Visual deterioration outcomes following optic pathway glioma treatment: a 12-year single institution retrospective study.

OBJECTIVE: The surgical treatment of optic pathway gliomas (OPG) remains controversial, with visual outcomes often unpredictable. The present study explored surgical and clinical factors influencing visual acuity (VA) after OPG treatment and developed anatomical subtypes correlated with clinical symptoms. METHODS: Children with OPG who underwent initial partial tumor resection at Beijing Tiantan Hospital from January 2011 to December 2022 were retrospectively analyzed. Multivariate logistic regression and random forest analyses were performed to identify risk factors for post-treatment VA deterioration and a decision tree model was created based on significant factors. RESULTS: A total of 140 patients were enrolled. Multivariate logistic regression analysis identified surgical approach and initial VA as independent predictors of post-treatment VA deterioration (P < 0.05). Surgical approach, initial VA, and extent of tumor resection were the most significant factors for risk assessment and were included in the decision tree model, with surgical approach as the most important "root" node. The model demonstrated good predictive performance, with area under the curve values of 0.75 and 0.66 for the training and test datasets, respectively. A simple anatomical classification was developed, which revealed clinical characteristic differences among OPG types. Meanwhile, a correlation analysis of post-treatment visual deterioration was performed for each of the three anatomical types. CONCLUSION: This study offers a predictive model for visual outcomes following initial tumor-reduction surgery in OPG patients, which may help in visual outcomes risk stratification. Additionally, the anatomical classification effectively indicates OPG growth direction, offering potential insights into clinical symptoms.

Novel machine-learning prediction tools for overall survival of patients with chondrosarcoma: Based on recursive partitioning analysis.

BACKGROUND: Chondrosarcoma (CHS), a bone malignancy, poses a significant challenge due to its heterogeneous nature and resistance to conventional treatments. There is a clear need for advanced prognostic instruments that can integrate multiple prognostic factors to deliver personalized survival predictions for individual patients. This study aimed to develop a novel prediction tool based on recursive partitioning analysis (RPA) to improve the estimation of overall survival for patients with CHS. METHODS: Data from the Surveillance, Epidemiology, and End Results (SEER) database were analyzed, including demographic, clinical, and treatment details of patients diagnosed between 2000 and 2018. Using C5.0 algorithm, decision trees were created to predict survival probabilities at 12, 24, 60, and 120 months. The performance of the models was assessed through confusion scatter plot, accuracy rate, receiver operator characteristic (ROC) curve, and area under ROC curve (AUC). RESULTS: The study identified tumor histology, surgery, age, visceral (brain/liver/lung) metastasis, chemotherapy, tumor grade, and sex as critical predictors. Decision trees revealed distinct patterns for survival prediction at each time point. The models showed high accuracy (82.40%-89.09% in training group, and 82.16%-88.74% in test group) and discriminatory power (AUC: 0.806-0.894 in training group, and 0.808-0.882 in test group) in both training and testing datasets. An interactive web-based shiny APP (URL: https://yangxg1209.shinyapps.io/chondrosarcoma_survival_prediction/) was developed, simplifying the survival prediction process for clinicians. CONCLUSIONS: This study successfully employed RPA to develop a user-friendly tool for personalized survival predictions in CHS. The decision tree models demonstrated robust predictive capabilities, with the interactive application facilitating clinical decision-making. Future prospective studies are recommended to validate these findings and further refine the predictive model.

An intersectional perspective on the sociodemographic and clinical factors influencing the status of not in Education, Employment, or training (NEET) in patients with first-episode psychosis (FEP).

PURPOSE: High rates of Not in Education, Employment or Training (NEET) are seen in people with first episode of psychosis (FEP). Sociodemographic and clinical factors were reported to be associated with NEET status in FEP patients. This study follows Intersectionality to examine the independent and additive effects, and most importantly the intersections of sociodemographic and clinical variables concerning NEET status in FEP patients. It was hypothesized that NEET status in FEP patients would be described by the intersection between at least two predictor variables. METHODS: Secondary analyses with chi-square tests, multiple logistic regression and Chi-squared Automatic Interaction Detection (CHAID) analyses were performed on 440 participants with FEP. RESULTS: Chi-square tests indicated that patient socioeconomic status and negative symptom severity were significantly and independently associated with their NEET status. Multiple logistic regression suggested additive effects of age (odds ratio = 1.61), patient socioeconomic status (odds ratio = 1.55) and negative symptom severity (odds ratio = 1.75) in predicting patients' NEET status. CHAID detected an intersection between patients' negative symptom severity and socioeconomic status in shaping their NEET status. CONCLUSION: This study explored how the NEET status of patients with FEP was explained not only by the separate effects of negative symptom severity and socioeconomic status but also by the unique intersections of their clinical and social identities. Findings indicated that functional outcomes of patients appear co-constructed by the intersections of multiple identities. Crucial clinical implications of complementing care for negative symptom severity with vocational resources to improve functional outcomes of patients are discussed.

Cost-Effectiveness of Low-Dose Computed Tomography Screenings for Lung Cancer in High-Risk Populations: A Markov Model.

Background: Domestic and foreign studies on lung cancer have been oriented to the medical efficacy of low-dose computed tomography (LDCT), but there is a lack of studies on the costs, value and cost-effectiveness of the treatment. There is a scarcity of conclusive evidence regarding the cost-effectiveness of LDCT within the specific context of Taiwan. This study is designed to address this gap by conducting a comprehensive analysis of the cost-effectiveness of LDCT and chest X-ray (CXR) as screening methods for lung cancer. Methods: Markov decision model simulation was used to estimate the cost-effectiveness of biennial screening with LDCT and CXR based on a health provider perspective. Inputs are based on probabilities, health status utility (quality-adjusted life years (QALYs)), costs of lung cancer screening, diagnosis, and treatment from the literatures, and expert opinion. A total of 1,000 simulations and five cycles of Markov bootstrapping simulations were performed to compare the incremental cost-utility ratio (ICUR) of these two screening strategies. Probability and one-way sensitivity analyses were also performed. Results: The ICUR of early lung cancer screening compared LDCT to CXR is $-24,757.65/QALYs, and 100% of the probability agree to adopt it under a willingness-to-pay (WTP) threshold of the Taiwan gross domestic product (GDP) per capita ($35,513). The one-way sensitivity analysis also showed that ICUR depends heavily on recall rate. Based on the prevalence rate of 39.7 lung cancer cases per 100,000 people in 2020, it could be estimated that LDCT screening for high-risk populations could save $17,154,115. Conclusion: LDCT can detect more early lung cancers, reduce mortality and is cost-saving than CXR in a long-term simulation of Taiwan's healthcare system. This study provides valuable insights for healthcare decision-makers and suggests analyzing cost-effectiveness for additional variables in future research.

Mapping and assessment of groundwater pollution risks in the main aquifer of the Mostaganem plateau (Northwest Algeria): utilizing the novel vulnerability index and decision tree model.

Water plays a pivotal role in socio-economic development in Algeria. However, the overexploitations of groundwater resources, water scarcity, and the proliferation of pollution sources (including industrial and urban effluents, untreated landfills, and chemical fertilizers, etc.) have resulted in substantial groundwater contamination. Preserving water irrigation quality has thus become a primary priority, capturing the attention of both scientists and local authorities. The current study introduces an innovative method to mapping contamination risks, integrating vulnerability assessments, land use patterns (as a sources of pollution), and groundwater overexploitation (represented by the waterhole density) through the implementation of a decision tree model. The resulting risk map illustrates the probability of contamination occurrence in the substantial aquifer on the plateau of Mostaganem. An agricultural region characterized by the intensive nutrients and pesticides use, the significant presence of septic tanks, widespread illegal dumping, and a technical landfill not compliant with environmental standards. The critical situation in the region is exacerbated by excessive groundwater pumping surpassing the aquifer's natural replenishment capacity (with 115 boreholes and 6345 operational wells), especially in a semi-arid climate featuring limited water resources and frequent drought. Vulnerability was evaluated using the DRFTID method, a derivative of the DRASTIC model, considering parameters such as depth to groundwater, recharge, fracture density, slope, nature of the unsaturated zone, and the drainage density. All these parameters are combined with analyses of inter-parameter relationship effects. The results show a spatial distribution into three risk levels (low, medium, and high), with 31.5% designated as high risk, and 56% as medium risk. The validation of this mapping relies on the assessment of physicochemical analyses in samples collected between 2010 and 2020. The results indicate elevated groundwater contamination levels in samples. Chloride exceeded acceptable levels by 100%, nitrate by 71%, calcium by 50%, and sodium by 42%. These elevated concentrations impact electrical conductivity, resulting in highly mineralized water attributed to anthropogenic agricultural pollution and septic tank discharges. High-risk zones align with areas exhibiting elevated nitrate and chloride concentrations. This model, deemed satisfactory, significantly enhances the sustainable management of water resources and irrigated land across various areas. In the long term, it would be beneficial to refine "vulnerability and risk" models by integrating detailed data on land use, groundwater exploitation, and hydrogeological and hydrochemical characteristics. This approach could improve vulnerability accuracy and pollution risk maps, particularly through detailed local data availability. It is also crucial that public authorities support these initiatives by adapting them to local geographical and climatic specificities on a regional and national scale. Finally, these studies have the potential to foster sustainable development at different geographical levels.

Development and validation of diagnosis model for inflammatory bowel diseases based on a serologic biomarker panel: A decision tree model study.

BACKGROUND AND STUDY AIMS: Currently, an increasing amount of experimental data is available on newly discovered biomarkers in inflammatory bowel diseases (IBD), but the role of these biomarkers is often questionable due to their limited sensitivity. Therefore, this study aimed to build a diagnostic tool incorporating a panel of serum biomarkers into a computational algorithm to identify patients with IBD and differentiate those with Crohn's disease (CD) from those with ulcerative colitis (UC). PATIENTS AND METHODS: We studied sera from 192 CD patients, 118 UC patients, 60 non-IBD controls and 60 healthy controls. Indirect immunofluorescence (IIF) assays were utilized to determine several serum biomarkers previously associated with IBD, and the decision tree algorithm was used to construct the diagnosis model. Performances of models were evaluated by prediction accuracy, precision, AUC and Matthews's correlation coefficient (MCC). The "Inflammatory Bowel Disease Multi-omics Database (IBDMDB)" cohorts were used to validate the model as external validation set. RESULTS: The prediction rates were determined and compared for decision tree models after each data was developed using C5.0, C&RT, QUEST and CHAID. The C5.0 and CHAID algorithms, which ranked top for the prediction rate in the IBD vs. non-IBD model and the CD vs. UC model, respectively, were utilized for final pattern analysis. The final decision tree model achieved higher classification accuracy than the approach based on conservative marker combinations (sensitivity 75.0% vs. 79.5%, specificity 93.8% vs. 78.3% for differentiating IBD from non-IBD; and sensitivity 84.3% vs. 73.4%, specificity 92.5% vs. 54.9% for differentiating CD from UC, respectively). The model prediction consistency was 93% (28/30) in the external validation set. CONCLUSION: The decision-tree-based approach used in this study, based on serum biomarkers, has shown to be a valid and useful approach to identifying IBD and differentiating CD from UC.

Exploring the Cost-Effectiveness of Newborn Screening for Metachromatic Leukodystrophy (MLD) in the UK.

Metachromatic leukodystrophy (MLD) is a fatal inherited lysosomal storage disease that can be detected through newborn bloodspot screening. The feasibility of the screening assay and the clinical rationale for screening for MLD have been previously demonstrated, so the aim of this study is to determine whether the addition of screening for MLD to the routine newborn screening program in the UK is a cost-effective use of National Health Service (NHS) resources. A health economic analysis from the perspective of the NHS and Personal Social Services was developed based on a decision-tree framework for each MLD subtype using long-term outcomes derived from a previously presented partitioned survival and Markov economic model. Modelling inputs for parameters related to epidemiology, test characteristics, screening and treatment costs were based on data from three major UK specialist MLD hospitals, structured expert opinion and published literature. Lifetime costs and quality-adjusted life years (QALYs) were discounted at 1.5% to account for time preference. Uncertainty associated with the parameter inputs was explored using sensitivity analyses. This health economic analysis demonstrates that newborn screening for MLD is a cost-effective use of NHS resources using a willingness-to-pay threshold appropriate to the severity of the disease; and supports the inclusion of MLD into the routine newborn screening programme in the UK.

Factors associated with a better treatment efficacy among psoriasis patients: a study based on decision tree model and logistic regression in Shanghai, China.

BACKGROUND: Many effective therapies for psoriasis are being applied in clinical practice in recent years, however, some patients still can't achieve satisfied effect even with biologics. Therefore, it is crucial to identify factors associated with the treatment efficacy among psoriasis patients. This study aims to explore factors influencing the treatment efficacy of psoriasis patients based on decision tree model and logistic regression. METHODS: We implemented an observational study and recruited 512 psoriasis patients in Shanghai Skin Diseases Hospital from 2021 to 2022. We used face-to-face questionnaire interview and physical examination to collect data. Influencing factors of treatment efficacy were analyzed by using logistic regression, and decision tree model based on the CART algorithm. The receiver operator curve (ROC) was plotted for model evaluation and the statistical significance was set at P < 0.05. RESULTS: The 512 patients were predominately males (72.1%), with a median age of 47.5 years. In this study, 245 patients achieved ≥ 75% improvement in psoriasis area and severity index (PASI) score in week 8 and was identified as treatment success (47.9%). Logistic regression analysis showed that patients with senior high school and above, without psoriasis family history, without tobacco smoking and alcohol drinking had higher percentage of treatment success in patients with psoriasis. The final decision tree model contained four layers with a total of seventeen nodes. Nine classification rules were extracted and five factors associated with treatment efficacy were screened, which indicated tobacco smoking was the most critical variable for treatment efficacy prediction. Model evaluation by ROC showed that the area under curve (AUC) was 0.79 (95%CI: 0.75 ~ 0.83) both for logistic regression model (0.80 sensitivity and 0.69 specificity) and decision tree model (0.77 sensitivity and 0.73 specificity). CONCLUSION: Psoriasis patients with higher education, without tobacco smoking, alcohol drinking and psoriasis family history had better treatment efficacy. Decision tree model had similar predicting effect with the logistic regression model, but with higher feasibility due to the nature of simple, intuitive, and easy to understand.

The Associations and Causal Relationships of Ovarian Cancer - Construction of a Prediction Model.

Purpose: To explore the risk and protective factors for developing ovarian cancer and construct a risk prediction model. Methods: Information related to patients diagnosed with ovarian cancer on the electronic medical record data platform of three tertiary hospitals in Guangdong Province from May 2018 to September 2023 was collected as the case group. Patients with non-ovarian cancer who attended the clinic during the same period were included in the control group. Logistic regression analysis was used to screen the independent variables and explore the factors associated with the development of ovarian cancer. An ovarian cancer risk prediction model was constructed using a decision tree C4.5 algorithm. The ROC and calibration curves were plotted, and the model was validated. Results: Logistic regression analysis identified independent risk and protective factors for ovarian cancer. The sample size was divided into training and test sets in a ratio of 7:3 for model construction and validation. The AUC of the training and test sets of the decision tree model were 0.961 (95% CI:0.944-0.978) and 0.902 (95% CI:0.840-0.964), respectively, and the optimal cut-off values and their coordinates were 0.532 (0.091, 0.957), and 0.474 (0.159, 0.842) respectively. The accuracies of the training and test sets were 93.3% and 84.2%, respectively, and their sensitivities were 95.7% and 84.2%, respectively. Conclusion: The constructed ovarian cancer risk prediction model has good predictive ability, which is conducive to improving the efficiency of early warning of ovarian cancer in high-risk groups.

Novel decision tree models predict the overall survival of patients with submandibular gland cancer.

BACKGROUND: While the accurate prediction of the overall survival (OS) in patients with submandibular gland cancer (SGC) is paramount for informed therapeutic planning, the development of reliable survival prediction models has been hindered by the rarity of SGC cases. The purpose of this study is to identify key prognostic factors for OS in SGC patients using a large database and construct decision tree models to aid the prediction of survival probabilities in 12, 24, 60 and 120 months. MATERIALS AND METHODS: We performed a retrospective cohort study using the Surveillance, Epidemiology and End Result (SEER) program. Demographic and peri-operative predictor variables were identified. The outcome variables overall survival at 12-, 24-, 60, and 120 months. The C5.0 algorithm was utilized to establish the dichotomous decision tree models, with the depth of tree limited within 4 layers. To evaluate the performances of the novel models, the receiver operator characteristic (ROC) curves were generated, and the metrics such as accuracy rate, and area under ROC curve (AUC) were calculated. RESULTS: A total of 1,705, 1,666, 1,543, and 1,413 SGC patients with a follow up of 12, 24, 60 and 120 months and exact survival status were identified from the SEER database. Predictor variables of age, sex, surgery, radiation, chemotherapy, tumor histology, summary stage, metastasis to distant lymph node, and marital status exerted substantial influence on overall survival. Decision tree models were then developed, incorporating these vital prognostic indicators. Favorable consistency was presented between the predicted and actual survival statuses. For the training dataset, the accuracy rates for the 12-, 24-, 60- and 120-month survival models were 0.866, 0.767, 0.737 and 0.797. Correspondingly, the AUC values were 0.841, 0.756, 0.725, and 0.774 for the same time points. CONCLUSIONS: Based on the most important predictor variables identified using the large, SEER database, decision tree models were established that predict OS of SGC patients. The models offer a more exhaustive evaluation of mortality risk and may lead to more personalized treatment strategies.

An Exploratory Approach of Clinically Useful Biomarkers of Cvid by Logistic Regression.

Despite advancements in genetic and functional studies, the timely diagnosis of common variable immunodeficiency (CVID) remains a significant challenge. This exploratory study was designed to assess the diagnostic performance of a novel panel of biomarkers for CVID, incorporating the sum of κ+λ light chains, soluble B-cell maturation antigen (sBCMA) levels, switched memory B cells (smB) and the VISUAL score. Comparative analyses utilizing logistic regression were performed against established gold-standard tests, specifically antibody responses. Our research encompassed 88 subjects, comprising 27 CVID, 23 selective IgA deficiency (SIgAD), 20 secondary immunodeficiency (SID) patients and 18 healthy controls. We established the diagnostic accuracy of sBCMA and the sum κ+λ, achieving sensitivity (Se) and specificity (Spe) of 89% and 89%, and 90% and 99%, respectively. Importantly, sBCMA showed strong correlations with all evaluated biomarkers (sum κ+λ, smB cell and VISUAL), whereas the sum κ+λ was uniquely independent from smB cells or VISUAL, suggesting its additional diagnostic value. Through a multivariate tree decision model, specific antibody responses and the sum κ+λ emerged as independent, signature biomarkers for CVID, with the model showcasing an area under the curve (AUC) of 0.946, Se 0.85, and Spe 0.95. This tree-decision model promises to enhance diagnostic efficiency for CVID, underscoring the sum κ+λ as a superior CVID classifier and potential diagnostic criterion within the panel.

A decision tree-based study of pulmonary tuberculosis diagnosis-related groups.

BACKGROUND: Globally, pulmonary tuberculosis is a significant public health and social problem. OBJECTIVE: We investigated the factors influencing the hospitalization cost of patients with pulmonary tuberculosis and grouped cases based on a decision tree model to provide a reference for enhancing the management of diagnosis-related groups (DRGs) of this disease. METHODS: The data on the first page of the medical records of patients with the primary diagnosis of pulmonary tuberculosis were extracted from the designated tuberculosis hospital. The influencing factors of hospitalization cost were determined using the Wilcoxon rank sum test and multiple linear stepwise regression analysis, and the included cases were grouped using the chi-squared automated interaction test decision tree model, with these influential factors used as classification nodes. In addition, the included cases were grouped according to the ZJ-DRG grouping scheme piloted in Zhejiang Province, and the differences between the two grouping methods were compared. RESULTS: The length of hospital stay, respiratory failure, sex, and age were the determining factors of the hospitalization cost of patients with pulmonary tuberculosis, and these factors were incorporated into the decision tree model to form eight case combinations. The reduction in variance (RIV) using this grouping method was 60.60%, the heterogeneity between groups was high, the coefficients of variance ranged from 0.29 to 0.47, and the intra-group difference was small. The patients were also divided into four groups based on the ZJ-DRG grouping scheme piloted in Zhejiang Province. The RIV using this grouping method was 55.24, the differences between groups were acceptable, the coefficients of variance were 1.00, 0.61, 0.77, and 0.87, respectively, and the intra-group difference was significant. CONCLUSION: When the pulmonary tuberculosis cases were grouped according to the duration of hospital stay, respiratory failure, and age, the results were rather reasonable, providing a reference for DRG management and cost control of this disease.

Machine learning models for prediction of postoperative venous thromboembolism in gynecological malignant tumor patients.

AIM: To identify risk factors that associated with the occurrence of venous thromboembolism (VTE) within 30 days after hysterectomy among gynecological malignant tumor patients, and to explore the value of machine learning (ML) models in VTE occurrence prediction. METHODS: A total of 1087 patients between January 2019 and January 2022 with gynecological malignant tumors were included in this single-center retrospective study and were randomly divided into the training dataset (n = 870) and the test dataset (n = 217). Univariate logistic regression analysis was used to identify risk factors that associated with the occurrence of postoperative VTE in the training dataset. Machine learning models (including decision tree (DT) model and logistic regression (LR) model) to predict the occurrence of postoperative VTE were constructed and internally validated. RESULTS: The incidence of developing 30-day postoperative VTE was 6.0% (65/1087). Age, previous VTE, length of stay (LOS), tumor stage, operative time, surgical approach, lymphadenectomy (LND), intraoperative blood transfusion and gynecologic Caprini (G-Caprini) score were identified as risk factors for developing postoperative VTE in gynecological malignant tumor patients (p < 0.05). The AUCs of LR model and DT model for predicting VTE were 0.722 and 0.950, respectively. CONCLUSION: The ML models, especially the DT model, constructed in our study had excellent prediction value and shed light upon its further application in clinic practice.

Classification tree obtained by artificial intelligence for the prediction of heart failure after acute coronary syndromes. / Árboles de clasificación obtenidos mediante inteligencia artificial para la predicción de insuficiencia cardiaca tras el síndrome coronario agudo.

BACKGROUND: Coronary heart disease is the leading cause of heart failure (HF), and tools are needed to identify patients with a higher probability of developing HF after an acute coronary syndrome (ACS). Artificial intelligence (AI) has proven to be useful in identifying variables related to the development of cardiovascular complications. METHODS: We included all consecutive patients discharged after ACS in two Spanish centers between 2006 and 2017. Clinical data were collected and patients were followed up for a median of 53months. Decision tree models were created by the model-based recursive partitioning algorithm. RESULTS: The cohort consisted of 7,097 patients with a median follow-up of 53months (interquartile range: 18-77). The readmission rate for HF was 13.6% (964 patients). Eight relevant variables were identified to predict HF hospitalization time: HF at index hospitalization, diabetes, atrial fibrillation, glomerular filtration rate, age, Charlson index, hemoglobin, and left ventricular ejection fraction. The decision tree model provided 15 clinical risk patterns with significantly different HF readmission rates. CONCLUSIONS: The decision tree model, obtained by AI, identified 8 leading variables capable of predicting HF and generated 15 differentiated clinical patterns with respect to the probability of being hospitalized for HF. An electronic application was created and made available for free.

Decision model for durable clinical benefit from front- or late-line immunotherapy alone or with chemotherapy in non-small cell lung cancer.

BACKGROUND: Predictive biomarkers and models of immune checkpoint inhibitors (ICIs) have been extensively studied in non-small cell lung cancer (NSCLC). However, evidence for many biomarkers remains inconclusive, and the opaqueness of machine learning models hinders practicality. We aimed to provide compelling evidence for biomarkers and develop a transparent decision tree model. METHODS: We consolidated data from 3,288 ICI-treated patients with NSCLC across real-world multicenter, public cohorts and the Choice-01 trial (ClinicalTrials.gov: NCT03856411). Over 50 features were examined for predicting durable clinical benefits (DCBs) from ICIs. Noteworthy biomarkers were identified to establish a decision tree model. Additionally, we explored the tumor microenvironment and peripheral CD8+ programmed death-1 (PD-1)+ T cell receptor (TCR) profiles. FINDINGS: Multivariate logistic regression analysis identified tumor histology, PD-ligand 1 (PD-L1) expression, tumor mutational burden, line, and regimen of ICI treatment as significant factors. Mutation subtypes of EGFR, KRAS, KEAP1, STK11, and disruptive TP53 mutations were associated with DCB. The decision tree (DT10) model, using the ten clinicopathological and genomic markers, showed superior performance in predicting DCB in the training set (area under the curve [AUC] = 0.82) and consistently outperformed other models in test sets. DT10-predicted-DCB patients manifested longer survival, an enriched inflamed tumor immune phenotype (67%), and higher peripheral TCR diversity, whereas the DT10-predicted-NDB (non-durable benefit) group showed an enriched desert immune phenotype (86%) and higher peripheral TCR clonality. CONCLUSIONS: The model effectively predicted DCB after front-/subsequent-line ICI treatment, with or without chemotherapy, for squamous and non-squamous lung cancer, offering clinicians valuable insights into efficacy prediction using cost-effective variables. FUNDING: This study was supported by the National Key R&D Program of China.

Thyroglobulin measurement is the most powerful outcome predictor in differentiated thyroid cancer: a decision tree analysis in a European multicenter series.

OBJECTIVES: An accurate prognostic assessment is pivotal to adequately inform and individualize follow-up and management of patients with differentiated thyroid cancer (DTC). We aimed to develop a predictive model for recurrent disease in DTC patients treated by surgery and 131I by adopting a decision tree model. METHODS: Age, sex, histology, T stage, N stage, risk classes, remnant estimation, thyroid-stimulating hormone (TSH), thyroglobulin (Tg), administered 131I activities and post-therapy whole body scintigraphy (PT-WBS) were identified as potential predictors and put into regression algorithm (conditional inference tree, c-tree) to develop a risk stratification model for predicting persistent/recurrent disease over time. RESULTS: The PT-WBS pattern identified a partition of the population into two subgroups (PT-WBS positive or negative for distant metastases). Patients with distant metastases exhibited lower disease-free survival (either structural, DFS-SD, and biochemical, DFS-BD, disease) compared to those without metastases. Meanwhile, the latter were further stratified into three risk subgroups based on their Tg values. Notably, Tg values >63.1 ng/mL predicted a shorter survival time, with increased DFS-SD for Tg values <63.1 and <8.9 ng/mL, respectively. A comparable model was generated for biochemical disease (BD), albeit different DFS were predicted by slightly different Tg cutoff values (41.2 and 8.8 ng/mL) compared to DFS-SD. CONCLUSIONS: We developed a simple, accurate and reproducible decision tree model able to provide reliable information on the probability of structurally and/or biochemically persistent/relapsed DTC after a TTA. In turn, the provided information is highly relevant to refine the initial risk stratification, identify patients at higher risk of reduced structural and biochemical DFS, and modulate additional therapies and the relative follow-up.

Development of interpretable machine learning models to predict in-hospital prognosis of acute heart failure patients.

AIMS: In recent years, there has been remarkable development in machine learning (ML) models, showing a trend towards high prediction performance. ML models with high prediction performance often become structurally complex and are frequently perceived as black boxes, hindering intuitive interpretation of the prediction results. We aimed to develop ML models with high prediction performance, interpretability, and superior risk stratification to predict in-hospital mortality and worsening heart failure (WHF) in patients with acute heart failure (AHF). METHODS AND RESULTS: Based on the Kyoto Congestive Heart Failure registry, which enrolled 4056 patients with AHF, we developed prediction models for in-hospital mortality and WHF using information obtained on the first day of admission (demographics, physical examination, blood test results, etc.). After excluding 16 patients who died on the first or second day of admission, the original dataset (n = 4040) was split 4:1 into training (n = 3232) and test datasets (n = 808). Based on the training dataset, we developed three types of prediction models: (i) the classification and regression trees (CART) model; (ii) the random forest (RF) model; and (iii) the extreme gradient boosting (XGBoost) model. The performance of each model was evaluated using the test dataset, based on metrics including sensitivity, specificity, area under the receiver operating characteristic curve (AUC), Brier score, and calibration slope. For the complex structure of the XGBoost model, we performed SHapley Additive exPlanations (SHAP) analysis, classifying patients into interpretable clusters. In the original dataset, the proportion of females was 44.8% (1809/4040), and the average age was 77.9 ± 12.0. The in-hospital mortality rate was 6.3% (255/4040) and the WHF rate was 22.3% (900/4040) in the total study population. In the in-hospital mortality prediction, the AUC for the XGBoost model was 0.816 [95% confidence interval (CI): 0.815-0.818], surpassing the AUC values for the CART model (0.683, 95% CI: 0.680-0.685) and the RF model (0.755, 95% CI: 0.753-0.757). Similarly, in the WHF prediction, the AUC for the XGBoost model was 0.766 (95% CI: 0.765-0.768), outperforming the AUC values for the CART model (0.688, 95% CI: 0.686-0.689) and the RF model (0.713, 95% CI: 0.711-0.714). In the XGBoost model, interpretable clusters were formed, and the rates of in-hospital mortality and WHF were similar among each cluster in both the training and test datasets. CONCLUSIONS: The XGBoost models with SHAP analysis provide high prediction performance, interpretability, and reproducible risk stratification for in-hospital mortality and WHF for patients with AHF.

Risk factors for methotrexate resistance in low-risk gestational trophoblastic neoplasia patients (FIGO score 0-4).

The challenge of methotrexate (MTX) resistance among low-risk gestational trophoblastic neoplasia (GTN) patients has always been prominent. Despite the International Federation of Gynaecology and Obstetrics (FIGO) score of 0-4 patients comprising the majority of low-risk GTN patients, a comprehensive exploration of the prevalence and risk factors associated with MTX resistance has been limited. Therefore, we aimed to identify associated risk factors in GTN patients with a FIGO score of 0-4. Between January 2005 and December 2020, 310 low-risk GTN patients received primary MTX chemotherapy in two hospitals, with 265 having a FIGO score of 0-4. In the FIGO 0-4 subgroup, 94 (35.5%) were resistant to MTX chemotherapy, and 34 (12.8%) needed multi-agent chemotherapy. Clinicopathologic diagnosis of postmolar choriocarcinoma (OR = 17.18, 95% CI: 4.64-63.70, P < 0.001) and higher pretreatment human chorionic gonadotropin concentration on a logarithmic scale (log-hCG concentration) (OR = 18.11, 95% CI: 3.72-88.15, P < 0.001) were identified as independent risk factors associated with MTX resistance according to multivariable logistic regression. The decision tree model and regression model were developed to predict the risk of MTX resistance in GTN patients with a FIGO score of 0-4. Evaluation of model discrimination, calibration and net benefit revealed the superiority of the decision tree model, which comprised clinicopathologic diagnosis and pretreatment hCG concentration. The patients in the high- and medium-risk groups of the decision tree model had a higher probability of MTX resistance. This study represents the investigation into MTX resistance in GTN patients with a FIGO score of 0-4 and disclosed a remission rate of approximately 65% with MTX chemotherapy. Higher pretreatment hCG concentration and clinicopathologic diagnosis of postmolar choriocarcinoma were independent risk factors associated with resistance to MTX chemotherapy. The decision tree model demonstrated enhanced predictive capabilities regarding the risk of MTX resistance and can serve as a valuable tool to guide the clinical treatment decisions for GTN patients with a FIGO score of 0-4.

Value-based pricing of cognitive behavioral therapy for depression in primary care: an economic evaluation.

OBJECTIVES: Value-based pricing (VBP) determines product prices based on their perceived benefits. In healthcare, VBP prices medical technologies considering health outcomes and other relevant factors. This study applies VBP using economic evaluation to provider-patient communication, taking cognitive behavioral therapy (CBT) for adult primary care patients with depressive disorders as a case study. METHODS: A 12-week decision-tree model was developed from the German social health insurance system's perspective, comparing CBT against the standard of care. The influence of an extended time horizon on VBP was assessed using a theoretical model and long-term data spanning 46 months. RESULTS: Using a willingness-to-pay threshold of €88,000 per quality-adjusted life year gained, the base-case 50-minute compensation rate for CBT was €45. Assuming long-term effects of CBT significantly affected the value-based compensation, increasing it to €226. CONCLUSIONS: This study showcases the potential of applying VBP to CBT. However, significant price variability is highlighted, contingent upon assumptions regarding CBT's long-term impacts.

Improving Prediction of Survival for Extremely Premature Infants Born at 23 to 29 Weeks Gestational Age in the Neonatal Intensive Care Unit: Development and Evaluation of Machine Learning Models.

BACKGROUND: Infants born at extremely preterm gestational ages are typically admitted to the neonatal intensive care unit (NICU) after initial resuscitation. The subsequent hospital course can be highly variable, and despite counseling aided by available risk calculators, there are significant challenges with shared decision-making regarding life support and transition to end-of-life care. Improving predictive models can help providers and families navigate these unique challenges. OBJECTIVE: Machine learning methods have previously demonstrated added predictive value for determining intensive care unit outcomes, and their use allows consideration of a greater number of factors that potentially influence newborn outcomes, such as maternal characteristics. Machine learning-based models were analyzed for their ability to predict the survival of extremely preterm neonates at initial admission. METHODS: Maternal and newborn information was extracted from the health records of infants born between 23 and 29 weeks of gestation in the Medical Information Mart for Intensive Care III (MIMIC-III) critical care database. Applicable machine learning models predicting survival during the initial NICU admission were developed and compared. The same type of model was also examined using only features that would be available prepartum for the purpose of survival prediction prior to an anticipated preterm birth. Features most correlated with the predicted outcome were determined when possible for each model. RESULTS: Of included patients, 37 of 459 (8.1%) expired. The resulting random forest model showed higher predictive performance than the frequently used Score for Neonatal Acute Physiology With Perinatal Extension II (SNAPPE-II) NICU model when considering extremely preterm infants of very low birth weight. Several other machine learning models were found to have good performance but did not show a statistically significant difference from previously available models in this study. Feature importance varied by model, and those of greater importance included gestational age; birth weight; initial oxygenation level; elements of the APGAR (appearance, pulse, grimace, activity, and respiration) score; and amount of blood pressure support. Important prepartum features also included maternal age, steroid administration, and the presence of pregnancy complications. CONCLUSIONS: Machine learning methods have the potential to provide robust prediction of survival in the context of extremely preterm births and allow for consideration of additional factors such as maternal clinical and socioeconomic information. Evaluation of larger, more diverse data sets may provide additional clarity on comparative performance.