ABSTRACT
Background: Laryngeal carcinoma is one of the most common types of head and neck tumors. The mortality rate in patients with laryngeal cancer has not declined in recent years. Previous studies have shown that laryngeal cancer mortality is related to the extent of laryngeal cancer, the proportion of lymph node metastases, treatment modalities, and postoperative lifestyle habits. Thus, early identifying patients at high risk of laryngeal cancer-specific death is of great clinical importance. However, in the presence of competing risk, the existing survival models based on Cox proportional hazards model may be biased in estimating tumor-specific mortality. In this study, we developed and validated a nomogram based on competitive risk analysis for patients with laryngeal cancer. Methods: We used SEER*Stat (Version 4.6.1) software to identify patients in the Surveillance, Epidemiology, and End Results (SEER) database who were diagnosed with laryngeal cancer between 2000 and 2019 as study subjects. The collected data included demographic data, the primary site of laryngeal cancer, the histological type of tumor, tumor size, and other variables. After excluding cases with missing information, the entire cohort was randomly split into a training cohort and a validation cohort at a 7:3 ratio. The training cohort was used in building the model while the validation cohort was used to validate the model. Univariate and multivariate Fine&Gray regression analyses were used to screen statistically significant variables, and the model performance was measured by establishing a consistency index, receiver operating characteristic curve (ROC), and calibration curves. Results: After excluding cases with missing information, 3,805 patients (2,264 in the training cohort and 1,141 in the validation cohort) were included in the study and followed for a median of 16 months. A total of 411 died of laryngeal cancer, and 2,104 patients died from other causes. Among 3,805 patients, the vast majority was male (80.9%), and Caucasian (77.2%), and aged 60-80 years old (58.4%). Conclusions: Advanced age and keratinized SCC are risk factors for laryngeal cancer-specific death. These high-risk patients should be given more attention and closer monitoring in clinical practice.
ABSTRACT
BACKGROUND: It is unclear about the cardiovascular and metabolic diseases (CMD) among Chinese patients with different clinical subtypes of obstructive sleep apnea (OSA). HYPOTHESIS: The prevalence of CMD varies among OSA patients of different clinical subtypes. METHODS: A total of 1483 Chinese patients with OSA were assessed to evaluate the existence of clinical subtypes of OSA using latent class analysis. We compared the differences in demographic characteristics and prevalence of CMD using ANOVA and χ2 tests. Associations between clinical subtypes and disease prevalence were assessed using adjusted logistic regression. RESULTS: We identified prevalent CMD in Chinese patients with the four subtypes of OSA: excessively sleepy (ES), moderately sleepy with disturbed sleep (ModSwDS), moderately sleepy (ModS), and minimally symptomatic (MinS). The ES subtype had a higher body mass index, average Epworth Sleepiness Scale score, Apnea-hypopnea index, and oxyhemoglobin saturation below 90% compared with the other subtypes (p < .05). The MinS subtype had the lowest mean ESS score (p < .05). We found a significant difference in the prevalence of CMD among the four subtypes, with the highest proportion of cases of CMD in the ES subtype. In adjusted models, significant associations with CMD were also found. ES, ModSwDS, ModS, and MinS subtypes are very high-risk, high-risk, medium-risk, and low-risk in prevalent CMD. CONCLUSIONS: This study identified four clinical subtypes of OSA in Chinese patients. Each clinical subtype corresponds with a different level of prevalence of CMD; this finding is helpful for the more precise treatment of patients with different clinical manifestations.
