Prediction Model For Patients Research Articles

Risk factor and prediction model development for severe radiation-induced oral mucositis in head and neck tumors.

Aim: This article aims to identify risk factors for severe radiation-induced oral mucositis (RIOM) in head and neck cancer (HNC) patients. In addition, we intend to establish a predictive model in patients undergoing intensity-modulated radiotherapy. Patients & methods: In this retrospective study, several HNC patients (n=179) treated at Zhejiang Provincial People's Hospital from January 2019 to June 2023 were considered. The recruited subjects were divided into modeling and validation groups. The experimental data on clinical characteristics and treatment were collected and analyzed to identify predictive factors for severe RIOM based on the logistic regression approach. Results: The results indicated that severe RIOM occurred in 55.3% of patients. Accordingly, significant predictors included smoking history, diabetes, concurrent chemotherapy, cumulative radiation dose and weight loss of ≥5% in relative to admission weight. A nomogram based on these factors was validated, showing excellent predictive accuracy. Conclusion: In summary, the predictive model could effectively identify high-risk patients for severe RIOM, enabling the design of targeted interventions and improving patient management during radiotherapy.

A predictive model in patients with chronic hydrocephalus following aneurysmal subarachnoid hemorrhage: a retrospective cohort study.

Our aim was to develop a nomogram that integrates clinical and radiological data obtained from computed tomography (CT) scans, enabling the prediction of chronic hydrocephalus in patients with aneurysmal subarachnoid hemorrhage (aSAH). A total of 318 patients diagnosed with subarachnoid hemorrhage (SAH) and admitted to the Department of Neurosurgery at the Affiliated People's Hospital of Jiangsu University between January 2020 and December 2022 were enrolled in our study. We collected clinical characteristics from the hospital's medical record system. To identify risk factors associated with chronic hydrocephalus, we conducted both univariate and LASSO regression models on these clinical characteristics and radiological features, accompanied with penalty parameter adjustments conducted through tenfold cross-validation. All features were then incorporated into multivariate logistic regression analyses. Based on these findings, we developed a clinical-radiological nomogram. To evaluate its discrimination performance, we conducted Receiver Operating Characteristic (ROC) curve analysis and calculated the Area Under the Curve (AUC). Additionally, we employed calibration curves, and utilized Brier scores as an indicator of concordance. Additionally, Decision Curve Analysis (DCA) was performed to determine the clinical utility of our models by estimating net benefits at various threshold probabilities for both training and testing groups. The study included 181 patients, with a determined chronic hydrocephalus prevalence of 17.7%. Univariate logistic regression analysis identified 11 potential risk factors, while LASSO regression identified 7 significant risk factors associated with chronic hydrocephalus. Multivariate logistic regression analysis revealed three independent predictors for chronic hydrocephalus following aSAH: Periventricular white matter changes, External lumbar drainage, and Modified Fisher Grade. A nomogram incorporating these factors accurately predicted the risk of chronic hydrocephalus in both the training and testing cohorts. The AUC values were calculated as 0.810 and 0.811 for each cohort respectively, indicating good discriminative ability of the nomogram model. Calibration curves along with Hosmer-Lemeshow tests demonstrated excellent agreement between predicted probabilities and observed outcomes in both cohorts. Furthermore, Brier scores (0.127 for the training and 0.09 for testing groups) further validated the predictive performance of our nomogram model. The DCA confirmed that this nomogram provides superior net benefit across various risk thresholds when predicting chronic hydrocephalus. The decision curve demonstrated that when an individual's threshold probability ranged from 5 to 62%, this model is more effective in predicting the occurrence of chronic hydrocephalus after aSAH. A clinical-radiological nomogram was developed to combine clinical characteristics and radiological features from CT scans, aiming to enhance the accuracy of predicting chronic hydrocephalus in patients with aSAH. This innovative nomogram shows promising potential in assisting clinicians to create personalized and optimal treatment plans by providing precise predictions of chronic hydrocephalus among aSAH patients.

Comparison of Cardiorespiratory Fitness Prediction Equations and Generation of New Predictive Model for Patients with Obesity.

Cardiorespiratory fitness (CRF) is a critical marker of overall health and a key predictor of morbidity and mortality, but the existing prediction equations for CRF are primarily derived from general populations and may not be suitable for patients with obesity. Predicted CRF from different non-exercise prediction equations was compared with measured CRF of patients with obesity who underwent maximal cardiopulmonary exercise testing (CPET). Multiple linear regression was used to develop a population-specific non-exercise CRF prediction model for treadmill exercise including age, sex, weight, height and physical activity level as determinants. 660 patients underwent CPET during the study period. Within the entire cohort, R2 values had a range of 0.24-0.46. Predicted CRF was statistically different from measured CRF for 19 included equations. Only 50% of patients were correctly classified into the measured CRF categories according to predicted CRF. A multiple model for CRF prediction (ml/min) was generated (R2 = 0.78) and validated using two cross-validation methods. Most used equations provide inaccurate estimates of CRF in patients with obesity, particularly in cases of severe obesity and low CRF. Therefore, a new prediction equation was developed and validated specifically for patients with obesity, offering a more precise tool for clinical CPET interpretation and risk stratification in this population.

Multimodal workflows optimally predict response to repetitive transcranial magnetic stimulation in patients with schizophrenia: a multisite machine learning analysis

The response variability to repetitive transcranial magnetic stimulation (rTMS) challenges the effective use of this treatment option in patients with schizophrenia. This variability may be deciphered by leveraging predictive information in structural MRI, clinical, sociodemographic, and genetic data using artificial intelligence. We developed and cross-validated rTMS response prediction models in patients with schizophrenia drawn from the multisite RESIS trial. The models incorporated pre-treatment sMRI, clinical, sociodemographic, and polygenic risk score (PRS) data. Patients were randomly assigned to receive active (N = 45) or sham (N = 47) rTMS treatment. The prediction target was individual response, defined as ≥20% reduction in pre-treatment negative symptom sum scores of the Positive and Negative Syndrome Scale. Our multimodal sequential prediction workflow achieved a balanced accuracy (BAC) of 94% (non-responders: 92%, responders: 95%) in the active-treated group and 50% in the sham-treated group. The clinical, clinical + PRS, and sMRI-based classifiers yielded BACs of 65%, 76%, and 80%, respectively. Apparent sadness, inability to feel, educational attainment PRS, and unemployment were most predictive of non-response in the clinical + PRS model, while grey matter density reductions in the default mode, limbic networks, and the cerebellum were most predictive in the sMRI model. Our sequential modelling approach provided superior predictive performance while minimising the diagnostic burden in the clinical setting. Predictive patterns suggest that rTMS responders may have higher levels of brain grey matter in the default mode and salience networks which increases their likelihood of profiting from plasticity-inducing brain stimulation methods, such as rTMS. The future clinical implementation of our models requires findings to be replicated at the international scale using stratified clinical trial designs.

Establishment and assessment of mortality risk prediction model in patients with sepsis based on early-stage peripheral lymphocyte subsets.

This study is aimed to explore the value of lymphocyte subsets in evaluating the severity and prognosis of sepsis. The counts of lymphocytes, CD3+ T cells, CD4+ T cells, CD8+ T cells, CD19+ B cells, and NK cells significantly decreased between day 1 and day 3 in both the survivor and the non-survivor groups. The peripheral lymphocyte subsets (PLS) at day 1 were not significantly different between the survivor and the non-survivor groups. However, at day 3, the counts of lymphocytes, CD3+ T cells, CD4+ T cells, and NK cells were remarkably lower in the non-survivor group. No significant differences in CD8+ T cells, or CD19+ B cells were observed. The PLS index was independently and significantly associated with the 28-day mortality risk in septic patients (OR: 3.08, 95% CI: 1.18-9.67). Based on these clinical parameters and the PLS index, we developed a nomograph for evaluating the individual mortality risk in sepsis. The area under the curve of prediction with the PLS index was significantly higher than that from the model with only clinical parameters (0.912 vs. 0.817). Our study suggests that the decline of PLS occurred in the early stage of sepsis. The new novel PLS index can be an independent predictor of 28-day mortality in septic patients. The prediction model based on clinical parameters and the PLS index has relatively high predicting ability.

Cervical Sagittal Alignment and Related Factor Analysis and Prediction Model in Patients Undergoing Revision Surgery After Anterior Cervical Fusion.

Patients with myelopathy or radiculopathy commonly undergo anterior cervical fusion surgery (ACFS), which has a notable failure rate on occasion. The goal of this study was to compare revision and nonrevision surgery patients in cervical sagittal alignment (CSA) subsequent to ACFS; additionally, to identify the best CSA parameters for predicting clinical outcome after ACFS; and furthermore, to create an equation model to assist surgeons in making decisions on patients undergoing ACFS. The data of 99 patients with symptomatic cervical myelopathy/radiculopathy who underwent ACFS were analyzed. Patients were divided into group A (underwent revision surgery after the first surgery failed) and group B (underwent only the first surgery). We measured and analyzed both preoperative and postoperative CSA parameters, including C2 slope, T1 slope, cervical lordosis C2-C7 (CL), C2-C7 sagittal vertical axis (C2C7 SVA), occiput-C2 lordosis angle (C0-C2), and chin brow vertical angle, and we further computed the correlation between the CSA parameters and created a prediction model. The (T1S-CL)-C2S mismatch differed significantly between groups A and B ([9.95 ± 9.95] 0 , [3.79 ± 6.58] 0 , P < 0.05, respectively). A significant correlation was observed between C2 slope and T1CL in group B relative to group A postoperatively (R 2 = 0.42 versus R 2 = 0.09, respectively). Compared with group B, patients in group A had significantly higher C2C7SVA values, more levels of fusion, and more smokers. The sensitivity, specificity, accuracy, and discrimination of the model were, respectively, 73.5%, 84%, 78.8%, and 85.65%. The causes of revision surgery in cervical myelopathic patients after anterior cervical corpectomy and fusion/anterior cervical diskectomy and fusion are multifactorial. (T1S-CL)-C2S mismatch and high C2C7SVA are the best cervical sagittal parameters that increase the odds of revision surgery, and the effect is more enhanced when comorbidities such as smoking, low bone-mineral density, and increased levels of fusion are taken into account.

Prognostic factors and predictive models for patients with lung large cell neuroendocrine carcinoma: Based on SEER database.

Lung Large cell neuroendocrine carcinoma (LCNEC) is a rare, aggressive, high-grade neuroendocrine carcinoma with a poor prognosis, mainly seen in elderly men. To date, we have found no studies on predictive models for LCNEC. We extracted data from the Surveillance, Epidemiology, and End Results (SEER) database of confirmed LCNEC from 2010 to 2018. Univariate and multivariate Cox proportional risk regression analyses were used to identify independent risk factors, and then we constructed a novel nomogram and assessed the predictive effectiveness by receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). A total of 2546 patients with LCNEC were included, excluding those diagnosed with autopsy or death certificate, tumor, lymph node, metastasis (TNM) stage, tumor grade deficiency, etc., and finally, a total of 743 cases were included in the study. After univariate and multivariate analyses, we concluded that the independent risk factors were N stage, intrapulmonary metastasis, bone metastasis, brain metastasis, and surgical intervention. The results of ROC curves, calibration curves, and DCA in the training and validation groups confirmed that the nomogram could accurately predict the prognosis. The nomogram obtained from our study is expected to be a useful tool for personalized prognostic prediction of LCNEC patients, which may help in clinical decision-making.

Pretreatment mortality risk prediction model in patients with polymyositis/dermatomyositis-associated interstitial lung disease

ObjectivesRisk prediction for patients with polymyositis/dermatomyositis-associated interstitial lung disease (PM/DM-ILD) is challenging due to heterogeneity in the disease course. We aimed to develop a mortality risk prediction model for PM/DM-ILD.MethodsThis...

Establishment and verification of invasion syndrome prediction model in patients with diabetes complicated with Klebsiella pneumoniae liver abscess

Objective: To analyze the correlative factors of invasion syndrome in patients with diabetes complicated with Klebsiella pneumoniae liver abscess, and to construct and verify the online nomographic prediction model. Methods: A case control study. The clinical data of 213 diabetic patients with Klebsiella pneumoniae liver abscess admitted to the Third Affiliated Hospital of Soochow University from January 1, 2015 to December 31, 2021 were retrospectively analyzed. The patients were divided into the training set (149 cases) and the test set (64 cases) by stratified random sampling method at a ratio of 7∶3. Synthetic minority over-sampling technique(SMOTE) was used to process the imbalanced data, then Lasso regression was used to screen out the optimal feature variables in the training set and multivariate logistic regression model was used to construct the prediction model of invasion syndrome in patients with diabetes complicated with Klebsiella pneumoniae liver abscess, and verify it in the training set and test set. Receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis (DCA) were used to evaluate the prediction efficiency of the model, and the simple and online interactive dynamic web page column graph was constructed. Results: Among the 213 patients, 60 were males and 153 were females, aged of (61.4±12.0) years. A total of 25(11.74%) diabetic patients with Klebsiella pneumoniae liver abscess developed invasion syndrome, which were included in divided into invasive K.pneumoniae liver abscesses syndrome (IKPLAS) group, and the other 188 cases were in without invasive K.pneumoniae liver abscesses syndrome (NIKPLAS) group. SMOTE algorithm was used for oversampling processing, so that the ratio of positive and negative samples was 1∶1. In the oversampling training set, 5 main risk factors were screened based on Lasso regression, namely fasting blood glucose (λ=0.063), hemoglobin (λ=-0.042), blood urea nitrogen (λ=-0.050), abscess size (λ=-0.025) and sequential organ failure assessment (SOFA) score (λ=0.450), respectively. Multivariate logistic regression model showed that fasting blood glucose (OR=1.20, 95%CI: 0.98-1.48, P=0.006), hemoglobin (OR=0.90, 95%CI: 0.86-0.95, P<0.001), blood urea nitrogen (OR=1.22, 95%CI: 1.03-1.43, P=0.017), abscess diameter (OR=0.76, 95%CI: 0.61-0.94, P=0.010), SOFA score (OR=3.08, 95%CI: 2.18-4.36, P<0.001) were associated with invasion syndrome in patients with diabetes complicated with Klebsiella pneumoniae liver abscess. The area under the curve of ROC in the training set was 0.966 (95%CI: 0.943-0.989), the sensitivity was 90.5%, and the specificity was 91.3%. The area under the curve of the validation set ROC was 0.946 (95%CI: 0.902-0.991), with a sensitivity of 79.6% and a specificity of 88.9%. The calibration curves drawn in the training set and the test set fit well with the ideal curve. DCA showed that the neomorph prediction model had a good clinical net benefit when predicting the risk of IKPLAS in patients with diabetes complicated with Klebsiella pneumoniae liver abscess was 0.10-0.40. Conclusions: Fasting blood glucose, hemoglobin, urea nitrogen, abscess size and SOFA score are the related factors for invasion syndrome in patients with diabetes complicated with Klebsiella pneumoniae liver abscess. The constructed column graph can effectively predict the risk of invasion syndrome in patients with diabetes complicated with Klebsiae pneumoniae liver abscess.

Establishment of a prognosis predictive model for liver cancer based on expression of genes involved in the ubiquitin-proteasome pathway.

The ubiquitin-proteasome pathway (UPP) has been proven to play important roles in cancer. To investigate the prognostic significance of genes involved in the UPP and develop a predictive model for liver cancer based on the expression of these genes. In this study, UPP-related E1, E2, E3, deubiquitylating enzyme, and proteasome gene sets were obtained from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, aiming to screen the prognostic genes using univariate and multivariate regression analysis and develop a prognosis predictive model based on the Cancer Genome Atlas liver cancer cases. Five genes (including autophagy related 10, proteasome 20S subunit alpha 8, proteasome 20S subunit beta 2, ubiquitin specific peptidase 17 like family member 2, and ubiquitin specific peptidase 8) were proven significantly correlated with prognosis and used to develop a prognosis predictive model for liver cancer. Among training, validation, and Gene Expression Omnibus sets, the overall survival differed significantly between the high-risk and low-risk groups. The expression of the five genes was significantly associated with immunocyte infiltration, tumor stage, and postoperative recurrence. A total of 111 differentially expressed genes (DEGs) were identified between the high-risk and low-risk groups and they were enriched in 20 and 5 gene ontology and KEGG pathways. Cell division cycle 20, Kelch repeat and BTB domain containing 11, and DDB1 and CUL4 associated factor 4 like 2 were the DEGs in the E3 gene set that correlated with survival. We have constructed a prognosis predictive model in patients with liver cancer, which contains five genes that associate with immunocyte infiltration, tumor stage, and postoperative recurrence.

Development of postoperative delirium prediction models in patients undergoing cardiovascular surgery using machine learning algorithms

Associations between delirium and postoperative adverse events in cardiovascular surgery have been reported and the preoperative identification of high-risk patients of delirium is needed to implement focused interventions. We aimed to develop and validate machine learning models to predict post-cardiovascular surgery delirium. Patients aged ≥ 40 years who underwent cardiovascular surgery at a single hospital were prospectively enrolled. Preoperative and intraoperative factors were assessed. Each patient was evaluated for postoperative delirium 7 days after surgery. We developed machine learning models using the Bernoulli naive Bayes, Support vector machine, Random forest, Extra-trees, and XGBoost algorithms. Stratified fivefold cross-validation was performed for each developed model. Of the 87 patients, 24 (27.6%) developed postoperative delirium. Age, use of psychotropic drugs, cognitive function (Mini-Cog < 4), index of activities of daily living (Barthel Index < 100), history of stroke or cerebral hemorrhage, and eGFR (estimated glomerular filtration rate) < 60 were selected to develop delirium prediction models. The Extra-trees model had the best area under the receiver operating characteristic curve (0.76 [standard deviation 0.11]; sensitivity: 0.63; specificity: 0.78). XGBoost showed the highest sensitivity (AUROC, 0.75 [0.07]; sensitivity: 0.67; specificity: 0.79). Machine learning algorithms could predict post-cardiovascular delirium using preoperative data.Trial registration: UMIN-CTR (ID; UMIN000049390).

External validation of the EuroSCORE I and II prediction models for postoperative 30-day mortality in patients with infective endocarditis: results from a nationwide prospective registry

Abstract Background The European System for Cardiac Operative Risk Evaluation (EuroSCORE) I and II are widely used risk models for predicting 30-day mortality after cardiac surgery [1,2]. It is suggested that both models overpredict 30-day mortality in specific groups of patients [3,4]. However, their performance in patients with infective endocarditis (IE) is unclear, since few IE patients were included in the development cohort (&amp;lt;500 patients for EuroSCORE II). Moreover, all previous studies that have evaluated the performance of EuroSCORE models in IE had small sample sizes [5-7]. Purpose To externally validate the EuroSCORE I and II prediction models in patients with IE in a nationwide prospective registry. Methods The Netherlands Heart Registration (NHR) is a national prospective registry of all patients undergoing cardiac surgeries and interventions in The Netherlands. It collects clinically relevant data, such as medical history and postoperative complications, and follows patients up to many years after surgery. In addition, the EuroSCORE I and II are recorded for each individual. For this study, we included all patients with IE who underwent cardiac surgery in the Netherlands between 2013 and 2021. The predictive performance of EuroSCORE I and II was assessed by discrimination and calibration. Discrimination was assessed by the area under the curve (AUC), and reflects how well the model discriminates between patients with and without the outcome of interest. Calibration describes the agreement between observed and predicted risks and was assessed by calibration in the large (average observed risk versus average predicted risk) and calibration plots. Results We included 2577 IE patients (median age 64 years (IQR 53 – 71), 76.2% male, 33.6% with previous cardiac surgery of which 94.3% previous valve surgery). Overall 30-day mortality was 10.2%. The mean EuroSCORE I and II were 27.9% (SD 22.1) and 14.9% (SD 16.3) respectively. The AUCs for EuroSCORE I and II were 0.73 (95% CI 0.70 – 0.76) and 0.72 (95% CI 0.69 – 0.75) (Figure 1). Calibration in the large showed that both models overpredict risks, with an observed to expected ratio of 0.37 for EuroSCORE I and 0.68 for EuroSCORE II. The calibration plot showed that EuroSCORE I overpredicts risks across the full range (Figure 2A). Calibration of the EuroSCORE II was reasonable in patients with low predicted probability, but overpredicts risks when the predicted probability was &amp;gt;20% (Figure 2B). Conclusion Both EuroSCORE I and II substantially overestimate 30-day mortality risk after cardiac surgery in IE patients. Use of EuroSCORE I and II may therefore lead to disadvantageous decision making if patients with a high predicted 30-day mortality risk are subsequently withheld from indicated surgical treatment (class I or IIa) [8]. Further research will recalibrate the EuroSCORE II for IE patients and assess the influence of its individual parameters on 30-day mortality in this population.Receiver operating characteristic curvesCalibration plots of EuroSCORE I and II

Abstract 18252: Comparison of an Artificial Intelligence Heart Failure Detection Model and Clinical Prediction Models in Patients With Heart Failure With Preserved Ejection Fraction (HFPEF)

Background: An artificial intelligence (AI) model to diagnose heart failure (HF) with preserved ejection fraction (HFpEF) based on deep learning of echocardiographic (TTE) images could improve under-recognition of disease. However, the diagnostic accuracy in high-risk patients, compared to existing clinical prediction models, remains uncertain. Methods: In this retrospective case-control study, we identified patients who received routine TTE at Beth Israel Deaconess Medical Center, Harvard Medical School. Cases were defined by presence of ICD-10 codes for HF (I50.X), LVEF≥50%, and grade II or III diastolic dysfunction. Controls were defined by absence of HF within 1 year of the index TTE. The discriminatory performance of an AI HFpEF model (EchoGo Heart Failure, Ultromics Ltd.) was compared to the clinical H2FPEF score. Results: Both cases (n=213, 75±12 years, 53% female, 29.4±7.1 kg/m2) and controls (n=226, 75±13 years, 54% female, 27.0±5.6 kg/m2) were high-risk for HFpEF, with high rates of hypertension, atrial fibrillation, diabetes, coronary artery, kidney, and pulmonary disease. The H2FPEF score demonstrated high sensitivity (mean [95% CI]; 97.3% [94.3, 100%]) and modest specificity (38.1% [25.9, 55.9%]). The AI HFpEF model had similar sensitivity (88.8% [84.0, 93.4%]) and slightly improved specificity (55.6% [47.7, 64.0%]). In the 265 patients (60.4%) with non-diagnostic H2FPEF scores, the AI HFpEF model successfully classified 144 patients as probable HFpEF ( Figure ). The H2FPEF score identified 31 (54.5%) whereas the AI HFpEF model identified 45 (78.9%) HFpEF patients who were hospitalized within 1 year of the TTE. Conclusions: An AI model to detect HFpEF improves upon identification of HFpEF amongst high-risk individuals, thus potentially improving the capacity to identify patients who might benefit from treatment.

Application of the Rotterdam postoperative cerebellar mutism syndrome prediction model in patients undergoing surgery for medulloblastoma in a single institution.

Postoperative cerebellar mutism syndrome (CMS) develops in up to 40% of children with medulloblastoma. The Rotterdam model (RM) has been reported to predict a 66% risk of CMS in patients with a score of ≥ 100. The aim of this study was to retrospectively apply the RM to an independent cohort of patients with newly diagnosed medulloblastoma and study the applicability of the RM in predicting postoperative CMS. Participants had to have their first tumor resection at the authors' institution and be enrolled in the SJMB12 protocol (NCT01878617). All participants underwent structured serial neurological evaluations before and then periodically after completing radiation therapy. Imaging was reviewed by the study neurologist who was blinded to CMS status when reviewing the scans and retrospectively applied RM score to each participant. Forty participants were included (14 females and 26 males). Four (10%) patients had CMS. The median age at tumor resection was 11.7 years (range 3.5-17.8 years). Tumor location was midline in 30 (75%), right lateral in 6 (15%), and left lateral in 4 (10%). The median Evans index was 0.3 (range 0.2-0.4), and 34 (85%) patients had an Evans index ≥ 0.3. Five participants required a ventricular shunt. The median tumor volume was 51.97 cm3 (range 20.13-180.58 cm3). Gross-total resection was achieved in 35 (87.5%) patients, near-total resection in 4 (10%), and subtotal in 1. The median RM score was 90 (range 25-145). Eighteen participants had an RM score of ≥ 100, and of these 16.7% (n = 3) had CMS. Of the 22 patients with an RM score < 100, 1 child developed CMS (4.5%, CI 0.1%-22.8%); 3 of the 18 patients with an RM score ≥ 100 developed CMS (16.7%, CI 3.6%-41.4%). The observed rate of CMS in the cohort of children with an RM score ≥ 100 was significantly lower than the observed rate in the original RM cohort (66.7%, CI 51%-80.0%, p < 0.001). A greater risk of CMS in patients with an RM score ≥ 100 could not be confirmed (p = 0.31). At the authors' institution, the incidence of CMS in patients who had an RM ≥ 100 was significantly lower than the RM cohort. These findings raise questions regarding generalizability of RM; however, fewer cases of CMS and a relatively small cohort limit this conclusion.

A predictive model for patients with local-regionally advanced oropharyngeal squamous cell carcinoma treated after cervical lymph node dissection.

To develop a nomogram to predict the cancer-specific survival of patients with local-regionally advanced oropharyngeal squamous cell carcinoma after cervical lymph node dissection. The clinical variables of patients confirmed as having oropharyngeal squamous cell carcinoma between 2008 and 2015 were retrieved from the Surveillance, Epidemiology and End Results database. Univariate and multivariate analysis were performed,followed by the construction of nomogramsfor CSS. Nomogram' accuracy was evaluated through the concordance index, calibration curves and decision curve analysis. A total of 1994 oropharyngeal squamous cell carcinoma patients who underwent surgery were included in this study. Sex, T-stage, American Joint Committee on Cancer-stage, positive lymph nodes, positive lymph node ratio, log odds of positive lymph nodes, and postoperative radiotherapy were selected to establish the nomogram for oropharyngeal squamous cell carcinoma. The concordance index of the nomogram was 0.747 (95% CI 0.714-0.780) in the training calibration cohort and 0.735 (95% CI 0.68-0.789) in the validationcohort and the time-dependent Area under the curve (> 0.7) indicated satisfactory discriminative ability of the nomogram. The calibration plot shows that there is a good consistency between the predictions of the nomogram and the actual observations in the training and validation cohorts. In addition, decision curve analysis showed that the nomogram was clinically useful and had a better ability to recognize patients at high risk than the American Joint Committee on Cancer tumor-node-metastasis staging. The predictive model has the potential to provide valuable guidance to clinicians in the treatment of patients with locoregionally advanced OPSCC confined to the cervical lymph nodes.

Evaluation of Trace Elements Levels and Construction of Auxiliary Prediction Model in Patients with Diabetes Ketoacidosis in Type 1 Diabetes

Evaluation of Trace Elements Levels and Construction of Auxiliary Prediction Model in Patients with Diabetes Ketoacidosis in Type 1 Diabetes

A nomogram to predict in-hospital mortality of gastrointestinal bleeding patients in the intensive care unit.

Gastrointestinal bleeding (GIB) is a common condition in clinical practice, and predictive models for patients with GIB have been developed. However, assessments of in-hospital mortality due to GIB in the intensive care unit (ICU), especially in critically ill patients, are still lacking. This study was designed to screen out independent predictive factors affecting in-hospital mortality and thus establish a predictive model for clinical use. This retrospective study included 1,442 patients with GIB who had been admitted to the ICU. They were selected from the Medical Information Mart for Intensive Care IV (MIMIC-IV) 1.0 database and divided into a training group and a validation group in a ratio of 7:3. The main outcome measure was in-hospital mortality. Least absolute shrinkage and section operator (LASSO) regression was used to screen out independent predictors and create a nomogram. LASSO regression picked out nine independent predictors: heart rate (HR), activated partial thromboplastin time (aPTT), acute physiology score III (APSIII), sequential organ failure assessment (SOFA), cerebrovascular disease, acute kidney injury (AKI), norepinephrine, vasopressin, and dopamine. Our model proved to have excellent predictive value with regard to in-hospital mortality (the area under the receiver operating characteristic curve was 0.906 and 0.881 in the training and validation groups, respectively), as well as a good outcome on a decision curve analysis to assess net benefit. Our model effectively predicts in-hospital mortality in patients with GIB, indicating that it may prove to be a valuable tool in future clinical practice.

Cardiovascular Health Behavior Prediction Model in Patients With Type 2 Diabetes.

To prevent cardiovascular disease in adults with type 2 diabetes, it is necessary to identify the factors that affect cardiovascular health behavior. The aim of this study was to verify the causal relationship between illness perception and diabetes knowledge as cognitive representations, depression as emotional representations, self-efficacy as cognitive coping, and oral health and cardiovascular health behaviors as behavioral coping strategies. A hypothetical model was established based on a literature review and the self-regulation model. In this study, a cross-sectional survey design was used, and the participants were adults 30 years or older who had been given a diagnosis of type 2 diabetes for more than 6 months and were recruited from South Korea. Data from 272 participants were collected through face-to-face interviews or Google surveys in 2021, and analyzed using SPSS 25.0 and AMOS 22.0. Oral health behavior (β = 0.26, P < .001) and self-efficacy (β = 0.16, P = .048) had a direct effect on cardiovascular health behavior, and illness perception (β = 0.24, P = .018) and depression (β = -0.25, P < .001) had an indirect effect through self-efficacy (β = 0.24, P = .016), with a total explanatory power of 19.1%. Oral health behavior was directly affected by illness perception and self-efficacy, and self-efficacy was directly affected by illness perception, diabetes knowledge, and depression (P < .05). To practice cardiovascular health behaviors in adults with diabetes, self-efficacy for diabetes management should be enhanced. To this end, it is necessary to increase illness perception and knowledge, and reduce depression through cardiovascular disease prevention education for them, and the importance of oral health behaviors should also be emphasized.

Predicting intensive care need for COVID-19 patients using deep learning on chest radiography.

Image-based prediction of coronavirus disease 2019 (COVID-19) severity and resource needs can be an important means to address the COVID-19 pandemic. In this study, we propose an artificial intelligence/machine learning (AI/ML) COVID-19 prognosis method to predict patients' needs for intensive care by analyzing chest X-ray radiography (CXR) images using deep learning. The dataset consisted of 8357 CXR exams from 5046 COVID-19-positive patients as confirmed by reverse transcription polymerase chain reaction (RT-PCR) tests for the SARS-CoV-2 virus with a training/validation/test split of 64%/16%/20% on a by patient level. Our model involved a DenseNet121 network with a sequential transfer learning technique employed to train on a sequence of gradually more specific and complex tasks: (1)fine-tuning a model pretrained on ImageNet using a previously established CXR dataset with a broad spectrum of pathologies; (2)refining on another established dataset to detect pneumonia; and (3)fine-tuning using our in-house training/validation datasets to predict patients' needs for intensive care within 24, 48, 72, and 96h following the CXR exams. The classification performances were evaluated on our independent test set (CXR exams of 1048 patients) using the area under the receiver operating characteristic curve (AUC) as the figure of merit in the task of distinguishing between those COVID-19-positive patients who required intensive care following the imaging exam and those who did not. Our proposed AI/ML model achieved an AUC (95% confidence interval) of 0.78 (0.74, 0.81) when predicting the need for intensive care 24h in advance, and at least 0.76 (0.73, 0.80) for 48h or more in advance using predictions based on the AI prognostic marker derived from CXR images. This AI/ML prediction model for patients' needs for intensive care has the potential to support both clinical decision-making and resource management.

Development and derivation of bacteremia prediction model in patients with hepatobiliary infection

IntroductionHepatobiliary infections are common in the emergency department (ED), and the mortality rate for this condition is high. A suitable bacteremia prediction model would support prompt identification of bacteremia and appropriate management of hepatobiliary infections in the ED. Therefore, we attempted to produce a bacteremia prediction model with both internal and external validation for hepatobiliary infections in the ED. MethodsPatients with hepatobiliary infection were extracted from retrospective cohort databases of two tertiary hospitals from January 2018 to December 2019 and from January 2016 to December 2019, respectively. Independent risk factors were determined using multivariable logistic regression in a developmental cohort. We assigned a weighted value to predictive factors and developed a prediction model, which was validated both internally and externally. We assessed discrimination using the area under the receiver operating characteristics curve (AUC). ResultsOne hospital cohort of 1568 patients was randomly divided into a developmental group of 927 patients (60%) and an internal validation group of 641 patients (40%), and 736 people from the other hospital cohort were used for external validation. Bacteremia rates were 20.5%, 18.1%, and 23.1% in the developmental, internal, and external validation cohorts, respectively. Nine significant factors were used for predicting bacteremia, including age, three vital signs, and five laboratory tests. After applying our bacteremia prediction rule to the validation cohort, 56.5% and 53.8% of the internal and external validation groups were classified as low-risk bacteremia groups (bacteremia rates: 8.6% and 13.9%, respectively). The AUCs were 0.727 (95% confidence interval [CI]: 0.686–0.767), 0.730 (95% CI: 0.679–0.781), and 0.715 (95% CI: 0.672–0.758) for the developmental, internal, and external validation cohorts, respectively. The sensitivity and specificity for internal validation/external validation was 73.2%/67.6% and 63.0%/60.2%, respectively. ConclusionA bacteremia prediction model for hepatobiliary infection might be useful to predict the risk of bacteremia. It might also reduce the need for blood culture in low-risk patients.