Background: Artificial intelligence (AI) has advanced substantially in recent years, transforming many industries and improving the way people live and work. In scientific research, AI can enhance the quality and efficiency of data analysis and publication. However, AI has also opened up the possibility of generating high-quality fraudulent papers that are difficult to detect, raising important questions about the integrity of scientific research and the trustworthiness of published papers. Objective: The aim of this study was to investigate the capabilities of current AI language models in generating high-quality fraudulent medical articles. We hypothesized that modern AI models can create highly convincing fraudulent papers that can easily deceive readers and even experienced researchers. Methods: This proof-of-concept study used ChatGPT (Chat Generative Pre-trained Transformer) powered by the GPT-3 (Generative Pre-trained Transformer 3) language model to generate a fraudulent scientific article related to neurosurgery. GPT-3 is a large language model developed by OpenAI that uses deep learning algorithms to generate human-like text in response to prompts given by users. The model was trained on a massive corpus of text from the internet and is capable of generating high-quality text in a variety of languages and on various topics. The authors posed questions and prompts to the model and refined them iteratively as the model generated the responses. The goal was to create a completely fabricated article including the abstract, introduction, material and methods, discussion, references, charts, etc. Once the article was generated, it was reviewed for accuracy and coherence by experts in the fields of neurosurgery, psychiatry, and statistics and compared to existing similar articles. Results: The study found that the AI language model can create a highly convincing fraudulent article that resembled a genuine scientific paper in terms of word usage, sentence structure, and overall composition. The AI-generated article included standard sections such as introduction, material and methods, results, and discussion, as well a data sheet. It consisted of 1992 words and 17 citations, and the whole process of article creation took approximately 1 hour without any special training of the human user. However, there were some concerns and specific mistakes identified in the generated article, specifically in the references. Conclusions: The study demonstrates the potential of current AI language models to generate completely fabricated scientific articles. Although the papers look sophisticated and seemingly flawless, expert readers may identify semantic inaccuracies and errors upon closer inspection. We highlight the need for increased vigilance and better detection methods to combat the potential misuse of AI in scientific research. At the same time, it is important to recognize the potential benefits of using AI language models in genuine scientific writing and research, such as manuscript preparation and language editing.

Methanol poisonings occur frequently globally, but reports of larger outbreaks where complete clinical and laboratory data are reported remain scarce. The objective of the present study was to report the data from the mass methanol poisoning in the Czech Republic in 2012 addressing the general epidemiology, treatment, and outcomes, and to present a protocol for the use of fomepizole ensuring that the antidote was provided to the most severely poisoned patients in the critical phase. A combined prospective and retrospective case series study of 121 patients with confirmed methanol poisoning. From a total of 121 intoxicated subjects, 20 died outside the hospital and 101 were hospitalized. Among them, 60 survived without, and 20 with visual/CNS sequelae, whereas 21 patients died. The total and hospital mortality rates were 34% and 21%, respectively. Multivariate regression analysis found pH < 7.0 (OR 0.04 (0.01-0.16), p < 0.001), negative serum ethanol (OR 0.08 (0.02-0.37), p < 0.001), and coma on admission (OR 29.4 (10.2-84.6), p < 0.001) to be the only independent parameters predicting death. Continuous hemodialysis was used more often than intermittent hemodialysis, but there was no significant difference in mortality rate between the two [29% (n = 45) vs 17% (n = 30), p = 0.23]. Due to limited stockpiles of fomepizole, ethanol was administered more often; no difference in mortality rate was found between the two [16% (n = 70) vs. 24% (n = 21), p = 0.39]. The effect of folate administration both on the mortality rate and on the probability of visual sequelae was not significant (both p > 0.05). Severity of metabolic acidosis, state of consciousness, and serum ethanol on admission were the only significant parameters associated with mortality. The type of dialysis or antidote did not appear to affect mortality. Recommendations that were issued for hospital triage of fomepizole administration allowed conservation of valuable antidote in this massive poisoning outbreak for those patients most in need.

Context. Visual disturbances due to the toxic effect of formic acid in acute methanol poisonings are generally transient. The subjective symptoms of visual toxicity may resolve within few weeks and fundoscopic signs of acute optic neuropathy subside within 1–2 months; therefore, the prevalence of long-term visual sequelae in the population of survivors of poisonings may be underestimated. Objective. To study the prevalence and character of long-term visual sequelae of acute methanol poisonings based on the data from the Czech mass methanol outbreak in 2012. Patients and methods. A total of 50 patients with confirmed methanol poisoning were included in this longitudinal cross-sectional study, median age: 48 (range, 23–73) years. The following tests were performed: optical coherence tomography or OCT with evaluation of the retinal nerve fibers layer (RNFL), visual evoked potentials (VEP), magnetic resonance imaging (MRI) of brain, complete ocular examination (visual acuity/field, color vision, contrast sensitivity, and fundus), neurological examinations, and biochemical tests. Results. Of 50 patients, 7/50 (14%) were discharged with diagnosed visual sequelae and 6/50 (12%) were discharged with both visual and central nervous system sequelae of poisoning. On the follow-up examination, 20/50 (40%) of the patients had long-term visual sequelae, with 8% of blindness. A total of 38% of the patients had abnormal (28% borderline) findings on RNFL, and 40% had abnormal (18% borderline) VEP. Among the patients discharged without detected visual sequelae, 8/37 (22%) had abnormal RNFL and VEP. Patients with visual sequelae had brain lesions more often (70% vs. 27%, p < 0.01). MRI identified optic nerve lesions in 2/20 cases with abnormal VEP only. The groups with and without visual sequelae differed in serum methanol, ethanol, HCO_3-, formate, pH, anion gap, and base deficit (all p < 0.01). Visual disturbances on admission and coma were more prevalent in the patients with visual sequelae (p < 0.05). Patients with positive serum ethanol on admission were 93% less likely to have optical axonal damage (OR: 0.07 (95% CI: 0.01–0.8); p < 0.05). No association was found between visual sequelae and type of antidote administered, mode of hemodialysis, or folate substitution. Pre-hospital administration of ethanol seemed beneficial: these patients were 90% less likely to have abnormal RNFL findings (OR: 0.10 (95% CI: 0.02–0.52); p < 0.01). Conclusions. The long-term visual sequelae were clearly underestimated on discharge, suggesting a significantly higher amount of patients with long-term sequelae than earlier reported. Thorough examinations before discharge and during follow-up will likely uncover a higher morbidity also after methanol poisonings in general.

Objective The use of nanotechnology is growing enormously and occupational physicians have an increasing interest in evaluating potential hazards and finding biomarkers of effect in workers exposed to nanoparticles. Methods A study was carried out with 36 workers exposed to (nano)TiO₂ pigment and 45 controls. Condensate (EBC) titanium and markers of oxidation of nucleic acids (including 8-hydroxy-2-deoxyguanosine (8-OHdG), 8-hydroxyguanosine (8-OHG), 5-hydroxymethyl uracil (5-OHMeU)) and proteins (such as o-tyrosine (o-Tyr), 3-chlorotyrosine (3-ClTyr) and 3-nitrotyrosine (3-NOTyr)) were analysed from samples of their exhaled breath. Results In the production workshops, the median total mass 2012 and 2013 TiO₂ concentrations were 0.65 and 0.40 mg/m³, respectively. The median numbers of concentrations measured by the scanning mobility particle sizer (SMPS) and aerodynamic particle sizer (APS) were 1.98×10⁴ and 2.32×10⁴ particles/cm³, respectively; and about 80% of those particles were smaller than 100 nm in diameter. In the research workspace, lower aerosol concentrations (0.16 mg/m³ and 1.32×10⁴ particles/cm³) were found. Titanium in the EBC was significantly higher in production workers (p<0.001) than in research workers and unexposed controls. Accordingly, most EBC oxidative stress markers, including in the preshift samples, were higher in production workers than in the two other groups. Multiple regression analysis confirmed an association between the production of TiO₂ and the levels of studied biomarkers. Conclusions The concentration of titanium in EBC may serve as a direct exposure marker in workers producing TiO₂ pigment; the markers of oxidative stress reflect the local biological effect of (nano)TiO₂ in the respiratory tract of the exposed workers.

Mass or cluster methanol poisonings are frequently reported from around the world. The comparative effectiveness of ethanol and fomepizole as antidotes for methanol poisoning is unknown due to the difficulty of performing a randomized controlled trial. During an outbreak of mass poisonings in the Czech Republic in 2012-2014, we compared the effects of antidotes on the frequency of health sequelae and mortality. The study was designed as a cross-sectional case series and quasi-case-control study. Patients with a diagnosis of methanol poisoning on admission to hospitals were identified for the study. Diagnosis was established when (i) a history of recent ingestion of illicit spirits was available and serum methanol was higher than 6.2 mmol/L (20 mg/dL), or (ii) there was a history/clinical suspicion of methanol poisoning, and serum methanol was above the limit of detection with at least two of the following: pH < 7.3, serum bicarbonate < 20 mmol/L, and anion gap or AG ≥ 20 mmol/L. Fomepizole was given as a bolus dose of 15 mg/kg i.v. diluted in isotonic saline, followed by 10 mg/kg every 12 h (every 4 h during hemodialysis); ethanol was administered both intravenously as a 10% solution in 5% glucose, and per os in boluses of 20% solution. Multivariate regression was applied to determine the effect of antidote on outcome. Additionally, for a retrospective quasi-case-control study, a control group of patients treated with ethanol, matched carefully on severity of poisoning and other key parameters, was selected. Data were obtained from 100 hospitalized patients with confirmed poisoning: 25 patients treated with fomepizole were compared with 68 patients receiving ethanol (seven patients did not receive any antidote). More severely acidotic (p < 0.001) and late-presenting (>12 h; p = 0.028) patients received fomepizole more often than ethanol, as reflected in the higher number of fomepizole-treated patients being intubated (p = 0.009). No association was found between the type of antidote and the survival in either the case series (p = 0.205) or the quasi-control groups (p = 0.705) in which patients were very closely matched to minimize confounding by allocation. In the multivariate analysis, positive serum ethanol (odds ratio [OR], 10.8; 95% confidence interval [CI], 2.9-39.9) and arterial blood pH (OR, 3.7; 95% CI, 1.3-10.5) on admission were the only independent variables for the survival. The median intensive care unit length of stay was 6 (range, 2-22) days in the fomepizole group and 4 (range, 1-33) days in the ethanol group (p = 0.131). There were no differences in the use of elimination techniques between the two groups (neither in the full material (n = 100), nor the case-control groups (n = 50)). This study on antidotes for methanol poisoning did not show any evidence of different clinical effectiveness. Although ethanol is generally associated with a higher incidence of complications, this study suggests that both antidotes are similarly effective and that ethanol should not be avoided on grounds of effectiveness.

Markers of oxidative stress and inflammation were analysed in the exhaled breath condensate (EBC) and urine samples of 14 workers (mean age 43 ± 7 years) exposed to iron oxide aerosol for an average of 10 ± 4 years and 14 controls (mean age 39 ± 4 years) by liquid chromatography-electrospray ionization-mass spectrometry/mass spectrometry (LC-ESI-MS/MS) after solid-phase extraction. Aerosol exposure in the workplace was measured by particle size spectrometers, a scanning mobility particle sizer (SMPS) and an aerodynamic particle sizer (APS), and by aerosol concentration monitors, P-TRAK and DustTRAK DRX. Total aerosol concentrations in workplace locations varied greatly in both time and space. The median mass concentration was 0.083 mg m(-3) (IQR 0.063-0.133 mg m(-3)) and the median particle concentration was 66 800 particles cm(-3) (IQR 16,900-86,900 particles cm(-3)). In addition, more than 80% of particles were smaller than 100 nm in diameter. Markers of oxidative stress, malondialdehyde (MDA), 4-hydroxy-trans-hexenale (HHE), 4-hydroxy-trans-nonenale (HNE), 8-isoProstaglandin F2α (8-isoprostane) and aldehydes C6-C12, in addition to markers of nucleic acid oxidation, including 8-hydroxy-2-deoxyguanosine (8-OHdG), 8-hydroxyguanosine (8-OHG), 5-hydroxymethyl uracil (5-OHMeU), and of proteins, such as o-tyrosine (o-Tyr), 3-chlorotyrosine (3-ClTyr), and 3-nitrotyrosine (3-NOTyr) were analysed in EBC and urine by LC-ESI-MS/MS. Almost all markers of lipid, nucleic acid and protein oxidation were elevated in the EBC of workers comparing with control subjects. Elevated markers were MDA, HNE, HHE, C6-C10, 8-isoprostane, 8-OHdG, 8-OHG, 5-OHMeU, 3-ClTyr, 3-NOTyr, o-Tyr (all p < 0.001), and C11 (p < 0.05). Only aldehyde C12 and the pH of samples did not differ between groups. Markers in urine were not elevated. These findings suggest the adverse effects of nano iron oxide aerosol exposure and support the utility of oxidative stress biomarkers in EBC. The analysis of urine oxidative stress biomarkers does not support the presence of systemic oxidative stress in iron oxide pigment production workers.