Readability, quality, and reliability of AI-generated ınformation on myofascial pain syndrome: A comparative analysis of ChatGPT, Gemini, and Perplexity.

Readability, quality, and reliability of AI-generated ınformation on myofascial pain syndrome: A comparative analysis of ChatGPT, Gemini, and Perplexity - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice PLoS One . 2026 Jun 4;21(6):e0350402. doi: 10.1371/journal.pone.0350402 Readability, quality, and reliability of AI-generated ınformation on myofascial pain syndrome: A comparative analysis of ChatGPT, Gemini, and Perplexity Yüksel Erkin 1 , Erkan Ozduran 2, * , İlhan Celil Özbek 3 , Volkan Hancı 4 Editor: Gauri Mankekar 5 Author information Article notes Copyright and License information 1 Dokuz Eylul University, Anesthesiology and Reanimation, Pain Medicine, Izmir, Turkey 2 Sivas Numune Hospital, Physical Medicine and Rehabilitation, Pain Medicine, Sivas, Turkey 3 Health Science University, Derince Training and Research Hospital, Physical Medicine and Rehabilitation, Kocaeli, Turkey 4 Dokuz Eylul University, Anesthesiology and Reanimation, Critical Care Medicine, Izmir, Turkey‌‌ 5 LSU Health Shreveport, UNITED STATES OF AMERICA ✉ * E-mail: erkanozduran@gmail.com Competing Interests: The authors have declared that no competing interests exist. Roles Yüksel Erkin : Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing Erkan Ozduran : Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing İlhan Celil Özbek : Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing Volkan Hancı : Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing Gauri Mankekar : Editor Received 2026 Mar 17; Accepted 2026 May 11; Collection date 2026. © 2026 Erkin et al This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. PMC Copyright notice PMCID: PMC13235886  PMID: 42241388 Abstract Patients seeking information about Myofascial Pain Syndrome (MPS), which affects a large segment of the population, are increasingly turning to AI-based chatbots as an alternative to traditional methods. However, the medical accuracy of the content offered by these digital platforms, as well as its suitability to the “grade 6 reading level” standard, which determines its comprehensibility by patients, is a critical point of uncertainty. This study aims to fill this significant gap in the literature by systematically comparing MPS content generated by different AI models using readability indices, reliability, and quality metrics. The 18 most relevant keywords, derived from 25 keywords identified via Google Trends data, were queried using ChatGPT (GPT-5.2), Gemini 3 Flash, and Perplexity (Sonar-4 Large) models. The readability of the generated responses was analyzed using six different indices (FRES, FKGL, GFOG, CLI, ARI, SMOG), while content quality was assessed using GQS and EQIP scales, and reliability using DISCERN and JAMA scales by two independent observers. The responses generated by all AI models examined were found to be statistically significantly more complex than the suggested 6th-grade reading level (p < 0.001). In inter-model comparisons, ChatGPT exhibited the easiest readability [lowest linguistic difficulty] scores, while Perplexity scored significantly higher than both ChatGPT and Gemini in content quality and reliability metrics (JAMA, DISCERN, GQS, EQIP) (p < 0.05). Correlation analysis revealed a strong and positive relationship between quality and reliability parameters. Artificial intelligence platforms have been observed to exhibit high potential in the production of medical information. However, linguistic barriers exceeding sixth-grade reading comprehension, along with reliability limitations of current models, prevent them from replacing professional medical consultation. Perplexity has been found superior in terms of academic quality, while ChatGPT has been found superior in terms of readability. Nevertheless, positioning these systems as complementary “secondary consultation mechanisms” supporting physician oversight in clinical decision-making processes is critically important for patient safety. Introduction Myofascial pain syndrome (MPS) is a common musculoskeletal pain disorder characterized by palpable trigger points and characteristic referred pain patterns in the skeletal muscles and associated facial structures [ 1 ]. Clinically, regional pain is often accompanied by muscle spasm, limited range of motion, and decreased functional capacity; this can significantly affect patients’ daily activities and quality of life [ 2 ]. While it can occur in all age groups, it is reported more frequently in individuals exposed to repetitive microtrauma, prolonged static posture, and occupational strain [ 3 ]. Epidemiological data show that approximately 85% of the general population experiences myofascial pain at some point in their lives; this high rate suggests that MPS is one of the leading causes of chronic musculoskeletal pain [ 4 ]. This prevalence indicates that the syndrome constitutes a significant health burden not only at the individual level but also at the societal level. In recent years, the use of artificial intelligence (AI) technologies in the healthcare field has increased significantly. With this development, AI-based systems have become more frequently used in the production of digital content and patient information related to chronic pain syndromes. In particular, chatbots have become platforms where individuals try to interpret their symptoms, search for information about possible diagnoses, and explore treatment options [ 5 ]. While studies examining the accuracy and reliability of AI-generated content in the fields of orthopedics, rheumatology, and chronic pain are increasing, the extent to which this content is scientifically adequate and suitable for patient safety remains unclear [ 6 – 8 ]. Incorrect or incomplete information presentation can lead to unnecessary anxiety, erroneous self-management attempts, and treatment adherence problems. Therefore, systematic and critical evaluation of AI-derived health information is clinically important. The accuracy of health-related information is as important as its comprehensibility by the target audience. Readability is assessed through criteria that indicate how well a text aligns with the reader’s educational level and language skills. For this purpose, formulas such as Flesch-Kincaid, Gunning Fog, and SMOG are frequently used, and texts are classified according to educational level. The American Medical Association (AMA), the National Institutes of Health ([NIH), and the U.S. Department of Health and Human Services recommend that patient education materials be prepared at a sixth-grade level or lower [ 9 , 10 ]. Given that MPS is a widespread condition affecting a broad segment of the population, it is particularly important that digital health information related to this disease is presented in a way that is understandable to everyone. Considering that individuals from different educational and socioeconomic backgrounds access this content, preparing the information in clear and simple language can be crucial for patients to accurately assess their disease process and treatment options [ 11 , 12 ]. Studies examining health content generated by AI-powered chatbots regarding MPS are quite limited. The current literature indicates a need for comprehensive and comparative evaluations in this area. This study aims to systematically evaluate the quality and readability of texts generated by different AI-powered chatbots regarding MPS. The results are expected to provide a clearer framework for understanding the quality of health information in the digital environment and contribute to the development of patient-centered information approaches in chronic pain management. Materials and methods Ethical Assessment: Since the study’s methodology did not require application to live subjects, it was evaluated outside the scope of studies requiring ethics committee approval, in accordance with similar scientific standards. Data collection and analysis were performed in accordance with the terms and conditions of the data source and institutional guidelines. The study was conducted in strict adherence to the principles of the Declaration of Helsinki, and all data were processed and analyzed in compliance with the relevant data protection regulations. Research design To prevent potential algorithmic biases and the influence of individual user preferences on the results during the data collection process, browser history and cookies were completely cleared before analysis. The research was conducted without the use of a VPN, via Google’s “Incognito Mode” tab, to minimize the impact of historical data. Global interest and geographical distribution data for the keyword “Myofascial Pain Syndrome” were obtained via the Google Trends platform on March 1, 2026. Accordingly, global data from 2004 to the present were scanned using the “most relevant” filter; the 25 most frequently used keywords and their regional concentrations were documented [ 13 ]. The reason for choosing Google as the primary data source in this study is that the platform offers the most comprehensive database with a global market share of 82.24% according to October 2025 data [ 14 ]. AI analysis protocol This research aims to qualitatively analyze the responses provided by three widely accessible AI platforms—ChatGPT, Perplexity, and Gemini—to keywords related to “myofascial pain syndrome” obtained from Google Trends data. The relevant terms were submitted to the systems as English queries [ 9 , 15 ]. To ensure methodological consistency, all models were evaluated using the same query expressions, and the initial response obtained for each keyword was recorded for analysis. To prevent possible systematic biases and the influence of previous queries on the models [contextual deviation], independent and new user sessions were opened for each keyword. The obtained outputs were classified and recorded according to the criteria of content accuracy, reliability, and readability. All raw data and model responses used in the analysis process are accessible through the digital archive determined in accordance with open science principles: ( https://figshare.com/articles/online_resource/myofascial_pain_syndrome/31583908?file=62548162 ). The study was based on the current free version of ChatGPT (GPT-5.2) as of December 9, 2025; free versions of Gemini 3 Flash and Perplexity (Sonar-4 Large) were also included in the comparative analysis to support the universality of the findings [ 9 , 16 ]. In our study, repeated/synonymous words and non-MPS keywords were determined as exclusion criteria, while English keywords related to MPS were determined as inclusion criteria [ 2 , 8 ]. Readability assessment of texts To ensure data consistency and reliability in the readability analysis of responses obtained from AI models, cross-checking was performed across two different independent platforms. Accordingly, all texts were simultaneously evaluated using both readabilityformulas.com (Calculator 1) and online-utility.org (Calculator 2) digital tools. The use of two different calculation interfaces and their average aims to minimize potential technical deviations arising from application variations of the metrics and to confirm the methodological validity of the obtained scores. The evaluation process was based on six different metrics whose reliability has been established in the academic literature: Flesch Readability Score [FRES], Flesch-Kincaid Class Level (FKGL), Gunning Fog Index (GFOG), Coleman-Liau Index (CLI), AutoReadability Index (ARI), and Gobbledygook Simple Criterion (SMOG). The simultaneous use of these formulas has made it possible to comprehensively analyze the suitability of the produced content for the general readership in terms of both word level and sentence structure [ 8 , 9 , 17 , 18 ]. The analysis results were reported using median values and minimum-maximum ranges to represent the overall comprehensibility level of the texts. The data obtained were compared with the “sixth-grade reading level” criterion recommended by the American Medical Association (AMA) and the National Institutes of Health (NIH) for public health information. Accordingly, while 80.0 points was accepted as the threshold value for the FRES metric; the sixth-grade [6.0] level was determined as the target achievement criterion for the FKGL, GFOG, CLI, ARI and SMOG indices [ 9 , 17 ]. Reliability assessment protocol Two main assessment tools were used to measure the quality and academic consistency of the information provided by AI systems. First, the Modified DISCERN Scale, which tests source reliability in five different dimensions, was applied [ 18 ]. Within this scale, parameters such as referencing, timeliness, level of transparency, approach to controversial issues, and objectivity were rated between 0–5 points. An increase in the total score was directly related to the high level of reliability and academic quality of the information provided [ 19 ]. The methodological validity of standardized measurement tools such as DISCERN and JAMA [Journal of the American Medical Association] is also supported by previous studies in the field [ 20 , 21 ]. The JAMA questionnaire was used as the second reliability criterion of the research. In this context, each response was checked against four basic publishing ethics principles: authorship, currency, disclosure, and attribution [ 22 ]. In the evaluation process based on JAMA criteria, a binary scoring system [0 or 1] was adopted for each criterion; the final score obtained [maximum 4 points] revealed the capacity of the relevant content to comply with academic standards. High scores were considered an indicator of the platform’s commitment to ethical publishing principles and scientific reliability norms [ 6 , 23 ]. Content quality assessment Global Quality Scale (GQS) analysis. To measure the quality of digital health content and the clinical benefit it provides to the user, the Global Quality Score (GQS), which has widespread validity in the literature, has been used [ 5 , 24 , 25 ]. This evaluation tool classifies materials on a five-point Likert scale ranging from 1 to 5. In the scoring system, a score of 1 represents low-quality and inadequate content that has no value for patients; a score of 5 represents high-quality material that is highly reliable, comprehensive, and provides direct clinical benefit to the patient. The scoring at the intermediate levels is structured according to the following criteria: 2 points indicate low quality but limited potential for use; 3 points indicate medium quality and limited information; and 4 points indicate high quality and clinically significant benefit [ 5 , 24 ]. The methodological reliability of the GQS instrument and its applicability in various digital health research have also been confirmed by previous scientific studies [ 26 ]. Another quality assessment tool used in the study is the EQIP [Ensuring Quality Information for Patients] scale, which allows for the systematic monitoring of the quality of medical texts from a patient perspective. This methodological instrument analyzes the content under review through a comprehensive set of 20 items; each item is rated as “yes”, “partially” or “no” with 1, 0.5 and 0 points, respectively. The final quality score is calculated as a percentage value obtained by dividing the total score by the number of applicable items and multiplying the result by 100 [ 27 ]. The percentage data resulting from this calculation classifies the qualitative level of the content into four hierarchical categories: 76–100% represents high-quality “well-written” texts; 51–75% represents content of “good quality with minor structural problems”; 26–50% represents materials containing “significant problems”; and 0–25% represents content with “severe problems” in terms of clinical reliability [ 8 ]. All reliability and quality assessments were carried out separately by two independent researchers (Y.E. and E.O.) familiar with the relevant scales; consistency between the evaluators was statistically analyzed. Statistical analysis All data generated during the study were processed and analyzed using IBM SPSS Statistics software (Version 24.0; IBM Corp., Armonk, NY, USA). Categorical variables were expressed as frequencies and percentages, whereas continuous variables were reported as medians with their corresponding ranges (minimum–maximum). Comparisons among the three AI models were performed using the non-parametric Kruskal–Wallis test. When statistically significant differences were detected, pairwise comparisons were conducted using the Mann–Whitney U test with Bonferroni correction to control for multiple comparisons. Additionally, comparisons between the observed readability scores and the recommended sixth-grade readability level (grade 6.0) were performed using the one-sample Wilcoxon signed-rank test. Correlations between quality and reliability metrics were assessed using Spearman’s rank correlation coefficient. For quality and reliability assessments, inter-observer agreement between the two independent evaluators [Y.E. and E.O.], who were also the authors of the study, was evaluated using the Intraclass Correlation Coefficient (ICC). Correlations between quality and reliability metrics were assessed using Spearman’s rank correlation coefficient. For quality and reliability assessments, inter-observer agreement between the two independent evaluators (Y.E. and E.O.), who were also the authors of the study, was evaluated using the Intraclass Correlation Coefficient (ICC) with 95% confidence intervals (CIs). ICC values were interpreted as follows: < 0.50 poor, 0.50–0.75 moderate, 0.75–0.90 good, and >0.90 excellent reliability. For all statistical analyses, a p-value of less than 0.05 was considered statistically significant. For post hoc pairwise comparisons, Bonferroni correction was applied; given that three pairwise comparisons were performed, the adjusted significance threshold was set at p < 0.0167. Reported p-values represent unadjusted values, and statistical significance was interpreted according to the Bonferroni-adjusted threshold. Results Google Trends data was analyzed to determine the trends of users searching for information about MPS on the Google platform. Within the scope of the research design, 18 unique queries were directed to each of three different AI models, resulting in a total of 54 datasets. These generated responses were individually examined in terms of readability, content quality, and reliability parameters, and the performance of each model was calculated based on the median values of these 18 responses. As a result of the analyses, it was determined that the most frequently used basic keywords were “muscle pain”, “icd 10 myofascial pain syndrome”, and “myofascial pain syndrome symptoms”. The basic terms that constitute the data pool of the study are: “muscle pain”, “icd 10 myofascial pain syndrome”, “myofascial pain syndrome symptoms”, “myofascial pain syndrome treatment”, “chronic myofascial pain”, “myofascial release”, “chronic myofascial pain syndrome”, “cervical myofascial pain syndrome”, “myofascial pain dysfunction syndrome”, “what is myofascial pain syndrome”, “myofascial trigger points”, “myofascial pain syndrome trigger points”, “myofascial pain syndrome causes”, “fibromyalgia and myofascial pain syndrome”, “myofascial MPSsage”, “myofascial pain syndrome diagnosis”, “fascia pain” and “myofascial pain syndrome disability”. However; The expressions “fibromyalgia”, “myofascial pain syndrome icd10”, “myofascial pain symptoms”, “myofascial pain treatment”, “chronic pain syndrome” and “neck pain” were excluded from the analysis process. This exclusion criterion is based on methodological grounds such as some terms being semantically repetitive, not directly aligning with the scope of the study, or AI systems producing off-focus responses to these queries. Eighteen keywords were evaluated in the study. Analyses conducted using Google Trends data revealed that the highest search volume for the term “myofascial pain syndrome” originated from the Philippines, Thailand, and the United States, respectively. Global geographic interest level and regional distribution data for this keyword are visualized in Fig 1 . These identified MPS-related keywords were submitted as query input to the Perplexity, ChatGPT, and Google Gemini platforms. Fig 1. Search interest in Myofascial Pain Syndrome across the countries: 2004-present (Based on Google Trends Dat‌‌a, https://www.google.com/trends ). Open in a new tab The readability performance of the response texts obtained from these AI systems was calculated using Calculator 1, Calculator 2, and the arithmetic mean of these two measurements. The obtained readability levels were subjected to a comparative analysis based on the “grade 6 reader level” standard, which is accepted as a reference for medical informational materials by the American Medical Association Foundation and the American Medical Association ( Tables 1–3 ). Table 1. Readability Scores (Calculator-1). CALCULATOR Statistics ChatGPT Gemini Perplexity ChatGPT C6thGRL (P)** † Gemini C6thGRL (P)*- † Perplexity C6thGRL (p)** † Between ChatGPT and Gemini (p)†† Between ChatGPT and Perplexity (p)†† Between Perplexity and Gemini (p)†† FRES 34.61 ± 11.41 (13.00–57.00) 53.28 ± 9.55 (32.00–75.00) 37.89 ± 8.07 (23.00–56.00) <.001 <.001 <.001 <0.001 0.281 0.281 GFOG 13.02 ± 1.66 (8.80–15.30) 10.38 ± 1.47 (7.90–13.80) 11.64 ± 1.29 (8.20–13.80) <.001 <.001 <.001 <0.001 0.005 0.005 FKGL 10.77 ± 1.50 (7.91–13.69) 9.64 ± 1.43 (6.33–12.13) 11.30 ± 1.32 (8.06–13.22) <.001 <.001 <.001 0.037 0.150 0.150 CLI 15.31 ± 2.06 (10.82–18.87) 12.25 ± 1.88 (7.15–15.53) 16.03 ± 1.62 (12.01–18.66) <.001 <.001 <.001 <0.001 0.229 0.229 SMOG 8.96 ± 1.10 (6.88–11.57) 8.85 ± 1.07 (6.97–11.01) 9.57 ± 0.98 (7.01–11.01) <.001 <.001 <.001 0.950 0.066 0.066 ARI 12.25 ± 1.57 (9.16–15.07) 11.32 ± 1.54 (7.11–13.70) 13.55 ± 1.56 (10.24–16.19) <.001 <.001 <.001 0.100 0.015 0.015 Open in a new tab Calculator 1: https://readabilityformulas.com/free-readability-formula-tests.php . Calculator 2: https://www.online-utility.org/english/readability_test_and_improve.jsp . Abbreviations: Flesch reading ease score (FRES), Gunning FOG (GFOG), Flesch-Kincaid Grade Level (FKGL), Simple Measure of Gobbledygook (SMOG), Coleman-Liau Index (CLI), Automated Readability Index (ARI) and Linsear Write (LW). **: C6thGRL (p): Comparison of the responses according to 6th grade reading level (p). †: Wilcoxon test. ††: Chi-Square test for categorical variables and Mann-Whitney U test for continuous variables. p values in bold are statistically significant . Pairwise comparisons were performed using the Mann–Whitney U test with Bonferroni correction. Reported p-values are unadjusted; statistical significance was interpreted using a Bonferroni-adjusted threshold of p < 0.0167. Bold p-values indicate statistically significant results. Table 2. Readability Scores (Calculator-2). CALCULATOR Statistics ChatGPT Gemini Perplexity ChatGPT C6thGRL (P)** † Gemini C6thGRL (P)*- † Perplexity C6thGRL (p)** † Between ChatGPT and Gemini (p)†† Between ChatGPT and Perplexity (p)†† Between Perplexity and Gemini (p)†† FRES 34.82 ± 11.36 (13.66–58.27) 48.88 ± 8.92 (26.55–65.21) 32.56 ± 8.76 (19.78–47.30) <.001 <.001 <.001 0.001 0.527 <0.001 GFOG 13.00 ± 1.92 (8.90–17.01) 12.50 ± 1.52 (10.77–15.72) 13.99 ± 1.89 (11.12–17.17) <.001 <.001 <.001 0.289 0.179 0.019 FKGL 10.87 ± 1.87 (7.66–15.35) 10.66 ± 1.50 (8.06–13.64) 12.79 ± 1.88 (8.90–15.67) <.001 <.001 <.001 0.752 0.005 0.002 CLI 13.54 ± 2.50 (8.18–18.12) 12.10 ± 1.92 (7.89–15.74) 16.31 ± 2.00 (11.84–19.25) <.001 <.001 <.001 0.052 0.001 <0.001 SMOG 11.94 ± 1.46 (9.20–15.87) 12.36 ± 1.21 (10.39–14.42) 13.81 ± 1.40 (11.32–16.19) <.001 <.001 <.001 0.255 0.001 0.005 ARI 9.80 ± 2.37 (5.19–15.10) 10.81 ± 2.02 (6.11–13.49) 13.96 ± 2.64 (8.67–18.14) <.001 <.001 <.001 0.076 <0.001 0.001 Open in a new tab Calculator 1: https://readabilityformulas.com/free-readability-formula-tests.php . Calculator 2: https://www.online-utility.org/english/readability_test_and_improve.jsp . Abbreviations: Flesch reading ease score (FRES), Gunning FOG (GFOG), Flesch-Kincaid Grade Level (FKGL), Simple Measure of Gobbledygook (SMOG), Coleman-Liau Index (CLI), Automated Readability Index (ARI) and Linsear Write (LW). **: C6thGRL (p): Comparison of the responses according to 6th grade reading level (p). †: Wilcoxon test. ††: Chi-Square test for categorical variables and Mann-Whitney U test for continuous variables. p values in bold are statistically significant . Pairwise comparisons were performed using the Mann–Whitney U test with Bonferroni correction. Reported p-values are unadjusted; statistical significance was interpreted using a Bonferroni-adjusted threshold of p < 0.0167. Bold p-values indicate statistically significant results. Table 3. Readability Scores (Average of Calculators). CALCULATOR Statistics ChatGPT Gemini Perplexity ChatGPT C6thGRL (P)** † Gemini C6thGRL (P)*- † Perplexity C6thGRL (p)** † Between ChatGPT and Gemini (p)†† Between ChatGPT and Perplexity (p)†† Between Perplexity and Gemini (p)†† FRES 34.72 ± 11.25 (13.33–57.64) 51.08 ± 9.11 (29.28–70.10) 35.22 ± 8.08 (21.75–50.67) <.001 <.001 <.001 <0.001 0.776 <0.001 GFOG 13.01 ± 1.73 (8.85–15.91) 11.44 ± 1.41 (9.39–14.63) 12.82 ± 1.45 (9.77–15.33) <.001 <.001 <.001 0.005 0.658 0.009 FKGL 10.82 ± 1.64 (8.03–14.52) 10.15 ± 1.39 (7.28–12.89) 12.04 ± 1.46 (9.35–14.34) <.001 <.001 <.001 0.217 0.018 0.001 CLI 14.43 ± 2.22 (9.50–18.50) 12.18 ± 1.88 (7.52–15.64) 16.17 ± 1.77 (11.93–18.96) <.001 <.001 <.001 0.003 0.015 <0.001 SMOG 10.45 ± 1.13 (8.44–13.72) 10.60 ± 1.07 (8.68–12.72) 11.69 ± 1.03 (9.64–13.47) <.001 <.001 <.001 0.752 <0.001 0.006 ARI 11.02 ± 1.87 (7.92–15.09) 11.07 ± 1.67 (7.14–13.56) 13.75 ± 1.91 (9.82–17.13) <.001 <.001 <.001 0.681 <0.001 <0.001 Open in a new tab Calculator 1: https://readabilityformulas.com/free-readability-formula-tests.php . Calculator 2: https://www.online-utility.org/english/readability_test_and_improve.jsp . Abbreviations: Flesch reading ease score (FRES), Gunning FOG (GFOG), Flesch-Kincaid Grade Level (FKGL), Simple Measure of Gobbledygook (SMOG), Coleman-Liau Index (CLI), Automated Readability Index (ARI) and Linsear Write (LW). **: C6thGRL (p): Comparison of the responses according to 6th grade reading level (p). †: Wilcoxon test. ††: Chi-Square test for categorical variables and Mann-Whitney U test for continuous variables. p values in bold are statistically significant . Pairwise comparisons were performed using the Mann–Whitney U test with Bonferroni correction. Reported p-values are unadjusted; statistical significance was interpreted using a Bonferroni-adjusted threshold of p < 0.0167. Bold p-values indicate statistically significant results. Readability analysis findings based on calculator 1 data During the text readability analysis conducted using Calculator 1 for 18 different responses to 18 questions, a statistically significant difference was found between the median readability values of all responses and the sixth-grade readability across all three AI chatbots (p < 0.001). Significant differences were found between ChatGPT and Gemini in the FRES (p < 0.001), GFOG (p < 0.001), and CLI (p < 0.001) formulas; between ChatGPT and Perplexity in the GFOG (p = 0.015) and ARI (p = 0.005) formulas; and between Gemini and Perplexity in the GFOG (p = 0.015) and ARI (p = 0.005) formulas ( Table 1 ). Readability analysis findings based on calculator 2 data During the text readability analysis using calculator 2 for 18 different answers to 18 questions, a statistically significant difference was found between the median readability values of all answers and the sixth-grade readability across all three AI chatbots (p < 0.001). Significant differences were detected between ChatGPT and Gemini in the FRES formula (p = 0.001), between ChatGPT and Perplexity in the FKGL formulas (p = 0.005), CLI formulas (p = 0.001), SMOG formulas (p < 0.001), and ARI formulas (p < 0.001), and between Gemini and Perplexity in the FRES formulas (p < 0.001), FKGL formulas (p = 0.002), CLI formulas (p < 0.001), SMOG formulas (p = 0.005), and ARI formulas (p = 0.001) ( Table 2 ). Text readability analysis based on the average of calculator 1 and 2 During the text readability analysis using the averages of calculators 1 and 2 for 18 different answers to 18 questions, a statistically significant difference was found between the median readability values of all answers and sixth-grade readability across all three AI chatbots (p < 0.001). Significant differences were detected between ChatGPT and Gemini in the FRES (p < 0.001), GFOG (p = 0.005), and CLI (p = 0.003) formulas; between Gemini and Perplexity in the FRES (p < 0.001), GFOG (p = 0.009), FKGL (p = 0.001), CLI (p < 0.001), SMOG (p = 0.006), and ARI (p < 0.001) formulas; and between ChatGPT and Perplexity in the CLI (p = 0.015), SMOG (p < 0.001), and ARI (p < 0.001) formulas. Interpretation of the statistical data revealed that ChatGPT-5 generated responses with the lowest linguistic complexity and highest accessibility; other models were hierarchically ranked as Gemini and Perplexity ( Table 3 ). Content Quality and Source Reliability Analysis Findings A detailed breakdown of the GQS, EQIP, DISCERN, and JAMA scores of the responses generated by AI-based chatbots is presented in Table 4 . In the overall evaluation, which included three different AI models, statistically significant differences were found in all JAMA, DISCERN, GQS, and EQIP parameters. In pairwise comparison analyses, Perplexity achieved significantly higher scores in the GQS (p = 0.001), JAMA (p < 0.001), mDISCERN (p = 0.001), and EQIP (p = 0.001) scales when compared to ChatGPT-5. Similarly, Perplexity also showed statistically better performance in the GQS (p = 0.001), JAMA (p < 0.001), mDISCERN (p < 0.001), and EQIP (p < 0.001) questionnaires according to the Gemini model. In the comparison between Gemini and ChatGPT, a significant difference was found only in the JAMA scoring (p < 0.001). These analyses demonstrated that Perplexity scored significantly higher than both ChatGPT and Gemini in all evaluation tools. Furthermore, Gemini scored higher than ChatGPT specifically in the JAMA criteria. In conclusion, Perplexity offers the highest level of quality and reliability; in the comparison between Gemini and ChatGPT, reliability favored Gemini only based on the JAMA results. Table 4. Comparison Of Quality and Reliability Ratings for the Responses from AI Chatbots. ChatGPT vs Perplexity ChatGPT vs Gemini Perplexity vs Gemini ChatGPT Perplexity P ChatGPT Gemini P Perplexity Gemini P GQS, n (%) 0.001 .736 0.001 1-point 2 (11,1%) 0 (0%) 2 (11,1%) 0 (0%) 0 (0%) 0 (0%) 2-point 3 (16,7%) 0 (0%) 3 (16,7%) 4 (22,2%) 0 (0%) 4 (22,2%) 3-point 7 (38,9%) 2 (11,1%) 7 (38,9%) 8 (44,4%) 2 (11,1%) 8 (44,4%) 4-point 6 (33,3%) 16 (88,9%) 6 (33,3%) 6 (33,3%) 16 (88,9%) 6 (33,3%) 5-point 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) JAMA, n (%) <.001 <.001* <.001 0-point 11 (61,1%) 0 (0%) 11 (61,1%) 3 (16,7%) 0 (0%) 3 (16,7%) 1-point 7 (38,9%) 0 (0%) 7 (38,9%) 9 (50,0%) 0 (0%) 9 (50,0%) 2-point 0 (0%) 8 (44,4%) 0 (0%) 6 (33,3%) 8 (44,4%) 6 (33,3%) 3-point 0 (0%) 10 (55,6%) 0 (0%) 0 (0%) 10 (55,6%) 0 (0%) 4-point 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) m DISCERN, n (%) 0.001 .781 <.001 1-point 1 (5,6%) 0 (0%) 1 (5,6%) 1 (5,6%) 0 (0%) 1 (5,6%) 2-point 11 (61,1%) 2 (11,1%) 11 (61,1%) 12 (66,7%) 2 (11,1%) 12 (66,7%) 3-point 4 (22,2%) 7 (38,9%) 4 (22,2%) 3 (16,7%) 7 (38,9%) 3 (16,7%) 4-point 2 (11,1%) 9 (50,0%) 2 (11,1%) 2 (11,1%) 9 (50,0%) 2 (11,1%) 5-point 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) EQIP, n(%) <.001 0.046 <.001 Serious problems with qood quality 7 (38,9%) 0 (0%) 7 (38,9%) 11 (61,1%) 0 (0%) 11 (61,1%) Good quality with minor problems 11 (61,1%) 18 (100%) 11 (61,1%) 7 (38,9%) 18 (100%) 7 (38,9%) Well written 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) Open in a new tab Abbreviations: mDISCERN, Modified DISCERN; JAMA, Journal of the American Medical Association benchmark criteria; GQS, Global Quality Score; EQIP, Ensuring Quality Information for Patients. Data are presented as median (minimum–maximum). Comparisons among the three AI models were performed using the Kruskal–Wallis test. Pairwise comparisons were conducted using the Mann–Whitney U test with Bonferroni correction when appropriate. Bold p-values indicate statistical significance. Pairwise comparisons were performed using the Mann–Whitney U test with Bonferroni correction. Reported p-values are unadjusted; statistical significance was interpreted using a Bonferroni-adjusted threshold of p < 0.0167. Bold p-values indicate statistically significant results. Inter-rater agreement was high in quality and reliability assessments. Intraclass Correlation Coefficient values were as follows: mDISCERN 0.761, JAMA 0.876, GQS 0.795, and EQIP 0.999, indicating good to excellent reliability according to predefined ICC interpretation thresholds. Correlation analysis The correlation analysis performed revealed a statistically strong and positive relationship between the survey metrics measuring reliability and content quality. This finding indicates that the quality (EQIP, GQS) and reliability (mDISCERN, JAMA) parameters methodologically support each other. In addition, significant correlation coefficients were found between different readability indices; this indicates that text complexity yields consistent results in measurements made with different formulas ( Table 5 ). Detailed data on the strengths and significance levels of these correlations are presented in the relevant table. Table 5. Spearman Correlation Analysis Between Variables (N = 54) r(p). Variables EQIP GQS JAMA mDiscern ARI FRES GFOG FKGL CLI SMOG EQIP 1.000 0.381 (0.004) 0.406 (0.002) 0.254 (0.064) 0.282 (0.039) −0.219 (0.112) 0.081 (0.560) 0.231 (0.093) 0.347 (0.010) 0.212 (0.124) GQS 0.381 (0.004) 1.000 0.371 (0.006) 0.417 (0.002) 0.399 (0.003) −0.327 (0.016) 0.308 (0.023) 0.386 (0.004) 0.398 (0.003) 0.384 (0.004) JAMA 0.406 (0.002) 0.371 (0.006) 1.000 0.414 (0.002) 0.449 (0.001) −0.082 (0.554) 0.028 (0.838) 0.333 (0.014) 0.332 (0.014) 0.431 (0.001) mDiscern 0.254 (0.064) 0.417 (0.002) 0.414 (0.002) 1.000 0.398 (0.003) −0.231 (0.093) 0.260 (0.058) 0.308 (0.023) 0.353 (0.009) 0.378 (0.005) ARI 0.282 (0.039) 0.399 (0.003) 0.449 (0.001) 0.398 (0.003) 1.000 −0.680 (<0.001) 0.675 (<0.001) 0.923 (<0.001) 0.851 (<0.001) 0.872 (<0.001) FRES −0.219 (0.112) −0.327 (0.016) −0.082 (0.554) −0.231 (0.093) −0.680 (<0.001) 1.000 −0.905 (<0.001) −0.852 (<0.001) −0.909 (<0.001) −0.632 (<0.001) GFOG 0.081 (0.560) 0.308 (0.023) 0.028 (0.838) 0.260 (0.058) 0.675 (<0.001) −0.905 (<0.001) 1.000 0.843 (<0.001) 0.786 (<0.001) 0.712 (<0.001) FKGL 0.231 (0.093) 0.386 (0.004) 0.333 (0.014) 0.308 (0.023) 0.923 (<0.001) −0.852 (<0.001) 0.843 (<0.001) 1.000 0.891 (<0.001) 0.893 (<0.001) CLI 0.347 (0.010) 0.398 (0.003) 0.332 (0.014) 0.353 (0.009) 0.851 (<0.001) −0.909 (<0.001) 0.786 (<0.001) 0.891 (<0.001) 1.000 0.751 (<0.001) SMOG 0.212 (0.124) 0.384 (0.004) 0.431 (0.001) 0.378 (0.005) 0.872 (<0.001) −0.632 (<0.001) 0.712 (<0.001) 0.893 (<0.001) 0.751 (<0.001) 1.000 Open in a new tab Flesch reading ease score(FRES), Flesch-Kincaid grade level(FKGL), Simple Measure of Gobbledygook(SMOG), Gunning FOG(GFOG), Coleman-Liau score(CL), automated readability index(ARI) ve Linsear Write(LW). JAMA: Journal of American Medical Association Benchmark Criteria, EQIP: ensuring quality information for patients, GQS (Global Quality Score)r, Spearman correlation coefficient; p, significance level. Spearman rank correlation analysis was used to evaluate the relationships between variables. Bold p-values indicate statistically significant correlations. Discussion This research reveals that the responses of leading AI-based language models (Perplexity, Gemini, and ChatGPT) to questions about MPS have a complexity exceeding the “6th-grade reading level” standard recommended by the U.S. Department of Health and Human Services (HHS) and the National Institutes of Health (NIH). Data from the GQS, EQIP, DISCERN, and JAMA scales demonstrate that ChatGPT exhibits the most optimized performance in terms of readability, followed by Gemini and Perplexity, respectively. While Perplexity achieved the highest scores in content quality and reliability parameters compared to the other models, a comparison between Gemini and ChatGPT showed that reliability was only in favor of Gemini based on JAMA criteria, with no significant difference in quality scores. The MPS content generated by these AI platforms was subjected to a comprehensive evaluation using DISCERN/JAMA tools for reliability and GQS/EQIP tools for quality and cognitive accessibility (readability). Our current study represents a unique and strategic contribution to the contemporary scientific literature as it is one of the pioneering studies that comparatively analyzes the responses of the three most popular major language models (LLMs) regarding MPS. While our study showed ChatGPT exhibited higher readability, it doesn’t guarantee high accuracy. In fact, when AI writes very fluently and clearly, it can create a ‘false sense of authority’ by making misinformation appear professional and trustworthy. This is a significant risk because people tend to believe and follow advice that is easy to read, even if it is wrong. If a text is easy to understand but has low reliability, high readability actually helps spread misinformation more effectively, making it more dangerous. Health literacy (HL) has become a key element in the development of modern health strategies and the enhancement of societal well-being. The World Health Organization (WHO) positions HL as one of the fundamental determinants of health and a priority public health issue on a global scale [ 28 ]. In this context, health literacy is defined as the totality of cognitive knowledge, motivational power, and functional competencies that determine individuals’ capacity to access medical information, understand it correctly, critically evaluate it, and apply it to daily life in the health-related decision-making processes they encounter throughout their lives. E-health literacy, first conceptualized by Norman and Skinner in 2006, is defined as the ability of individuals to search for, analyze, critically evaluate, and apply information obtained from digital media to solve their health problems [ 29 ]. It is known that patients with this competency take a more proactive role in managing their own treatment processes, and this directly correlates with high-quality health outcomes [ 30 , 31 ]. On the other hand, a limited level of e-health literacy; It has been found to be associated with negative clinical outcomes such as increased hospitalization rates, inadequate access to preventive health services, and increased treatment costs [ 32 ]. Current data show that a significant portion of the adult population has low health literacy and has difficulty comprehending medical texts, especially those above the “6th grade level”. The findings from our current study reveal that the responses produced by all three AI systems examined significantly exceed this proposed accessibility threshold and may therefore constitute a linguistic barrier for the general user base. In the current literature, a limited number of studies evaluating online information about MPS have revealed that the readability levels of these texts are above the recommended standard and not of sufficient quality. In a study evaluating 151 websites on Google related to MPS, it was found that 56 of the websites offered high-quality information, with higher quality scores in scientific publications and professional websites. The authors emphasized that the presence of data on difficult readability scores in high-quality websites is concerning in terms of health literacy [ 33 ]. In a study similar to our methodology, where the 17 most popular keywords related to MPS on Google Trends were asked to ChatGPT-4 and the responses were analyzed, it was found that although the chatbot provided accessible content, its current quality and readability did not meet the comprehensive standards required in healthcare. The authors highlighted the structural limitations of ChatGPT-5 and similar AI-based systems in replacing professional medical consultation processes; they academically reaffirmed the central role of patient-physician interaction in accurate diagnosis and effective treatment planning. The findings show that the data provided by digital information sources alone cannot be sufficient in clinical decision-making processes and cannot replace the professional expertise and human trust that are at the heart of medical intervention [ 34 ]. There are a limited number of studies in the literature on the readability, quality, and reliability scores of online materials related to MPS. Looking at studies in this field on other musculoskeletal issues: In a study examining the responses to the 10 most popular Google Trending questions about fibromyalgia syndrome in the last two years using 6 different AI chatbots (ChatGPT-3.5, ChatGPT-4o, Copilot AI, Perplexity AI, Gemini AI, and ChatSonic AI), the authors emphasized that although platforms like Gemini, Copilot, and Perplexity AI offer relatively higher content quality, the detected reference inconsistencies, low readability levels, and risks of misinformation clearly show that these systems need improvement. In this context, it was stated that clinicians should actively guide fibromyalgia patients to approach AI-generated health content cautiously and to critically filter this data. In the same study, the authors found that Gemini, rather than Perplexity, produced the highest quality data, unlike in our study. This can be explained by the fact that FMS, unlike MPS, is a widespread, systemic condition, the artificial intelligence models used in their studies are different, and the results were expressed with a smaller sample size compared to our study [ 35 ]. In a study on patellar tendinopathy, it was found that although ChatGPT-4 produced comprehensive answers to patient-focused queries, this information lacked accuracy and was complex enough to create a linguistic barrier for readers below university level [ 36 ]. In a study using ChatGPT4o, Perplexity AI, Bing CoPilot, and Claude2, which examined the answers to questions about patellar instability and medial patellofemoral ligament reconstruction, it was found that although all chatbots gave satisfactory answers, Google Gemini performed statistically worse in terms of accuracy compared to the other four LLMs [ 37 ]. In a study examining the most popular questions about Ankylosing Spondylitis, it was shown that the easiest-to-read answers were obtained from Gemini, while the highest quality and reliability scores were obtained from Perplexity, similar to our study [ 6 ]. The data from our current study confirm that AI-generated content related to MPS follows a trend parallel to similar studies in the literature in terms of both readability and content quality, and that despite the potential of models such as ChatGPT and Gemini to produce comprehensive responses, these texts generally have a university-level language structure and constitute a serious linguistic barrier for individuals with low e-health literacy [ 35 – 37 ]. Similarly, in our study, it was found that the responses of all three AI models remained above the targeted 6th-grade reading level, indicating insufficient accessibility. On the other hand, when examined in terms of content quality and source reliability, it was observed that Perplexity exhibited the highest scores, consistent with the study in the literature, but no model reached a “perfection” that would fully meet clinical standards [ 6 ]. The identified citation inconsistencies and inaccuracies prove that the information provided by AI is still suffering from a serious “quality problem” and is far from replacing professional medical consultation. Consequently, it is vital for clinicians to actively guide patients to critically examine this data in the face of the complex and controllable content offered by digital information sources. Limitations Some methodological limitations should be considered when interpreting the findings of this research; the study was limited to a narrow timeframe, such as March 2026, and only included English keywords and free versions of models identified via Google Trends. Conducting queries without using a VPN may lead to regional server variations affecting the responses; the lack of control over hyperparameters (temperature, maximum token, etc.) of the AI models and the constantly updated stochastic nature of the systems point to temporal variability in the results. Consequently, long-term studies encompassing different languages and a wider range of models are essential to verify the reliability and accessibility of AI-derived medical content. Another limitation that can be highlighted is that although our study included scales used in studies with similar methodologies in the literature, it resulted in high correlation levels among scales that question similar criteria. Using scales that question different criteria in future studies will prevent this situation. Strengths of the Study The most fundamental contribution of this research to the literature is that it is a pioneering study that examines the data of AI-based chatbots on MPS not only in terms of readability but also through multidimensional criteria such as accuracy, consistency, source citation quality, and content reliability. Unlike the existing literature, the evaluation of 3 different popular and widely used AI platforms within the same methodological framework has revealed the current capacities and structural deficiencies of these technologies in health communication in a much more transparent way. This multidimensional and comparative approach provides strategic data that can guide the development of AI-supported health information systems in both clinical practices and future academic projects. Conclusion While AI ecosystems (Perplexity, ChatGPT, Gemini]) examined specifically for complex clinical conditions like MPS demonstrate revolutionary potential in medical knowledge production, our current findings prove that these technologies have not yet fully passed the “accessibility” and “absolute reliability” tests. Our research revealed that the generated content created a linguistic barrier by remaining above the universally accepted 6th-grade reading level; in this process, ChatGPT distinguished itself from its competitors in terms of cognitive accessibility (readability), while Perplexity stood out in terms of academic quality thanks to its source-based validation mechanism. In the final evaluation, AI should be positioned not as the central element of clinical decision-making mechanisms, but as a “secondary advisory mechanism” supporting the process, provided it passes through the physician’s professional filter. It is vital for patient safety and accurate diagnostic strategies that AI remains a complementary instrument that reinforces human-centered expertise with digital data, rather than replacing it. To ensure AI acts as a reliable secondary advisory mechanism, specific safeguards are necessary. We propose mandatory disclaimer labels to distinguish AI from professional advice, a ‘clinician-in-the-loop’ validation process to verify content accuracy before it reaches patients, and transparent citations to evidence-based literature. These measures are essential to mitigate ‘fluency bias,’ ensuring that high readability does not lead patients to trust low-quality or unverified medical information. Supporting information S1 File. mas020326. (SAV) pone.0350402.s001.sav (18KB, sav) Data Availability All relevant data required to replicate the findings of this study are fully available without restriction. The minimal anonymized dataset, including all raw prompt queries and model responses analyzed during the study, has been deposited in the public repository Figshare. It can be freely and directly accessed via the following direct link: https://figshare.com/articles/online_resource/myofascial_pain_syndrome/31583908?file=62548162 . Additionally, the complete dataset (raw data) has been uploaded directly to the journal system as a Supporting information file. Funding Statement The author(s) received no specific funding for this work. References 1. Steen JP, Jaiswal KS, Kumbhare D. Myofascial Pain Syndrome: An Update on Clinical Characteristics, Etiopathogenesis, Diagnosis, and Treatment. Muscle Nerve. 2025;71(5):889–910. doi: 10.1002/mus.28377 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Barbero M, Schneebeli A, Koetsier E, Maino P. Myofascial pain syndrome and trigger points: evaluation and treatment in patients with musculoskeletal pain. Curr Opin Support Palliat Care. 2019;13(3):270–6. doi: 10.1097/SPC.0000000000000445 [ DOI ] [ PubMed ] [ Google Scholar ] 3. Phan V, Shah J, Tandon H, Srbely J, DeStefano S, Kumbhare D, et al. Myofascial Pain Syndrome: A Narrative Review Identifying Inconsistencies in Nomenclature. PM R. 2020;12(9):916–25. doi: 10.1002/pmrj.12290 [ DOI ] [ PubMed ] [ Google Scholar ] 4. Jafri MS. Mechanisms of myofascial pain. International Scholarly Research Notices. 2014;2014:523924. doi: 10.1155/2014/523924 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Ozduran E, Hancı V, Erkin Y, Özbek İC, Abdulkerimov V. Assessing the readability, quality and reliability of responses produced by ChatGPT, Gemini, and Perplexity regarding most frequently asked keywords about low back pain. PeerJ. 2025;13:e18847. doi: 10.7717/peerj.18847 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Kara M, Ozduran E, Kara MM, Özbek İC, Hancı V. Evaluating the readability, quality, and reliability of responses generated by ChatGPT, Gemini, and Perplexity on the most commonly asked questions about Ankylosing spondylitis. PLoS One. 2025;20(6):e0326351. doi: 10.1371/journal.pone.0326351 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Özbek İC. Evaluation of artificial intelligence-supported osteoarthritis information texts: Content quality and readability analysis. Turkiye Klinikleri Journal of Physical Medicine and Rehabilitation Sciences. 2025;28(1). [ Google Scholar ] 8. Ozduran E, Akkoc I, Büyükçoban S, Erkin Y, Hanci V. Readability, reliability and quality of responses generated by ChatGPT, gemini, and perplexity for the most frequently asked questions about pain. Medicine (Baltimore). 2025;104(11):e41780. doi: 10.1097/MD.0000000000041780 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Gül Ş, Erdemir İ, Hanci V, Aydoğmuş E, Erkoç YS. How artificial intelligence can provide information about subdural hematoma: Assessment of readability, reliability, and quality of ChatGPT, BARD, and perplexity responses. Medicine (Baltimore). 2024;103(18):e38009. doi: 10.1097/MD.0000000000038009 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Özbek İC, Hancı V, Özduran E. Digital guidance: quality and readability analysis of artificial intelligence-generated spondyloarthropathy texts. Turk J Osteoporos. 2025;31(1):12–8. doi: 10.4274/tod.galenos.2024.76743 [ DOI ] [ Google Scholar ] 11. Şan AU. YouTube as a Source of Patient Information on Myofascial Pain Syndrome. Aktuelle Rheumatologie. 2021;46(06):577–82. doi: 10.1055/a-1652-1351 [ DOI ] [ Google Scholar ] 12. Bagcier F, Yurdakul OV, Temel MH. Quality and readability of online information on myofascial pain syndrome. J Bodyw Mov Ther. 2021;25:61–6. doi: 10.1016/j.jbmt.2020.11.001 [ DOI ] [ PubMed ] [ Google Scholar ] 13. Pan A, Musheyev D, Bockelman D, Loeb S, Kabarriti AE. Assessment of Artificial Intelligence Chatbot Responses to Top Searched Queries About Cancer. JAMA Oncol. 2023;9(10):1437–40. doi: 10.1001/jamaoncol.2023.2947 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Statista. Accessed: 1 March 2026. Available from: https://www.statista.com/statistics/216573/worldwide-market-share-of-search-engines/ 15. Strzalkowski P, Strzalkowska A, Chhablani J, Pfau K, Errera M-H, Roth M, et al. Evaluation of the accuracy and readability of ChatGPT-4 and Google Gemini in providing information on retinal detachment: a multicenter expert comparative study. Int J Retina Vitreous. 2024;10(1):61. doi: 10.1186/s40942-024-00579-9 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Ömür Arça D, Erdemir İ, Kara F, Shermatov N, Odacioğlu M, İbişoğlu E, et al. Assessing the readability, reliability, and quality of artificial intelligence chatbot responses to the 100 most searched queries about cardiopulmonary resuscitation: An observational study. Medicine (Baltimore). 2024;103(22):e38352. doi: 10.1097/MD.0000000000038352 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Hancı V, Ergün B, Gül Ş, Uzun Ö, Erdemir İ, Hancı FB. Assessment of readability, reliability, and quality of ChatGPT®, BARD®, Gemini®, Copilot®, Perplexity® responses on palliative care. Medicine. 2024;103(33):e39305. doi: 10.1097/MD.0000000000039305 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Özduran E, Hanci V. Evaluating the readability, quality and reliability of online information on Behçet’s disease. Reumatismo. 2022;74(2):10.4081/reumatismo.2022.1495. doi: 10.4081/reumatismo.2022.1495 [ DOI ] [ PubMed ] [ Google Scholar ] 19. Erkin Y, Hanci V, Ozduran E. Evaluation of the reliability and quality of YouTube videos as a source of information for transcutaneous electrical nerve stimulation. PeerJ. 2023;11:e15412. doi: 10.7717/peerj.15412 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Erkin Y, Hanci V, Ozduran E. Evaluating the readability, quality and reliability of online patient education materials on transcutaneuous electrical nerve stimulation (TENS). Medicine (Baltimore). 2023;102(16):e33529. doi: 10.1097/MD.0000000000033529 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Silberg WM, Lundberg GD, Musacchio RA. Assessing, controlling, and assuring the quality of medical information on the Internet: Caveant lector et viewor--Let the reader and viewer beware. JAMA. 1997;277(15):1244–5. [ PubMed ] [ Google Scholar ] 22. Charnock D, Shepperd S, Needham G, Gann R. DISCERN: an instrument for judging the quality of written consumer health information on treatment choices. J Epidemiol Community Health. 1999;53(2):105–11. doi: 10.1136/jech.53.2.105 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Özduran E, Hanci V. Youtube as a source of information about stroke rehabilitation during the COVID-19 pandemic. NeuroAsia. 2023;28(4):907–15. doi: 10.54029/2023kif [ DOI ] [ Google Scholar ] 24. Gunduz ME, Matis GK, Ozduran E, Hanci V. Evaluating the Readability, Quality, and Reliability of Online Patient Education Materials on Spinal Cord Stimulation. Turk Neurosurg. 2024;34(4):588–99. doi: 10.5137/1019-5149.JTN.42973-22.3 [ DOI ] [ PubMed ] [ Google Scholar ] 25. Ozduran E, Hanci V. Evaluating the Readability, Quality, and Reliability of Online Information on Sjogren’s Syndrome. Indian Journal of Rheumatology. 2023;18(1):16–25. doi: 10.4103/injr.injr_56_22 [ DOI ] [ Google Scholar ] 26. Bernard A, Langille M, Hughes S, Rose C, Leddin D, Veldhuyzen van Zanten S. A systematic review of patient inflammatory bowel disease information resources on the World Wide Web. Am J Gastroenterol. 2007;102(9):2070–7. doi: 10.1111/j.1572-0241.2007.01325.x [ DOI ] [ PubMed ] [ Google Scholar ] 27. Ladhar S, Koshman SL, Yang F, Turgeon R. Evaluation of Online Written Medication Educational Resources for People Living With Heart Failure. CJC Open. 2022;4(10):858–65. doi: 10.1016/j.cjco.2022.07.004 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Zerilli I, Bulfone G, Capizzello D, Gambera A, Fazzino V, Sudano M, et al. Understanding Health Literacy and eHealth Literacy in Nursing Students: A Cross-Sectional Cluster Analysis. Nurs Rep. 2026;16(2):52. doi: 10.3390/nursrep16020052 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Xie L, Zhang S, Xin M, Zhu M, Lu W, Mo PK-H. Electronic health literacy and health-related outcomes among older adults: A systematic review. Prev Med. 2022;157:106997. doi: 10.1016/j.ypmed.2022.106997 [ DOI ] [ PubMed ] [ Google Scholar ] 30. Neter E, Brainin E. eHealth literacy: extending the digital divide to the realm of health information. J Med Internet Res. 2012;14(1):e19. doi: 10.2196/jmir.1619 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Hayat TZ, Brainin E, Neter E. With Some Help From My Network: Supplementing eHealth Literacy With Social Ties. J Med Internet Res. 2017;19(3):e98. doi: 10.2196/jmir.6472 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Berkman ND, Sheridan SL, Donahue KE, Halpern DJ, Crotty K. Low health literacy and health outcomes: an updated systematic review. Ann Intern Med. 2011;155(2):97–107. doi: 10.7326/0003-4819-155-2-201107190-00005 [ DOI ] [ PubMed ] [ Google Scholar ] 33. Bagcier F, Yurdakul OV, Temel MH. Quality and readability of online information on myofascial pain syndrome. J Bodyw Mov Ther. 2021;25:61–6. doi: 10.1016/j.jbmt.2020.11.001 [ DOI ] [ PubMed ] [ Google Scholar ] 34. Ateş Demiroglu SB, Özpolat Bulut Ö, Bağcier F. AI-powered insights: Analyzing ChatGPT’s responses on myofascial pain syndrome. J Bodyw Mov Ther. 2025;44:558–63. doi: 10.1016/j.jbmt.2025.05.043 [ DOI ] [ PubMed ] [ Google Scholar ] 35. Zure M, Menekşeoğlu AK. Assessment of the artificial intelligence-generated fibromyalgia information: beyond the hype. Arch Rheumatol. 2025;40(3):358–64. doi: 10.5152/ArchRheumatol.2025.11149 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Deng J, Li L, Oosterhof JJ, Malliaras P, Silbernagel KG, Breda SJ, et al. ChatGPT is a comprehensive education tool for patients with patellar tendinopathy, but it currently lacks accuracy and readability. Musculoskelet Sci Pract. 2025;76:103275. doi: 10.1016/j.msksp.2025.103275 [ DOI ] [ PubMed ] [ Google Scholar ] 37. Vivekanantha P, Cohen D, Slawaska-Eng D, Nagai K, Tarchala M, Matache B, et al. Evaluating the performance of five large language models in answering Delphi consensus questions relating to patellar instability and medial patellofemoral ligament reconstruction. BMC Musculoskelet Disord. 2025;26(1):1022. doi: 10.1186/s12891-025-09227-1 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] PLoS One. doi: 10.1371/journal.pone.0350402.r001 Decision Letter 0 Gauri Mankekar Author information Copyright and License information Roles Gauri Mankekar : Academic Editor © 2026 Gauri Mankekar This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. PMC Copyright notice 30 Apr 2026 -->PONE-D-26-12800-->-->Artificial Intelligence in Myofascial Pain Syndrome: A Comparative Analysis of Readability, Quality, and Reliability Among ChatGPT, Gemini, and Perplexity-->-->PLOS One Dear Dr. Ozduran, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. -->-->ACADEMIC EDITOR: Please see reviewer 1's comments, revise the manuscript appropriately and resubmit. Please submit your revised manuscript by Jun 14 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org . When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript:--> A letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. --> If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols . As the corresponding author, your ORCID iD is verified in the submission system and will appear in the published article. PLOS supports the use of ORCID, and we encourage all coauthors to register for an ORCID iD and use it as well. Please encourage your coauthors to verify their ORCID iD within the submission system before final acceptance, as unverified ORCID iDs will not appear in the published article. Only the individual author can complete the verification step; PLOS staff cannot verify ORCID iDs on behalf of authors. We look forward to receiving your revised manuscript. Kind regards, Gauri Mankekar, MD,PhD,FACS Academic Editor PLOS One Journal Requirements: When submitting your revision, we need you to address these additional requirements. 1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf 2. Please note that PLOS One has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, we expect all author-generated code to be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse. 3. In your Methods section, please include additional information about your dataset and ensure that you have included a statement specifying whether the collection and analysis method complied with the terms and conditions for the source of the data. 4. We note that your Data Availability Statement is currently as follows: [All relevant data are within the manuscript and its Supporting Information files] Please confirm at this time whether or not your submission contains all raw data required to replicate the results of your study. Authors must share the “minimal data set” for their submission. PLOS defines the minimal data set to consist of the data required to replicate all study findings reported in the article, as well as related metadata and methods ( https://journals.plos.org/plosone/s/data-availability#loc-minimal-data-set-definition ). For example, authors should submit the following data: - The values behind the means, standard deviations and other measures reported; - The values used to build graphs; - The points extracted from images for analysis. Authors do not need to submit their entire data set if only a portion of the data was used in the reported study. If your submission does not contain these data, please either upload them as Supporting Information files or deposit them to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of recommended repositories, please see https://journals.plos.org/plosone/s/recommended-repositories . If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent. If data are owned by a third party, please indicate how others may request data access. 5. When completing the data availability statement of the submission form, you indicated that you will make your data available on acceptance. We strongly recommend all authors decide on a data sharing plan before acceptance, as the process can be lengthy and hold up publication timelines. Please note that, though access restrictions are acceptable now, your entire data will need to be made freely accessible if your manuscript is accepted for publication. This policy applies to all data except where public deposition would breach compliance with the protocol approved by your research ethics board. If you are unable to adhere to our open data policy, please kindly revise your statement to explain your reasoning and we will seek the editor's input on an exemption. Please be assured that, once you have provided your new statement, the assessment of your exemption will not hold up the peer review process. 6. We note that Figure 1 in your submission contain [map/satellite] images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright . We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission: 1. You may seek permission from the original copyright holder of Figure 1 to publish the content specifically under the CC BY 4.0 license. We recommend that you contact the original copyright holder with the Content Permission Form ( http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf ) and the following text: “I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 ( http://creativecommons.org/licenses/by/4.0/ ). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.” Please upload the completed Content Permission Form or other proof of granted permissions as an ""Other"" file with your submission. In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].” 2. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only. The following resources for replacing copyrighted map figures may be helpful: USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/ The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/ Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/ Landsat: http://landsat.visibleearth.nasa.gov/ USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/# Natural Earth (public domain): http://www.naturalearthdata.com/ 7. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise. 8. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice. [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions --> Comments to the Author 1. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. --> Reviewer #1: Yes Reviewer #2: Partly ********** -->2. Has the statistical analysis been performed appropriately and rigorously? --> Reviewer #1: Yes Reviewer #2: Yes ********** -->3. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.--> Reviewer #1: Yes Reviewer #2: Yes ********** -->4. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.--> Reviewer #1: Yes Reviewer #2: Yes ********** -->5. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)--> Reviewer #1: This is a well-structured and timely manuscript. The topic is relevant, the methodology is reasonably rigorous, and the findings have clear clinical implications. Below is a constructive, peer-review-style critique, organized by section, with specific strengths, weaknesses, and suggestions for improvement. Overall Assessment 1. Abstract & Material and Methods :Critical Error in Model Version Reporting • Issue: You state you used "ChatGPT (GPT-5.2)" as of December 9, 2025. GPT-5.2 does not exist as of March 2026. OpenAI has released GPT-4, GPT-4 Turbo, GPT-4o, and o1-series models, but not GPT-5.2. This is a serious factual error that undermines credibility. • Recommendation: Immediately verify and correct the model version. Most likely you used GPT-4 or GPT-4 Turbo. If you used a version released in late 2025, specify the exact model name (e.g., "GPT-4o-2024-11-20"). 2. Results section: Statistical Analysis Issues • Issue A: You report p-values for pairwise comparisons (e.g., "p=0.015" for ChatGPT vs. Perplexity on GFOG) but do not state whether you applied Bonferroni correction (mentioned in Methods) to all pairwise tests. With 3 models, there are 3 comparisons per metric. At α=0.05, Bonferroni-adjusted significance would be p < 0.0167. Your reported p=0.015 would be significant, but p=0.005 would also be significant. However, some comparisons (e.g., Table 2, ChatGPT vs. Perplexity, FKGL p=0.005) are fine. Clarify which p-values are corrected. • Issue B: You state "statistically significant difference was found between the median readability values of all answers and the sixth-grade readability" (p<0.001). What statistical test was used for this? It appears you compared a single median value against a fixed threshold (6.0). Was this a one-sample Wilcoxon signed-rank test? This must be explicitly stated. • Issue C: Table 5 (correlation) reports many coefficients >0.9 (e.g., JAMA vs. mDISCERN = 0.934). Such high correlations suggest potential collinearity or that scales measure nearly identical constructs. This should be discussed as a limitation. 3. Material and Methods :Missing Essential Information • Issue A: You mention "two independent researchers (Y.E. and E.O.)" but these initials are not defined in the manuscript (no author list provided). Are these authors or external reviewers? Clarify. • Issue B: No information on inter-observer agreement calculation beyond ICC values. What was the confidence interval for each ICC? What constitutes "good to excellent"? Provide benchmarks (e.g., <0.5 poor, 0.5-0.75 moderate, >0.75 good). • Issue C: You state "browser history and cookies were completely cleared" and used "Incognito" mode. However, you did not mention using a VPN to control for geographic variation in search results. Since Google Trends data showed varying interest by country (Philippines, Thailand, USA), server location could bias results. This is acknowledged in Limitations, but it should also be stated in Methods that no VPN was used. 4.Abstract and Discussion: Writing and Clarity Problems • Issue A: Abstract is overly dense. The sentence beginning "While AI platforms exhibit high potential..." is 58 words long. Break it into shorter sentences. • Issue B: You introduce an undefined acronym "MPSs Learning Achievement" in the Discussion (first paragraph). This appears to be an error. Delete. • Issue C: The phrase "MPS [MPSsive Activity Syndrome]" appears in the Conclusion. This is incorrect. MPS stands for Myofascial Pain Syndrome throughout the manuscript. Remove the erroneous expansion. • Issue D: In the Introduction, page 2: "limited range of motionan decreased functional capacity" – missing space and likely typo ("and"). • Issue E: Results section: "The research focused on eighteen key keywords" – redundant. "Eighteen keywords" suffices. 5. Discussion section: Interpretation and Discussion Weaknesses • Issue A: You state "ChatGPT exhibits the most optimized performance in terms of readability" but readability is not necessarily a virtue if it comes at the cost of accuracy. ChatGPT had lower quality/reliability scores. This trade-off is mentioned but not adequately explored. Add a paragraph discussing whether easier readability might actually increase risk if the content is less reliable. • Issue B: You cite a study on fibromyalgia (reference 35) but do not mention whether those findings align with or contradict your own. Specifically, that study found Gemini offered higher quality for fibromyalgia, whereas you found Perplexity superior for MPS. Discuss possible reasons (different condition, different version of models). • Issue C: The conclusion that AI should be a "secondary advisory mechanism" is reasonable but vague. What specific safeguards do you propose? (e.g., mandatory disclaimer labels, AI-generated text being reviewed by a human clinician before patient access, integration with EHRs?) 6. Minor Concerns Suggested Title Revision Current title is very long (20+ words). Consider shortening while preserving key elements: Option A: "Readability, Quality, and Reliability of AI-Generated Information on Myofascial Pain Syndrome: A Comparative Analysis of ChatGPT, Gemini, and Perplexity" Option B: "Are AI Chatbots Readable and Reliable for Myofascial Pain Syndrome Information? A Comparison of Three Large Language Models" Reviewer #2: I would term this study as a prototype of studies to come when it concerns AI and health care. No physical patents were studied or involved in this study. The study involves the comparison of various systems with the benchmark of 6th grade literacy. This is highly contextual. The study acknowledges its limitations. E Health Literacy is becoming very common and clarity of AI programs will mature, evolve and become more sophisticated in a short span of time. Ultimately human interfaces will decide which AI programs will be most referred to and consulted. ********** -->6. PLOS authors have the option to publish the peer review history of their article ( what does this mean? ). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .--> Reviewer #1: Yes: Mohammed Elrabie Ahmed Reviewer #2: Yes: Christopher de Souza ********** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation . NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications. --> PLoS One. 2026 Jun 4;21(6):e0350402. doi: 10.1371/journal.pone.0350402.r002 Author response to Decision Letter 1 Article notes Copyright and License information Collection date 2026. PMC Copyright notice 5 May 2026 The response file was uploaded separately to the referees as a word document. Attachment Submitted filename: Response to reviewers.docx pone.0350402.s002.docx (37.2KB, docx) PLoS One. doi: 10.1371/journal.pone.0350402.r003 Decision Letter 1 Gauri Mankekar Author information Copyright and License information Roles Gauri Mankekar : Academic Editor © 2026 Gauri Mankekar This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. PMC Copyright notice 13 May 2026 Readability, Quality, and Reliability of AI-Generated Information on Myofascial Pain Syndrome: A Comparative Analysis of ChatGPT, Gemini, and Perplexity PONE-D-26-12800R1 Dear Dr. Ozduran, We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements. Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication. An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support . If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. Kind regards, Gauri Mankekar, MD,PhD,FACS Academic Editor PLOS One Additional Editor Comments (optional): Reviewers' comments: PLoS One. doi: 10.1371/journal.pone.0350402.r004 Acceptance letter Gauri Mankekar Author information Copyright and License information Roles Gauri Mankekar : Academic Editor © 2026 Gauri Mankekar This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. PMC Copyright notice PONE-D-26-12800R1 PLOS One Dear Dr. Ozduran, I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team. At this stage, our production department will prepare your paper for publication. This includes ensuring the following: * All references, tables, and figures are properly cited * All relevant supporting information is included in the manuscript submission, * There are no issues that prevent the paper from being properly typeset You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps. Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing . If we can help with anything else, please email us at customercare@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Dr. Gauri Mankekar Academic Editor PLOS One Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials S1 File. mas020326. (SAV) pone.0350402.s001.sav (18KB, sav) Attachment Submitted filename: Response to reviewers.docx pone.0350402.s002.docx (37.2KB, docx) Data Availability Statement All relevant data required to replicate the findings of this study are fully available without restriction. The minimal anonymized dataset, including all raw prompt queries and model responses analyzed during the study, has been deposited in the public repository Figshare. It can be freely and directly accessed via the following direct link: https://figshare.com/articles/online_resource/myofascial_pain_syndrome/31583908?file=62548162 . Additionally, the complete dataset (raw data) has been uploaded directly to the journal system as a Supporting information file. Articles from PLOS One are provided here courtesy of PLOS ACTIONS View on publisher site PDF (555.5 KB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Osteopathic Manipulative Treatment and Its Relationship to Autonomic Nervous System Activity#1151311
• PhenoDM1 (Myotonic Dystrophy Type 1 Natural History Study)#1033145
• Human versus artificial intelligence: evaluating ChatGPT's performance in conducting published systematic reviews with meta-analysis in chronic pain research.#162136
• Muscle regeneration failure may lead to clinical features of myopathy in anti-IgLON5 disease#449047
• Peripheral Nerve Injections for CRPS#877900