Background: Large language models (LLMs) have gained prominence since the release of ChatGPT in late 2022. Objective: The aim of this study was to assess the accuracy of citations and references generated by ChatGPT (GPT-3.5) in two distinct academic domains: the natural sciences and humanities. Methods: Two researchers independently prompted ChatGPT to write an introduction section for a manuscript and include citations; they then evaluated the accuracy of the citations and Digital Object Identifiers (DOIs). Results were compared between the two disciplines. Results: Ten topics were included, including 5 in the natural sciences and 5 in the humanities. A total of 102 citations were generated, with 55 in the natural sciences and 47 in the humanities. Among these, 40 citations (72.7%) in the natural sciences and 36 citations (76.6%) in the humanities were confirmed to exist (P=.42). There were significant disparities found in DOI presence in the natural sciences (39/55, 70.9%) and the humanities (18/47, 38.3%), along with significant differences in accuracy between the two disciplines (18/55, 32.7% vs 4/47, 8.5%). DOI hallucination was more prevalent in the humanities (42/55, 89.4%). The Levenshtein distance was significantly higher in the humanities than in the natural sciences, reflecting the lower DOI accuracy. Conclusions: ChatGPT’s performance in generating citations and references varies across disciplines. Differences in DOI standards and disciplinary nuances contribute to performance variations. Researchers should consider the strengths and limitations of artificial intelligence writing tools with respect to citation accuracy. The use of domain-specific models may enhance accuracy.

Purpose: In this study, we aimed to develop a novel liver function and inflammatory markers-based nomogram to predict recurrence-free survival (RFS) for AFP-negative (< 20 ng/mL) HCC patients after curative resection. Patients and Methods: A total of 166 pathologically confirmed AFP-negative HCC patients were included at the Ningbo Medical Center Lihuili Hospital. A LASSO regression analysis was used for data dimensionality reduction and element selection. Univariate and multivariate Cox regression analyses were performed to identify the independent risk factors relevant to RFS. Finally, clinical nomogram prediction model for RFS of HCC was established. Nomogram performance was assessed via internal validation and calibration curve statistics. Receiver operating characteristic (ROC) and decision curve analysis (DCA) curve were used to validate the performance and clinical utility of the nomogram. Results: Multivariate Cox regression analysis indicated that ALBI grade (hazard ratio, [HR] = 2.624, 95% confidence interval [CI]: 1.391– 4.949, P = 0.003), INR (HR = 2.605, 95% CI: 1.061– 6.396, P = 0.037), MLR (HR = 1.769, 95% CI: 1.073– 2.915, P = 0.025) and MVI (HR = 4.726, 95% CI: 2.365– 9.444, P < 0.001) were independent prognostic factors of RFS. Nomogram with independent factors was established and achieved a better concordance index of 0.753 (95% CI: 0.672– 0.834) for predicting RFS. The ROC found that the area under curve (AUC) was consistent with the C-index and the sensitivity was 85.4%. The risk score calculated by nomogram could divide AFP-negative HCC patients into high-, moderate- and low-risk groups (P < 0.05). DCA analysis revealed that the nomogram could augment net benefits and exhibited a wider range of threshold probabilities by the risk stratification than the AJCC T and BCLC stage in the prediction of AFP-negative HCC recurrence. Conclusion: The ALBI grade- and MLR-based nomogram prognostic model for RFS showed high predictive accuracy in AFP-negative HCC patients after surgical resection.

Background: Inflammation plays a significant role in tumour development, progression, and metastasis. In this study, we focused on comparing the predictive potential of inflammatory markers for overall survival (OS), recurrence-free survival (RFS), and 1- and 2-year RFS in hepatocellular carcinoma (HCC) patients. Methods: A total of 360 HCC patients were included in this study. A LASSO regression analysis model was used for data dimensionality reduction and element selection. Univariate and multivariate Cox regression analyses were performed to identify the independent risk factors for HCC prognosis. Nomogram prediction models were established and decision curve analysis (DCA) was conducted to determine the clinical utility of the nomogram model. Results: Multivariate Cox regression analysis indicated that the prognostic nutritional index (PNI) and neutrophil-to-lymphocyte ratio (NLR) were independent prognostic factors of OS, and aspartate aminotransferase-to-platelet ratio (APRI) was a common independent prognostic factor among RFS, 1-year RFS, and 2-year RFS. The systemic inflammation response index (SIRI) was an independent prognostic factor for 1-year RFS in HCC patients after curative resection. Nomograms established and achieved a better concordance index of 0.772(95% CI: 0.730-0.814), 0.774(95% CI: 0.734-0.815), 0.809(95% CI: 0.766-0.852), and 0.756(95% CI: 0.696-0.816) in predicting OS, RFS, 1-year RFS, and 2-year RFS respectively. The risk scores calculated by nomogram models divided HCC patients into high-, moderate- and low-risk groups (P < 0.05). DCA analysis revealed that the nomogram models could augment net benefits and exhibited a wider range of threshold probabilities in the prediction of HCC prognosis. Conclusions: The nomograms showed high predictive accuracy for OS, RFS, 1-year RFS, and 2-year RFS in HCC patients after surgical resection. The nomograms could be useful clinical tools to guide a rational and personalized treatment approach and prognosis judgement.

The presence of microvascular invasion (MVI) is a critical determinant of early hepatocellular carcinoma (HCC) recurrence and prognosis. We developed a nomogram model integrating clinical laboratory examinations and radiological imaging results from our clinical database to predict microvascular invasion presence at preoperation in HCC patients. 242 patients with pathologically confirmed HCC at the Ningbo Medical Centre Lihuili Hospital from September 2015 to January 2021 were included in this study. Baseline clinical laboratory examinations and radiological imaging results were collected from our clinical database. LASSO regression analysis model was used to construct data dimensionality reduction and elements selection. Multivariate logistic regression analysis was performed to identify the independent risk factors associated with MVI and finally a nomogram for predicting MVI presence of HCC was established. Nomogram performance was assessed via internal validation and calibration curve statistics. Decision curve analysis (DCA) was conducted to determine the clinical usefulness of the nomogram model by quantifying the net benefits along with the increase in threshold probabilities. Survival analysis indicated that the probability of overall survival (OS) and recurrence-free survival (RFS) were significantly different between patients with MVI and without MVI (P < 0.05). Histopathologically identified MVI was found in 117 of 242 patients (48.3%). The preoperative factors associated with MVI were large tumor diameter (OR = 1.271, 95%CI: 1.137–1.420, P < 0.001), AFP level greater than 20 ng/mL (20–400 vs. ≤ 20, OR = 2.025, 95%CI: 1.056–3.885, P = 0.034; > 400 vs. ≤ 20, OR = 3.281, 95%CI: 1.661–6.480, P = 0.001), total bilirubin level greater than 23 umol/l (OR = 2.247, 95%CI: 1.037–4.868, P = 0.040). Incorporating tumor diameter, AFP and TB, the nomogram achieved a better concordance index of 0.725 (95%CI: 0.661–0.788) in predicting MVI presence. Nomogram analysis showed that the total factor score ranged from 0 to 160, and the corresponding risk rate ranged from 0.20 to 0.90. The DCA showed that if the threshold probability was > 5%, using the nomogram to diagnose MVI could acquire much more benefit. And the net benefit of the nomogram model was higher than single variable within 0.3–0.8 of threshold probability. In summary, the presence of MVI is an independent prognostic risk factor for RFS. The nomogram detailed here can preoperatively predict MVI presence in HCC patients. Using the nomogram model may constitute a usefully clinical tool to guide a rational and personalized subsequent therapeutic choice.

Observational studies inconsistently reported the relationship between vitamin C intake and risk of pancreatic cancer. We conducted a meta-analysis of published case-control and cohort studies to quantify the association. Potentially eligible studies were found on PubMed and EMBASE databases through May 31, 2015. A random-effects model was assigned to compute summary point estimates with corresponding 95% confidence intervals (CIs). Subgroup and meta-regression analyses were also performed to explore sources of heterogeneity. Our final analyses included 20 observational studies comprising nearly 5 thousand cases of pancreatic cancer. When comparing the highest with the lowest categories of vitamin C intake, the summary odds ratio/relative risk for case-control studies (14 studies), cohort studies (6 studies) and all studies combined was 0.58 (95% CI: 0.52–0.66), 0.93 (95% CI: 0.78–1.11) and 0.66 (95% CI: 0.58–0.75), respectively. The difference in the findings between case-control and cohort studies was statistically significant (P < .001). Possible publication bias was shown in the meta-analysis of case-control studies. There is insufficient evidence to conclude any relationship between vitamin C intake and risk of pancreatic cancer. The strong inverse association observed in case-control studies may be affected by biases (eg, recall and selection biases) that particularly affect case-control studies and/or potential publication bias. Future prospective studies of vitamin C intake and pancreatic cancer are needed.