What is the quality of surgery-related information on the internet? Lessons learned from a standardized evaluation of 10 common operations.

Background: Although there is high-quality information on the Internet, it is difficult for patients to identify high-quality Web sites from those with inaccurate or misleading information. Our goal was to determine specific characteristics of Web search results that yield high-quality information and can be discerned easily by patients.

Study Design: A validated rating system was used to evaluate surgical Web sites for appropriateness and adequacy. Web sites were identified using three search term types (technical, descriptive, and layperson) for 10 common surgical procedures. The top three sponsored (paid) and unsponsored (unpaid) Web site matches were identified. The search and analysis were repeated 1 month later.

Results: One hundred forty-five Web sites were retrieved: 90 unsponsored and 55 sponsored. Unsponsored sites had higher mean composite scores than sponsored Web sites (50.6% versus 25%, p < 0.0001). Searches using layperson terms had lower mean composite scores compared with those using technical terms (36.9% versus 47.5%, p < 0.02). Professional Web sites had the highest mean composite scores (66.3%); legal Web sites had the lowest (6.3%). On regression analysis, unsponsored Web sites were associated with higher composite scores (p < 0.0001); number 1 match results (p < 0.02) and using layperson search terms (p < 0.052) were associated with lower mean composite scores. Repeat search results demonstrated no significant differences, except number 3 match results were no longer significant.

Conclusions: To optimize patients' Web searches, surgeons should recommend unsponsored sites; suggest professional society sites, if available; and provide technical search terms. But information on some topics, such as risks of not undergoing surgery, remains poor and requires discussion between the surgeon and patient.