Quality Of Cancer Information Research Articles

Using ChatGPT to evaluate cancer myths and misconceptions: artificial intelligence and cancer information

Data about the quality of cancer information that chatbots and other artificial intelligence systems provide are limited. Here, we evaluate the accuracy of cancer information on ChatGPT compared with the National Cancer Institute's (NCI's) answers by using the questions on the "Common Cancer Myths and Misconceptions" web page. The NCI's answers and ChatGPT answers to each question were blinded, and then evaluated for accuracy (accurate: yes vs no). Ratings were evaluated independently for each question, and then compared between the blinded NCI and ChatGPT answers. Additionally, word count and Flesch-Kincaid readability grade level for each individual response were evaluated. Following expert review, the percentage of overall agreement for accuracy was 100% for NCI answers and 96.9% for ChatGPT outputs for questions 1 through 13 (ĸ = ‒0.03, standard error = 0.08). There were few noticeable differences in the number of words or the readability of the answers from NCI or ChatGPT. Overall, the results suggest that ChatGPT provides accurate information about common cancer myths and misconceptions.

Facilitating patient communication through understanding their social media use: A comparison by age groups.

71 Background: Social media and internet is increasingly used by patients for cancer education, which can affect provider-patient communication. Usage habits of the adolescent-young adult (AYA; aged < 40 years), adult (age 40- < 65 years), and geriatric cancer populations (age 65+ years) are likely different. Methods: Using age-specific sampling, cancer patients across all disease sites cross-sectionally were asked to complete a survey of demographics, health status, and social media/online resource use for cancer education. Clinical information was abstracted. Results: Of 429 approached, 320 participated (126 AYA, 128 adults, 66 elderly). Males comprised 44%; 72% had post-secondary education; 31% had household incomes of > $100,000. Elderly patients were most likely to refuse participation (33% of elderly approached vs 16% AYA; p < 0.001), with the most common reason being "I do not use internet resources/don't plan on using them"(96% of all elderly refusals with available data). Among respondents, the proportion who utilized the internet for cancer education was 76%, 76% and 70% in AYA, adults, and elderly, respectively (p > 0.5). The use of social media tools in respondents was 49%, 40%, and 36%, respectively (p = 0.16 across age groups). While 75% of patients felt they could judge the quality of cancer-related information on the internet (no differences by age group, p > 0.5), a significantly lower 43% (p < 0.001) felt similarly confident to judge the quality of social media; AYA patients (49%) were numerically more likely to feel confident than seniors (36%; p = 0.16). Elderly were less likely to want online health record access (p = 0.015), treatment option (p = 0.042) and side effect education (p < 0.001), future care plan (p < 0.001) and wellness programs compared to others (p < 0.001). Conclusions: Although cancer patients used social media frequently, confidence is lacking on the quality of cancer information obtained (across all age groups), while elderly perceive fewer benefits of using online/social media related to their cancer. Guidelines for patients on how to assess quality and appropriately use social media could help facilitate patient-provider communication.

A qualitative study of Canadian Aboriginal women’s beliefs about “credible” cancer information on the internet

Criteria for evaluating the quality of cancer information on the Internet include source credibility and accuracy and currency of content. Cultural relevance of cancer resources is often overlooked in assessments of quality of Internet Web sites. Interviews with senior Aboriginal women (n = 25) were conducted in Ontario Canada to determine their beliefs about "high quality" and "credible" cancer prevention resources. Participants did not regard online cancer information from the medical community to be completely credible. They recommended that cancer resources include contact information for traditional healers in addition to local cancer agencies. Cultural appropriateness of cancer information should be assessed. Web resources considered credible according to published criteria may not be as relevant for Aboriginal populations.

A retrospective study of the accuracy of cancer information in Ontario daily newspapers.

Studies suggest that the mass media is a common source of cancer information for the public. However, the quality of cancer information through various print outlets has not been extensively investigated. To assess the accuracy of cancer information in a retrospective sample of Ontario daily newspapers as well as to determine the amount of mobilizing (enabling) information about community resources for cancer. Of 1,027 articles on cancer for 1991, drawn from the 5 highest and 5 lowest circulating newspapers, a random 30% sample (306 articles) was obtained. Only 40 articles had traceable citations (journal name, name of researcher, name of educational institution, or direct quotes from identifiable sources). Of these, 47.5% had misleading titles and 55% included erroneous information or omitted important study results. Only 13 (6.9%) included mobilizing information regarding prevention, diagnosis, treatment or support. Cancer information in newspapers contain frequent inaccuracies and fail to provide mobilizing information. Inaccurate newspaper information about cancer is of concern if the public relies on this channel for at least part of their health knowledge.

Cancer registration in the European community.

A survey of population-based cancer registries in the European Community (EC) is reported: 82 (92%) of 89 registries replied. In 3 of the 12 EC countries (Denmark, Netherlands, UK), general cancer registration applies to the entire population; national registration for childhood neoplasms also applies in 3 countries (FRG, Ireland, UK). About 350,000 new cancers were recorded each year during 1982-1985 in the 32% of the EC population which is covered by general registration. It is estimated that during this period about 988,000 new cancers (excluding non-melanoma skin cancer) occurred each year in the 10 EC countries for which incidence data are available. Most registries report 90-95% completeness of cancer registration, but there is evidence of imprecision in these estimates. Risk factors for cancer are recorded in 43 registries, in particular occupation (38), but recording of occupation is often incomplete. The use of international or compatible systems of coding occupation would facilitate screening of cancer registration data for occupational hazards in the European Community. The quality of cancer information in the Community would be greatly improved by the systematic and widespread use of simple methods to estimate completeness of registration and by allowing registries confidential access to death certificates.