Belz and colleagues present GREAT, a seven-item clinical instrument for predicting electroconvulsive therapy response in unipolar depression, with promising discriminatory validity (AUC 0.841). We identify three methodological gaps – absent calibration, limited sample representativeness and unquantified incremental value – that must be addressed before GREAT can guide individual clinical decisions.

Slotted blade technology is a passive flow control strategy that can effectively suppress the boundary layer separation within compressors. To reduce the iteration time of the traditional Design-Experiment-Design method, this study innovatively proposes a fast and universal three-dimensional design method for the slotted blade technology, enabling slot modeling completion within 1 s. Furthermore, combined with machine learning (ML), the mapping relationships between eight design parameters and two key aerodynamic performances – compressor design point efficiency ($\eta$DE) and stator total pressure recovery coefficient at the near-stall point ($\sigma ^{*}_{NS}$) – were pioneeringly established. In this study, the prediction performances of six models were compared: one-dimensional convolutional neural network (1D-CNN), random forest (RF), support vector regression (SVR), Gaussian process regression (GPR), multi-layer perceptron (MLP) and long short-term memory network (LSTM). The results indicate that 1D-CNN achieves the highest prediction accuracy: for the $\eta$DE, the mean absolute error (MAE) and coefficient of determination (R2) are 0.041 and 0.987, respectively; for the $\sigma ^{*}_{NS}$, the MAE and R² are 0.479 × 10−3 and 0.955, respectively. Notably, the computational time of the 1D-CNN model is 99.11% less than that of the computational fluid dynamics (CFD). The Shapley Additive exPlanations (SHAP) method was employed to reveal the effects of design parameters on the compressor aerodynamic performance. Notably, the slot outlet axial position (Zout) exerts the most significant influence on the $\eta$DE, while the slot outlet radial position close to the casing (R1_out) has the strongest impact on the $\sigma ^{*}_{NS}$. This study provides theoretical support and valuable references for the intelligent design of slotted blade technology.

Adequate protein intake is crucial for preventing frailty and sarcopenia in older adults, but conventional dietary assessments are time-consuming. Developing a rapid screening tool for identifying those at risk of low protein intake is essential; however, no such tool exists for Asian populations. This study developed a prediction model to identify older adults in Japan at risk of low protein intake using a simple food intake questionnaire. The model was developed using data from 4,085 older adults aged ≥65 years from the 2013 and 2017 National Health and Nutrition Surveys. Protein intake was assessed using a one-day dietary record with a semi-weighted method. A multivariable logistic regression model was developed to predict low protein intake (<1.0 g/kg adjusted body weight/day), incorporating sex, age, adjusted body weight, and food intake frequency of nine food groups. Model performance was evaluated using the area under the receiver operator characteristic (ROC) curve and a calibration plot, both adjusted for optimism through bootstrap resampling. Participants had a mean age of 74.1 years (standard deviation = 6.6), and 55.5% of all participants were female. The prevalence of low protein intake was 21.8%. Internal validation showed a bootstrap optimism-corrected ROC area of 0.695 and a calibration slope of 0.921. We developed a 12-item prediction model for identifying older adults at risk of low protein intake. The model provides a practical tool to identify older adults at high risk of low protein intake and is expected to be used in public health settings.

A predictive column chart was developed to assess the risk of primary liver cancer (PLC) in hepatitis B patients. Data from 107 PLC patients and 107 controls were used as the training set, with 92 patients as the validation set. An additional 446 patients from other hospitals, including 15 with PLC, formed the external validation group. Multivariate logistic regression identified gender, BMI, alcohol consumption, diabetes, family history of liver cancer, cirrhosis, and HBV DNA load as independent risk factors. The model showed strong discrimination with AUCs of 0.882 and 0.859 in the training and validation sets, respectively, and good calibration (Hosmer–Lemeshow χ² = 2.648, P = 0.954; χ² = 4.117, P = 0.846). Decision curve analysis (DCA) confirmed clinical benefit within a risk threshold of 0.07–0.95. In the external validation group, the model maintained discrimination (AUC = 0.863) and calibration (Hosmer–Lemeshow χ² = 7.999, P = 0.434), with DCA showing net benefit across 0.14–0.95. These results indicate the column chart is a reliable tool for PLC risk prediction in hepatitis B patients.

To enhance the capacity for early and effective management of genital tract infections at primary and secondary levels of the healthcare system, we developed a prediction model, validated internally to help predict individual risk of self-reported genital tract infections (sGTIs) at the community level in Ghana. The study involved 32973 men and women aged 15–49 years from three rounds of the Ghana Demographic Health Survey, from 2003 to 2014. The outcomes were sGTIs. We applied the least absolute shrinkage and selection operator (LASSO) penalized regression with a 10-fold cross-validation model to 11 predictors based on prior review of the literature. The bootstrapping technique was also employed as a sensitivity analysis to produce a robust model. We further employed discriminant and calibration analyses to evaluate the performance of the model. Statistical significance was set at P-value <0.05. The mean±standard deviation age was 29.1±9.7 years with female preponderance (60.7%). The prevalence of sGTIs within the period was 11.2% (95% CI = 4.5–17.8) and it ranged from 5.4% (95% CI = 4.8–5.86) in 2003 to 17.5% (95% CI = 16.4–18.7) in 2014. The LASSO regression model retained all 11 predictors. The model’s ability to discriminate between those with sGTIs and those without sGTIs was approximately 73.50% (95% CI = 72.50–74.26) from the area under the curve with bootstrapping technique. There was no evidence of miscalibration from the calibration belt plot with bootstrapping (test statistic = 17.30; P-value = 0.060). The model performance was judged to be good and acceptable. In the absence of clinical measurement, this prediction tool can be used to identify individuals aged 15–49 years with a high risk of sGTIs at the community level in Ghana. Frontline healthcare staff can use this tool for screening and early detection. We, therefore, propose external validation of the model to confirm its generalizability and reliability in different population.

The objective was to develop equations to predict carcass weight (CW), use CW to predict empty body weight (EBW); and carcass gain (CG) to predict empty body weight gain (EBWG) and retained energy (RE) in hair sheep. To generate the prediction models, a data set was composed of individual measurements from 569 sheep encompassing intact males (n = 416), castrated males (n = 51), and females (n = 102). Validation analyses were performed by using the Model Evaluation System (MES). The prediction equations for CW, EBW, and EBWG were not influenced by sex class (P > 0.05), and the following equations were generated, respectively: CW (kg) = − 0.234 (±1.1358) + 0.485 (±0.0387) × FBW; EBW (kg) = 1.367 (±0.5472) + 1.681 (±0.0210) × CW and EBWG (kg) = 0.004 (±0.0026) + 1.679 (±0.0758) × CG. There was an effect of sex class on the intercept (P = 0.0013) of the relationship between RE and CG: RE (MJ/day) = 1.448 (±0.0657) × EBW0.75 × CG0.797 (±0.0399); RE (MJ/day) = 1.522 (±0.0699) × EBW0.75 × CG0.797 (±0.0399) and RE (MJ/day) = 1.827 (±0.0739) × EBW0.75 × CG0.797 (±0.0399) for intact males, castrated males and females, respectively. This study highlights the importance of incorporating carcass information into EBW, EBWG, and RE predictions. Replacing empty body weight gain with carcass gain might be a suitable alternative to estimate the retained energy of hair sheep. In addition, the generated equations will provide support for meat production systems in carcass weight prediction.

To develop a machine learning model and nomogram to predict the probability of persistent virus shedding (PVS) in hospitalized patients with coronavirus disease 2019 (COVID-19), the clinical symptoms and signs, laboratory parameters, cytokines, and immune cell data of 429 patients with nonsevere COVID-19 were retrospectively reviewed. Two models were developed using the Akaike information criterion (AIC). The performance of these two models was analyzed and compared by the receiver operating characteristic (ROC) curve, calibration curve, net reclassification index (NRI), and integrated discrimination improvement (IDI). The final model included the following independent predictors of PVS: sex, C-reactive protein (CRP) level, interleukin-6 (IL-6) level, the neutrophil-lymphocyte ratio (NLR), monocyte count (MC), albumin (ALB) level, and serum potassium level. The model performed well in both the internal validation (corrected C-statistic = 0.748, corrected Brier score = 0.201) and external validation datasets (corrected C-statistic = 0.793, corrected Brier score = 0.190). The internal calibration was very good (corrected slope = 0.910). The model developed in this study showed high discriminant performance in predicting PVS in nonsevere COVID-19 patients. Because of the availability and accessibility of the model, the nomogram designed in this study could provide a useful prognostic tool for clinicians and medical decision-makers.

Predicting the need for hospitalisation of patients with coronavirus disease 2019 (COVID-19) is important for preventing healthcare disruptions. This observational study aimed to use the COVID-19 Registry Japan (COVIREGI-JP) to develop a simple scoring system to predict respiratory failure due to COVID-19 using only underlying diseases and symptoms. A total of 6873 patients with COVID-19 admitted to Japanese medical institutions between 1 June 2020 and 2 December 2020 were included and divided into derivation and validation cohorts according to the date of admission. We used multivariable logistic regression analysis to create a simple risk score model, with respiratory failure as the outcome for young (18–39 years), middle-aged (40–64 years) and older (≥65 years) groups, using sex, age, body mass index, medical history and symptoms. The models selected for each age group were quite different. Areas under the receiver operating characteristic curves for the simple risk score model were 0.87, 0.79 and 0.80 for young, middle-aged and elderly derivation cohorts, and 0.81, 0.80 and 0.67 in the validation cohorts. Calibration of the model was good. The simple scoring system may be useful in the appropriate allocation of medical resources during the COVID-19 pandemic.

Numerous predictive models for the risk of type 2 diabetes mellitus (T2DM) exist, but a minority of them has implemented nutrition data so far, even though the significant effect of nutrition on the pathogenesis, prevention and management of T2DM has been established. Thus, in the present study, we aimed to build a predictive model for the risk of T2DM that incorporates nutrition data and calculates its predictive performance. We analysed cross-sectional data from 1591 individuals from the population-based Cooperative Health Research in the Region of Augsburg (KORA) FF4 study (2013–14) and used a bootstrap enhanced elastic net penalised multivariate regression method in order to build our predictive model and select among 193 food intake variables. After selecting the significant predictor variables, we built a logistic regression model with these variables as predictors and T2DM status as the outcome. The values of area under the receiver operating characteristic (ROC) curve, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy of our predictive model were calculated. Eleven out of the 193 food intake variables were selected for inclusion in our model, which yielded a value of area under the ROC curve of 0⋅79 and a maximum PPV, NPV and accuracy of 0⋅37, 0⋅98 and 0⋅91, respectively. The present results suggest that nutrition data should be implemented in predictive models to predict the risk of T2DM, since they improve their performance and they are easy to assess.

Epidemic forecasting provides an opportunity to predict geographic disease spread and counts when an outbreak occurs and plays a key role in preventing or controlling their adverse impact. However, conventional prediction models based on complex mathematical modelling rely on the estimation of model parameters, which yields unreliable and unsustainable results. Herein, we proposed a simple model for predicting the epidemic transmission dynamics based on nonlinear regression of the epidemic growth rate and iterative methods, which is applicable to the progression of the COVID-19 outbreak under the strict control measures of the Chinese government. Our model yields reliable and accurate results as confirmed by the available data: we predicted that the total number of infections in mainland China would be 91 253, and the maximum number of beds required for hospitalised patients would be 62 794. We inferred that the inflection point (when the growth rate turns from positive to negative) of the epidemic across China would be mid-February, and the end of the epidemic would be in late March. This model is expected to contribute to resource allocation and planning in the health sector while providing a theoretical basis for governments to respond to future global health crises or epidemics.

Most of the existing prediction models for COVID-19 lack validation, are inadequately reported or are at high risk of bias, a reason which has led to discourage their use. Few existing models have the potential to be extensively used by healthcare providers in low-resource settings since many require laboratory and imaging predictors. Therefore, we sought to develop and validate a multivariable prediction model of death in Mexican patients with COVID-19, by using demographic and patient history predictors. We conducted a national retrospective cohort study in two different sets of patients from the Mexican COVID-19 Epidemiologic Surveillance Study. Patients with a positive reverse transcription-polymerase chain reaction for SARS-CoV-2 and complete unduplicated data were eligible. In total, 83 779 patients were included to develop the scoring system through a multivariable Cox regression model; 100 000, to validate the model. Eight predictors (age, sex, diabetes, chronic obstructive pulmonary disease, immunosuppression, hypertension, obesity and chronic kidney disease) were included in the scoring system called PH-Covid19 (range of values: −2 to 25 points). The predictive model has a discrimination of death of 0.8 (95% confidence interval (CI) 0.796–0.804). The PH-Covid19 scoring system was developed and validated in Mexican patients to aid clinicians to stratify patients with COVID-19 at risk of fatal outcomes, allowing for better and efficient use of resources.