Understanding Key Audit Matters: How Readability Boosts Analyst Forecast Precision in Indonesia's Early Implementation

PurposeThe authors examine the association between two important audit partner characteristics and the readability of key audit matters (KAMs) disclosed in the audit reports. Specifically, the authors examine how the readability of KAMs is associated with audit partner tenure and workload.Design/methodology/approachThe authors conduct the study in the audit context of Norway and applied the Flesch reading ease scale to measure the readability levels of reported KAMs in the audit reports of companies listed on the Oslo Stock Exchange. Panel data estimation techniques are applied in estimating how partner tenure and workload are associated with the readability of KAMs. In addition, several robustness tests including different measures of KAMs readability and subsample analyses are performed.FindingsThe authors find that audit partner tenure and workload have significant associations with the level of KAMs readability. Specifically, the results show that the reported KAMs become more readable as the audit partner tenure increases but are less readable for partners with more workload. These results appear stronger in subsamples of KAMs typically noted to be more complex and associated with higher risks.Research limitations/implicationsAs KAMs represent the most significant issues in financial statements audit, these results provide important insights to stakeholders on the potential impact of audit partner tenure and workload on KAMs readability. Less readable KAMs could derail stakeholders' desire to bridge the information gap between auditors and users of the audit report. The uniqueness of this study lies in its focus on audit partner characteristics as opposed to the audit firm.Practical implicationsExcessive audit partner workload impairs KAMs readability.Originality/valueAs KAMs represent the most significant issues in financial statements audit, these results provide important insights to stakeholders on the potential impact of audit partner tenure and workload on KAMs readability. Less readable KAMs could derail stakeholders' desire to bridge the information gap between auditors and users of the audit report. The uniqueness of this study lies in its focus on audit partner characteristics as opposed to the audit firm.

The implementation of ISA 701 on Key Audit Matters (KAM) aims to increase the transparency and communicative value of auditor reports and reduce audit expectation gaps. However, empirical findings suggest that KAM disclosure is increasingly dominated by boilerplate textual practices in the form of redaction uniformity and narrative repetition, which have the potential to degrade the informative value of KAM and make it merely a form of formal compliance. This study aims to systematically synthesize the latest empirical literature to assess the impact of KAM boilerplate on the communicative value of auditor reports and to identify the factors that encourage its occurrence. The research method used is a Systematic Literature Review (SLR) of 18 reputable journal articles indexed in Scopus and published between 2021 and 2025. The results of the synthesis show that the majority of studies found that KAM boilerplate practices significantly lower communicative value, weaken market reactions, and risk creating pseudo-transparency. However, some studies have shown that KAM remains informative when auditors face high audit and litigation risk and actively apply professional judgment. The determinants of KAM boilerplate were identified at five main levels: auditor characteristics, the routine and organizational structure of KAP, client characteristics, the regulatory and institutional context, and linguistic factors. This study concludes that the main problem with KAM lies in implementation practices rather than the design of the ISA 701 standard.

Background: Patients frequently consult the internet for health information. Our aim was to perform an Internet-based readability and quality control study using recognised quality scoring systems to assess the patient information available online relating to anaesthesia for total hip and knee replacement surgery.Methods: Online patient information relating to anaesthesia for total hip and knee replacement was identified using Google, Bing and Yahoo with search terms‘hip replacement anaesthetic’,‘knee replacement anaesthetic.’ Readability was assessed using Flesch Reading Ease (FRE), Flesch-Kincaid grade level (FKGL) and Gunning Fog Index (GFI). Quality was assessed using DISCERN instrument, Health On the Net Foundation seal, and Information Standard mark.Results: 32 websites were analysed. 25% were HONcode certified, 15.6% had the Information Standard. Mean FRE was 55.2±12.8. Mean FKGL was 8.6±1.9. Six websites (18.8%) had the recommended 6th-grade readability level. Mean of 10.4±2.6 years of formal education was required to read the websites. Websites with Information Standard were easier to read: FKGL (6.2 vs. 9,P < 0.001), GFI (8.8 vs. 10.7,P = 0.04), FRE score (64.2 vs. 9,P = 0.02). Mean DISCERN score was low: 40.3 ± 13.Conclusions: Overall, most websites were poor quality with reading levels too high for the target audience. Information Standard NHS quality mark was associated with improved readability, however along with HONcode were not found to have a statistically significant correlation with quality. Based on this study, we would encourage healthcare professionals to be judicious in the websites they recommend to patients, and to consider both the readability and quality of the information provided.

How auditors identify and report key audit matters - An organizational routines perspective

Aim: This study aimed to evaluate and compare the readability and quality of information in responses generated by artificial intelligence (AI) models to patients' frequently asked questions about systemic isotretinoin, a medication commonly prescribed in dermatology. Materials and Methods:Thirty-four frequently asked questions from patients using isotretinoin were prepared by a team of dermatology specialists.These questions were posed to three AI-based text-generation tools (ChatGPT, Gemini 2.0, and Copilot), and the responses were analyzed.The resulting texts were compared in terms of readability levels [Flesch Reading Ease score (FRES), Flesch-Kincaid Grade Level (FKGL), Simple Measure of Gobbledygook (SMOG), Gunning Fog index (GFOG), Coleman-Liau index (CLI), and Automated Readability index (ARI)], sentence lengths, and content quality, which was evaluated by dermatologists.Results: None of the AI models achieved the optimal readability threshold (FRES 60).Readability metrics differed significantly among models.Gemini produced responses that were significantly less readable and more complex than those produced by ChatGPT and Copilot across all readability indices, including FRES, FKGL, SMOG, GFOG, CLI, and ARI; post-hoc analyses confirmed differences between Gemini and the other models.Sentence counts also differed significantly, with Gemini generating longer responses than Copilot.In contrast, Likert-based quality scores and response appropriateness were comparable across models, with no statistically significant differences observed. Conclusion:This study demonstrates that AI models produce academic responses that are difficult for those unfamiliar with medical terminology to understand, and can generate outputs with variable readability in health-related content.These findings highlight the need for careful evaluation of AI-based content for use in healthcare.

This study aims to explore the factors influencing the disclosure of Key Audit Matters (KAM) in audit reports, using the Audit Expectation Gap Theory as a theoretical framework. Based on a literature review of 15 articles published between 2020 and 2025, it was found that factors such as company characteristics (size and complexity), auditor characteristics (experience and specialization), and corporate governance significantly influence the quality and quantity of Key Audit Matters disclosure. Larger and more complex companies tend to disclose more Key Audit Matters, while auditors from large firms provide more detailed disclosures. Furthermore, disclosures that are too generic or use overly formal and non-specific language risk exacerbating the expectation gap. Therefore, it is essential for regulators and auditors to continually improve guidelines and evaluate the implementation of Key Audit Matters to meet transparency and accountability goals.

Background Osteoarthritis is the most common cause of disability in people over 65 years old. The readability of of online osteoarthritis information has never been assessed. A 2003 study found the quality of online osteoarthritis information to be poor. This study reviews the quality of online information regarding osteoarthritis in 2018 using three validated scoring systems. Readability is reviewed for the first time, again using three validated tools. Methods The term osteoarthritis was searched across the three most popular English language search engines. The first 25 pages from each search engine were analysed. Duplicate pages, websites featuring paid advertisements, inaccessible pages (behind a pay wall, not available for geographical reasons) and non-text pages were excluded. Readability was measured using Flesch Reading Ease Score (FRES), Flesch-Kincaid Grade Level (FKGL) and Gunning-Fog Index (GFI). Website quality was scored using the the Journal of the American Medical Association (JAMA) benchmark criteria and DISCERN criteria. Presence or absence of HONcode certification, age of content, content producer and author characteristics were noted. Results 37 unique websites were suitable for analysis. Readability varied by assessment tool from 8th to 12th grade level. This compares with the recommended 7- 8th grade level. One (2.7%) website met all four JAMA Criteria. Mean DISCERN quality of information for OA websites was “fair”, comparing favourably with the “poor” grading of a 2003 study. HONCode endorsed websites (43.2%) were of a statistically significantly higher quality. Conclusion Quality of online health information for OA is “fair”. 2.7% of websites met JAMA benchmark criteria for quality. Readability was equal to or more difficult than recommendations. HONcode certification was indicative of higher quality, but not readability. Disclosures K. Murray None. T. Murray None. C. Low None. A. O'Rourke None. D.J. Veale None.

Aim: The aim of this study was to compare three large language models (LLMs); ChatGPT-5 (OpenAI), Gemini 2.5 Pro (Google), and DeepSeek-V3.1 (Hangzhou DeepSeek AI), in terms of their ability to provide information on questions asked about avulsion, a type of traumatic dental injury. Methods: In the study, 25 open-ended questions prepared based on the guidelines of the International Association of Dental Traumatology (IADT) were directed to large language models in their "Think / DeepThink / Reasoning" modes. The responses were evaluated by two experienced experts using a five-point Likert scale for accuracy and a three-point Likert scale for completeness (exhaustiveness). Readability was analyzed using the Flesch Reading Ease Score (FRES), Flesch–Kincaid Grade Level (FKGL), Gunning Fog Index (GFI), Simplified Measure of Gobbledygook (SMOG), and Coleman–Liau Index (CLI). Inter-rater reliability was assessed, and ANOVA/Kruskal-Wallis and post-hoc Dunn-Bonferroni tests were used for group comparisons, based on the data's normality. Results: The results showed that all three models demonstrated high evaluator agreement in terms of both accuracy and completeness. Significant differences were observed between the models in terms of accuracy scores (p=0.016). In subgroup analyses, the ChatGPT-5 model's accuracy score was significantly higher than that of the DeepSeek-V3.1 model (p=0.020). Significant differences were observed between the models in terms of completeness scores (p=0.006). In the subgroup analyses, the completeness scores of the ChatGPT-5 and Gemini 2.5 Pro models were significantly higher than those of the DeepSeek-V3.1 model (p=0.007 and p=0.046, respectively). In terms of readability, the DeepSeek-V3.1 model achieved higher FRES scores and lower FKGL, GFI, and SMOG scores. Overall, ChatGPT-5 demonstrated the highest performance in terms of accuracy and completeness, while DeepSeek-V3.1 achieved the highest readability values. Conclusion: The results show that ChatGPT-5 is the most reliable model in terms of clinical accuracy and content integrity, whereas DeepSeek-V3.1 produces more linguistically understandable responses. However, no model can completely replace human expertise. These systems should be considered tools to support clinical decision-making processes and should be used under physician supervision, especially in pediatric emergencies.

Introduction: This study evaluates the ability of Generative Pre-trained Transformer 4 (GPT-4) to generate clear and accessible explanations for commonly performed gastroenterology (GI) procedures to enhance patient understanding. Methods: Seven common GI procedures were selected, and GPT-4 was prompted to produce patient-friendly educational material for each. Readability was assessed using established metrics, including the Flesch-Kincaid Grade Level, Gunning Fog Index, and Flesch Reading Ease. These metrics were analyzed through the online platform Readable.com, a widely used tool for readability analysis that applies validated algorithms to assess text comprehensibility. For comparison, procedural educational materials were sourced from credible websites of a renowned medical institution and a professional GI organization, both recognized for their evidence-based and patient-centered content. A paired t-test was conducted to statistically compare the readability scores between GPT-4-generated materials and the reference materials, determining whether GPT-4 content met or exceeded the readability standards of these established resources. Results: GPT-4 demonstrated significantly superior readability metrics compared to both the renowned institution and professional GI organization websites. The Flesch-Kincaid Grade Level was significantly lower for GPT-4 instructions, indicating excellent readability at approximately a 5th to 6th-grade level compared to 8th to 9th grade for the renowned institution (5.5 ± 0.47 vs 8.4 ± 1.24, P = 0.004). Similarly, the Gunning Fog Index was notably lower for GPT-4, suggesting clearer text (7.2 ± 0.51 vs 10.8 ± 1.67, P = 0.004 for the renowned institution; 7.3 ± 0.32 vs 9.2 ± 0.367, P = 0.024 for GI organization). The Flesch Reading Ease score was significantly higher for GPT-4, reflecting easier readability (74.4 ± 2.5 vs 55.8 ± 7.03, P = 0.002 for renowned institution; 73.95 ± 1.35 vs 68.2 ± 1.03, P = 0.002 for GI organization). Conclusion: Results indicate that GPT-4-generated GI procedure instructions can be more accessible and easier to understand for the public. Our study highlights the potential of artificial intelligence language tools in serving as an avenue to improve patient procedural understanding and address disparities in health literacy. Further validation through qualitative assessment, including surveys involving both clinical and non-clinical assessors, is needed to ensure the reliability and practical utility of these materials.

Patients are turning to the internet to access educational materials to help them make healthcare decisions, making readability an important factor. This cross-sectional study assessed the readability of online patient education materials for myopia management treatments that have regulatory approval. The top 10 Google search results from May 2024 for freely available online patient information on myopia management modalities and regulatory-approved products in Canada and Australia were analysed for readability. The modalities included orthokeratology, myopia control spectacle lenses, myopia control soft contact lenses and atropine. The products included MiYOSMART® [HOYA®], Stellest® [Essilor®], MyoCare® [ZEISS], MiSight® 1 day [CooperVision®], ACUVUE® Abiliti® 1-Day [Johnson & Johnson], NaturalVue® Multifocal 1 Day [VTI], ACUVUE® Abiliti® Overnight [Johnson & Johnson] and Eikance [Aspen Pharmacare Australia]. These searches gave 120 results. Readability was assessed with Flesch Reading Ease Score (FRES), Flesch-Kincaid Grade Level (FKGL), Gunning Fog Index (GFI), Simple Measure of Gobbledygook (SMOG) Index and Coleman Liau Index (CLI). Additionally, websites were scored on Journal of the American Medical Association (JAMA) benchmark criteria. Statistical analysis was performed with two-tailed tests. Of 120 websites, none met the recommended sixth-grade reading level across all readability indices. There were 13 websites meeting at least one readability index, 10 being product-related. There were seven websites satisfying all four JAMA benchmarks, while the majority met one. There was a weak positive relationship between product search rank and readability (SMOG p = 0.02, GFI p = 0.02) and a weak negative relationship between JAMA benchmarks and readability for both modality (CLI p = 0.045) and product (CLI p = 0.049). Online information about myopia management is generally written above the recommended sixth-grade reading level and does not meet all JAMA benchmarks. Websites that appear as top search results do not necessarily have easier readability. The readability of online patient education materials may influence access to treatment and outcomes.

PurposeThe purpose of this paper is to examine whether corporate social media information helps improve analysts’ forecast accuracy.Design/methodology/approachThis study uses hand-collected information on S&P 500 firms’ official Facebook pages and uses posts and reactions to such posts to measure corporate Facebook information. Multivariate regression models are estimated to test the relationship between analysts’ forecast accuracy and corporate Facebook information.FindingsThe results indicate that analysts forecast accuracy is unresponsive to posts. However, analyst forecast errors are decreasing in reactions to posts. These findings are robust to the inclusion of control variables, firm and time fixed effects, and alternative specifications of forecast errors and different pre-forecast time windows.Research limitations/implicationsThis study has some limitations. It focuses only on the S&P 500 firms, which are large and generally provide better information to the market. The sample period coincides with the early period of the corporate Facebook culture. However, more recent data sets are likely to provide stronger results.Practical implicationsThe findings of this study provide support for “information generation” role of social media and show that reactions to corporate Facebook posts are the new and unique information generated from corporate social media activities, which help information intermediaries in improving their forecasting accuracy.Originality/valueThis study makes an important contribution to the literature by separating the information dissemination role of social media from information generation role and establishes the first evidence on how corporate social media information affects forecast accuracy of financial analysts.

Research Type: Level 5 - Case report, Expert opinion, Personal observation Introduction/Purpose: The internet has become a primary resource for patients seeking information on health-related topics, including the diagnosis and management of musculoskeletal injuries. However, the quality and readability of this information vary widely. Ankle sprains, a common injury, are frequently researched and discussed online. This study aimed to assess the quality, reliability, and readability of online resources related to ankle sprains. By evaluating the trustworthiness of these resources, this study aims to provide insights into how online content on ankle sprains can be improved, ensuring it serves as a reliable and accessible source of information for patients. Methods: Searches were conducted across three search engines (Google, Yahoo, and Bing) using the terms “sprained ankle,” “twisted ankle,” and “torn ankle.” The first 20 websites from each search were selected, excluding those intended for healthcare professionals. Quality assessment was conducted using the DISCERN instrument, JAMA benchmark, and AAOS quality rating guidelines. Readability was evaluated with the Flesch-Kincaid Reading Grade Level (FKGL) and Flesch Reading Ease Score (FRE). Statistical analysis was performed using one-way analysis of variance (ANOVA). Results: A total of 49 websites were reviewed for quality and readability. The quality ratings were poor across all tools: the DISCERN instrument yielded a mean score of 40.08 ± 9.0 (range, 25.67-59.67 out of 80), the JAMA benchmark had a mean score of 1.65 ± 1.19 (range, 0-4 out of 4), and the AAOS quality rating resulted in a mean score of 11.4 ± 4.61 (range, 3-22 out of 24). Readability assessments revealed a mean Flesch Reading Ease (FRE) score of 58.8 ± 11.94 (range, 26.2-80.2) and a mean Flesch-Kincaid Reading Grade Level (FKGL) of 9.49 ± 2.97 (range, 5.3-21.5), indicating moderate readability at a high school to early college level. Conclusion: The online sources on ankle sprains had poor quality and readability, which may contribute to the spread of misinformation. To address this issue, efforts should be made to improve the quality of content through evidence-based information and enhance readability by simplifying language to better match the general public’s understanding. Future initiatives may include collaboration with healthcare organizations to improve online resources and ensure they are accessible and reliable for patients seeking accurate information.

6544 Background: Informed consent forms (ICF) play an integral role in the consent process for clinical trials (CT) by serving as a guide for discussion between the research team and study subjects. Our aims were to compare the quality of ICF among different 1) trial phases, 2) funding sources, 3) oncology subspecialties, 4) disease settings and 5) intervention modalities. Methods: ICF for prospectively conducted CT approved at the Princess Margaret Hospital from 2005–2007 were examined for their content on benefits and risks, alternatives, voluntary participation and confidentiality. Readability was assessed with the Flesch Reading Ease (FRE) score (range 0 to 100 with higher scores denoting greater reading ease) and Reading Grade Level (RGL). Results: Of the 262 evaluable CT, ICF contained an average of 3,982 words, 379 sentences and 10.5 pages. The mean FRE score and RGL were 61.2 and 7.4, respectively. All ICF explicitly stated that the CT was investigational. Only 2 (1%) promised direct personal benefits, 16 (6%) suggested the chance of cure or prolonged survival and 89 (34%) indicated a potential for tumor response. Conversely, 239 (91%) mentioned the risk of serious harms, 217 (83%) admitted that some side effects could be unknown or unpredictable and 126 (48%) reported hospitalization or death as a possibility. Alternatives to participation, right to withdraw from study and data confidentiality were addressed in 242 (92%), 254 (97%) and 260 (99%) ICF, respectively. Multivariate comparisons of benefit and risk descriptions and readability were tabulated (see table). ICF for hematology CT were most likely to highlight major risks (p=0.02), while readability was better in phase I CT and in those which were conducted by medical oncology, sponsored by governmental agencies and involved systemic therapy (all p<0.05). Conclusions: ICF provided a realistic overview of the benefits and risks of CT, although the risk of hospitalization or fatality was underreported. While readability was acceptable in most ICF, these documents remained lengthy. Parameter N Description of Benefits and Risks Readability % of CFs Stating CA May Improve p-value % of CFs Stating Risk of Death p-value Mean FRE p-value Mean RGL p-value Phase I I/II II III 57 25 90 90 26.3 24.0 33.3 42.2 0.12 43.9 44.0 48.9 51.1 0.28 62.9 58.5 61.2 60.8 0.002 7.1 7.8 7.3 7.5 0.001 Funding Academic Gov't Industry Unknown 36 39 158 29 38.9 41.0 29.1 42.9 0.27 27.8 51.3 56.3 25.0 0.16 60.4 64.1 60.8 60.8 0.003 7.5 6.9 7.5 7.4 <0.001 Specialty MedOnc Heme RadOnc Surgery 101 60 52 49 38.6 26.7 42.3 24.5 0.14 51.5 70.0 28.9 34.7 0.02 62.4 61.2 60.2 59.8 0.003 7.2 7.4 7.5 7.6 0.002 Setting Diagnostic Curative Metastatic 28 73 161 32.1 34.3 34.2 0.95 17.9 38.4 57.8 0.23 59.8 60.3 61.8 0.47 7.6 7.5 7.3 0.56 Modality SystemicTx Other 193 69 34.2 33.3 0.20 87.3 12.7 0.06 61.8 59.6 0.02 7.3 7.7 0.001 %: percentage; CFs: consent forms; CA: cancer; FRE: Flesch Reading Ease score; RGL: Reading Grade Level; Gov't: government; MedOnc: medical oncology; Heme: hematology; RadOnc: radiation oncology; SystemicTx: systemic therapy; p-values listed are based on multivariate analyses. No significant financial relationships to disclose.

BackgroundInformation leaflets in research studies should be age-appropriate to be understood, however the formal readability of children’s participant information leaflets (PILs) for research studies has not been assessed.MethodsA single-centre cross-sectional study assessing paediatric PILs. Six readability tests were applied (Gunning Fog Index (GFI), Simple Measure of Gobbledygook (SMOG), Flesch Kincaid Grade Level (FKGL), Coleman–Liau Index (CLI), Automated Readability Index (ARI) and Flesch Reading Ease score (FRE). Results were compared between age groups, and whether the PIL was from either a commercially sponsored or investigator led study.Results191 paediatric PILs were included. Age categories; <10 years (n = 65), ≤12 (n = 73), ≤15 (n = 73) and ≥16 (n = 61); were used for analysis. There were 39 commercial PILs and 226 non-commercial PILs. For the ≤10 and ≤12 age bands, all 6 median readability scores exceeded the target age group (thus hard to read, p < 0.005), and there was no difference in readability scores between these two age bands. Four scores from the readability tests were considered age-appropriate in the ≤15 year category, and all median scores were age-appropriate in the ≥16 years age groups. Readability scores for children’s PILs were significantly higher in commercially sponsored versus non-commercial studies (P < 0.005).ConclusionImprovements are required to make children’s PILs readable for the target audience, particularly in commercially sponsored research studies.ImpactPaediatric participant information leaflets may not be readable in research studies, especially in younger age groups.PILs for children participating in commercially sponsored studies were less readable than non-commercial studies.Research teams writing PILs for a paediatric study need to consider the use of readability tools to ensure that the information they are providing is readable by the target audience.

Background Autoimmune hepatitis (AIH) is a chronic immune-mediated liver disease that requires long-term management, in which effective patient education plays a critical role. With the rapid development of large language models (LLMs), AI-generated health information is increasingly accessed by patients; however, the readability, quality, and educational suitability of LLM-generated AIH-related content remain insufficiently evaluated. Methods Five widely used LLMs—ChatGPT, Doubao, DeepSeek, Wenxin Yiyan, and Tongyi Qianwen—were assessed based on their responses to 20 frequently asked AIH patient education questions covering five thematic categories. Text readability was evaluated using multiple indices, including the Automated Readability Index, Flesch Reading Ease Score, Gunning Fog Index, Flesch–Kincaid Grade Level, Coleman–Liau Index, SMOG, and Linsear Write formula. Information quality and educational suitability were assessed using the Global Quality Score (GQS) and the Chinese version of the Patient Education Materials Assessment Tool (C-PEMAT). Clinical Intent Alignment (CIA) was used to evaluate the coverage of guideline-defined medical key points based on the 2025 EASL Clinical Practice Guidelines. Inter-rater reliability was analyzed using Cohen’s kappa, and comparative and correlation analyses were performed. Results Significant differences were observed among the LLMs in readability, information quality, and educational suitability (all p &amp;lt; 0.05). ChatGPT achieved the highest GQS and C-PEMAT scores, followed by Doubao and DeepSeek, whereas Wenxin Yiyan and Tongyi Qianwen showed lower performance and greater variability. CIA analysis indicated comparable coverage of guideline-defined clinical intent across models. Readability varied significantly across content themes, with texts related to disease mechanisms and diagnostic processes exhibiting higher linguistic complexity. Correlation analysis demonstrated moderate associations between GQS and grade-level readability indices, whereas C-PEMAT and CIA showed weak correlations with traditional readability metrics. Conclusion Substantial variability exists among LLMs in generating AIH patient education materials. Model selection critically influences information quality and educational suitability, whereas content theme primarily affects linguistic complexity. Although most models produced moderate-to-good quality information, relatively high readability levels suggest that further simplification may be needed for general patient populations. A multidimensional evaluation framework integrating readability, quality, educational suitability, and clinical intent alignment is essential for the responsible use of LLMs in AIH patient education.