Increasing numbers of people gather information about health issues on the internet, especially via Wikipedia.1,2 However, Wikipedia’s open editing policy and the lack of assessment of contributors’ competence raises concerns about information quality.3,4 Therefore, the American Academy of Ophthalmology launched EyeWiki, which is open to contributions from all ophthalmologists anywhere in the world, in an effort to improve information quality and accuracy.
Rhegmatogenous retinal detachment (RRD) is a serious ocular disease that might motivate affected patients and their relatives to search for information online. The aim of this study is to evaluate the correctness of information about RRD provided by different sources and to compare the results with PubMed, which serves as a benchmark for all medical issues.
Using 17 predefined questions (Table), 3 vitreoretinal surgeons (D.I., W.W., and A.W.) independently evaluated the major medical sites that appear when one searches for information on RRD, including the National Eye Institute, Wikipedia, WebMD, Mayo Clinic, MedicineNet, and EyeWiki. No approval of this study by a local ethics committee was required, because no patient data was evaluated.
The 17 questions were composed by the authors based on the frequently asked questions of patients they had encountered. The reproducibility of questions was not tested. The search results were scored from 0 to 3, depending on completeness of provided information (no information, 0; some rudimentary information, 1; additional but incomplete information, 2; and complete information sufficient for understanding RRD, 3). The observers scored each question independently, and the arithmetic mean of scores for each question was calculated. The mean (SD) score for each source is displayed. The significance level was not defined before data collection. Since the data were ordinal, the intersource differences were calculated with the Kruskal-Wallis analysis of variance and Mann-Whitney U test as a post hoc test. The interobserver reproducibility was measured with an intraclass correlation coefficient (Spearman correlation).
The scores were similar among the 3 observers (r = 0.94 between W.W. and D.I., r = 0.92 between D.I. and A.W., and r = 0.95 between W.W. and A.W.; P < .001). The Table displays the scores of freely available sources. PubMed hypothetically scored 3 points for each question, resulting in a mean (SD) score of 3 (0), with an unmeasurable 95% CI and a total score of 51 points. Among freely available sources, Wikipedia scored 2.4 (1.1) points (95% CI, 1.8-2.9 points); MedicineNet, 1.5 (1.1) points (95% CI, 0.9-2.0 points); EyeWiki, 1.4 (1.2) points (95% CI, 0.8-2.0 points); National Eye Institute, 1.3 (1.3) points (95% CI, 0.6-1.9 points); Mayo Clinic, 0.9 (1.3) points (95% CI, 0.3-1.6 points); and WebMD, 0.8 (1.3) points (95% CI, 0.2-1.5 points).
Overall, there was a significant difference among the sources. Including PubMed, the mean (SD) score was 27.4 (9.4) points (95% CI, 27.40-27.42 points; P < .001); excluding PubMed, it was 20.0 (5.1) points (95% CI, 20.00-20.02 points; P = .01). Wikipedia scored 10.7 points less than PubMed (mean [SD] difference 0.6 [1.1] points [95% CI, 0.1-1.2 points]; P = .02) but 15 points more than MedicineNet (mean difference 0.9 [1.4] points [95% CI, 0.1-1.6 points]; P = .01). A definitive difference among other freely available sources was not identified.
Our evaluation suggests that Wikipedia provides less information than PubMed but substantially more information than other freely available sources. Moreover, Wikipedia’s containment of expert knowledge seemed good, except for recommendations on which surgical method would be preferred. EyeWiki provided details about proliferative vitreoretinopathy, including its occurrence rate and outcomes. No pictures and no general information (eg, age and sex) were provided, suggesting that EyeWiki presumes a certain level of knowledge and thus might not be appropriate for patients. One decade ago, Wikipedia was shown4 to be inferior to other internet sources regarding other medical issues, in particular osteosarcoma. However, Wikipedia appears to be improving, potentially because of a professional responsibility for correctness.4,5
The inherent limitation of this study is that all questions were composed and assessed by the authors. However, the questions adhered to patients’ frequently asked questions.
This evaluation shows that, among freely available internet sources, Wikipedia provides the most complete and correct answers regarding RRD. Wikipedia might thus be considered a valuable source of information on RRD for patients at this time.
Corresponding Author: Domagoj Ivastinovic, MD, Department of Ophthalmology, Medical University Graz, Auenbruggerplatz 4, 8036 Graz, Austria (domagoj.ivastinovic@medunigraz.at).
Published Online: October 4, 2018. doi:10.1001/jamaophthalmol.2018.4682
Author Contributions: Drs Ivastinovic and Wackernagel had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Ivastinovic, Wackernagel.
Acquisition, analysis, or interpretation of data: Ivastinovic, Wedrich.
Drafting of the manuscript: Ivastinovic.
Critical revision of the manuscript for important intellectual content: Ivastinovic, Wedrich, Wackernagel.
Statistical analysis: Ivastinovic, Wackernagel.
Administrative, technical, or material support: Ivastinovic.
Supervision: Wedrich.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. No disclosures were reported.