Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration

External Validation of the International Prognosis Prediction Model of IgA Nephropathy

BACKGROUND

The International IgA Nephropathy (IgAN) Network developed and validated two prognostic prediction models for IgAN, one incorporating a race parameter. These models could anticipate the risk of a 50% reduction in estimated glomerular filtration rate (eGFR) or progression to end-stage renal disease (ESRD) subsequent to an IgAN diagnosis via renal biopsy. This investigation aimed to validate the International IgA Nephropathy Prediction Tool (IIgANPT) within a contemporary Chinese cohort.

METHODS

Within this study,185 patients diagnosed with IgAN via renal biopsy at the Center for Kidney Disease, Second Affiliated Hospital of Nanjing Medical University, between January 2012 and December 2021, were encompassed. Each patient's risk of progression was assessed utilizing the IIgANPT formula. The primary outcome, a 50% decline in eGFR or progression to ESRD, was examined. Two predictive models, one inclusive and the other exclusive of a race parameter, underwent evaluation via receiver-operating characteristic (ROC) curves, subgroup survival analyses, calibration plots, and decision curve analyses.

RESULTS

The median follow-up duration within our cohort spanned 5.1 years, during which 18 patients encountered the primary outcome. The subgroup survival curves exhibited distinct separations, and the comparison of clinical and histological characteristics among the risk subgroups revealed significant differences. Both models demonstrated outstanding discrimination, evidenced by the areas under the ROC curve at five years: 0.882 and 0.878. Whether incorporating the race parameter or not, both prediction models exhibited acceptable calibration. Decision curve analysis affirmed the favorable clinical utility of both models.

CONCLUSIONS

Both prognostic risk evaluation models for IgAN exhibited remarkable discrimination, sound calibration, and acceptable clinical utility.

External validation of the prediction model of intradialytic hypotension: a multicenter prospective cohort study

OBJECTIVE

Intradialytic hypotension (IDH) is a common and serious complication in patients with Maintenance Hemodialysis (MHD). The purpose of this study is to externally verify three IDH risk prediction models recently developed by Ma et al. and recalibrate, update and present the optimal model to improve the accuracy and applicability of the model in clinical environment.

METHODS

A multicenter prospective cohort study of patients from 11 hemodialysis centers in Sichuan Province, China, was conducted using convenience sampling from March 2022 to July 2022, with a follow-up period of 1 month. Model performance was assessed by: (1) Discrimination: Evaluated through the computation of the Area Under Curve (AUC) and its corresponding 95% confidence intervals. (2) Calibration: scrutinized through visual inspection of the calibration plot and utilization of the Brier score. (3) The incremental value of risk prediction and the utility of updating the model were gauged using NRI (Net Reclassification Improvement) and IDI (Integrated Discrimination Improvement). Decision Curve Analysis (DCA) was employed to evaluate the clinical benefit of updating the model.

RESULTS

The final cohort comprised 2235 individuals undergoing maintenance hemodialysis, exhibiting a 14.6% occurrence rate of IDH. The externally validated Area Under the Curve (AUC) values for the three original prediction models were 0.746 (95% CI: 0.718 to 0.775), 0.709 (95% CI: 0.679 to 0.739), and 0.735 (95% CI: 0.706 to 0.764) respectively. Conversely, the AUC value for the recalibrated and updated columnar plot model reached 0.817 (95% CI: 0.791 to 0.842), accompanied by a Brier score of 0.081. Furthermore, Decision Curve Analysis (DCA) exhibited a net benefit within the threshold probability range of 15.2% to 87.1%.

CONCLUSION

Externally validated, recalibrated, updated, and presented IDH prediction models may serve as a valuable instrument for evaluating IDH risk in clinical practice. Furthermore, they hold the potential to guide clinical providers in discerning individuals at risk and facilitating judicious clinical intervention decisions.

A Nomogram Based on Conventional Ultrasound Radiomics for Differentiating Between Radial Scar and Invasive Ductal Carcinoma of the Breast

ABSTRACT

We aimed to develop and validate a nomogram based on conventional ultrasound (CUS) radiomics model to differentiate radial scar (RS) from invasive ductal carcinoma (IDC) of the breast. In total, 208 patients with histopathologically diagnosed RS or IDC of the breast were enrolled. They were randomly divided in a 7:3 ratio into a training cohort (n = 145) and a validation cohort (n = 63). Overall, 1316 radiomics features were extracted from CUS images. Then a radiomics score was constructed by filtering unstable features and using the maximum relevance minimum redundancy algorithm and the least absolute shrinkage and selection operator logistic regression algorithm. Two models were developed using data from the training cohort: one using clinical and CUS characteristics (Clin + CUS model) and one using clinical information, CUS characteristics, and the radiomics score (radiomics model). The usefulness of nomogram was assessed based on their differentiating ability and clinical utility. Nine features from CUS images were used to build the radiomics score. The radiomics nomogram showed a favorable predictive value for differentiating RS from IDC, with areas under the curve of 0.953 and 0.922 for the training and validation cohorts, respectively. Decision curve analysis indicated that this model outperformed the Clin + CUS model and the radiomics score in terms of clinical usefulness. The results of this study may provide a novel method for noninvasively distinguish RS from IDC.

Development, evaluation and validation of machine learning models to predict hospitalizations of patients with coronary artery disease within the next 12 months

BACKGROUND

Improved survival of patients after acute coronary syndromes, population growth, and overall life expectancy rise have led to a significant increase in the proportion of patients with stable coronary artery disease (CAD), creating a significant load on the entire healthcare system. The disease often progresses with the development of many complications while significantly increasing the likelihood of hospitalization. Developing and applying a machine learning model for predicting hospitalizations of patients with CAD to an inpatient medical facility will allow for close monitoring of high-risk patients, early preventive interventions, and optimized medical care.

AIMS

Development and external validation of personalized models for predicting the preventable hospitalizations of patients with stable CAD and its complications using ML algorithms and data of real-world clinical practice.

METHODS

135,873 depersonalized electronic health records of 49,103 patients with stable CAD were included in the study. Anthropometric measurements, physical examination results, laboratory, instrumental, anamnestic, and socio-demographic data, widely used in routine medical practice, were considered as potential predictors, a total of 73 features. Logistic regression, decision tree-based methods including gradient boosting (AdaBoost, LightGBM, XGBoost, CatBoost) and bagging (RandomForest and ExtraTrees), discriminant analysis (LinearDiscriminant, QuadraticDiscriminant), and naive Bayes classifier were compared. External validation was performed on the data of a separate region.

RESULTS

The best results and stability to external validation data were shown by the CatBoost model with an AUC of 0.875 (95% CI 0.865-0.885) for the internal testing and 0.872 (95% CI 0.856-0.886) for the external validation. The best model showed good performance evaluated through AUROC, Brier score and standardized net benefit (for the target NPV threshold) for the validation dataset that was only slightly similar to the train data.

CONCLUSION

The metrics of the best model were superior to previously published studies. The results of external validation demonstrated the relative stability of the model to new data from another region that confirms the possibility of the model's application in real clinical practice.

Contesting the conventional wisdom of periodontal risk assessment

Over the years, several reviews of periodontal risk assessment tools have been published. However, major misunderstandings still prevail in repeated attempts to use these tools for prognostic risk prediction. Here we review the principles of risk prediction and discuss the value and the challenges of using prediction models in periodontology. Most periodontal risk prediction models have not been properly developed according to guidance given for the risk prediction model development. This shortcoming has led to several problems, including the creation of arbitrary risk scores. These scores are often labelled as 'high risk' without explicit boundaries or thresholds for the underlying continuous risk estimates of patient-important outcomes. Moreover, it is apparent that prediction models are often misinterpreted as causal models by clinicians and researchers although they cannot be used as such. Additional challenges like the critical assessment of transportability and applicability of these prediction models, as well as their impact on clinical practice and patient outcomes, are not considered in the literature. Nevertheless, these instruments are promoted with claims regarding their ability to deliver more individualized and precise periodontitis treatment and prevention, purportedly resulting in improved patient outcomes. However, people with or without periodontitis deserve proper information about their risk of developing patient-important outcomes such as tooth loss or pain. The primary objective of disseminating such information should not be to emphasize assumed treatment efficacy, hype individualization of care, or promote business interests. Instead, the focus should be on providing individuals with locally validated and regularly updated predictions of specific risks based on readily accessible and valid key predictors (e.g. age and smoking).

Development and internal validation of a clinical risk score to predict incident renal and pulmonary tumours in people with tuberous sclerosis complex

OBJECTIVE

This study aims to develop and internally validate a clinical risk score to predict incident renal angiomyolipoma (AML) and pulmonary lymphangioleiomyomatosis (LAM) in people with tuberous sclerosis complex (TSC).

STUDY DESIGN

Data from 2420 participants in the TSC Alliance Natural History Database were leveraged for these analyses. Logistic regression was used to predict AML and LAM development using 10 early-onset clinical manifestations of TSC as potential predictors, in addition to sex and genetic mutation. For our models, we divided AML into three separate outcomes: presence or absence of AML, unilateral or bilateral and whether any are ≥3 cm in diameter. The resulting regression models were turned into clinical risk scores which were then internally validated using bootstrap resampling, measuring discrimination and calibration.

RESULTS

The lowest clinical risk scores predicted a risk of AML and LAM of 1% and 0%, while the highest scores predicted a risk of 99% and 73%, respectively. Calibration was excellent for all three AML outcomes and good for LAM. Discrimination ranged from good to strong. C-statistics of 0.84, 0.83, 0.83 and 0.92 were seen for AML, bilateral AML, AML with a lesion≥3 cm and LAM, respectively.

CONCLUSION

Our work is an important step towards identifying individuals who could benefit from preventative strategies as well as more versus less frequent screening imaging. We expect that our work will allow for more personalised medicine in people with TSC. External validation of the risk scores will be important to confirm the robustness of our findings.

Development and validation of machine learning models to predict MDRO colonization or infection on ICU admission by using electronic health record data

BACKGROUND

Multidrug-resistant organisms (MDRO) pose a significant threat to public health. Intensive Care Units (ICU), characterized by the extensive use of antimicrobial agents and a high prevalence of bacterial resistance, are hotspots for MDRO proliferation. Timely identification of patients at high risk for MDRO can aid in curbing transmission, enhancing patient outcomes, and maintaining the cleanliness of the ICU environment. This study focused on developing a machine learning (ML) model to identify patients at risk of MDRO during the initial phase of their ICU stay.

METHODS

Utilizing patient data from the First Medical Center of the People's Liberation Army General Hospital (PLAGH-ICU) and the Medical Information Mart for Intensive Care (MIMIC-IV), the study analyzed variables within 24 h of ICU admission. Machine learning algorithms were applied to these datasets, emphasizing the early detection of MDRO colonization or infection. Model efficacy was evaluated by the area under the receiver operating characteristics curve (AUROC), alongside internal and external validation sets.

RESULTS

The study evaluated 3,536 patients in PLAGH-ICU and 34,923 in MIMIC-IV, revealing MDRO prevalence of 11.96% and 8.81%, respectively. Significant differences in ICU and hospital stays, along with mortality rates, were observed between MDRO positive and negative patients. In the temporal validation, the PLAGH-ICU model achieved an AUROC of 0.786 [0.748, 0.825], while the MIMIC-IV model reached 0.744 [0.723, 0.766]. External validation demonstrated reduced model performance across different datasets. Key predictors included biochemical markers and the duration of pre-ICU hospital stay.

CONCLUSIONS

The ML models developed in this study demonstrated their capability in early identification of MDRO risks in ICU patients. Continuous refinement and validation in varied clinical contexts remain essential for future applications.

Machine learning-based decision support model for selecting intra-arterial therapies for unresectable hepatocellular carcinoma: A national real-world evidence-based study

IMPORTANCE

Intra-arterial therapies(IATs) are promising options for unresectable hepatocellular carcinoma(HCC). Stratifying the prognostic risk before administering IAT is important for clinical decision-making and for designing future clinical trials.

OBJECTIVE

To develop and validate a machine learning(ML)-based decision support model(MLDSM) for recommending IAT modalities for unresectable HCC.

DESIGN, SETTING, AND PARTICIPANTS

Between October 2014 and October 2022, a total of 2,959 patients with HCC who underwent initial IATs were enroled retrospectively from 13 tertiary hospitals. These patients were divided into the training cohort (n = 1700), validation cohort (n = 428), and test cohort (n = 200).

MAIN OUTCOMES AND MEASURES

Thirty-two clinical variables were input, and five supervised ML algorithms, including eXtreme Gradient Boosting (XGBoost), Categorical Gradient Boosting (CatBoost), Gradient Boosting Decision Tree (GBDT), Light Gradient Boosting Machine (LGBM) and Random Forest (RF), were compared using the areas under the receiver operating characteristic curve (AUC) with the DeLong test.

RESULTS

A total of 1856 patients were assigned to the IAT alone Group(I-A), and 1103 patients were assigned to the IAT combination Group(I-C). The 12-month death rates were 31.9% (352/1103) in the I-A group and 50.4% (936/1856) in the I-C group. For the test cohort, in the I-C group, the CatBoost model achieved the best discrimination when 30 variables were input, with an AUC of 0.776 (95% confidence intervals [CI], 0.833-0.868). In the I-A group, the LGBM model achieved the best discrimination when 24 variables were input, with an AUC of 0.776 (95% CI, 0.833-0.868). According to the decision trees, BCLC grade, local therapy, and diameter as top three variables were used to guide clinical decisions between IAT modalities.

CONCLUSIONS AND RELEVANCE

The MLDSM can accurately stratify prognostic risk for HCC patients who received IATs, thus helping physicians to make decisions about IAT and providing guidance for surveillance strategies in clinical practice.

Primary care seeking among adults with chronic neck and low back pain in Norway: A prospective study from the HUNT study linked to Norwegian primary healthcare registry

BACKGROUND

To describe the frequency of primary care seeking for neck or back-related conditions among people with chronic neck and low back pain and to develop prediction models of primary care seeking and frequent visits.

METHODS

We included participants of the Trøndelag Health Study (HUNT4, 2017-19) in Norway who self-reported chronic neck and/or low back pain in the preceding year, and extracted data of primary care visits from the Norwegian primary healthcare registry. We investigated a total of 23 potential predictors and used multivariable logistic regression models to predict primary care seeking for neck or back-related conditions and frequent visits by healthcare provider (i.e., the highest quartile of number of visits).

RESULTS

Among the 15,352 HUNT4 participants with chronic neck and/or low back pain, 6231 participants (40.6%) sought primary care for neck or back-related conditions (median = 5 visits, IQR 2-15) within 2 years after the study. Participants who consulted physical therapists sought care the most frequently (median = 10 visits, IQR 3-26). Discrimination of the best-fit prediction model of primary care seeking and frequent visits by healthcare provider, assessed by C-statistic, ranged from 0.66-0.76. Participants who made frequent primary care visits in the preceding year were highly likely to continue frequent care seeking in the following 2 years.

CONCLUSIONS

Around 40% of people seek primary care for chronic neck and low back pain, and frequent care seeking may continue for years. Future studies should investigate strategies to reduce recurrent primary care visits, especially seeking physical therapist care, and promote self-management of chronic pain.

SIGNIFICANCE

People with chronic neck and low back pain who seek physical therapist care had the highest frequency of care seeking, underscoring the significant burden on healthcare systems. The high frequency of visits and associated healthcare expenditures highlight the critical need for effective and valuable primary care for chronic pain management. To mitigate recurrent visits and reduce costs, it is essential to provide patients with evidence-based treatments and self-management interventions.

MAATrica: a measure for assessing consistency and methods in medicinal and nutraceutical chemistry papers

The growing number of scientific papers and document sources underscores the need for methods capable of evaluating the quality of publications. Researchers who are looking for relevant papers for their studies need ways to assess the scientific value of these documents. One approach involves using semantic search engines that can automatically extract important knowledge from the growing body of text. In this study, we introduce a new metric called "MAATrica," which serves as the foundation for an innovative method designed to evaluate research papers. MAATrica offers a new way to analyze and categorize text, focusing on the consistency of research documents in the life sciences, particularly in the fields of medicinal and nutraceutical chemistry. This method utilizes semantic descriptions to cover in silico experiments, as well as in vitro and in vivo essays. Created to aid in evaluation processes like peer review, MAATrica uses toolkits and semantic applications to build the proposed measure, identify scientific entities, and gather information. We have applied MAATrica to roughly 90,000 papers and present our findings here.

Development and external validation of machine learning-based models to predict patients with cellulitis developing sepsis during hospitalisation

OBJECTIVE

Cellulitis is the most common cause of skin-related hospitalisations, and the mortality of patients with sepsis remains high. Some stratification models have been developed, but their performance in external validation has been unsatisfactory. This study was designed to develop and compare different models for predicting patients with cellulitis developing sepsis during hospitalisation.

DESIGN

This is a retrospective cohort study.

SETTING

This study included both the development and the external-validation phases from two independent large cohorts internationally.

PARTICIPANTS AND METHODS

A total of 6695 patients with cellulitis in the Medical Information Mart for Intensive care (MIMIC)-IV database were used to develop models with different machine-learning algorithms. The best models were selected and then externally validated in 2506 patients with cellulitis from the YiduCloud database of our university. The performances and robustness of selected models were further compared in the external-validation group by area under the curve (AUC), diagnostic accuracy, sensitivity, specificity and diagnostic OR.

PRIMARY OUTCOME MEASURES

The primary outcome of interest in this study was the development based on the Sepsis-3.0 criteria during hospitalisation.

RESULTS

Patient characteristics were significantly different between the two groups. In internal validation, XGBoost was the best model, with an AUC of 0.780, and AdaBoost was the worst model, with an AUC of 0.585. In external validation, the AUC of the artificial neural network (ANN) model was the highest, 0.830, while the AUC of the logistic regression (LR) model was the lowest, 0.792. The AUC values changed less in the boosting and ANN models than in the LR model when variables were deleted.

CONCLUSIONS

Boosting and neural network models performed slightly better than the LR model and were more robust in complex clinical situations. The results could provide a tool for clinicians to detect hospitalised patients with cellulitis developing sepsis early.

Construction and validation of a risk prediction model for postoperative ICU admission in patients with colorectal cancer: clinical prediction model study

BACKGROUND

Transfer to the ICU is common following non-cardiac surgeries, including radical colorectal cancer (CRC) resection. Understanding the judicious utilization of costly ICU medical resources and supportive postoperative care is crucial. This study aimed to construct and validate a nomogram for predicting the need for mandatory ICU admission immediately following radical CRC resection.

METHODS

Retrospective analysis was conducted on data from 1003 patients who underwent radical or palliative surgery for CRC at Ningxia Medical University General Hospital from August 2020 to April 2022. Patients were randomly assigned to training and validation cohorts in a 7:3 ratio. Independent predictors were identified using the least absolute shrinkage and selection operator (LASSO) and multivariate logistic regression in the training cohort to construct the nomogram. An online prediction tool was developed for clinical use. The nomogram's calibration and discriminative performance were assessed in both cohorts, and its clinical utility was evaluated through decision curve analysis (DCA).

RESULTS

The final predictive model comprised age (P = 0.003, odds ratio [OR] 3.623, 95% confidence interval [CI] 1.535-8.551); nutritional risk screening 2002 (NRS2002) (P = 0.000, OR 6.129, 95% CI 2.920-12.863); serum albumin (ALB) (P = 0.013, OR 0.921, 95% CI 0.863-0.982); atrial fibrillation (P = 0.000, OR 20.017, 95% CI 4.191-95.609); chronic obstructive pulmonary disease (COPD) (P = 0.009, OR 8.151, 95% CI 1.674-39.676); forced expiratory volume in 1 s / Forced vital capacity (FEV1/FVC) (P = 0.040, OR 0.966, 95% CI 0.935-0.998); and surgical method (P = 0.024, OR 0.425, 95% CI 0.202-0.891). The area under the curve was 0.865, and the consistency index was 0.367. The Hosmer-Lemeshow test indicated excellent model fit (P = 0.367). The calibration curve closely approximated the ideal diagonal line. DCA showed a significant net benefit of the predictive model for postoperative ICU admission.

CONCLUSION

Predictors of ICU admission following radical CRC resection include age, preoperative serum albumin level, nutritional risk screening, atrial fibrillation, COPD, FEV1/FVC, and surgical route. The predictive nomogram and online tool support clinical decision-making for postoperative ICU admission in patients undergoing radical CRC surgery.

TRIAL REGISTRATION

Despite the retrospective nature of this study, we have proactively registered it with the Chinese Clinical Trial Registry. The registration number is ChiCTR2200062210, and the date of registration is 29/07/2022.

Development of the DONOR prediction model on the risk of hypertensive complications in oocyte donation pregnancy: study protocol for a multicentre cohort study in the Netherlands

INTRODUCTION

Oocyte donation (OD) pregnancy is accompanied by a high incidence of hypertensive complications, with serious consequences for mother and child. Optimal care management, involving early recognition, optimisation of suitable treatment options and possibly eventually also prevention, is in high demand. Prediction of patient-specific risk factors for hypertensive complications in OD can provide the basis for this. The current project aims to establish the first prediction model on the risk of hypertensive complications in OD pregnancy.

METHODS AND ANALYSIS

The present study is conducted within the DONation of Oocytes in Reproduction project. For this multicentre cohort study, at least 541 OD pregnancies will be recruited. Baseline characteristics and obstetric data will be collected. Additionally, one sample of maternal peripheral blood and umbilical cord blood after delivery or a saliva sample from the child will be obtained, in order to determine the number of fetal-maternal human leucocyte antigen mismatches. Following data collection, a multivariate logistic regression model will be developed for the binary outcome hypertensive complication 'yes' and 'no'. The Prediction model Risk Of Bias ASsessment Tool will be used as guide to minimise the risk of bias. The study will be reported in line with the 'Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis' guideline. Discrimination and calibration will be determined to assess model performance. Internal validation will be performed using the bootstrapping method. External validation will be performed with the 'DONation of Oocytes in Reproduction individual participant data' dataset.

ETHICS AND DISSEMINATION

This study is approved by the Medical Ethics Committee LDD (Leiden, Den Haag, Delft), with protocol number P16.048 and general assessment registration (ABR) number NL56308.058.16. Further results will be shared through peer-reviewed journals and international conferences.

A systematic review of fall prediction models for community-dwelling older adults: comparison between models based on research cohorts and models based on routinely collected data

BACKGROUND

Prediction models can identify fall-prone individuals. Prediction models can be based on either data from research cohorts (cohort-based) or routinely collected data (RCD-based). We review and compare cohort-based and RCD-based studies describing the development and/or validation of fall prediction models for community-dwelling older adults.

METHODS

Medline and Embase were searched via Ovid until January 2023. We included studies describing the development or validation of multivariable prediction models of falls in older adults (60+). Both risk of bias and reporting quality were assessed using the PROBAST and TRIPOD, respectively.

RESULTS

We included and reviewed 28 relevant studies, describing 30 prediction models (23 cohort-based and 7 RCD-based), and external validation of two existing models (one cohort-based and one RCD-based). The median sample sizes for cohort-based and RCD-based studies were 1365 [interquartile range (IQR) 426-2766] versus 90 441 (IQR 56 442-128 157), and the ranges of fall rates were 5.4% to 60.4% versus 1.6% to 13.1%, respectively. Discrimination performance was comparable between cohort-based and RCD-based models, with the respective area under the receiver operating characteristic curves ranging from 0.65 to 0.88 versus 0.71 to 0.81. The median number of predictors in cohort-based final models was 6 (IQR 5-11); for RCD-based models, it was 16 (IQR 11-26). All but one cohort-based model had high bias risks, primarily due to deficiencies in statistical analysis and outcome determination.

CONCLUSIONS

Cohort-based models to predict falls in older adults in the community are plentiful. RCD-based models are yet in their infancy but provide comparable predictive performance with no additional data collection efforts. Future studies should focus on methodological and reporting quality.

An explainable machine learning framework for predicting the risk of buprenorphine treatment discontinuation for opioid use disorder among commercially insured individuals

OBJECTIVES

Buprenorphine is an effective evidence-based medication for opioid use disorder (OUD). Yet premature discontinuation undermines treatment effectiveness, increasing the risk of mortality and overdose. We developed and evaluated a machine learning (ML) framework for predicting buprenorphine care discontinuity within 12 months following treatment initiation.

METHODS

This retrospective study used United States (US) 2018-2021 MarketScan commercial claims data of insured individuals aged 18-64 who initiated buprenorphine between July 2018 and December 2020 with no buprenorphine prescriptions in the previous six months. We measured buprenorphine prescription discontinuation gaps of ≥30 days within 12 months of initiating treatment. We developed predictive models employing logistic regression, decision tree classifier, random forest, extreme gradient boosting, Adaboost, and random forest-extreme gradient boosting ensemble. We applied recursive feature elimination with cross-validation to reduce dimensionality and identify the most predictive features while maintaining model robustness. For model validation, we used several statistics to evaluate performance, such as C-statistics and precision-recall curves. We focused on two distinct treatment stages: at the time of treatment initiation and one and three months after treatment initiation. We employed SHapley Additive exPlanations (SHAP) analysis that helped us explain the contributions of different features in predicting buprenorphine discontinuation. We stratified patients into risk subgroups based on their predicted likelihood of treatment discontinuation, dividing them into decile subgroups. Additionally, we used a calibration plot to analyze the reliability of the models.

RESULTS

A total of 30,373 patients initiated buprenorphine and 14.98% (4551) discontinued treatment. C-statistic varied between 0.56 and 0.76 for the first-stage models including patient-level demographic and clinical variables. Inclusion of proportion of days covered (PDC) measured after one month and three months following treatment initiation significantly increased the models' discriminative power (C-statistics: 0.60 to 0.82). Random forest (C-statistics: 0.76, 0.79 and 0.82 with baseline predictors, one-month PDC and three-months PDC, respectively) outperformed other ML models in discriminative performance in all stages (C-statistics: 0.56 to 0.77). Most influential risk factors of discontinuation included early stage medication adherence, age, and initial days of supply.

CONCLUSION

ML algorithms demonstrated a good discriminative power in identifying patients at higher risk of buprenorphine care discontinuity. The proposed framework may help healthcare providers optimize treatment strategies and deliver targeted interventions to improve buprenorphine care continuity.

Developing and validating a prediction model of live birth following single vitrified-warmed blastocyst transfer

RESEARCH QUESTION

Can the developed clinical prediction model offer an accurate estimate of the likelihood of live birth, involving blastocyst morphology and vitrification day after single vitrified-warmed blastocyst transfer (SVBT), and therefore assist clinicians and patients?

STUDY DESIGN

Retrospective cohort study conducted at a Spanish university-based reproductive medicine unit (2017-2021) including consecutive vitrified-warmed blastocysts from IVF cycles. A multivariable logistic regression incorporated key live birth predictors: vitrification day, embryo score, embryo ploidy status and clinically relevant variables, i.e. maternal age.

RESULTS

The training set involved 1653 SVBT cycles carried out between 2017 and 2020; 592 SVBT cycles from 2021 constituted the external validation dataset. The model revealed that female age and embryo characteristics, including overall quality and blastulation day, is linked to live birth rate in SVBT cycles. Stratification by vitrification day and quality (from day-5A to day-6 C blastocysts) applied to genetically tested and untested embryos. The model's area under the curve was 0.66 (95% CI 0.64 to 0.69) during development and 0.65 (95% CI 0.61 to 0.70) in validation, denoting moderate discrimination. Calibration plots showed strong agreement between predicted and observed probabilities.

CONCLUSION

By incorporating essential predictors such as vitrification day, embryo morphology grade, age and preimplantation genetic testing for aneuploidy usage, this predictive model offers valuable guidance to clinicians and patients, enabling accurate forecasts of live birth rates for any given vitrified blastocyst within SVBT cycles. Additionally, it serves as a potentially indispensable laboratory tool, aiding in selecting the most promising blastocysts for optimal outcomes.

Equations to estimate human milk intake in infants aged 7 to 10 months: prediction models from a cross-sectional study

BACKGROUND

Obtaining valid estimates of nutrient intake in infants is currently limited by the difficulties of accurately measuring human milk intake. Current methods are either unsuitable for large-scale studies (i.e., the gold standard dose-to-mother stable isotope technique) or use set amounts, regardless of known variability in individual intake.

OBJECTIVES

This cross-sectional study aimed to develop equations to predict human milk intake using simple measures and to carry out external validation of existing methods against the gold standard technique.

METHODS

Data on human milk intake were obtained using the dose-to-mother stable isotope technique in 157 infants aged 7-10 mo and their mothers. Predictive equations were developed using questionnaire and anthropometric data (Model 1) and additional dietary data (Model 2) using lasso regression. Bland-Altman plots and intraclass correlation coefficients (ICC) also assessed the validity of existing methods (FITS and ALSPAC studies).

RESULTS

The strongest univariate predictors of human milk intake in infants of 8.3 mo on average (46% female) were infant age, infant body mass index (BMI), number of breastfeeds a day, infant formula consumption, and energy from complementary food intake. Mean [95% confidence interval (CI)] differences in predicted versus measured human milk intake [mean (SD): 762 (257) mL/day] were 0.0 mL/day (-26, 26) for Model 1 (ICC 0.74) and 0.5 mL/day (-21, 22) for Model 2 (ICC 0.83). Corresponding differences were -197 mL/day (-233, -161; ICC 0.32) and -175 mL/day (-216, -134; ICC 0.41) for the methods used by FITS and ALSPAC, respectively.

CONCLUSIONS

The Human Milk Intake Level Calculation provides substantial improvements on existing methods to estimate human milk intake in infants aged 7-10 mo, while utilizing data commonly collected in nutrition surveys. Although further validation in an external sample is recommended, these equations can be used to estimate human milk intake at this age with some confidence. This clinical trial was registered at http://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=379436) as ACTRN12620000459921.

Developing and validating clinical prediction models in hepatology - An overview for clinicians

Prediction models are everywhere in clinical medicine. We use them to assign a diagnosis or a prognosis, and there have been continuous efforts to develop better prediction models. It is important to understand the fundamentals of prediction modelling, thus, we herein describe nine steps to develop and validate a clinical prediction model with the intention of implementing it in clinical practice: Determine if there is a need for a new prediction model; define the purpose and intended use of the model; assess the quality and quantity of the data you wish to develop the model on; develop the model using sound statistical methods; generate risk predictions on the probability scale (0-100%); evaluate the performance of the model in terms of discrimination, calibration, and clinical utility; validate the model using bootstrapping to correct for the apparent optimism in performance; validate the model on external datasets to assess the generalisability and transportability of the model; and finally publish the model so that it can be implemented or validated by others.

Predicting Recurrence-Free Survival After Upfront Surgery in Resectable Pancreatic Ductal Adenocarcinoma: A Preoperative Risk Score Based on CA 19-9, CT, and 18F-FDG PET/CT

OBJECTIVE

To develop and validate a preoperative risk score incorporating carbohydrate antigen (CA) 19-9, CT, and fluorine-18-fluorodeoxyglucose (18F-FDG) PET/CT variables to predict recurrence-free survival (RFS) after upfront surgery in patients with resectable pancreatic ductal adenocarcinoma (PDAC).

MATERIALS AND METHODS

Patients with resectable PDAC who underwent upfront surgery between 2014 and 2017 (development set) or between 2018 and 2019 (test set) were retrospectively evaluated. In the development set, a risk-scoring system was developed using the multivariable Cox proportional hazards model, including variables associated with RFS. In the test set, the performance of the risk score was evaluated using the Harrell C-index and compared with that of the postoperative pathological tumor stage.

RESULTS

A total of 529 patients, including 335 (198 male; mean age ± standard deviation, 64 ± 9 years) and 194 (103 male; mean age, 66 ± 9 years) patients in the development and test sets, respectively, were evaluated. The risk score included five variables predicting RFS: tumor size (hazard ratio [HR], 1.29 per 1 cm increment; P < 0.001), maximal standardized uptake values of tumor ≥ 5.2 (HR, 1.29; P = 0.06), suspicious regional lymph nodes (HR, 1.43; P = 0.02), possible distant metastasis on 18F-FDG PET/CT (HR, 2.32; P = 0.03), and CA 19-9 (HR, 1.02 per 100 U/mL increment; P = 0.002). In the test set, the risk score showed good performance in predicting RFS (C-index, 0.61), similar to that of the pathologic tumor stage (C-index, 0.64; P = 0.17).

CONCLUSION

The proposed risk score based on preoperative CA 19-9, CT, and 18F-FDG PET/CT variables may have clinical utility in selecting high-risk patients with resectable PDAC.

Comparison of Simplified Palliative Prognostic Index and Palliative Performance Scale in Patients with Advanced Cancer in a Home Palliative Care Setting

Objective: The Palliative Performance Scale (PPS) has been reported to be as accurate as Palliative Prognostic Index (PPI). PPS is a component of the simplified PPI (sPPI). It is unknown whether PPS is as accurate as sPPI. This study aimed to compare the prognostic performance of the PPS and sPPI in patients with advanced cancer in a home palliative care setting in South Korea. Methods: This was a secondary analysis of a prospective cohort study that included Korean patients with advanced cancer who received home-based palliative care. We used the medical records maintained by specialized palliative care nurses. We computed the prognostic performance of PPS and sPPI using the area under the receiver operating characteristic curve (AUROC) and calibration plots for the 3- and 6-week survival. Results: A total of 80 patients were included, with a median overall survival of 47.0 days. The AUROCs of PPS were 0.71 and 0.69 at the 3- and 6-week survival predictions, respectively. The AUROCs of sPPI were 0.87 and 0.73 at the 3- and 6-week survival predictions, respectively. The calibration plot demonstrated satisfactory agreement across all score ranges for both the PPS and sPPI. Conclusions: This study showed that the sPPI assessed by nurses was more accurate than the PPS in a home palliative care setting in predicting the 3-week survival in patients with advanced cancer. The PPS can be used for a quick assessment.

A pilot clinical risk model to predict polymyxin-induced nephrotoxicity: a real-world, retrospective cohort study

OBJECTIVES

Polymyxin-induced nephrotoxicity (PIN) is a major safety concern and challenge in clinical practice, which limits the clinical use of polymyxins. This study aims to investigate the risk factors and to develop a scoring tool for the early prediction of PIN.

METHODS

Data on critically ill patients who received intravenous polymyxin B or colistin sulfate for over 24 h were collected. Logistic regression with the least absolute shrinkage and selection operator (LASSO) was used to identify variables that are associated with outcomes. The eXtreme Gradient Boosting (XGB) classifier algorithm was used to further visualize factors with significant differences. A prediction model for PIN was developed through binary logistic regression analysis and the model was assessed by temporal validation and external validation. Finally, a risk-scoring system was developed based on the prediction model.

RESULTS

Of 508 patients, 161 (31.6%) patients developed PIN. Polymyxin type, loading dose, septic shock, concomitant vasopressors and baseline blood urea nitrogen (BUN) level were identified as significant predictors of PIN. All validation exhibited great discrimination, with the AUC of 0.742 (95% CI: 0.696-0.787) for internal validation, of 0.708 (95% CI: 0.605-0.810) for temporal validation and of 0.874 (95% CI: 0.759-0.989) for external validation, respectively. A simple risk-scoring tool was developed with a total risk score ranging from -3 to 4, corresponding to a risk of PIN from 0.79% to 81.24%.

CONCLUSIONS

This study established a prediction model for PIN. Before using polymyxins, the simple risk-scoring tool can effectively identify patients at risk of developing PIN within a range of 7% to 65%.

Prognoses Associated With Palliative Performance Scale Scores in Modern Palliative Care Practice

Importance

The Palliative Performance Scale (PPS) is one of the most widely used prognostic tools for patients with serious illness. However, current prognostic estimates associated with PPS scores are based on data that are over a decade old.

Objective

To generate updated prognostic estimates by PPS score, care setting, and illness category, and examine how well PPS predicts short- and longer-term survival.

Design, Setting, and Participants

This prognostic study was conducted at a large academic medical center with robust inpatient and outpatient palliative care practices using electronic health record data linked with data from California Vital Records. Eligible participants included patients who received a palliative care consultation between January 1, 2018, and December 31, 2020. Data analysis was conducted from November 2022 to February 2024.

Exposure

Palliative care consultation with a PPS score documented.

Main Outcomes and Measures

The primary outcomes were predicted 1-, 6-, and 12-month mortality and median survival of patients by PPS score in the inpatient and outpatient settings, and performance of the PPS across a range of survival times. In subgroup analyses, mortality risk by PPS score was estimated in patients with cancer vs noncancer illnesses and those seen in-person vs by video telemedicine in the outpatient setting.

Results

Overall, 4779 patients (mean [SD] age, 63.5 [14.8] years; 2437 female [51.0%] and 2342 male [49.0%]) had a palliative care consultation with a PPS score documented. Of these patients, 2276 were seen in the inpatient setting and 3080 were seen in the outpatient setting. In both the inpatient and outpatient settings, 1-, 6-, and 12-month mortality were higher and median survival was shorter for patients with lower PPS scores. Prognostic estimates associated with PPS scores were substantially longer (2.3- to 11.7-fold) than previous estimates commonly used by clinicians. The PPS had good ability to discriminate between patients who lived and those who died in the inpatient setting (integrated time-dependent area under the curve [iAUC], 0.74) but its discriminative ability was lower in the outpatient setting (iAUC, 0.67). The PPS better predicted 1-month survival than longer-term survival. Mortality rates were higher for patients with cancer than other serious illnesses at most PPS levels.

Conclusions and Relevance

In this prognostic study, prognostic estimates associated with PPS scores were substantially longer than previous estimates commonly used by clinicians. Based on these findings, an online calculator was updated to assist clinicians in reaching prognostic estimates that are more consistent with modern palliative care practice and specific to the patient's setting and diagnosis group.

An Appraisal of the Quality of Development and Reporting of Predictive Models in Neurosurgery: A Systematic Review

BACKGROUND AND OBJECTIVES

Significant evidence has indicated that the reporting quality of novel predictive models is poor because of confounding by small data sets, inappropriate statistical analyses, and a lack of validation and reproducibility. The Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) statement was developed to increase the generalizability of predictive models. This study evaluated the quality of predictive models reported in neurosurgical literature through their compliance with the TRIPOD guidelines.

METHODS

Articles reporting prediction models published in the top 5 neurosurgery journals by SCImago Journal Rank-2 (Neurosurgery, Journal of Neurosurgery, Journal of Neurosurgery: Spine, Journal of NeuroInterventional Surgery, and Journal of Neurology, Neurosurgery, and Psychiatry) between January 1st, 2018, and January 1st, 2023, were identified through a PubMed search strategy that combined terms related to machine learning and prediction modeling. These original research articles were analyzed against the TRIPOD criteria.

RESULTS

A total of 110 articles were assessed with the TRIPOD checklist. The median compliance was 57.4% (IQR: 50.0%-66.7%). Models using machine learning-based models exhibited lower compliance on average compared with conventional learning-based models (57.1%, 50.0%-66.7% vs 68.1%, 50.2%-68.1%, P = .472). Among the TRIPOD criteria, the lowest compliance was observed in blinding the assessment of predictors and outcomes (n = 7, 12.7% and n = 10, 16.9%, respectively), including an informative title (n = 17, 15.6%) and reporting model performance measures such as confidence intervals (n = 27, 24.8%). Few studies provided sufficient information to allow for the external validation of results (n = 26, 25.7%).

CONCLUSION

Published predictive models in neurosurgery commonly fall short of meeting the established guidelines laid out by TRIPOD for optimal development, validation, and reporting. This lack of compliance may represent the minor extent to which these models have been subjected to external validation or adopted into routine clinical practice in neurosurgery.

Risk factors and a predictive model for left ventricular hypertrophy in young adults with salt-sensitive hypertension

Salt-sensitive hypertension is common among individuals with essential hypertension, and the prevalence of left ventricular hypertrophy (LVH) has increased. However, data from early identification of the risk of developing LVH in young adults with salt-sensitive hypertension are lacking. Thus, the present study aimed to design a nomogram for predicting the risk of developing LVH in young adults with salt-sensitive hypertension. A retrospective analysis of 580 patients with salt-sensitive hypertension was conducted. The training set consisted of 70% (n = 406) of the patients, while the validation set consisted of the remaining 30% (n = 174). Based on multivariate analysis of the training set, predictors for LVH were extracted to develop a nomogram. Discrimination curves, calibration curves, and clinical utility were employed to assess the predictive performance of the nomogram. The final simplified nomogram model included age, sex, office systolic blood pressure, duration of hypertension, abdominal obesity, triglyceride-glucose index, and estimated glomerular filtration rate (eGFR). In the training set, the model demonstrated moderate discrimination, as indicated by an area under the receiver operating characteristic (ROC) curve of 0.863 (95% confidence interval: 0.831-0.894). The calibration curve exhibited good agreement between the predicted and actual probabilities of LVH in the training set. Additionally, the validation set further confirmed the reliability of the prediction nomogram. In conclusions, the simplified nomogram, which consists of seven routine clinical variables, has shown good performance and clinical utility in identifying young adults with salt-sensitive hypertension who are at high risk of LVH at an early stage.

Predictions for functional outcome and mortality in acute ischaemic stroke following successful endovascular thrombectomy

Background

Accurate outcome predictions for patients who had ischaemic stroke with successful reperfusion after endovascular thrombectomy (EVT) may improve patient treatment and care. Our study developed prediction models for key clinical outcomes in patients with successful reperfusion following EVT in an Australian population.

Methods

The study included all patients who had ischaemic stroke with occlusion in the proximal anterior cerebral circulation and successful reperfusion post-EVT over a 7-year period. Multivariable logistic regression and Cox regression models, incorporating bootstrap and multiple imputation techniques, were used to identify predictors and develop models for key clinical outcomes: 3-month poor functional status; 30-day, 1-year and 3-year mortality; survival time.

Results

A total of 978 patients were included in the analyses. Predictors associated with one or more poor outcomes include: older age (ORs for every 5-year increase: 1.22-1.40), higher premorbid functional modified Rankin Scale (ORs: 1.31-1.75), higher baseline National Institutes of Health Stroke Scale (ORs: 1.05-1.07) score, higher blood glucose (ORs: 1.08-1.19), larger core volume (ORs for every 10 mL increase: 1.10-1.22), pre-EVT thrombolytic therapy (ORs: 0.44-0.56), history of heart failure (outcome: 30-day mortality, OR=1.87), interhospital transfer (ORs: 1.42 to 1.53), non-rural/regional stroke onset (outcome: functional dependency, OR=0.64), longer onset-to-groin puncture time (outcome: 3-year mortality, OR=1.08) and atherosclerosis-caused stroke (outcome: functional dependency, OR=1.68). The models using these predictors demonstrated moderate predictive abilities (area under the receiver operating characteristic curve range: 0.752-0.796).

Conclusion

Our models using real-world predictors assessed at hospital admission showed satisfactory performance in predicting poor functional outcomes and short-term and long-term mortality for patients with successful reperfusion following EVT. These can be used to inform EVT treatment provision and consent.

Predicting the individualized risk of human immunodeficiency virus infection among sexually active women in Ethiopia using a nomogram: prediction model development and validation

Introduction

Women are more vulnerable to HIV infection due to biological and socioeconomic reasons. Developing a predictive model for these vulnerable populations to estimate individualized risk for HIV infection is relevant for targeted preventive interventions. The objective of the study was to develop and validate a risk prediction model that allows easy estimations of HIV infection risk among sexually active women in Ethiopia.

Methods

Data from the 2016 Ethiopian Demographic and Health Survey, which comprised 10,253 representative sexually active women, were used for model development. Variables were selected using the least absolute shrinkage and selection operator (LASSO). Variables selected by LASSO were incorporated into the multivariable mixed-effect logistic regression model. Based on the multivariable model, an easy-to-use nomogram was developed to facilitate its applicability. The performance of the nomogram was evaluated using discrimination and calibration abilities, Brier score, sensitivity, and specificity. Internal validation was carried out using the bootstrapping method.

Results

The model selected seven predictors of HIV infection, namely, age, education, marital status, sex of the household head, age at first sex, multiple sexual partners during their lifetime, and residence. The nomogram had a discriminatory power of 89.7% (95% CI: 88.0, 91.5) and a calibration p-value of 0.536. In addition, the sensitivity and specificity of the nomogram were 74.1% (95% CI: 68.4, 79.2) and 80.9% (95% CI: 80.2, 81.7), respectively. The internally validated model had a discriminatory ability of 89.4% (95% CI: 87.7, 91.1) and a calibration p-value of 0.195. Sensitivity and specificity after validation were 72.9% (95% CI: 67.2, 78.2) and 80.1% (95% CI: 79.3, 80.9), respectively.

Conclusion

A new prediction model that quantifies the individualized risk of HIV infection has been developed in the form of a nomogram and internally validated. It has very good discriminatory power and good calibration ability. This model can facilitate the identification of sexually active women at high risk of HIV infection for targeted preventive measures.

Framework for personalized prediction of treatment response in relapsing-remitting multiple sclerosis: a replication study in independent data

BACKGROUND

Individualizing and optimizing treatment of relapsing-remitting multiple sclerosis patients is a challenging problem, which would benefit from a clinically valid decision support. Stühler et al. presented black box models for this aim which were developed and internally evaluated in a German registry but lacked external validation.

METHODS

In patients from the French OFSEP registry, we independently built and validated models predicting being free of relapse and free of confirmed disability progression (CDP), following the methodological roadmap and predictors reported by Stühler. Hierarchical Bayesian models were fit to predict the outcomes under 6 disease-modifying treatments given the individual disease course up to the moment of treatment change. Data was temporally split on 2017, and models were developed in patients treated earlier (n = 5517). Calibration curves, discrimination, mean squared error (MSE) and relative percentage of root MSE (RMSE%) were assessed by external validation of models in more-recent patients (n = 3768). Non-Bayesian fixed-effects GLMs were also applied and their outcomes were compared to these of the Bayesian ones. For both, we modelled the number of on-therapy relapses with a negative binomial distribution, and CDP occurrence with a binomial distribution.

RESULTS

The performance of our temporally-validated relapse model (MSE: 0.326, C-Index: 0.639) is potentially superior to that of Stühler's (MSE: 0.784, C-index: 0.608). Calibration plots revealed miscalibration. Our CDP model (MSE: 0.072, C-Index: 0.777) was also better than its counterpart (MSE: 0.131, C-index: 0.554). Results from non-Bayesian fixed-effects GLM models were similar to the Bayesian ones.

CONCLUSIONS

The relapse and CDP models rebuilt and externally validated in independent data could compare and strengthen the credibility of the Stühler models. Their model-building strategy was replicable.

Development and external validation of a prognostic model for time to readmission or death in multimorbid patients

OBJECTIVE

To develop and externally validate a prognostic model built on important factors predisposing multimorbid patients to all-cause readmission and/or death. In addition to identify patients who may benefit most from a comprehensive clinical pharmacist intervention.

METHODS

A multivariable prognostic model was developed based on data from a randomised controlled trial investigating the effect of pharmacist-led medicines management on readmission rate in multimorbid, hospitalised patients. The derivation set comprised 386 patients randomised in a 1:1 manner to the intervention group, i.e. with a pharmacist included in their multidisciplinary treatment team, or the control group receiving standard care at the ward. External validation of the model was performed using data from an independent cohort, in which 100 patients were randomised to the same intervention, or standard care. The setting was an internal medicines ward at a university hospital in Norway.

RESULTS

The number of patients who were readmitted or had died within 18 months after discharge was 297 (76.9 %) in the derivation set, i.e. the randomized controlled trial, and 69 (71.1 %) in the validation set, i.e. the independent cohort. Charlson comorbidity index (CCI; low, moderate or high), previous hospital admissions within the previous six months and heart failure were the strongest prognostic factors and were included in the final model. The efficacy of the pharmaceutical intervention did not prove significant in the model. A prognostic index (PI) was constructed to estimate the hazard of readmission or death (low, intermediate or high-risk groups). Overall, the external validation replicated the result. We were unable to identify a subgroup of the multimorbid patients with better efficacy of the intervention.

CONCLUSIONS

A prognostic model including CCI, previous admissions and heart failure can be used to obtain valid estimates of risk of readmission and death in patients with multimorbidity.

Development and validation of a community acquired sepsis-worsening score in the adult emergency department: a prospective cohort: the CASC score

BACKGROUND

Sepsis is a leading cause of death and serious illness that requires early recognition and therapeutic management to improve survival. The quick-SOFA score helps in its recognition, but its diagnostic performance is insufficient. To develop a score that can rapidly identify a community acquired septic situation at risk of clinical complications in patients consulting the emergency department (ED).

METHODS

We conducted a monocentric, prospective cohort study in the emergency department of a university hospital between March 2016 and August 2018 (NCT03280992). All patients admitted to the emergency department for a suspicion of a community-acquired infection were included. Predictor variables of progression to septic shock or death within the first 90 days were selected using backward stepwise multivariable logistic regression to develop a clinical score. Receiver operating characteristic (ROC) curves were constructed to determine the discriminating power of the area under the curve (AUC). We also determined the threshold of our score that optimized the performance required for a sepsis-worsening score. We have compared our score with the NEWS-2 and qSOFA scores.

RESULTS

Among the 21,826 patients admitted to the ED, 796 patients were suspected of having community-acquired infection and 461 met the sepsis criteria; therefore, these patients were included in the analysis. The median [interquartile range] age was 72 [54-84] years, 248 (54%) were males, and 244 (53%) had respiratory symptoms. The clinical score ranged from 0 to 90 and included 8 variables with an area under the ROC curve of 0.85 (confidence interval [CI] 95% 0.81-0.89). A cut-off of 26 yields a sensitivity of 88% (CI 95% 0.79-0.93), a specificity of 62% (CI 95% 57-67), and a negative predictive value of 95% (CI 95% 91-97). The area under the ROC curve for our score was 0.85 (95% CI, 0.81-0.89) versus 0.73 (95% CI, 0.68-0.78) for qSOFA and 0.66 (95% CI, 0.60-0.72) for NEWS-2.

CONCLUSIONS

Our study provides an accurate clinical score for identifying septic patients consulting the ED early at risk of worsening disease. This score could be implemented at admission.

Computational simulation of the potential improvement in clinical outcomes of cardiovascular diseases with the use of a personalized predictive medicine approach

IMPORTANCE AND OBJECTIVES

The current medical paradigm of evidence-based medicine relies on clinical guidelines derived from randomized clinical trials (RCTs), but these guidelines often overlook individual variations in treatment effects. Approaches have been proposed to develop models predicting the effects of individualized management, such as predictive allocation, individualizing treatment allocation. It is currently unknown whether widespread implementation of predictive allocation could result in better population-level outcomes over guideline-based therapy. We sought to simulate the potential effect of predictive allocation using data from previously conducted RCTs.

METHODS AND RESULTS

Data from 3 RCTs (positive trial, negative trial, trial stopped for futility) in pediatric cardiology were used in a computational simulation study to quantify the potential benefits of a personalized approach based on predictive allocation. Outcomes were compared when using a universal approach vs predictive allocation where each patient was allocated to the treatment associated with the lowest predicted probability of negative outcome. Compared to results from RCTs, predictive allocation yielded absolute risk reductions of 13.8% (95% confidence interval [CI] -1.9 to 29.5), 13.9% (95% CI 4.5-23.2), and 15.6% (95% CI 1.5-29.6), respectively, corresponding to a number needed to treat of 7.3, 7.2, and 6.4. The net benefit of predictive allocation was directly proportional to the performance of the prediction models and disappeared as model performance degraded below an area under the curve of 0.55.

DISCUSSION

These findings highlight that predictive allocation could result in improved group-level outcomes, particularly when highly predictive models are available. These findings will need to be confirmed in simulations of other trials with varying conditions and eventually in RCTs of predictive vs guideline-based treatment allocation.

Machine Learning and External Validation of the IDENTIFY Risk Calculator for Patients with Haematuria Referred to Secondary Care for Suspected Urinary Tract Cancer

BACKGROUND

The IDENTIFY study developed a model to predict urinary tract cancer using patient characteristics from a large multicentre, international cohort of patients referred with haematuria. In addition to calculating an individual's cancer risk, it proposes thresholds to stratify them into very-low-risk (<1%), low-risk (1-<5%), intermediate-risk (5-<20%), and high-risk (≥20%) groups.

OBJECTIVE

To externally validate the IDENTIFY haematuria risk calculator and compare traditional regression with machine learning algorithms.

DESIGN, SETTING, AND PARTICIPANTS

Prospective data were collected on patients referred to secondary care with new haematuria. Data were collected for patient variables included in the IDENTIFY risk calculator, cancer outcome, and TNM staging. Machine learning methods were used to evaluate whether better models than those developed with traditional regression methods existed.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

The area under the receiver operating characteristic curve (AUC) for the detection of urinary tract cancer, calibration coefficient, calibration in the large (CITL), and Brier score were determined.

RESULTS AND LIMITATIONS

There were 3582 patients in the validation cohort. The development and validation cohorts were well matched. The AUC of the IDENTIFY risk calculator on the validation cohort was 0.78. This improved to 0.80 on a subanalysis of urothelial cancer prevalent countries alone, with a calibration slope of 1.04, CITL of 0.24, and Brier score of 0.14. The best machine learning model was Random Forest, which achieved an AUC of 0.76 on the validation cohort. There were no cancers stratified to the very-low-risk group in the validation cohort. Most cancers were stratified to the intermediate- and high-risk groups, with more aggressive cancers in higher-risk groups.

CONCLUSIONS

The IDENTIFY risk calculator performed well at predicting cancer in patients referred with haematuria on external validation. This tool can be used by urologists to better counsel patients on their cancer risks, to prioritise diagnostic resources on appropriate patients, and to avoid unnecessary invasive procedures in those with a very low risk of cancer.

PATIENT SUMMARY

We previously developed a calculator that predicts patients' risk of cancer when they have blood in their urine, based on their personal characteristics. We have validated this risk calculator, by testing it on a separate group of patients to ensure that it works as expected. Most patients found to have cancer tended to be in the higher-risk groups and had more aggressive types of cancer with a higher risk. This tool can be used by clinicians to fast-track high-risk patients based on the calculator and investigate them more thoroughly.

Automated Pupillometry Is Able to Discriminate Patients with Acute Stroke from Healthy Subjects: An Observational, Cross-Sectional Study

BACKGROUND

Automated pupillometry (AP) is a handheld, non-invasive tool that is able to assess pupillary light reflex dynamics and is useful for the detection of intracranial hypertension. Limited evidence is available on acute ischemic stroke (AIS) patients. The primary objective was to evaluate the ability of AP to discriminate AIS patients from healthy subjects (HS). Secondly, we aimed to compute a predictive score for AIS diagnosis based on clinical, demographic, and AP variables.

METHODS

We included 200 consecutive patients admitted to a comprehensive stroke center who underwent AP assessment through NPi-200 (NeurOptics®) within 72 h of stroke onset and 200 HS. The mean values of AP parameters and the absolute differences between the AP parameters of the two eyes were considered in the analyses. Predictors of stroke diagnosis were identified through univariate and multivariate logistic regressions; we then computed a nomogram based on each variable's β coefficient. Finally, we developed a web app capable of displaying the probability of stroke diagnosis based on the predictive algorithm.

RESULTS

A high percentage of pupil constriction (CH, p < 0.001), a low constriction velocity (CV, p = 0.002), and high differences between these two parameters (p = 0.036 and p = 0.004, respectively) were independent predictors of AIS. The highest contribution in the predictive score was provided by CH, the Neurological Pupil Index, CV, and CV absolute difference, disclosing the important role of AP in the discrimination of stroke patients.

CONCLUSIONS

The results of our study suggest that AP parameters, and in particular, those concerning pupillary constriction, may be useful for the early diagnosis of AIS.

Comparing penalization methods for linear models on large observational health data

OBJECTIVE

This study evaluates regularization variants in logistic regression (L1, L2, ElasticNet, Adaptive L1, Adaptive ElasticNet, Broken adaptive ridge [BAR], and Iterative hard thresholding [IHT]) for discrimination and calibration performance, focusing on both internal and external validation.

MATERIALS AND METHODS

We use data from 5 US claims and electronic health record databases and develop models for various outcomes in a major depressive disorder patient population. We externally validate all models in the other databases. We use a train-test split of 75%/25% and evaluate performance with discrimination and calibration. Statistical analysis for difference in performance uses Friedman's test and critical difference diagrams.

RESULTS

Of the 840 models we develop, L1 and ElasticNet emerge as superior in both internal and external discrimination, with a notable AUC difference. BAR and IHT show the best internal calibration, without a clear external calibration leader. ElasticNet typically has larger model sizes than L1. Methods like IHT and BAR, while slightly less discriminative, significantly reduce model complexity.

CONCLUSION

L1 and ElasticNet offer the best discriminative performance in logistic regression for healthcare predictions, maintaining robustness across validations. For simpler, more interpretable models, L0-based methods (IHT and BAR) are advantageous, providing greater parsimony and calibration with fewer features. This study aids in selecting suitable regularization techniques for healthcare prediction models, balancing performance, complexity, and interpretability.

Predictive modeling of ambulatory outcomes after spinal cord injury using machine learning

STUDY DESIGN

Retrospective multi-site cohort study.

OBJECTIVES

To develop an accurate machine learning predictive model using predictor variables from the acute rehabilitation period to determine ambulatory status in spinal cord injury (SCI) one year post injury.

SETTING

Model SCI System (SCIMS) database between January 2000 and May 2019.

METHODS

Retrospective cohort study using data that were previously collected as part of the SCI Model System (SCIMS) database. A total of 4523 patients were analyzed comparing traditional models (van Middendorp and Hicks) compared to machine learning algorithms including Elastic Net Penalized Logistic Regression (ENPLR), Gradient Boosted Machine (GBM), and Artificial Neural Networks (ANN).

RESULTS

Compared with GBM and ANN, ENPLR was determined to be the preferred model based on predictive accuracy metrics, calibration, and variable selection. The primary metric to judge discrimination was the area under the receiver operating characteristic curve (AUC). When compared to the van Middendorp all patients (0.916), ASIA A and D (0.951) and ASIA B and C (0.775) and Hicks all patients (0.89), ASIA A and D (0.934) and ASIA B and C (0.775), ENPLR demonstrated improved AUC for all patients (0.931), ASIA A and D (0.965) ASIA B and C (0.803).

CONCLUSIONS

Utilizing artificial intelligence and machine learning methods are feasible for accurately classifying outcomes in SCI and may provide improved sensitivity in identifying which individuals are less likely to ambulate and may benefit from augmentative strategies, such as neuromodulation. Future directions should include the use of additional variables to further refine these models.

Systematic Review and Meta-Analysis of Prehospital Machine Learning Scores as Screening Tools for Early Detection of Large Vessel Occlusion in Patients With Suspected Stroke

BACKGROUND

Enhanced detection of large vessel occlusion (LVO) through machine learning (ML) for acute ischemic stroke appears promising. This systematic review explored the capabilities of ML models compared with prehospital stroke scales for LVO prediction.

METHODS AND RESULTS

Six bibliographic databases were searched from inception until October 10, 2023. Meta-analyses pooled the model performance using area under the curve (AUC), sensitivity, specificity, and summary receiver operating characteristic curve. Of 1544 studies screened, 8 retrospective studies were eligible, including 32 prehospital stroke scales and 21 ML models. Of the 9 prehospital scales meta-analyzed, the Rapid Arterial Occlusion Evaluation had the highest pooled AUC (0.82 [95% CI, 0.79-0.84]). Support Vector Machine achieved the highest AUC of 9 ML models included (pooled AUC, 0.89 [95% CI, 0.88-0.89]). Six prehospital stroke scales and 10 ML models were eligible for summary receiver operating characteristic analysis. Pooled sensitivity and specificity for any prehospital stroke scale were 0.72 (95% CI, 0.68-0.75) and 0.77 (95% CI, 0.72-0.81), respectively; summary receiver operating characteristic curve AUC was 0.80 (95% CI, 0.76-0.83). Pooled sensitivity for any ML model for LVO was 0.73 (95% CI, 0.64-0.79), specificity was 0.85 (95% CI, 0.80-0.89), and summary receiver operating characteristic curve AUC was 0.87 (95% CI, 0.83-0.89).

CONCLUSIONS

Both prehospital stroke scales and ML models demonstrated varying accuracies in predicting LVO. Despite ML potential for improved LVO detection in the prehospital setting, application remains limited by the absence of prospective external validation, limited sample sizes, and lack of real-world performance data in a prehospital setting.

Development of a predictive nomogram for in-hospital death risk in multimorbid patients with hepatocellular carcinoma undergoing Palliative Locoregional Therapy

Patients diagnosed with hepatocellular carcinoma (HCC) often present with multimorbidity, significantly contributing to adverse outcomes, particularly in-hospital mortality. This study aimed to develop a predictive nomogram to assess the impact of comorbidities on in-hospital mortality risk in HCC patients undergoing palliative locoregional therapy. We retrospectively analyzed data from 345 hospitalized HCC patients who underwent palliative locoregional therapy between January 2015 and December 2022. The nomogram was constructed using independent risk factors such as length of stay (LOS), hepatitis B virus (HBV) infection, hypertension, chronic obstructive pulmonary disease (COPD), anemia, thrombocytopenia, liver cirrhosis, hepatic encephalopathy (HE), N stage, and microvascular invasion. The model demonstrated high predictive accuracy with an AUC of 0.908 (95% CI: 0.859-0.956) for the overall dataset, 0.926 (95% CI: 0.883-0.968) for the training set, and 0.862 (95% CI: 0.728-0.994) for the validation set. Calibration curves indicated a strong correlation between predicted and observed outcomes, validated by statistical tests. Decision curve analysis (DCA) and clinical impact curves (CIC) confirmed the model's clinical utility in predicting in-hospital mortality. This nomogram offers a practical tool for personalized risk assessment in HCC patients undergoing palliative locoregional therapy, facilitating informed clinical decision-making and improving patient management.

Machine learning evaluation of immune infiltrate through digital tumour score allows prediction of survival outcome in a pooled analysis of three international stage III colon cancer cohorts

BACKGROUND

T-cell immune infiltrates are robust prognostic variables in localised colon cancer. Evaluation of prognosis using artificial intelligence is an emerging field. We evaluated whether machine learning analysis improved prediction of patient outcome in comparison with analysis of T cell infiltrate only or in association with clinical variables.

METHODS

We used data from two phase III clinical trials (Prodige-13 and PETACC08) and one retrospective Italian cohort (HARMONY). Cohorts were split into training (N = 692), internal validation (N = 297) and external validation (N = 672) sets. Tumour slides were stained with CD3mAb. CD3 Machine Learning (CD3ML) score was computed using graphical parameters within the tumour tiles obtained from CD3 slides. CD3 infiltrates in tumour core and invasive margin were automatically detected. Associations of CD3 infiltrates and CD3ML with 5-year Disease-Free Survival (DFS) were examined using univariate and multivariable survival models by Cox regression.

FINDINGS

CD3 density both in the invasive margin and the tumour core were significantly associated with DFS in the different sets. Similarly, CD3ML score was significantly associated with DFS in all sets. CD3 assessment did not provide added value on top of CD3ML assessment (Likelihood Ratio Test (LRT), p = 0.13). In contrast, CD3ML improved prediction of DFS when combined with a clinical risk stage (LRT, p = 0.001). Stratified by clinical risk score (High or Low), patients with low CD3ML score had better DFS.

INTERPRETATION

In all tested sets, machine learning analysis of tumour cells improved prediction of prognosis compared to clinical parameters. Adding tumour-infiltrating lymphocytes assessment did not improve prognostic determination.

FUNDING

This research received no external funding.

Development of a diagnostic nomogram for alpha-fetoprotein-negative hepatocellular carcinoma based on serological biomarkers

BACKGROUND

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide. Serum biomarkers play an important role in the early diagnosis and prognosis of HCC. Because a certain percentage of HCC patients are negative for alpha-fetoprotein (AFP), the diagnosis of AFP-negative HCC is essential to improve the detection rate of HCC.

AIM

To establish an effective model for diagnosing AFP-negative HCC based on serum tumour biomarkers.

METHODS

A total of 180 HCC patients were enrolled in this study. The expression levels of GP73, des-γ-carboxyprothrombin (DCP), CK18-M65, and CK18-M30 were detected by a fully automated chemiluminescence analyser. The variables were selected by logistic regression analysis. Several models were constructed using stepwise backward logistic regression. The performance of the models was compared using the C statistic, integrated discrimination improvement, net reclassification improvement, and calibration curves. The clinical utility of the nomogram was assessed using decision curve analysis (DCA).

RESULTS

The results showed that the expression levels of GP73, DCP, CK18-M65, and CK18-M30 were significantly greater in AFP-negative HCC patients than in healthy controls (P < 0.001). Multivariate logistic regression analysis revealed that GP73, DCP, and CK18-M65 were independent factors for diagnosing AFP-negative HCC. By comparing the diagnostic performance of multiple models, we included GP73 and CK18-M65 as the model variables, and the model had good discrimination ability (area under the curve = 0.946) and good goodness of fit. The DCA curves indicated the good clinical utility of the nomogram.

CONCLUSION

Our study identified GP73 and CK18-M65 as serum biomarkers with certain application value in the diagnosis of AFP-negative HCC. The diagnostic nomogram based on CK18-M65 combined with GP73 demonstrated good performance and effectively identified high-risk groups of patients with HCC.

Development and validation of a prediction score to assess the risk of depression in primary care

BACKGROUND

Major depression is the most frequent psychiatric disorder and primary care is a crucial setting for its early recognition. This study aimed to develop and validate the DEP-HScore as a tool to predict depression risk in primary care and increase awareness and investigation of this condition among General Practitioners (GPs).

METHODS

The DEP-HScore was developed using data from the Italian Health Search Database (HSD). A cohort of 903,748 patients aged 18 years or older was selected and followed until the occurrence of depression, death or end of data availability (December 2019). Demographics, somatic signs/symptoms and psychiatric/medical comorbidities were entered in a multivariate Cox regression to predict the occurrence of depression. The coefficients formed the DEP-HScore for individual patients. Explained variance (pseudo-R2), discrimination (AUC) and calibration (slope estimating predicted-observed risk relationship) assessed the prediction accuracy.

RESULTS

The DEP-HScore explained 18.1 % of the variation in occurrence of depression and the discrimination value was equal to 67 %. With an event horizon of three months, the slope and intercept were not significantly different from the ideal calibration.

LIMITATIONS

The DEP-HScore has not been tested in other settings. Furthermore, the model was characterized by limited calibration performance when the risk of depression was estimated at the 1-year follow-up.

CONCLUSIONS

The DEP-HScore is reliable tool that could be implemented in primary care settings to evaluate the risk of depression, thus enabling prompt and suitable investigations to verify the presence of this condition.

Development of a diagnostic nomogram for alpha-fetoprotein-negative hepatocellular carcinoma based on serological biomarkers

BACKGROUND

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide. Serum biomarkers play an important role in the early diagnosis and prognosis of HCC. Because a certain percentage of HCC patients are negative for alpha-fetoprotein (AFP), the diagnosis of AFP-negative HCC is essential to improve the detection rate of HCC.

AIM

To establish an effective model for diagnosing AFP-negative HCC based on serum tumour biomarkers.

METHODS

A total of 180 HCC patients were enrolled in this study. The expression levels of GP73, des-γ-carboxyprothrombin (DCP), CK18-M65, and CK18-M30 were detected by a fully automated chemiluminescence analyser. The variables were selected by logistic regression analysis. Several models were constructed using stepwise backward logistic regression. The performance of the models was compared using the C statistic, integrated discrimination improvement, net reclassification improvement, and calibration curves. The clinical utility of the nomogram was assessed using decision curve analysis (DCA).

RESULTS

The results showed that the expression levels of GP73, DCP, CK18-M65, and CK18-M30 were significantly greater in AFP-negative HCC patients than in healthy controls (P < 0.001). Multivariate logistic regression analysis revealed that GP73, DCP, and CK18-M65 were independent factors for diagnosing AFP-negative HCC. By comparing the diagnostic performance of multiple models, we included GP73 and CK18-M65 as the model variables, and the model had good discrimination ability (area under the curve = 0.946) and good goodness of fit. The DCA curves indicated the good clinical utility of the nomogram.

CONCLUSION

Our study identified GP73 and CK18-M65 as serum biomarkers with certain application value in the diagnosis of AFP-negative HCC. The diagnostic nomogram based on CK18-M65 combined with GP73 demonstrated good performance and effectively identified high-risk groups of patients with HCC.

Diagnostic test accuracy of externally validated convolutional neural network (CNN) artificial intelligence (AI) models for emergency head CT scans - A systematic review

BACKGROUND

The surge in emergency head CT imaging and artificial intelligence (AI) advancements, especially deep learning (DL) and convolutional neural networks (CNN), have accelerated the development of computer-aided diagnosis (CADx) for emergency imaging. External validation assesses model generalizability, providing preliminary evidence of clinical potential.

OBJECTIVES

This study systematically reviews externally validated CNN-CADx models for emergency head CT scans, critically appraises diagnostic test accuracy (DTA), and assesses adherence to reporting guidelines.

METHODS

Studies comparing CNN-CADx model performance to reference standard were eligible. The review was registered in PROSPERO (CRD42023411641) and conducted on Medline, Embase, EBM-Reviews and Web of Science following PRISMA-DTA guideline. DTA reporting were systematically extracted and appraised using standardised checklists (STARD, CHARMS, CLAIM, TRIPOD, PROBAST, QUADAS-2).

RESULTS

Six of 5636 identified studies were eligible. The common target condition was intracranial haemorrhage (ICH), and intended workflow roles auxiliary to experts. Due to methodological and clinical between-study variation, meta-analysis was inappropriate. The scan-level sensitivity exceeded 90 % in 5/6 studies, while specificities ranged from 58,0-97,7 %. The SROC 95 % predictive region was markedly broader than the confidence region, ranging above 50 % sensitivity and 20 % specificity. All studies had unclear or high risk of bias and concern for applicability (QUADAS-2, PROBAST), and reporting adherence was below 50 % in 20 of 32 TRIPOD items.

CONCLUSION

0.01 % of identified studies met the eligibility criteria. The evidence on the DTA of CNN-CADx models for emergency head CT scans remains limited in the scope of this review, as the reviewed studies were scarce, inapt for meta-analysis and undermined by inadequate methodological conduct and reporting. Properly conducted, external validation remains preliminary for evaluating the clinical potential of AI-CADx models, but prospective and pragmatic clinical validation in comparative trials remains most crucial. In conclusion, future AI-CADx research processes should be methodologically standardized and reported in a clinically meaningful way to avoid research waste.

Development and Validation of a Scoring System (SAGA Score) to Predict Weight Loss in Community-Dwelling, Self-Supported Older Adults

This retrospective cohort study explored the prevalence of substantial weight loss (≥10% per year) in independent older individuals in order to develop and validate a scoring system for high-risk group identification and targeted intervention against malnutrition. We used insurance claims and the Kokuho Database (KDB), a nationwide repository of Japanese-specific health checkups and health assessments for the older people. The study included 12,882 community-dwelling individuals aged 75 years and older who were self-supported in their activities of daily living in Saga Prefecture, Japan. Health evaluations and questionnaires categorized weight-loss factors into organic, physiological, psychological, and non-medical domains. The resulting scoring system (SAGA score), incorporating logistic regression models, predicted ≥ 10% annual weight-loss risk. The results revealed a 1.7% rate of annual substantial weight loss, with the SAGA score effectively stratifying the participants into low-, intermediate-, and high-risk categories. The high-risk category exhibited a weight-loss rate of 17.6%, highlighting the utility of this scoring system for targeted prevention. In conclusion, the validated SAGA score is a crucial tool for identifying individuals at high risk of significant weight loss, enabling tailored interventions and social support benefiting both older individuals and their relatives.

Development and validation of an interpretable model for predicting sepsis mortality across care settings

There are numerous prognostic predictive models for evaluating mortality risk, but current scoring models might not fully cater to sepsis patients' needs. This study developed and validated a new model for sepsis patients that is suitable for any care setting and accurately forecasts 28-day mortality. The derivation dataset, gathered from 20 hospitals between September 2019 and December 2021, contrasted with the validation dataset, collected from 15 hospitals from January 2022 to December 2022. In this study, 7436 patients were classified as members of the derivation dataset, and 2284 patients were classified as members of the validation dataset. The point system model emerged as the optimal model among the tested predictive models for foreseeing sepsis mortality. For community-acquired sepsis, the model's performance was satisfactory (derivation dataset AUC: 0.779, 95% CI 0.765-0.792; validation dataset AUC: 0.787, 95% CI 0.765-0.810). Similarly, for hospital-acquired sepsis, it performed well (derivation dataset AUC: 0.768, 95% CI 0.748-0.788; validation dataset AUC: 0.729, 95% CI 0.687-0.770). The calculator, accessible at https://avonlea76.shinyapps.io/shiny_app_up/ , is user-friendly and compatible. The new predictive model of sepsis mortality is user-friendly and satisfactorily forecasts 28-day mortality. Its versatility lies in its applicability to all patients, encompassing both community-acquired and hospital-acquired sepsis.

Association of weight-adjusted waist index with all-cause mortality among non-Asian individuals: a national population-based cohort study

INTRODUCTION

The Weight-Adjusted Waist Index (WWI) is a new indicator of obesity that is associated with all-cause mortality in Asian populations. Our study aimed to investigate the linear and non-linear associations between WWI and all-cause mortality in non-Asian populations in the United States, and whether WWI was superior to traditional obesity indicators as a predictor of all-cause mortality.

METHODS

We conducted a cohort study using data from the 2011-2018 National Health and Nutrition Examination Survey (NHANES), involving 18,592 participants. We utilized Cox proportional hazard models to assess the association between WWI, BMI, WC, and the risk of all-cause mortality, and performed subgroup analyses and interaction tests. We also employed a receiver operating characteristics (ROC) curve study to evaluate the effectiveness of WWI, BMI, and WC in predicting all-cause mortality.

RESULTS

After adjusting for confounders, WWI, BMI, and WC were positively associated with all-cause mortality. The performance of WWI, BMI, and WC in predicting all-cause mortality yielded AUCs of 0.697, 0.524, and 0.562, respectively. The data also revealed a U-shaped relationship between WWI and all-cause mortality. Race and cancer modified the relationship between WWI and all-cause mortality, with the relationship being negatively correlated in African Americans and cancer patients.

CONCLUSIONS

In non-Asian populations in the United States, there is a U-shaped relationship between WWI and all-cause mortality, and WWI outperforms BMI and WC as a predictor of all-cause mortality. These findings may contribute to a better understanding and prediction of the relationship between obesity and mortality, and provide support for effective obesity management strategies.

CT-based quantification of intratumoral heterogeneity for predicting pathologic complete response to neoadjuvant immunochemotherapy in non-small cell lung cancer

Objectives

To investigate the prediction of pathologic complete response (pCR) in patients with non-small cell lung cancer (NSCLC) undergoing neoadjuvant immunochemotherapy (NAIC) using quantification of intratumoral heterogeneity from pre-treatment CT image.

Methods

This retrospective study included 178 patients with NSCLC who underwent NAIC at 4 different centers. The training set comprised 108 patients from center A, while the external validation set consisted of 70 patients from center B, center C, and center D. The traditional radiomics model was contrasted using radiomics features. The radiomics features of each pixel within the tumor region of interest (ROI) were extracted. The optimal division of tumor subregions was determined using the K-means unsupervised clustering method. The internal tumor heterogeneity habitat model was developed using the habitats features from each tumor sub-region. The LR algorithm was employed in this study to construct a machine learning prediction model. The diagnostic performance of the model was evaluated using criteria such as area under the receiver operating characteristic curve (AUC), accuracy, specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV).

Results

In the training cohort, the traditional radiomics model achieved an AUC of 0.778 [95% confidence interval (CI): 0.688-0.868], while the tumor internal heterogeneity habitat model achieved an AUC of 0.861 (95% CI: 0.789-0.932). The tumor internal heterogeneity habitat model exhibits a higher AUC value. It demonstrates an accuracy of 0.815, surpassing the accuracy of 0.685 achieved by traditional radiomics models. In the external validation cohort, the AUC values of the two models were 0.723 (CI: 0.591-0.855) and 0.781 (95% CI: 0.673-0.889), respectively. The habitat model continues to exhibit higher AUC values. In terms of accuracy evaluation, the tumor heterogeneity habitat model outperforms the traditional radiomics model, achieving a score of 0.743 compared to 0.686.

Conclusion

The quantitative analysis of intratumoral heterogeneity using CT to predict pCR in NSCLC patients undergoing NAIC holds the potential to inform clinical decision-making for resectable NSCLC patients, prevent overtreatment, and enable personalized and precise cancer management.

Development and validation of a LASSO prediction model for cisplatin induced nephrotoxicity: a case-control study in China

BACKGROUND

Early identification of high-risk individuals with cisplatin-induced nephrotoxicity (CIN) is crucial for avoiding CIN and improving prognosis. In this study, we developed and validated a CIN prediction model based on general clinical data, laboratory indications, and genetic features of lung cancer patients before chemotherapy.

METHODS

We retrospectively included 696 lung cancer patients using platinum chemotherapy regimens from June 2019 to June 2021 as the traing set to construct a predictive model using Absolute shrinkage and selection operator (LASSO) regression, cross validation, and Akaike's information criterion (AIC) to select important variables. We prospectively selected 283 independent lung cancer patients from July 2021 to December 2022 as the test set to evaluate the model's performance.

RESULTS

The prediction model showed good discrimination and calibration, with AUCs of 0.9217 and 0.8288, sensitivity of 79.89% and 45.07%, specificity of 94.48% and 94.81%, in the training and test sets respectively. Clinical decision curve analysis suggested that the model has value for clinical use when the risk threshold ranges between 0.1 and 0.9. Precision-Recall (PR) curve shown in recall interval from 0.5 to 0.75: precision gradually declines with increasing Recall, up to 0.9.

CONCLUSIONS

Predictive models based on laboratory and demographic variables can serve as a beneficial complementary tool for identifying high-risk populations with CIN.

Data set terminology of deep learning in medicine: a historical review and recommendation

Medicine and deep learning-based artificial intelligence (AI) engineering represent two distinct fields each with decades of published history. The current rapid convergence of deep learning and medicine has led to significant advancements, yet it has also introduced ambiguity regarding data set terms common to both fields, potentially leading to miscommunication and methodological discrepancies. This narrative review aims to give historical context for these terms, accentuate the importance of clarity when these terms are used in medical deep learning contexts, and offer solutions to mitigate misunderstandings by readers from either field. Through an examination of historical documents, including articles, writing guidelines, and textbooks, this review traces the divergent evolution of terms for data sets and their impact. Initially, the discordant interpretations of the word 'validation' in medical and AI contexts are explored. We then show that in the medical field as well, terms traditionally used in the deep learning domain are becoming more common, with the data for creating models referred to as the 'training set', the data for tuning of parameters referred to as the 'validation (or tuning) set', and the data for the evaluation of models as the 'test set'. Additionally, the test sets used for model evaluation are classified into internal (random splitting, cross-validation, and leave-one-out) sets and external (temporal and geographic) sets. This review then identifies often misunderstood terms and proposes pragmatic solutions to mitigate terminological confusion in the field of deep learning in medicine. We support the accurate and standardized description of these data sets and the explicit definition of data set splitting terminologies in each publication. These are crucial methods for demonstrating the robustness and generalizability of deep learning applications in medicine. This review aspires to enhance the precision of communication, thereby fostering more effective and transparent research methodologies in this interdisciplinary field.