Clinical Prediction Model Research Articles

Preoperative prediction of HER-2 expression status in breast cancer based on MRI radiomics model

Objective: This study aims to explore the predictive value of T2-weighted imaging (T2WI), apparent diffusion coefficient (ADC), and early-delayed phases enhanced magnetic resonance imaging (DCE-MRI) radiomics prediction model in determining human epidermal growth factor receptor 2 status in breast cancer. Methods: A retrospective study was conducted, involving 187 patients with confirmed breast cancer by postsurgical pathology at Zhenjiang First People's Hospital during January 2021 and May 2023. Immunohistochemistry or fluorescence in situ hybridization was used to determine the HER-2 status of these patients, with 48 cases classified as HER-2 positive and 139 cases as HER-2 negative. The training set was used to construct the prediction models and the validation set was used to verify the prediction models. Layers of T2WI, ADC, and early-delayed phase DCE-MRI images were used to delineate the volumeof interest and 960 radiomic features were extracted from each case using Pyradiomic. After screening and dimensionality reduction by intraclass correlation coefficient, Pearson correlation analysis, least absolute shrinkage, and selection operator, the radiomics labels were established. Logistic regression analysis was used to construct the T2WI radiomics model, ADC radiomics model, DCE-2 radiomics model, DCE-6 radiomics model, and the joint sequence radiomics model to predict the HER-2 expression status of breast cancer, respectively. Based on the clinical, pathological, and MRI image characteristics of patients, univariate and multivariate logistic regression analysis wasused to construct a clinicopathological MRI feature model. The radscore of every patient and the clinicopathological MRI features which were statistically significant after screening were used to construct a nomogram model. The receiver operating characteristic (ROC) curve was used to evaluate the predictive performance of each model and the decision curve analysis wasused to evaluate the clinical usefulness. Results: The T2WI, ADC, DCE-2, DCE-6, and joint sequence radiomics models, the clinicopathological MRI feature model, and the nomogram model were successfully constructed to predict the expression status of HER-2 in breast cancer. ROC analysis showed that in the training set and validation set, the areas under the curve (AUC) of the T2WI radiomics model were 0.797 and 0.760, of the ADC radiomics model were 0.776 and 0.634, of the DCE-2 radiomics model were 0.804 and 0.759, of the DCE-6 radiomics model were 0.869 and 0.798, of the combined sequence radiomics model were 0.908 and 0.847, of the clinicopathological MRI feature model were 0.703 and 0.693, and of the nomogram model were 0.938 and 0.859, respectively. In the training set, the combined sequence radiomics model outperformed the clinicopathological features model (P＜0.001). In the training and validation sets, the nomogram outperformed the clinicopathological features model (P＜0.05). In addition, the diagnostic performance of the nomogram was better than that of the four single-modality radiomics models in the training cohort (P＜0.05) and was better than that of DCE-2 and ADC models in the validation cohort (P＜0.05). Decision curve analysis indicated that the value of individualized prediction models was higher than clinical and pathological prediction models in clinical practice. The calibration curve showed that the multimodal radiomics model had a high consistency with the actual results in predicting HER-2 expression. Conclusions: T2WI, ADC and early-delayed phase DCE-MRI imaging histology models for HER-2 expression status in breast cancer are expected to provide a non-invasive virtual pathological basis for decision-making on preoperative neoadjuvant regimens in breast cancer.

#1763 Risk factors and clinical prediction models for osteoporosis in predialysis chronic kidney disease patients

Abstract Background and Aims The research concerning osteoporosis among individuals with predialysis chronic kidney disease (CKD) has been relatively neglected in existing literature. Furthermore, the risk factors for osteoporosis in this demographic have not been adequately investigated. The objective of this study was to explore the risk factors associated with osteoporosis in predialysis CKD patients and to construct a predictive model capable of estimating the likelihood of osteoporosis in these individuals. Method A retrospective study was conducted on 326 predialysis CKD patients. Relevant clinical examination results and measurements of bone mineral density (BMD) using dual-energy X-ray absorptiometry were collected. Binary logistic regression was utilized to explore the risk factors associated with osteoporosis and to develop predictive models. Results In male predialysis CKD patients, age represents a risk factor, while 25-(OH)-Vitamin D acts as a protective factor against osteoporosis. Among female patients, age, underweight status, NBAP, and advanced CKD (G5) are significant risk factors for osteoporosis, whereas PTH serves as a protective factor. In pre-dialysis CKD patients in stages CKD1-4, age, underweight status, and NBAP are identified as risk factors for osteoporosis. For G5 predialysis patients, advanced age and NBAP pose a high risk of osteoporosis, while 25-(OH)-Vitamin D and PTH exhibit protective effects. Additionally, nomogram models illustrating the risk of osteoporosis in pre-dialysis patients of different genders and renal function stages were created. Conclusion Age, underweight, severity of renal dysfunction, NBAP, PTH, and 25-(OH)-vitamin D are strongly associated with osteoporosis in predialysis CKD patients.

Anoikis-related gene signatures in colorectal cancer: implications for cell differentiation, immune infiltration, and prognostic prediction

Colorectal cancer (CRC) is a malignant tumor originating from epithelial cells of the colon or rectum, and its invasion and metastasis could be regulated by anoikis. However, the key genes and pathways regulating anoikis in CRC are still unclear and require further research. The single cell transcriptome dataset GSE221575 of GEO database was downloaded and applied to cell subpopulation type identification, intercellular communication, pseudo time cell trajectory analysis, and receptor ligand expression analysis of CRC. Meanwhile, the RNA transcriptome dataset of TCGA, the GSE39582, GSE17536, and GSE17537 datasets of GEO were downloaded and merged into one bulk transcriptome dataset. The differentially expressed genes (DEGs) related to anoikis were extracted from these data sets, and key marker genes were obtained after feature selection. A clinical prognosis prediction model was constructed based on the marker genes and the predictive effect was analyzed. Subsequently, gene pathway analysis, immune infiltration analysis, immunosuppressive point analysis, drug sensitivity analysis, and immunotherapy efficacy based on the key marker genes were conducted for the model. In this study, we used single cell datasets to determine the anoikis activity of cells and analyzed the DEGs of cells based on the score to identify the genes involved in anoikis and extracted DEGs related to the disease from the transcriptome dataset. After dimensionality reduction selection, 7 marker genes were obtained, including TIMP1, VEGFA, MYC, MSLN, EPHA2, ABHD2, and CD24. The prognostic risk model scoring system built by these 7 genes, along with patient clinical data (age, tumor stage, grade), were incorporated to create a nomogram, which predicted the 1-, 3-, and 5-years survival of CRC with accuracy of 0.818, 0.821, and 0.824. By using the scoring system, the CRC samples were divided into high/low anoikis-related prognosis risk groups, there are significant differences in immune infiltration, distribution of immune checkpoints, sensitivity to chemotherapy drugs, and efficacy of immunotherapy between these two risk groups. Anoikis genes participate in the differentiation of colorectal cancer tumor cells, promote tumor development, and could predict the prognosis of colorectal cancer.

Development and internal validation of a clinical prediction model for osteopenia in Chinese middle-aged and elderly men: a prospective cohort study

BackgroundEarly identification of patients at risk of osteopenia is an essential step in reducing the population at risk for fractures. We aimed to develop and validate a prediction model for osteopenia in Chinese middle-aged and elderly men that provides individualized risk estimates.MethodsIn this prospective cohort study, 1109 patients who attend regular physical examinations in the Second Medical Centre of Chinese PLA General Hospital were enrolled from 2015.03 to 2015.09. The baseline risk factors included dietary habits, exercise habits, medical histories and medication records. Osteopenia during follow-up were collected from Electronic Health Records (EHRs) and telephone interviews. Internal validation was conducted using bootstrapping to correct the optimism. The independent sample T-test analysis, Mann_Whitney U test, Chi-Square Test and multivariable Cox regression analysis were utilized to identify predictive factors for osteopenia in Chinese middle-aged and elderly men. A nomogram based on the seven variables was built for clinical use. Concordance index (C-index), receiver operating characteristic curve (ROC), decision curve analysis (DCA) and calibration curve were used to evaluate the efficiency of the nomogram.ResultsThe risk factors included in the prediction model were bone mineral density at left femoral neck (LNBMD), hemoglobin (Hb), serum albumin (ALB), postprandial blood glucose (PBG), fatty liver disease (FLD), smoking and tea consumption. The C-index for the risk nomogram was 0.773 in the prediction model, which presented good refinement. The AUC of the risk nomogram at different time points ranged from 0.785 to 0.817, exhibiting good predictive ability and performance. In addition, the DCA showed that the nomogram had a good clinical application value. The nomogram calibration curve indicated that the prediction model was consistent.ConclusionsOur study provides a novel nomogram and a web calculator that can effectively predict the 7-year incidence risk of osteopenia in Chinese middle-aged and elderly men. It is convenient for clinicians to prevent fragility fractures in the male population.

Clinical prediction model for MODY type diabetes mellitus in children

BACKGROUND: MODY (maturity-onset diabetes of the young) is a rare monogenic form of diabetes mellitus, the gold standard of diagnosis is mutations detection in the genes responsible for the development of this form diabetes. Genetic test is expensive and takes a lot of time. The diagnostic criteria for MODY are well known. The development of clinical decision support system (CDSS) which allows physicians based on clinical data to determine who should have molecular genetic testing is relevant.AIM: Provided a retrospective analysis of clinical data of the patients with T1DM and MODY, from 0 to 18 years old, regardless of the duration of the disease to develop the model. Based on clinical data, a feedforward neural network (NN) was implemented - a multilayer perceptron.MATERIALS AND METHODS: Development of the most effective algorithm for predicting MODY in children based on available clinical indicators of 1710 patients with diabetes under the age of 18 years using a multilayer feedforward neural network.RESULTS: The sample consisted of 1710 children under the age of 18 years with T1DM (78%) and MODY (22%) diabetes. For the final configuration of NS the following predictors were selected: gender, age at passport age, age at the diagnosis with DM, HbA1c, BMI SDS, family history of DM, treatment. The performance (quality) assessment of the NN was carried out on a test sample (the area under the ROC (receiver operating characteristics) curve reached 0.97). The positive predictive value of PCPR was achieved at a cut-off value of 0.40 (predicted probability of MODY diabetes 40%). At which the sensitivity was 98%, specificity 93%, PCR with prevalence correction was 78%, and PCR with prevalence correction was 99%, the overall accuracy of the model was 94%.Based on the NN model, a CDSS was developed to determine whether a patient has MODY diabetes, implemented as an application.CONCLUSION: The clinical prediction model MODY developed in this work based on the NN, uses the clinical characteristic available for each patient to determine the probability of the patient having MODY. The use of the developed model in clinical practice will assist in the selection of patients for diagnostic genetic testing for MODY, which will allow for the efficient allocation of healthcare resources, the selection of personalized treatment and patient monitoring.

A nomogram-based prediction model for dysphagia in patients with chronic obstructive pulmonary disease: A cross-sectional study.

To investigate the prevalence of dysphagia in patients with COPD, identify the risk factors for dysphagia, develop a visual clinical prediction model and quantitatively predict the probability of developing dysphagia. Patients with COPD are at high risk of dysphagia, which is strongly linked to the acute exacerbation of their condition. The use of effective tools to predict its risk may contribute to the early identification and treatment of dysphagia in patients with COPD. A cross-sectional design. From July 2021 to April 2023, we enrolled 405 patients with COPD for this study. The clinical prediction model was constructed according to the results of a univariate analysis and a logistic regression analysis, evaluated by discrimination, calibration and decision curve analysis and visualized by a nomogram. This study was reported using the TRIPOD checklist. In total, 405 patients with COPD experienced dysphagia with a prevalence of 59.01%. A visual prediction model was constructed based on age, whether combined with cerebrovascular disease, chronic pulmonary heart disease, acute exacerbation of COPD, home noninvasive positive pressure ventilation, dyspnoea level and xerostomia level. The model exhibited excellent discrimination at an AUC of .879. Calibration curve analysis indicated a good agreement between experimental and predicted values, and the decision curve analysis showed a high clinical utility. The model we devised may be used in clinical settings to predict the occurrence of dysphagia in patients with COPD at an early stage. The model can help nursing staff to calculate the risk probability of dysphagia in patients with COPD, formulate personalized preventive care measures for high-risk groups as soon as possible to achieve early prevention or delay of dysphagia and its related complications and improve the prognosis. No patient or public contribution.

Serrated polyposis syndrome: defining the epidemiology and predicting the risk of dysplasia.

Serrated polyposis syndrome is the most common polyposis syndrome that has neoplastic potential. However, the natural history, genetic basis, and risk of dysplasia and neoplasia of serrated polyposis syndrome are incompletely understood. The objective of this study is to define the epidemiology of serrated polyposis syndrome. Using this data, we aim to evaluate candidate variables for predicting the risk of dysplasia and neoplasia in sessile serrated lesions found in serrated polyposis syndrome patients. Finally, we aim to use this data to create and evaluate clinical prediction models for accuracy in predicting dysplastic sessile serrated lesions in serrated polyposis syndrome patients. This was a regional Australian single-centre retrospective cohort study. Data was prospectively collected data from the clinical record database of a regional Australian gastroenterology practice. All patients undergoing colonoscopy at Port Macquarie Gastroenterology between January 2015 and September 2021 were screened for this study. Collected data included patient demographic, endoscopic, and histopathological findings. Clinical and endoscopic multivariate logistic regression models were created to predict dysplastic sessile serrated lesions. Model performance was examined using the area under the receiver operating curve. In total 8401 patients underwent a colonoscopy procedure during the study period. Serrated polyposis syndrome was diagnosed in 247, representing a prevalence of 2.94% (mean age 67.15 years, 62.75% female). Logistic regression identified; older age at serrated polyposis syndrome diagnosis, a personal history of colorectal cancer, size of the largest sessile serrated lesions removed, and total sessile serrated lesions count as predictors of dysplastic sessile serrated lesions. The clinical and endoscopic model had an area under the receiver operating curve of 0.75. Serrated polyposis syndrome is more common than previously described. The clinical and endoscopic variables identified in logistic regression have acceptable accuracy in predicting the risk of dysplasia, however other populations need to be studied to achieve generalisability and improve model performance.

Risk Factor Analysis for Growth Arrest in Paediatric Physeal Fractures-A Prospective Study.

Background: Fractures through the physis account for 18-30% of all paediatric fractures, leading to growth arrest in up to 5.5% of cases. We have limited knowledge to predict which physeal fractures result in growth arrest and subsequent deformity or limb length discrepancy. The purpose of this study is to identify factors associated with physeal growth arrest to improve patient outcomes. Methods: This prospective cohort study was designed to develop a clinical prediction model for growth arrest after physeal injury. Patients ≤ 18 years old presenting within four weeks of injury were enrolled if they had open physes and sustained a physeal fracture of the humerus, radius, ulna, femur, tibia or fibula. Patients with prior history of same-site fracture or a condition known to alter bone growth or healing were excluded. Demographic data, potential prognostic indicators, and radiographic data were collected at baseline, during healing, and at one- and two-years post-injury. Results: A total of 332 patients had at least six months of follow-up or a diagnosis of growth arrest within six months of injury. In a comparison analysis, patients who developed growth arrest were more likely to be older (12.8 years vs. 9.4 years) and injured on the right side (53.0% vs. 45.7%). Initial displacement and angulation rates were higher in the growth arrest group (59.0% vs. 47.8% and 47.0% vs. 38.8%, respectively), but the amount of angulation was similar (27.0° vs. 28.4°). Rates of growth arrest were highest in distal femoral fractures (86%). Conclusions: The incidence of growth arrest in this patient population appears higher than the past literature reports at 30.1%. However, there may be variances in diagnostic criteria for growth arrest, and the true incidence may be lower. A number of patients were approaching skeletal maturity, and any growth arrest is likely to have less clinical significance in these cases. Further prospective long-term follow-up is required to determine risk factors, incidence, and true clinical impact of growth arrest when it does occur.

Machine learning for clinical outcome prediction in cerebrovascular and endovascular neurosurgery: systematic review and meta-analysis

BackgroundMachine learning (ML) may be superior to traditional methods for clinical outcome prediction. We sought to systematically review the literature on ML for clinical outcome prediction in cerebrovascular and endovascular...

Risk factors and clinical prediction models for prolonged mechanical ventilation after heart valve surgery

ObjectivesProlonged mechanical ventilation (PMV) is a common complication following cardiac surgery linked to unfavorable patient prognosis and increased mortality. This study aimed to search for the factors associated with the occurrence of PMV after valve surgery and to develop a risk prediction model.MethodsThe patient cohort was divided into two groups based on the presence or absence of PMV post-surgery. Comprehensive preoperative and intraoperative clinical data were collected. Univariate and multivariate logistic regression analyses were employed to identify risk factors contributing to the incidence of PMV. Based on the logistic regression results, a clinical nomogram was developed.ResultsThe study included 550 patients who underwent valve surgery, among whom 62 (11.27%) developed PMV. Multivariate logistic regression analysis revealed that age (odds ratio [OR] = 1.082, 95% confidence interval [CI] = 1.042–1.125; P < 0.000), current smokers (OR = 1.953, 95% CI = 1.007–3.787; P = 0.047), left atrial internal diameter index (OR = 1.04, 95% CI = 1.002–1.081; P = 0.041), red blood cell count (OR = 0.49, 95% CI = 0.275–0.876; P = 0.016), and aortic clamping time (OR = 1.031, 95% CI = 1.005–1.057; P < 0.017) independently influenced the occurrence of PMV. A nomogram was constructed based on these factors. In addition, a receiver operating characteristic (ROC) curve was plotted, with an area under the curve (AUC) of 0.782 and an accuracy of 0.884.ConclusionAge, current smokers, left atrial diameter index, red blood cell count, and aortic clamping time are independent risk factors for PMV in patients undergoing valve surgery. Furthermore, the nomogram based on these factors demonstrates the potential for predicting the risk of PMV in patients following valve surgery.

Machine Learning-Based Predictive Models for Patients with Venous Thromboembolism: A Systematic Review.

Venous thromboembolism (VTE) is a chronic disorder with a significant health and economic burden. Several VTE-specific clinical prediction models (CPMs) have been used to assist physicians in decision-making but have several limitations. This systematic review explores if machine learning (ML) can enhance CPMs by analyzing extensive patient data derived from electronic health records. We aimed to explore ML-CPMs' applications in VTE for risk stratification, outcome prediction, diagnosis, and treatment. Three databases were searched: PubMed, Google Scholar, and IEEE electronic library. Inclusion criteria focused on studies using structured data, excluding non-English publications, studies on non-humans, and certain data types such as natural language processing and image processing. Studies involving pregnant women, cancer patients, and children were also excluded. After excluding irrelevant studies, a total of 77 studies were included. Most studies report that ML-CPMs outperformed traditional CPMs in terms of receiver operating area under the curve in the four clinical domains that were explored. However, the majority of the studies were retrospective, monocentric, and lacked detailed model architecture description and external validation, which are essential for quality audit. This review identified research gaps and highlighted challenges related to standardized reporting, reproducibility, and model comparison. ML-CPMs show promise in improving risk assessment and individualized treatment recommendations in VTE. Apparently, there is an urgent need for standardized reporting and methodology for ML models, external validation, prospective and real-world data studies, as well as interventional studies to evaluate the impact of artificial intelligence in VTE.

Construction and validation of clinical prediction model of acupuncture and moxibustion for Bell's palsy

To establish and validate a clinical prediction model of acupuncture and moxibustion for Bell's palsy so as to provide a tool for predicting the effect of acupuncture and moxibustion on Bell's palsy. A total of 269 patients with Bell's palsy were collected from department of acupuncture, moxibustion and tuina, Shengli Oilfield Central Hospital, neurology department, Shenxian County Central Hospital and department of rehabilitation medicine, Dongying Municipal Hospital of TCM from June 2018 to June 2023. All of these cases were treated with acupuncture and moxibustion. Of them, 182 cases, from department of acupuncture, moxibustion and tuina, Shengli Oilfield Central Hospital and neurology department, Shenxian County Central Hospital, were randomized into a training group (128 cases) and an internal validation group (54 cases); 87 cases from department of rehabilitation medicine, Dongying Municipal Hospital of TCM were assigned to an external validation group. The clinical data of all of the cases were extracted from the electronic medical record information platform. Using SPSS25.0 and R4.2.3, through univariate and multivariate Logistic regression analysis, the independent factors influencing the effects of acupuncture and moxibustion on Bell's palsy were identified. By means of internal and external validations, the receiver operating characteristic curve (ROC), the goodness-of-fit curve (GFC) and the decision curve analysis (DCA) were plotted. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of the model were calculated; and its comprehensive performance was evaluated. The results of the multivariate Logistic regression analysis showed that the independent factors for the unsatisfactory effect on Bell's palsy were advanced age, severe symptoms before treatment, no use of steroids within 72 h of onset, and lack of acupuncture-moxibustion therapy during the acute phase or single acupuncture-moxibustion protocol (P<0.05, P<0.01). Based on these factors, nomogram model and online columnar plot prediction tool (https://bmuchen.shinyapps.io/dynnomapp/) were established. The area under the ROC curve of the model was 0.921 (95% CI: 0.877, 0.966), 0.876 (95% CI: 0.787, 0.966), and 0.846 (95% CI: 0.766, 0.926) in the training group, the internal validation group, and the external validation group, respectively, indicating good predictive value. The model showed a satisfactory calibration curve alignment. The decision threshold in the range of 0 to 0.8 provided clinical benefits for participants. The model exhibited the sensitivity from 65.9% to 88.0%, the specificity ranging from 77.3% to 90.7%, the accuracy from 77.8% to 85.9%, the positive predictive value from 83.3% to 90.1%, and the negative predictive value from 70.8% to 78.7%. The comprehensive evaluation indicated a satisfactory clinical application value of the model. The clinical prediction model of acupuncture and moxibustion for Bell's palsy is valuable in its practice and promotion to a certain extent. The predicted results are conductive to clinicians' judgement of the effect of acupuncture and moxibustion for this disease and making effective and high-quality clinical decisions, as well as formulating the optimal therapeutic regimen.

Development and validation of a clinical prediction model for glioma grade using machine learning.

Histopathological evaluation is currently the gold standard for grading gliomas; however, this technique is invasive. This study aimed to develop and validate a diagnostic prediction model for glioma by employing multiple machine learning algorithms to identify risk factors associated with high-grade glioma, facilitating the prediction of glioma grading. Data from 1114 eligible glioma patients were obtained from The Cancer Genome Atlas (TCGA) database, which was divided into a training set (n= 781) and a test set (n= 333). Fifty machine learning algorithms were employed, and the optimal algorithm was selected to construct a prediction model. The performance of the machine learning prediction model was compared to the clinical prediction model in terms of discrimination, calibration, and clinical validity to assess the performance of the prediction model. The area under the curve (AUC) values of the machine learning prediction models (training set: 0.870 vs. 0.740, test set: 0.863 vs. 0.718) were significantly improved from the clinical prediction models. Furthermore, significant improvement in discrimination was observed for the Integrated Discrimination Improvement (IDI) (training set: 0.230, test set: 0.270) and Net Reclassification Index (NRI) (training set: 0.170, test set: 0.170) from the clinical prognostic model. Both models showed a high goodness of fit and an increased net benefit. A strong prediction accuracy model can be developed using machine learning algorithms to screen for high-grade glioma risk predictors, which can serve as a non-invasive prediction tool for preoperative diagnostic grading of glioma.

Emotion-Induced Frontal Alpha Asymmetry as a Candidate Predictor of Relapse After Discontinuation of Antidepressant Medication

BackgroundOne in 3 patients relapse after antidepressant discontinuation. Thus, the prevention of relapse after achieving remission is an important component in the long-term management of major depressive disorder. However, no clinical or other predictors are established. Frontal reactivity to sad mood as measured by functional magnetic resonance imaging has been reported to relate to relapse independently of antidepressant discontinuation and is an interesting candidate predictor. MethodsPatients (n = 56) who had remitted from a depressive episode while taking antidepressants underwent electroencephalography (EEG) recording during a sad mood induction procedure prior to gradually discontinuing their medication. Relapse was assessed over a 6-month follow-up period. Thirty five healthy control participants were also tested. Current source density of the EEG power in the alpha band (8–13 Hz) was extracted and alpha asymmetry was computed by comparing the power across 2 hemispheres at frontal electrodes (F5 and F6). ResultsSad mood induction was robust across all groups. Reactivity of alpha asymmetry to sad mood did not distinguish healthy control participants from patients with remitted major depressive disorder on medication. However, the 14 (25%) patients who relapsed during the follow-up period after discontinuing medication showed significantly reduced reactivity in alpha asymmetry compared with patients who remained well. This EEG signal provided predictive power (69% out-of-sample balanced accuracy and a positive predictive value of 0.75). ConclusionsA simple EEG-based measure of emotional reactivity may have potential to contribute to clinical prediction models of antidepressant discontinuation. Given the very small sample size, this finding must be interpreted with caution and requires replication in a larger study.

Improved clinical data imputation via classical and quantum determinantal point processes

Imputing data is a critical issue for machine learning practitioners, including in the life sciences domain, where missing clinical data is a typical situation and the reliability of the imputation is of great importance. Currently, there is no canonical approach for imputation of clinical data and widely used algorithms introduce variance in the downstream classification. Here we propose novel imputation methods based on determinantal point processes (DPP) that enhance popular techniques such as the multivariate imputation by chained equations and MissForest. Their advantages are twofold: improving the quality of the imputed data demonstrated by increased accuracy of the downstream classification and providing deterministic and reliable imputations that remove the variance from the classification results. We experimentally demonstrate the advantages of our methods by performing extensive imputations on synthetic and real clinical data. We also perform quantum hardware experiments by applying the quantum circuits for DPP sampling since such quantum algorithms provide a computational advantage with respect to classical ones. We demonstrate competitive results with up to 10 qubits for small-scale imputation tasks on a state-of-the-art IBM quantum processor. Our classical and quantum methods improve the effectiveness and robustness of clinical data prediction modeling by providing better and more reliable data imputations. These improvements can add significant value in settings demanding high precision, such as in pharmaceutical drug trials where our approach can provide higher confidence in the predictions made.

Improved clinical data imputation via classical and quantum determinantal point processes

Imputing data is a critical issue for machine learning practitioners, including in the life sciences domain, where missing clinical data is a typical situation and the reliability of the imputation is of great importance. Currently, there is no canonical approach for imputation of clinical data and widely used algorithms introduce variance in the downstream classification. Here we propose novel imputation methods based on determinantal point processes (DPP) that enhance popular techniques such as the multivariate imputation by chained equations and MissForest. Their advantages are twofold: improving the quality of the imputed data demonstrated by increased accuracy of the downstream classification and providing deterministic and reliable imputations that remove the variance from the classification results. We experimentally demonstrate the advantages of our methods by performing extensive imputations on synthetic and real clinical data. We also perform quantum hardware experiments by applying the quantum circuits for DPP sampling since such quantum algorithms provide a computational advantage with respect to classical ones. We demonstrate competitive results with up to 10 qubits for small-scale imputation tasks on a state-of-the-art IBM quantum processor. Our classical and quantum methods improve the effectiveness and robustness of clinical data prediction modeling by providing better and more reliable data imputations. These improvements can add significant value in settings demanding high precision, such as in pharmaceutical drug trials where our approach can provide higher confidence in the predictions made.

Don't be misled: 3 misconceptions about external validation of clinical prediction models

Clinical prediction models provide risks of health outcomes that can inform patients and support medical decisions. However, most models never make it to actual implementation in practice. A commonly heard reason for this lack of implementation is that prediction models are often not externally validated. While we generally encourage external validation, we argue that an external validation is often neither sufficient nor required as an essential step before implementation. As such, any available external validation should not be perceived as a license for model implementation. We clarify this argument by discussing 3 common misconceptions about external validation. We argue that there is not one type of recommended validation design, not always a necessity for external validation, and sometimes a need for multiple external validations. The insights from this paper can help readers to consider, design, interpret, and appreciate external validation studies.

Risk factor analysis and model construction for fish bone foreign body in children's pharynx

ObjectivesThis study aimed to assess the risk factors for predicting the presence of fish bone foreign bodies and to develop a risk prediction model. MethodsData of 1405 children who underwent video-guided laryngoscope for suspected fish bone foreign body ingestion were retrospectively analyzed. Multi-factor logistic regression analyses were performed to analyze the risk factors for the presence of fish bone foreign body in patients, and a risk prediction model was established based on the results of the logistic regression analysis. ResultsThe results of the statistical analysis showed the presence of an ulcerated surface increased the risk of having a fishbone foreign body in the pharynx by approximately 55.36-fold (95 % confidence interval (CI): 15.78–194.24), followed by a clear chief complaint site, which increased the risk of having a fishbone foreign body in the pharynx by approximately 7.963-fold (95 % CI: 4.820–13.15), and a tingling sensation, which increased the risk of having a fishbone foreign body by approximately 7-fold (95 % CI: 3.483, 14.233). A clinical prediction model (nomogram) was developed and its validation was performed using receiver operating characteristic (ROC) curve analysis, in which an area under the curve (AUC) value of 0.808 indicated that the model had a great prediction capability. ConclusionThe predictive capability of a logistic regression model for the detection of fish bone foreign bodies following ingestion is significant. Clinicians can concentrate on monitoring these risk factors and implementing appropriate interventions to reduce the risks of patients presenting with fish bone foreign bodies.

Efficacy and Safety of Direct Oral Anticoagulants for Acute Treatment of Venous Thromboembolism in Older Adults: A Network Meta-Analysis of Randomised Controlled Trials.

This systematic review and network meta-analysis aimed to evaluate the efficacy and safety of direct oral anticoagulants (DOACs) in adults aged 75 and over undergoing acute venous thromboembolism (VTE) treatment. PubMed, Embase and the CENTRAL were searched up to 25 December 2023. The incidence of VTE recurrence and bleeding events was assessed. Employing a frequentist network meta-analysis approach, interventions not directly compared could be indirectly assessed through the 95% confidence interval (CI), enhancing the interpretability of the search results. The surface under the cumulative ranking curves (SUCRA) was utilized to generate the relative ranking probabilities for each group. Our study, analysing 6 randomised controlled trials with 3665 patients, compares direct oral anticoagulants (DOACs) and vitamin K antagonists (VKAs) in adults aged 75 and over with acute venous thromboembolism. Edoxaban reduces VTE recurrence risk compared with VKAs (risk ratio [RR] .50, 95% CI 0.27 - .95), while apixaban significantly decreases bleeding risk compared with VKAs (RR .23, 95% CI 0.08 - .69), edoxaban (RR .28, 95% CI 0.09 - .86) and rivaroxaban (RR .28, 95% CI 0.09 - .86). Despite low overall evidence quality, apixaban consistently ranks highest for both efficacy and safety. Findings underscore the nuanced efficacy-safety balance in this population, emphasizing cautious interpretation due to evidence limitations. Apixaban emerges as a favourable choice for acute VTE treatment in the elderly, displaying reduced bleeding risk compared to other treatments while maintaining comparable efficacy. Future studies should explore diverse anticoagulants efficacy and safety in older populations. Additionally, clinical prediction models tailored to geriatric cohorts are crucial for guiding treatment duration decisions.

Clinical Response Predictive Model for Omalizumab in Moderate-to-Severe Asthma Patients

Objective. Our study aimed to develop a predictive model for evaluating the clinical response of omalizumab treatment in moderate-to-severe asthma patients. Methods. This single-center, prospective study collected patients who meet the diagnostic criteria for moderate-to-severe bronchial asthma set by the National Asthma Prevention and Treatment Group in 2016 in the first hospital affiliated with Soochow University. Patients recruited were treated with omalizumab once per four weeks; at the beginning of each injection, blood eosinophils and the level of total serum IgE (IU/mL) were tested. After four injections of omalizumab, asthma control test (ACT), the 15-item Mini Asthma Quality of Life Questionnaire (MiniAQLQ), global treatment effectiveness (GETE), and lung function of all patients were evaluated in the 16th week. We used the selection operator method to build a logistic model and evaluated the clinical response of omalizumab in these patients. Results. This study included 108 moderate-to-severe patients (aged 39.86 ± 14.59 years). Eighty-nine patients finished treatment for 16 weeks, and 74 patients (83.1%) had an excellent or good response. The serum level of total IgE increased significantly after injection of omalizumab, while blood eosinophils count decreased significantly from baseline. Using the GETE as a clinical outcome, several clinical variables were significant predictors of clinical response. The corrected AUC and Brier scores were 0.872 and 0.111, which showed good discrimination. Significant variables included age, weight, family allergic history, acute exacerbations, the ratio of total serum IgE level at the 4th week to the baseline level, forced expiratory volume in one second/forced vital capacity (FEV1/FVC), and commodities of rhinitis. Using the improvement in maximal expiratory flow 25% of the measured value to the predicted value (MEF25%pre) as clinical outcome, the significant variables included weight, duration of asthma, use of oral corticosteroids (OCS), total serum IgE level at the 4th week, and history of rhinitis. Its corrected AUC and Brier scores were 0.674 and 0.225 after internal validation. Conclusion. Omalizumab treatment remarkably improved asthma control and pulmonary function in Chinese patients with moderate-to-severe asthma. The response prediction model we developed provides convenient approaches to help identify better clinical response patients to omalizumab treatment.