information is enormous" (p. 1663). This has led to the need for quality assurance instruments that measure authorship, attribution, disclosure, scientific quality, accessibility, timeliness and readability. Yet, were such instruments to be developed, health care professionals must still assume responsibility in critically evaluating health information found on the Internet.
Others (12) agree with Mcleod and go on to suggest that "the time has come to discuss vigorously how a basic set of standards can be developed and applied in an electronic context. Such standards are built on a foundation of accountability - principles that amount to a framework for critical thinking, allowing professionals to reasonably judge whether what they are reading is credible, reasonable and useful and to make informed decisions about how to apply this information to the real world" (p. 1244).
|
|
| ||
|
[Health Informatics] |
[Neurology] |
