Search Results (630)

Background/Objectives: Antimicrobial photodynamic therapy (aPDT) is a widely investigated adjunct for managing oral biofilm-associated bacterial and fungal infections, but poor photosensitizer penetration, aggregation, biofilm hypoxia, and protocol variability limit its clinical efficacy. Methods: This review was registered with PROSPERO (CRD420251207522) and conducted according to PRISMA 2020 guidelines. PubMed/MEDLINE, Embase, Scopus, and the Cochrane Library were searched for English-language RCTs. Eligible studies evaluated aPDT combined with physicochemical, delivery-based, light-based, biological, or host-modulating adjuncts in patients with oral biofilm-related bacterial or fungal disease, using an appropriate comparator. Data were synthesized narratively due to substantial clinical and methodological heterogeneity, and risk of bias was assessed using RoB 2. Results: Fourteen RCTs were included, predominantly addressing periodontal and peri-implant conditions, with additional studies involving orthodontic biofilms; no eligible RCT evaluated a fungal oral infection. Surfactant-based enhancement (e.g., sodium dodecyl sulfate) produced formulation- and protocol-dependent effects, with benefits reported for methylene blue and curcumin but not riboflavin. Nanocarrier-based photosensitizer delivery, evaluated in combination with Aloe vera vehicles, improved selected clinical outcomes without consistently superior microbiological effects; however, the independent contribution of the nanocarrier could not be isolated from the potential bioactive effects of the vehicle. An oxygen-releasing adjunct showed no significant between-group clinical benefit. Repeated dual-light irradiation improved bleeding on probing and pocket-related outcomes in some but not all trials. Probiotics, vitamin D, and low-level laser therapy showed inconsistent or preliminary benefits, each supported by limited trial numbers. Risk-of-bias assessments predominantly indicated low risk or some concerns, with one study judged at high overall risk. Conclusions: Enhancement of aPDT should not be regarded as a single therapeutic concept; its clinical benefit is strategy-, formulation-, and context-dependent. Current randomized evidence does not support a universally superior adjunct, highlighting the need for standardized, adequately powered trials with mechanism-oriented comparator designs. Full article

This study investigates Villa Ögel, a rare surviving example of Suadiye’s modernist garden-villa era (1940s–1950s), to illuminate its architectural, historical, and socio-cultural significance amid Istanbul’s rapid urban transformation. Suadiye is a coastal residential district on the Anatolian side of Istanbul, with Villa Ögel located in Şaşkınbakkal, a smaller micro-geographical area within the broader Suadiye neighborhood, representing a specific locality within its wider urban and historical context. As neoliberal policies and post-1999 earthquake “renewal” initiatives erased nearly all traces of Suadiye’s early Republican heritage, dating from 1923 to 1950—a formative period of the Republic of Türkiye marked by state-led modernization, changing social structures, and new approaches to domestic architecture and urban development—Villa Ögel’s precarious survival offers a lens through which to examine the tensions between modernist preservation and profit-driven urbanization. The research is contextualized within Suadiye’s evolution from a 19th-century summer resort to a densely urbanized neighborhood, marked by the replacement of garden villas with high-rises. Combining archival analysis, published sources, informal oral communication, and architectural surveys, the study reconstructs the villa’s history, design, and materiality. Key findings reveal its hybrid construction—featuring Tuzla’s “S&P Tousla” bricks and reinforced concrete—and its role as a socio-cultural artifact: designed by naval engineer Kazım Ögel, the villa hosted non-Muslim summer tenants and reflected gendered social norms, evidenced by the educational disparity between Kazım and his wife, Aliye Sakibe Hanım. The study concludes that Villa Ögel embodies a transitional architectural language, bridging Ottoman-era kiosks and modernist apartments, while its deterioration illustrates limitations in the recognition and protection of modernist heritage in Istanbul. Its hierarchical layout, period-specific materials, and recollections enrich understandings of elite domesticity in mid-20th-century Istanbul. Its documentation as one of the rare surviving garden-villa examples in Şaşkınbakkal offers insights into construction practices, urban memory, and the limited recognition of modernist heritage. The research identifies legal protection, emergency stabilization, and community engagement as potential measures for addressing the building’s current condition and status, while positioning the villa as a surviving material reference within a substantially transformed urban landscape and contributing to architectural historiography on vulnerable modernist sites, urban change, and contemporary heritage debates. Full article

Digitizing oral cultural heritage necessitates computational methods capable of bridging the acoustic, orthographic, and semantic gaps between historical and contemporary language forms. This study proposes a cascaded Automatic Speech Recognition–Neural Machine Translation–Knowledge Graph (ASR–NMT–KG) framework designed to transcribe archaic Thai recitations, translate them into modern Thai, and subsequently convert their semantic content into structured knowledge. The framework was evaluated utilizing the Thai Ancient Medical Speech Corpus (TAMSC), which comprises 14000 original and augmented utterances from ten certified Thai Traditional Medicine practitioners. The domain-adapted ASR model achieved a word error rate (WER) of 8.3%, a character error rate (CER) of 4.8%, and an Archaic Orthography Error Rate (AOER) of 6.2%. The mT5-Large translation model obtained a BLEU score of 51.7, while the Knowledge Graph construction module achieved an F₁-score of 0.89. Furthermore, the WER exhibited a strong negative association with translation quality (Pearson’s r = −0.87, p < 0.001), providing empirical evidence of cross-stage performance dependency within the cascaded pipeline. These results indicate that the cascaded architecture offers an accurate, modular, and interpretable approach for preserving and semantically exploring low-resource historical language materials. Full article

Background/Objectives: Large language models (LLMs) are increasingly used in clinical decision-support, yet their sycophantic behavior in response to user-provided misleading contextual cues remains underexplored in image-based disciplines such as oral and maxillofacial radiology. This study compared sycophantic behavior across five LLMs using image-based and text-based diagnostic scenarios in oral and maxillofacial radiology. Methods: ChatGPT-5.3, Google Gemini 3, Grok 4, Claude Sonnet 4.6, and DeepSeek V3.2 were evaluated on 200 scenarios (100 image-based and 100 text-based) under three prompt conditions: neutral, misleading-cue, and authoritative-cue. The resulting 3000 responses were analyzed using generalized estimating equations (GEE). Results: Diagnostic accuracy ranged from 53.5% to 65.0% under the neutral condition and declined to 28.5–44.0% under the authoritative-cue condition. The authoritative-cue condition had significantly higher odds of sycophancy than the misleading-cue condition (adjusted odds ratio [AOR] = 2.92; 95% confidence interval [CI], 2.52–3.37). Modality had no significant main effect on diagnostic accuracy or sycophancy outcomes; however, the Model × Modality interaction was significant for the overall sycophancy rate (p = 0.014). Under the authoritative-cue condition, self-reported confidence did not differ significantly between correct and incorrect responses (p = 0.678). Conclusions: LLMs are susceptible to misleading contextual cues in oral and maxillofacial radiology diagnostic scenarios, particularly when misinformation is presented as an authoritative cue. The dissociation between self-reported confidence and accuracy indicates that self-reported confidence scores alone are unreliable indicators of diagnostic performance. LLMs should therefore serve as supportive tools requiring expert verification rather than as independent decision-makers in clinical practice. Full article

Large-scale surveys often use cognitive test total scores to describe group differences, but the construct meaning of such scores depends on task content, administration, language experience, and opportunities to learn. The China Family Panel Studies (CFPS) word test is an oral Mandarin character reading task with education-dependent entry and stopping rules. We analyzed 16,689 adults followed from 2010 to 2018 using the full survey strata, primary sampling units, and longitudinal weights, and contrasted baseline-adjusted, gain score, geographic overlap, propensity overlap, floor exclusion, nonlinear, and measurement error sensitivity results. The fully adjusted 2018 contrast relative to Mandarin-speaking homes was −0.023 SD (95% CI [−0.077, 0.030]) for Sinitic-dialect homes and −0.292 SD (95% CI [−0.430, −0.154]) for the pooled minority-language response category. The latter estimate was reduced to −0.183 SD after exact ethnicity code fixed effects and to −0.081 SD (95% CI [−0.291, 0.129]) under overlap weighting; only 49 respondents fell in the 1st–99th percentile propensity score intersection. A covariate-adjusted gain score model yielded a smaller minority-language contrast (−0.180 SD, 95% CI [−0.356, −0.005]). Correcting the baseline under plausible reliability values of 0.85–0.95 produced point estimates from −0.226 to −0.277 SD. Floor exclusions reduced but did not eliminate the negative contrast. These results show that the pooled minority-language coefficient is strongly estimand- and support-dependent. They support a constrained interpretation of the observed CFPS score rather than a direct inference about broad crystallized intelligence or a uniform effect of minority-language use. Full article

Background/Objectives. Electronic cigarettes allow smokers to consume nicotine, a highly addictive substance, without burning tobacco. Since their commercialization in early 2004, electronic cigarettes have become popular in the Asian market and then worldwide. Initially conceived as an alternative to combustible smoking, they have gradually established themselves as a mass consumption product, raising both hopes for the reduction of the risks caused by burned cigarettes and concerns about their use among non-smokers, especially young people. Currently, because of their availability, accessibility to non-smokers, including young people, and their manageability, electronic cigarettes are causing more concern than hope among health professionals. This narrative review aimed to synthesize current evidence on the different chemical components generated by electronic cigarette devices and their effects on oral health. We specifically sought to (i) compile and discuss the available evidence regarding the effects of each chemical component in the e-cigarette aerosols on oral health outcomes, and (ii) feature the risks associated with e-cigarette use and how to ovoid/prevent them. Methods: We reviewed relevant English-language literature on electronic cigarette design, use, chemical generation, and adverse effects on oral health through targeted searches of PubMed, Elsevier, Oxford Academic, BioMed Central, MDPI, Scopus, and Google Scholar from 2008 to 2026. Evidence was narratively synthesized to shed light on different oral health concerns reported following electronic cigarette use. Results: This narrative review highlights several adverse effects of electronic cigarettes on oral health, including functional deregulation of the soft (tongue, gums, salivary glands) tissues, hard tissues (teeth, dental restorative materials), and the oral microbiota. These effects were attributed to the different chemicals generated when electronic cigarettes heat-vape e-liquids. These chemicals damage oral tissues/cells, leading to apoptosis, oxidative stress, and DNA damage that may lead to oral cancer. These effects may reduce oral cell growth and increase tissue inflammation, which may delay wound healing. The reviewed studies identified several limitations, including the need for additional investigations involving electronic cigarette use and long-term follow-up. Conclusions: The available in vitro and in vivo studies clearly show the potential harm of vaping on oral health. This review offers a good overview to electronic cigarette users, oral health professionals, and scientists about the potential oral harm effects of electronic cigarettes. The available studies suggest the need for oral health professionals’ support in providing information and awareness about the harmful effects of electronic cigarettes. Thus, the health system and professional orders should develop information and awareness tools to help oral health professionals in their interventions with e-cigarette users. Full article

Background/Objectives: Dental black stains (BSs) are a form of extrinsic tooth discoloration of bacterial origin, with a prevalence ranging from 2.4% to 20%. As scaling and polishing remain the standard of care, recurrence is frequent. Therefore, this systematic review assessed alternative or adjunctive treatments, such as probiotics, chemical agents, and photodynamic therapy, for managing BSs. The primary objective was to evaluate clinical changes in BS recurrence and severity, while the secondary objective was to analyze microbiological changes. Methods: Conducted following PRISMA 2020 guidelines and registered on PROSPERO (CRD420261357137), this review searched PubMed, Cochrane Library, Scopus, Web of Science, and Dentistry & Oral Sciences Source through EBSCOhost as well as clinical trial registers (ClinicalTrials.gov and ISRCTN) and grey literature, up to 30 April 2026, with no language or date restrictions. We included randomized and non-randomized clinical trials. Two independent reviewers performed study selection and data extraction; a third resolved disagreements. We assessed risk of bias using RoB 2 and ROBINS-I V2, and quality of evidence using GRADE. Results: Of 716 records identified, eight clinical trials (five randomized clinical trials and three non-randomized clinical trials were included), with 250 patients enrolled and 227 assessed at follow-up. Four intervention categories were investigated: probiotics, antimicrobial mouthwashes (0.1%, 0.12% chlorhexidine and 0.1% chlorine dioxide), enzyme-containing toothpaste, and antimicrobial photodynamic therapy (aPDT). Probiotics may reduce BS indices, though the effect was inconsistent over time. In a single small, randomized study, chlorine dioxide showed a short-term reduction in bacterial viability greater than chlorhexidine. aPDT and enzyme-containing toothpaste did not show a significant benefit in the included studies. High risk of bias, heterogeneity, and the small sample prevented quantitative synthesis. Conclusions: Evidence supporting alternative strategies for managing recurrent dental BSs or reducing chromogenic bacterial load remains insufficient. Larger RCTs with standardized outcomes and longer follow-up are needed. Full article

Developmental Language Disorder (DLD) and developmental dyslexia are among the most prevalent neurodevelopmental disorders affecting children’s language, literacy, academic achievement, and social participation. Although traditionally conceptualised as distinct disorders of oral and written language, respectively, increasing evidence suggests substantial overlap in their underlying developmental mechanisms while preserving distinguishable clinical phenotypes. This critical narrative review synthesises evidence from epidemiological, longitudinal, cognitive–linguistic, behavioural genetic, and clinical studies examining the relationship between DLD and developmental dyslexia. Particular attention is given to prevalence, comorbidity, shared risk factors, developmental trajectories, and the implications of categorical versus dimensional models for clinical assessment and intervention. The evidence indicates that DLD and developmental dyslexia share multiple cognitive and linguistic liabilities, including weaknesses in phonological processing, verbal memory, processing efficiency, and broader oral language abilities. These vulnerabilities interact with genetic, environmental, educational, and orthographic influences, producing heterogeneous developmental profiles and frequent comorbidity. At the same time, each disorder retains characteristic clinical features that justify their continued diagnostic distinction. The available evidence supports a hybrid conceptualisation in which DLD and developmental dyslexia are viewed as distinct but partially overlapping neurodevelopmental disorders within a multidimensional language–literacy framework. Building on previous work concerning diagnostic differentiation, comorbidity, and shared developmental mechanisms, this review integrates epidemiological, developmental, cognitive–linguistic, genetic, cross-linguistic, and clinical evidence within a clinically oriented two-dimensional language–literacy framework intended to generate testable hypotheses and inform future research on assessment and intervention. Full article

Given the growing linguistic diversity in U.S. schools, there is a need to better understand how early dual language abilities contribute to later academic achievement. This study investigates English and Spanish oral language proficiencies at kindergarten as predictors of academic performance in third, fourth, fifth, and sixth grades among second-generation Latino children of immigrants (N = 228). Children’s early English and Spanish abilities were directly assessed using the Woodcock Language Proficiency Battery–Revised, and achievement outcomes were obtained from state standardized exams (MCAS) administered in English for English Language Arts (ELA) and Mathematics. A series of multiple regression models controlling for poverty, gender, and maternal education documented that English proficiency at kindergarten predicted ELA and Math performance across all four grade levels. Importantly, early Spanish proficiency uniquely contributed to later achievement above and beyond English, predicting ELA across all four grade levels and Math outcomes in third, fourth, and fifth grades. Overall, models explained moderate proportions of variance in academic performance (R² = 0.20–0.35). Both English and Spanish proficiencies at school entry meaningfully predict long-term academic outcomes, drawing attention to the educational value of supporting the development of children’s dual language abilities. These findings support evidence-based practices and policies that leverage emergent bilingual children’s full linguistic repertoire to promote equitable academic success. Full article

This scoping review maps multilingual education practices implemented in European Early Childhood Education and Care (ECEC) and early-primary settings for children primarily aged 3–7. Mixed-age ECEC samples extending below age 3 were also eligible when the participating group served children within the target range. Following PRISMA-ScR guidance, peer-reviewed studies published between 2020 and 2025 and indexed in Scopus or Web of Science were examined. Fifteen studies met the inclusion criteria and were analysed in relation to their theoretical foundations, pedagogical practices, reported benefits, and challenges. The reported practices drew on diverse frameworks, including translanguaging and plurilingualism, pluralistic and language-awareness approaches, bilingual and CLIL education, emergent literacy, oral-language intervention, and home–ECEC collaboration. Practices included storytelling, shared reading, language comparison, play-based and multimodal activities, songs and rhymes, vocabulary and phonics instruction, translanguaging interaction, and parent-supported literacy activities. Across the included studies, reported benefits concerned engagement, inclusion, oral language, vocabulary, narrative competence, emergent literacy, metalinguistic awareness, attitudes towards linguistic diversity, identity, agency, and home–school continuity. Reported challenges included uneven effects, implementation constraints, fidelity issues, persistent monolingual ideologies, and limited integration of children’s home or heritage languages. The findings offer tentative implications for early-years pedagogy and educator preparation. Further longitudinal, comparative and implementation-focused research would help strengthen the evidence base for sustainable multilingual pedagogies in early education. Full article

Background/Objective: As large language models (LLMs) are increasingly used to obtain oral health information, their ability to provide accurate and comprehensive responses to patient-focused periodontal questions derived from commonly reported misconceptions is increasingly relevant to patient education. This study aimed to evaluate the performance of contemporary LLMs in answering such questions with respect to accuracy, comprehensiveness, readability, and temporal consistency. Methods: Following a comprehensive review of the existing literature, structured patient-focused periodontal questions derived from commonly reported misconceptions were presented to seven different LLMs using a standardized prompt, and their performance was compared. Responses were evaluated for accuracy and comprehensiveness using a Likert scale, while readability was assessed using the Simple Measure of Gobbledygook (SMOG), the Flesch Reading Ease (FRE) score, and the Flesch–Kincaid Grade Level (FKGL). Results: The evaluated LLMs demonstrated uniformly high performance across the benchmark, with no individual response scoring below the scale midpoint (score < 3) for accuracy or comprehensiveness, confirming the absence of clinically unsafe or misleading information. Although statistically significant inter-model differences were observed for accuracy (p < 0.001, Kendall’s W = 0.197) and comprehensiveness (p < 0.001, Kendall’s W = 0.257), the rank-based effect sizes were small, indicating that relative rankings reflect minor numerical variations rather than clinically substantial performance gaps. Regarding readability, absolute metrics across all models revealed a high level of structural and linguistic difficulty that exceeds layperson health literacy requirements. Mean FRE scores across models ranged from 39.5 ± 9.1 to 52.1 ± 12.0 (indicating “difficult” to “fairly difficult” reading ease), mean FKGL scores ranged from 9.1 ± 2.2 to 10.8 ± 1.3 (requiring a 9th to 11th-grade education level), and mean SMOG scores ranged from 11.8 ± 1.6 to 14.0 ± 2.2. Statistically significant differences with large effect sizes were observed among models for FRE (p < 0.001, partial η_p² = 0.308, η_G² = 0.178), FKGL (p < 0.001, partial η_p² = 0.162, η_G² = 0.104) and SMOG (p < 0.001, partial η_p² = 0.217, η_G² = 0.140), with DeepSeek V4 and Gemini 3.1 exhibiting relatively more favorable readability profiles. In the temporal consistency analysis, significant differences in change scores were observed only for SMOG (p = 0.004), whereas no significant temporal changes occurred for accuracy, comprehensiveness, FRE, or FKGL. Conclusions: Contemporary LLMs provide uniformly accurate and clinically safe responses to patient-focused periodontal misconceptions, with scores consistently concentrated at the ceiling. However, absolute readability metrics demonstrate that AI-generated outputs consistently exceed standard public health literacy comprehension levels, presenting a critical barrier for general patient communication that necessitates targeted optimization. Full article

Background: Since the 1990s, the United States has been facing the recurring problem of widespread misuse, addiction and overdoses involving opioid compounds. This phenomenon is known as the opioid crisis or the opioid epidemic. Every day, around 200 Americans die due to opioid use. Nowadays the country is facing the fourth wave, dominated by fentanyl combined with stimulants. The problem of opioid abuse is not confined to the United States alone, but is a global issue. Objective: In this paper, we do not focus on fentanyl, but on another painkiller, oxycodone, as it is often overlooked in discussions of the current crisis, whilst its sale and use are becoming increasingly common. In Poland alone, the number of prescriptions increased by 50% between 2019 and 2023, and nearly 200 Poles have died as a result of an opioid use disorder. This narrative review examines oxycodone at the intersection of pain management, non-medical use, and toxicological investigation, with emphasis on its pharmacology, clinical applications, abuse liability, overdose risk, and detection in biological matrices. Methods: PubMed and Google Scholar were searched using terms related to oxycodone, toxicology, addiction, the opioid crisis, detection methods, biological matrices, analgesic effects, pharmacokinetics, metabolism, overdose, and hospital treatment. English- and Polish-language publications from January 2020 to July 2026 were prioritised, with the search extended to 2000 where necessary. No formal systematic-review protocol was applied. Results: By interacting with opioid receptors, oxycodone relieves moderate to severe pain. It is therefore used as an analgesic for various types of pain—intraoperative, postsurgical or chronic pain, including cancer-related pain. It has good bioavailability following oral administration and accumulates in tissue less than, for example, morphine. A further advantage is the availability of a variety of routes of administration: in addition to oral administration (including through immediate- and extended-release formulations), oxycodone can also be administered, inter alios, intravenously, intranasally, and rectally. On the other hand, oxycodone causes a range of adverse effects—it can lead to respiratory depression and constipation and has high potential of addiction. Its use is associated with a risk of developing tolerance, and discontinuation may trigger withdrawal symptoms. Oxycodone’s high utility for medical purposes makes it a widely used drug in hospital settings. At the same time, there is an increasing number of cases of inappropriate oxycodone use. This results in the need for analytical methods to monitor concentrations in hospitalised patients and to investigate concentrations in cases of suspected poisoning. The toxicological methods used for detection of oxycodone include both screening tests, such as immunoenzymatic assays, and confirmatory tests—i.e., liquid chromatography/mass spectrometry. The process of choosing an appropriate method begins at the stage of selecting the diagnostic sample. Both standard samples, such as blood (including dried blood spots—a novelty) and urine, and alternative samples—oral fluid, hair, nails and vitreous humour—are useful, depending on the patient’s condition, their availability and the time elapsed since the substance was ingested. Conclusions: Oxycodone remains an important analgesic but carries substantial risks of misuse, addiction, and fatal toxicity. Safe prescribing and monitoring, recognition of drug interactions and individual variability, and integrated interpretation of individual case circumstances, appropriate biological samples, and confirmatory mass-spectrometric findings are essential in clinical and forensic practice. Full article

Background and Objectives: Penile prosthesis infection usually ends in device removal, whereas unnecessarily broad or prolonged prophylaxis causes toxicity and promotes resistance. We assessed whether the systemic perioperative antibacterial or antifungal strategy, dose or duration is associated with infection, infection-related explantation or reoperation or antimicrobial toxicity after primary or revision implantation. Materials and Methods: PubMed, Europe PMC, OpenAlex and ClinicalTrials.gov were searched from inception to 29 August 2026 using two complementary strategies that combined penile-prosthesis terms with antimicrobial-prophylaxis terms and, separately, with infection or explantation terms, without language or date limits, supplemented by related-article expansion and reference-list checking. Comparative primary studies were eligible. Risk of bias was appraised with RoB 2 and ROBINS-I, and random-effects models were fitted where independent cohorts reported compatible data. Results: Of 4947 records, 11 studies were included: one small randomised trial and ten non-randomised studies. The pooled crude frequency of infection or infection-related explantation was 2.50% (95% CI 1.78–3.50; I2 = 89%), a descriptive summary across heterogeneous outcome definitions rather than an estimate of the infection rate under any regimen. Postoperative oral antibiotics were not associated with lower risk (pooled odds ratio 1.16, 95% CI 0.89–1.53). A higher or weight-based gentamicin dose did not reduce infection (pooled odds ratio 0.75, 95% CI 0.26–2.18) but was associated with more acute kidney injury in the single study that reported comparative toxicity. Guideline-defined regimens carried higher adjusted risk in two overlapping cohorts and no difference in two others. Antifungal prophylaxis was associated with lower infection in two overlapping analyses but not in a revision-only cohort, an exploratory finding that has not been independently replicated. An explicit dose was reported by only three studies, a numerical duration by four and the interval between administration and incision by none, which limits comparison between regimens, and all non-randomised studies had serious risk of bias. Conclusions: Current evidence identifies no superior antibacterial regimen and does not support routine prolonged prophylaxis. The exploratory antifungal signal and the preliminary renal safety signal with weight-based gentamicin warrant prospective, standardised multicentre evaluation before either informs routine practice. Full article

Background/Objectives: Large language models (LLMs) are increasingly investigated in dentistry and oral and maxillofacial surgery, but comparative evidence regarding specialist-level response quality, clinical content, and potential safety concerns remains limited. This study compared four contemporary LLM interfaces in third molar surgery complication scenarios. Methods: Thirty open-ended specialist-level scenarios were submitted once to ChatGPT 5.4 Thinking, DeepSeek-V3.2, Gemini 3.1 Pro, and Microsoft Copilot (Work and Learning mode) in isolated single-turn sessions, yielding 120 responses. Two oral and maxillofacial surgeons, blinded to interface identity, independently assessed responses using a task-adapted five-point Global Quality Score (GQS). An additional post hoc criterion-referenced evaluation assessed diagnosis, management, escalation/referral, potentially harmful recommendations, safety-critical omissions, and potential clinical risk. Readability and externally observable response latency were also evaluated. Results: Among the sampled outputs, mean GQSs were 4.53 for Gemini, 4.40 for ChatGPT, 4.30 for DeepSeek, and 3.92 for Copilot (p < 0.001). Mean Clinical Content Scores were 4.80/5 for ChatGPT, 4.77 for Gemini, 4.75 for DeepSeek, and 3.90 for Copilot (p < 0.001). All responses were rated as recognizing the target complication. Within the 120 sampled outputs, one Copilot response met the study-specific criterion for a potentially harmful recommendation, and two Copilot responses met the predefined criteria for safety-critical omissions. Within this benchmark, the sampled Copilot outputs had higher study-specific potential clinical-risk scores than the sampled outputs from the other evaluated interfaces (p < 0.001). Readability and externally observable response latency also differed across the sampled interface outputs. Conclusions: In this single-generation benchmark, the sampled outputs showed different profiles in specialist-rated quality, clinical content, potential clinical risk, readability, and externally observable response latency under the specific interface configurations tested. These findings apply to the observed responses from the single synchronized testing session and should not be interpreted as establishing superiority of any underlying LLM, independent diagnostic accuracy, clinical equivalence, clinical safety, effectiveness, or readiness for autonomous decision support. Full article

Background/Objectives: Orally administered herbal preparations have gained attention as complementary approaches in supportive cancer care; however, their clinical efficacy and safety remain uncertain. This systematic review critically evaluated evidence from randomized clinical trials investigating these interventions in patients with cancer. Methods: The review followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and was prospectively registered in PROSPERO. PubMed/MEDLINE, ScienceDirect, Springer Nature Link, CENTRAL, Scopus, LILACS, ClinicalTrials.gov, and WHO ICTRP were searched for English-language studies published from January 1990 through 8 September 2026. Risk of bias was assessed using RoB 2, and outcome-level certainty was evaluated using GRADE. Results: Eight randomized clinical trials met the eligibility criteria and were included in the qualitative synthesis. The studies evaluated Gamiguibi-tang, Rikkunshito, holy basil (Ocimum tenuiflorum), Alcea digitata–Malva sylvestris, Annona muricata, Malva sylvestris–Althaea digitata, Jing Si herbal tea, and Renshen Yangrong Tang across heterogeneous cancer populations. Individual trials reported potentially favorable findings for sleep, fatigue, chemotherapy-induced anorexia, anxiety, salivary function, xerostomia, gastrointestinal symptoms, and quality of life. However, the Jing Si herbal tea trial found no statistically significant between-group differences in fatigue or quality of life, and exploratory ex vivo cytotoxicity findings did not demonstrate clinical antitumor efficacy. Meta-analysis was not feasible because of substantial clinical and methodological heterogeneity. All evaluated outcomes were rated as very-low-certainty evidence, primarily because of risk of bias and serious imprecision. No serious intervention-related adverse events were identified in the published reports, but adverse-event monitoring was inconsistent and no trial was adequately powered to establish safety. Conclusions: The available evidence is insufficient to establish definitive efficacy or safety or to support routine clinical implementation. Adequately powered multicenter trials using botanically authenticated and chemically characterized preparations, clinically meaningful outcomes, interaction assessment, and long-term pharmacovigilance are required. Full article