The Internet is a major resource for health information, but the quality of the information on different Web sites has not been systematically studied. Sandvik evaluated the quality of Internet information on female urinary incontinence. He chose this subject because of the need for public awareness that urinary incontinence can be treated and because some patients may not broach the subject with their physicians and may seek impersonal sources of information.

Internet sites for female urinary incontinence were found by using seven general search engines, such as AltaVista, Yahoo, Excite, Hotbox and Galaxy, and nine medical search engines, such as Medsite, OMNI, Cliniweb and Medical World Search. The quality of each site was evaluated on the basis of the timeliness of the data, the declaration of ownership and authorship, mention of sources of information, the balance of the content and the ease of the Web site's navigability and interactivity.

The quality of the content was further assessed by means of a four-point scale (zero = not mentioned; 1 = mentioned; 2 = briefly explained; and 3 = comprehensively explained). This type of assessment was made for each of 14 key items, such as the epidemiology of incontinence, a description of the different types of incontinence, drugs that may cause incontinence, diagnostic procedures and treatment options such as pelvic floor exercises, biofeedback, alpha-adrenergic agonists, estrogen, electrostimulation, vaginal cones and pessaries, bladder neck support, urethral plugs and shields, collagen and other bulking agents, surgery and pads.

Interactive sites were evaluated by their responses to a fictitious patient created by the author. This patient described herself as a 63-year-old woman who had urinary incontinence for several years. She was seeking advice on what she could do to stop the problem.

A total of 75 sites were evaluated, with 25 sites each in the categories of professional institutions such as hospitals, universities and clinics, commercial companies and organizations such as societies, foundations and journals.

The mean score for quality was highest for the organization sites. The mean total quality score was 24.2 for organizations, 21.6 for professional sites and 17.1 for commercial sites. Statistically significant differences in the quality scores existed between professional and commercial sites and between organizational and commercial sites. The most common deficiencies related to the source and currency of the information and in the declaration of authorship. Commercial sites accounted for 14 of the 15 sites that lacked balance.

The author found that the most informative sites were easily found with general search engines. With respect to the interactive portion of the assessment, the fictitious patient received 66 e-mail messages, 32 of them within 24 hours. Few provided comprehensive information about specific treatments. Twelve of the responders warned the patient that her urinary incontinence could be related to the antihypertensive agent she was taking. The most common advice, given by 33 of the 66 respondents, was to consult a physician. Twenty-three directed the patient to other Web sites, 16 offered to send printed material and 15 offered to arrange treatment.

The author concludes that excellent information concerning female urinary incontinence may be found on the Internet. Although none of the Web sites provided comprehensive information, valuable advice and support, particularly from interactive sites, was obtainable.