Accuracy and Readability of Artificial Intelligence Chatbot Responses to Vasectomy-Related Questions: Public Beware.

Purpose Artificial intelligence (AI) has rapidly gained popularity with the growth of ChatGPT (OpenAI, San Francisco, USA) and other large-language model chatbots, and these programs have tremendous potential to impact medicine. One important area of consequence in medicine and public health is that patients may use these programs in search of answers to medical questions. Despite the increased utilization of AI chatbots by the public, there is little research to assess the reliability of ChatGPT and alternative programs when queried for medical information. This study seeks to elucidate the accuracy and readability of AI chatbots in answering patient questions regarding urology. As vasectomy is one of the most common urologic procedures, this study investigates AI-generated responses to frequently asked vasectomy-related questions. For this study, five popular and free-to-access AI platforms were utilized to undertake this investigation. Methods Fifteen vasectomy-related questions were individually queried to five AI chatbots from November-December 2023: ChatGPT (OpenAI, San Francisco, USA), Bard (Google Inc., Mountainview, USA) Bing (Microsoft, Redmond, USA) Perplexity (Perplexity AI Inc., San Francisco, USA), and Claude (Anthropic, San Francisco, USA). Responses from each platform were graded by two attending urologists, two urology research faculty, and one urological resident physician using a Likert (1-6) scale: (1-completely inaccurate, 6-completely accurate) based on comparison to existing American Urological Association guidelines. Flesch-Kincaid Grade levels (FKGL) and Flesch Reading Ease scores (FRES) (1-100) were calculated for each response. To assess differences in Likert, FRES, and FKGL, Kruskal-Wallis tests were performed using GraphPad Prism V10.1.0 (GraphPad, San Diego, USA) with Alpha set at 0.05. Results Analysis shows that ChatGPT provided the most accurate responses across the five AI chatbots with an average score of 5.04 on the Likert scale. Subsequently, Microsoft Bing (4.91), Anthropic Claude (4.65), Google Bard (4.43), and Perplexity (4.41) followed. All five chatbots were found to score, on average, higher than 4.41 corresponding to a score of at least "somewhat accurate." Google Bard received the highest Flesch Reading Ease score (49.67) and lowest Grade level (10.1) when compared to the other chatbots. Anthropic Claude scored 46.7 on the FRES and 10.55 on the FKGL. Microsoft Bing scored 45.57 on the FRES and 11.56 on the FKGL. Perplexity scored 36.4 on the FRES and 13.29 on the FKGL. ChatGPT had the lowest FRES of 30.4 and highest FKGL of 14.2. Conclusion This study investigates the use of AI in medicine, specifically urology, and it helps to determine whether large-language model chatbots can be reliable sources of freely available medical information. All five AI chatbots on average were able to achieve at least "somewhat accurate" on a 6-point Likert scale. In terms of readability, all five AI chatbots on average had Flesch Reading Ease scores of less than 50 and were higher than a 10th-grade level. In this small-scale study, there were several significant differences identified between the readability scores of each AI chatbot. However, there were no significant differences found among their accuracies. Thus, our study suggests that major AI chatbots may perform similarly in their ability to be correct but differ in their ease of being comprehended by the general public.

Carlson JA ，Cheng RZ ，Lange A ，Nagalakshmi N ，Rabets J ，Shah T ，Sindhwani P ... -  《Cureus》

Assessing the Readability of Patient Education Materials on Cardiac Catheterization From Artificial Intelligence Chatbots: An Observational Cross-Sectional Study.

Artificial intelligence (AI) is a burgeoning new field that has increased in popularity over the past couple of years, coinciding with the public release of large language model (LLM)-driven chatbots. These chatbots, such as ChatGPT, can be engaged directly in conversation, allowing users to ask them questions or issue other commands. Since LLMs are trained on large amounts of text data, they can also answer questions reliably and factually, an ability that has allowed them to serve as a source for medical inquiries. This study seeks to assess the readability of patient education materials on cardiac catheterization across four of the most common chatbots: ChatGPT, Microsoft Copilot, Google Gemini, and Meta AI. A set of 10 questions regarding cardiac catheterization was developed using website-based patient education materials on the topic. We then asked these questions in consecutive order to four of the most common chatbots: ChatGPT, Microsoft Copilot, Google Gemini, and Meta AI. The Flesch Reading Ease Score (FRES) was used to assess the readability score. Readability grade levels were assessed using six tools: Flesch-Kincaid Grade Level (FKGL), Gunning Fog Index (GFI), Coleman-Liau Index (CLI), Simple Measure of Gobbledygook (SMOG) Index, Automated Readability Index (ARI), and FORCAST Grade Level. The mean FRES across all four chatbots was 40.2, while overall mean grade levels for the four chatbots were 11.2, 13.7, 13.7, 13.3, 11.2, and 11.6 across the FKGL, GFI, CLI, SMOG, ARI, and FORCAST indices, respectively. Mean reading grade levels across the six tools were 14.8 for ChatGPT, 12.3 for Microsoft Copilot, 13.1 for Google Gemini, and 9.6 for Meta AI. Further, FRES values for the four chatbots were 31, 35.8, 36.4, and 57.7, respectively. This study shows that AI chatbots are capable of providing answers to medical questions regarding cardiac catheterization. However, the responses across the four chatbots had overall mean reading grade levels at the 11th-13th-grade level, depending on the tool used. This means that the materials were at the high school and even college reading level, which far exceeds the recommended sixth-grade level for patient education materials. Further, there is significant variability in the readability levels provided by different chatbots as, across all six grade-level assessments, Meta AI had the lowest scores and ChatGPT generally had the highest.

Behers BJ ，Vargas IA ，Behers BM ，Rosario MA ，Wojtas CN ，Deevers AC ，Hamad KM ... -  《Cureus》

Performance of Artificial Intelligence Chatbots on Glaucoma Questions Adapted From Patient Brochures.

Introduction With the potential for artificial intelligence (AI) chatbots to serve as the primary source of glaucoma information to patients, it is essential to characterize the information that chatbots provide such that providers can tailor discussions, anticipate patient concerns, and identify misleading information. Therefore, the purpose of this study was to evaluate glaucoma information from AI chatbots, including ChatGPT-4, Bard, and Bing, by analyzing response accuracy, comprehensiveness, readability, word count, and character count in comparison to each other and glaucoma-related American Academy of Ophthalmology (AAO) patient materials. Methods Section headers from AAO glaucoma-related patient education brochures were adapted into question form and asked five times to each AI chatbot (ChatGPT-4, Bard, and Bing). Two sets of responses from each chatbot were used to evaluate the accuracy of AI chatbot responses and AAO brochure information, and the comprehensiveness of AI chatbot responses compared to the AAO brochure information, scored 1-5 by three independent glaucoma-trained ophthalmologists. Readability (assessed with Flesch-Kincaid Grade Level (FKGL), corresponding to the United States school grade levels), word count, and character count were determined for all chatbot responses and AAO brochure sections. Results Accuracy scores for AAO, ChatGPT, Bing, and Bard were 4.84, 4.26, 4.53, and 3.53, respectively. On direct comparison, AAO was more accurate than ChatGPT (p=0.002), and Bard was the least accurate (Bard versus AAO, p<0.001; Bard versus ChatGPT, p<0.002; Bard versus Bing, p=0.001). ChatGPT had the most comprehensive responses (ChatGPT versus Bing, p<0.001; ChatGPT versus Bard p=0.008), with comprehensiveness scores for ChatGPT, Bing, and Bard at 3.32, 2.16, and 2.79, respectively. AAO information and Bard responses were at the most accessible readability levels (AAO versus ChatGPT, AAO versus Bing, Bard versus ChatGPT, Bard versus Bing, all p<0.0001), with readability levels for AAO, ChatGPT, Bing, and Bard at 8.11, 13.01, 11.73, and 7.90, respectively. Bing responses had the lowest word and character count. Conclusion AI chatbot responses varied in accuracy, comprehensiveness, and readability. With accuracy scores and comprehensiveness below that of AAO brochures and elevated readability levels, AI chatbots require improvements to be a more useful supplementary source of glaucoma information for patients. Physicians must be aware of these limitations such that patients are asked about existing knowledge and questions and are then provided with clarifying and comprehensive information.

Yalla GR ，Hyman N ，Hock LE ，Zhang Q ，Shukla AG ，Kolomeyer NN ... -  《Cureus》

Dr. Google to Dr. ChatGPT: assessing the content and quality of artificial intelligence-generated medical information on appendicitis.

Generative artificial intelligence (AI) chatbots have recently been posited as potential sources of online medical information for patients making medical decisions. Existing online patient-oriented medical information has repeatedly been shown to be of variable quality and difficult readability. Therefore, we sought to evaluate the content and quality of AI-generated medical information on acute appendicitis. A modified DISCERN assessment tool, comprising 16 distinct criteria each scored on a 5-point Likert scale (score range 16-80), was used to assess AI-generated content. Readability was determined using the Flesch Reading Ease (FRE) and Flesch-Kincaid Grade Level (FKGL) scores. Four popular chatbots, ChatGPT-3.5 and ChatGPT-4, Bard, and Claude-2, were prompted to generate medical information about appendicitis. Three investigators independently scored the generated texts blinded to the identity of the AI platforms. ChatGPT-3.5, ChatGPT-4, Bard, and Claude-2 had overall mean (SD) quality scores of 60.7 (1.2), 62.0 (1.0), 62.3 (1.2), and 51.3 (2.3), respectively, on a scale of 16-80. Inter-rater reliability was 0.81, 0.75, 0.81, and 0.72, respectively, indicating substantial agreement. Claude-2 demonstrated a significantly lower mean quality score compared to ChatGPT-4 (p = 0.001), ChatGPT-3.5 (p = 0.005), and Bard (p = 0.001). Bard was the only AI platform that listed verifiable sources, while Claude-2 provided fabricated sources. All chatbots except for Claude-2 advised readers to consult a physician if experiencing symptoms. Regarding readability, FKGL and FRE scores of ChatGPT-3.5, ChatGPT-4, Bard, and Claude-2 were 14.6 and 23.8, 11.9 and 33.9, 8.6 and 52.8, 11.0 and 36.6, respectively, indicating difficulty readability at a college reading skill level. AI-generated medical information on appendicitis scored favorably upon quality assessment, but most either fabricated sources or did not provide any altogether. Additionally, overall readability far exceeded recommended levels for the public. Generative AI platforms demonstrate measured potential for patient education and engagement about appendicitis.

Ghanem YK ，Rouhi AD ，Al-Houssan A ，Saleh Z ，Moccia MC ，Joshi H ，Dumon KR ，Hong Y ，Spitz F ，Joshi AR ，Kwiatt M ... -  《-》

Assessment of readability, reliability, and quality of ChatGPT®, BARD®, Gemini®, Copilot®, Perplexity® responses on palliative care.

There is no study that comprehensively evaluates data on the readability and quality of "palliative care" information provided by artificial intelligence (AI) chatbots ChatGPT®, Bard®, Gemini®, Copilot®, Perplexity®. Our study is an observational and cross-sectional original research study. In our study, AI chatbots ChatGPT®, Bard®, Gemini®, Copilot®, and Perplexity® were asked to present the answers of the 100 questions most frequently asked by patients about palliative care. Responses from each 5 AI chatbots were analyzed separately. This study did not involve any human participants. Study results revealed significant differences between the readability assessments of responses from all 5 AI chatbots (P < .05). According to the results of our study, when different readability indexes were evaluated holistically, the readability of AI chatbot responses was evaluated as Bard®, Copilot®, Perplexity®, ChatGPT®, Gemini®, from easy to difficult (P < .05). In our study, the median readability indexes of each of the 5 AI chatbots Bard®, Copilot®, Perplexity®, ChatGPT®, Gemini® responses were compared to the "recommended" 6th grade reading level. According to the results of our study answers of all 5 AI chatbots were compared with the 6th grade reading level, statistically significant differences were observed in the all formulas (P < .001). The answers of all 5 artificial intelligence robots were determined to be at an educational level well above the 6th grade level. The modified DISCERN and Journal of American Medical Association scores was found to be the highest in Perplexity® (P < .001). Gemini® responses were found to have the highest Global Quality Scale score (P < .001). It is emphasized that patient education materials should have a readability level of 6th grade level. Of the 5 AI chatbots whose answers about palliative care were evaluated, Bard®, Copilot®, Perplexity®, ChatGPT®, Gemini®, their current answers were found to be well above the recommended levels in terms of readability of text content. Text content quality assessment scores are also low. Both the quality and readability of texts should be brought to appropriate recommended limits.

Hancı V ，Ergün B ，Gül Ş ，Uzun Ö ，Erdemir İ ，Hancı FB ... -  《-》