Development of prognostic models for predicting 90-day neurological function and mortality after cardiac arrest

BACKGROUND

The survivors of cardiac arrest experienced vary extent of hypoxic ischemic brain injury causing mortality and long-term neurologic disability. However, there is still a need to develop robust and reliable prognostic models that can accurately predict these outcomes.

OBJECTIVES

To establish reliable models for predicting 90-day neurological function and mortality in adult ICU patients recovering from cardiac arrest.

METHODS

We enrolled patients who had recovered from cardiac arrest at Binhaiwan Central Hospital of Dongguan, from January 2018 to July 2021. The study's primary outcome was 90-day neurological function, assessed and divided into two categories using the Cerebral Performance Category (CPC) scale: either good (CPC 1-2) or poor (CPC 3-5). The secondary outcome was 90-day mortality. We analyzed the relationships between risk factors and outcomes individually. A total of four models were developed: two multivariable logistic regression models (models 1 and 2) for predicting neurological function, and two Cox regression models (models 3 and 4) for predicting mortality. Models 2 and 4 included new neurological biomarkers as predictor variables, while models 1 and 3 excluded. We evaluated calibration, discrimination, clinical utility, and relative performance to establish superiority between the models.

RESULTS

Model 1 incorporates variables such as gender, site of cardiopulmonary resuscitation (CPR), total CPR time, and acute physiology and chronic health evaluation II (APACHE II) score, while model 2 includes gender, site of CPR, APACHE II score, and serum level of ubiquitin carboxy-terminal hydrolase L1 (UCH-L1). Model 2 outperforms model 1, showcasing a superior area under the receiver operating characteristic curve (AUC) of 0.97 compared to 0.83. Additionally, model 2 exhibits improved accuracy, sensitivity, and specificity. The decision curve analysis confirms the net benefit of model 2. Similarly, models 3 and 4 are designed to predict 90-day mortality. Model 3 incorporates the variables such as site of CPR, total CPR time, and APACHE II score, while model 4 includes APACHE II score, total CPR time, and serum level of UCH-L1. Model 4 outperforms model 3, showcasing an AUC of 0.926 and a C-index of 0.830. The clinical decision curve analysis also confirms the net benefit of model 4.

CONCLUSIONS

By integrating new neurological biomarkers, we have successfully developed enhanced models that can predict 90-day neurological function and mortality outcomes more accurately.

MRI radiomics in head and neck cancer from reproducibility to combined approaches

The clinical applicability of radiomics in oncology depends on its transferability to real-world settings. However, the absence of standardized radiomics pipelines combined with methodological variability and insufficient reporting may hamper the reproducibility of radiomic analyses, impeding its translation to clinics. This study aimed to identify and replicate published, reproducible radiomic signatures based on magnetic resonance imaging (MRI), for prognosis of overall survival in head and neck squamous cell carcinoma (HNSCC) patients. Seven signatures were identified and reproduced on 58 HNSCC patients from the DB2Decide Project. The analysis focused on: assessing the signatures' reproducibility and replicating them by addressing the insufficient reporting; evaluating their relationship and performances; and proposing a cluster-based approach to combine radiomic signatures, enhancing the prognostic performance. The analysis revealed key insights: (1) despite the signatures were based on different features, high correlations among signatures and features suggested consistency in the description of lesion properties; (2) although the uncertainties in reproducing the signatures, they exhibited a moderate prognostic capability on an external dataset; (3) clustering approaches improved prognostic performance compared to individual signatures. Thus, transparent methodology not only facilitates replication on external datasets but also advances the field, refining prognostic models for potential personalized medicine applications.

Prognostic Model Construction of Disulfidptosis-Related Genes and Targeted Anticancer Drug Research in Pancreatic Cancer

Pancreatic cancer stands as one of the most lethal malignancies, characterized by delayed diagnosis, high mortality rates, limited treatment efficacy, and poor prognosis. Disulfidptosis, a recently unveiled modality of cell demise induced by disulfide stress, has emerged as a critical player intricately associated with the onset and progression of various cancer types. It has emerged as a promising candidate biomarker for cancer diagnosis, prognosis assessment, and treatment strategies. In this study, we have effectively established a prognostic risk model for pancreatic cancer by incorporating multiple differentially expressed long non-coding RNAs (DElncRNAs) closely linked to disulfide-driven cell death. Our investigation delved into the nuanced relationship between the DElncRNA-based predictive model for disulfide-driven cell death and the therapeutic responses to anticancer agents. Our findings illuminate that the high-risk subgroup exhibits heightened susceptibility to the small molecule compound AZD1208, positioning it as a prospective therapeutic agent for pancreatic cancer. Finally, we have elucidated the underlying mechanistic potential of AZD1208 in ameliorating pancreatic cancer through its targeted inhibition of the peroxisome proliferator-activated receptor-γ (PPARG) protein, employing an array of comprehensive analytical methods, including molecular docking and molecular dynamics (MD) simulations. This study explores disulfidptosis-related genes, paving the way for the development of targeted therapies for pancreatic cancer and emphasizing their significance in the field of oncology. Furthermore, through computational biology approaches, the drug AZD1208 was identified as a potential treatment targeting the PPARG protein for pancreatic cancer. This discovery opens new avenues for exploring targets and screening drugs for pancreatic cancer.

GRADE concept paper 8: judging the certainty of discrimination performance estimates of prognostic models in a body of validation studies

BACKGROUND

Prognostic models incorporate multiple prognostic factors to estimate the likelihood of future events for individual patients based on their prognostic factor values. Evaluating these models crucially involves conducting studies to assess their predictive performance, like discrimination. Systematic reviews and meta-analyses of these validation studies play an essential role in selecting models for clinical practice.

METHODS

In this paper, we outline 3 thresholds to determine the target for certainty rating in the discrimination of prognostic models, as observed across a body of validation studies.

RESULTS AND CONCLUSION

We propose 3 thresholds when rating the certainty of evidence about a prognostic model's discrimination. The first threshold amounts to rating certainty in the model's ability to classify better than random chance. The other 2 approaches involve setting thresholds informed by other mechanisms for classification: clinician intuition or an alternative prognostic model developed for the same disease area and outcome. The choice of threshold will vary based on the context. Instead of relying on arbitrary discrimination cut-offs, our approach positions the observed discrimination within an informed spectrum, potentially aiding decisions about a prognostic model's practical utility.

External validation of six COVID-19 prognostic models for predicting mortality risk in older populations in a hospital, primary care, and nursing home setting

OBJECTIVES

To systematically evaluate the performance of COVID-19 prognostic models and scores for mortality risk in older populations across three health-care settings: hospitals, primary care, and nursing homes.

STUDY DESIGN AND SETTING

This retrospective external validation study included 14,092 older individuals of ≥70 years of age with a clinical or polymerase chain reaction-confirmed COVID-19 diagnosis from March 2020 to December 2020. The six validation cohorts include three hospital-based (CliniCo, COVID-OLD, COVID-PREDICT), two primary care-based (Julius General Practitioners Network/Academisch network huisartsgeneeskunde/Network of Academic general Practitioners, PHARMO), and one nursing home cohort (YSIS) in the Netherlands. Based on a living systematic review of COVID-19 prediction models using Prediction model Risk Of Bias ASsessment Tool for quality and risk of bias assessment and considering predictor availability in validation cohorts, we selected six prognostic models predicting mortality risk in adults with COVID-19 infection (GAL-COVID-19 mortality, 4C Mortality Score, National Early Warning Score 2-extended model, Xie model, Wang clinical model, and CURB65 score). All six prognostic models were validated in the hospital cohorts and the GAL-COVID-19 mortality model was validated in all three healthcare settings. The primary outcome was in-hospital mortality for hospitals and 28-day mortality for primary care and nursing home settings. Model performance was evaluated in each validation cohort separately in terms of discrimination, calibration, and decision curves. An intercept update was performed in models indicating miscalibration followed by predictive performance re-evaluation.

MAIN OUTCOME MEASURE

In-hospital mortality for hospitals and 28-day mortality for primary care and nursing home setting.

RESULTS

All six prognostic models performed poorly and showed miscalibration in the older population cohorts. In the hospital settings, model performance ranged from calibration-in-the-large -1.45 to 7.46, calibration slopes 0.24-0.81, and C-statistic 0.55-0.71 with 4C Mortality Score performing as the most discriminative and well-calibrated model. Performance across health-care settings was similar for the GAL-COVID-19 model, with a calibration-in-the-large in the range of -2.35 to -0.15 indicating overestimation, calibration slopes of 0.24-0.81 indicating signs of overfitting, and C-statistic of 0.55-0.71.

CONCLUSION

Our results show that most prognostic models for predicting mortality risk performed poorly in the older population with COVID-19, in each health-care setting: hospital, primary care, and nursing home settings. Insights into factors influencing predictive model performance in the older population are needed for pandemic preparedness and reliable prognostication of health-related outcomes in this demographic.

Disulfidptosis-Related lncRNA for the Establishment of Novel Prognostic Signature and Therapeutic Response Prediction to Endometrial Cancer

Disulfidptosis, a newly discovered cellular death mechanism initiated by disulfide stress, features elevated expression of SLC7A11 and restricted glucose availability, rendering it a possible therapeutic target for cancer. Endometrial cancer of the uterine corpus (ECUC) ranks among prevalent gynecological malignancies. Long non-coding RNAs (lncRNAs) have been implicated in ECUC's metabolic pathways, invasive capabilities, and metastatic processes. Yet, the prognostic implications of Disulfidptosis-Linked lncRNAs (DLLs) in ECUC remain ambiguous. Transcriptome and clinical datasets related to ECUC were sourced from The Cancer Genome Atlas (TCGA), while genes linked with disulfidptosis were identified from existing literature. A panel of ten DLLs was discerned through least absolute shrinkage and selection operator (LASSO) coupled with Cox regression methods to formulate and validate risk prognostic models. We engineered a nomogram for ECUC patient prognosis forecasting and further examined the model via gene set enrichment analysis (GSEA), principal component analysis (PCA), gene set analysis (GSA), immune profiling, and sensitivity to antineoplastic agents. Prognostic models employing a set of ten DLLs (including AC005034.2, AC020765.2, AL158071.4, AL161663.2, AP000787.1, CR392039.3, EMSLR, SEC24B-AS1, Z69733.1, Z94721.3) were established. Based on median risk values, patient samples were stratified into high- and low-risk cohorts, revealing notable differences in survival across both training and validation datasets. The risk scores, when amalgamated with clinical variables, acted as standalone predictors of prognosis. GSEA findings indicated that the high-risk category predominantly aligned with pathways like extracellular matrix interactions and cell adhesion molecules, suggesting a likely association with metastatic potential. Concurrently, we scrutinized disparities in immune function and tumor mutational burden across risk categories and identified anticancer drugs with likely efficacy. In summary, a set of ten DLLs proved useful in forecasting patient outcomes and holds potential for informing targeted therapeutic approaches in ECUC.

A comparative study of explainable ensemble learning and logistic regression for predicting in-hospital mortality in the emergency department

This study addresses the challenges associated with emergency department (ED) overcrowding and emphasizes the need for efficient risk stratification tools to identify high-risk patients for early intervention. While several scoring systems, often based on logistic regression (LR) models, have been proposed to indicate patient illness severity, this study aims to compare the predictive performance of ensemble learning (EL) models with LR for in-hospital mortality in the ED. A cross-sectional single-center study was conducted at the ED of Imam Reza Hospital in northeast Iran from March 2016 to March 2017. The study included adult patients with one to three levels of emergency severity index. EL models using Bagging, AdaBoost, random forests (RF), Stacking and extreme gradient boosting (XGB) algorithms, along with an LR model, were constructed. The training and validation visits from the ED were randomly divided into 80% and 20%, respectively. After training the proposed models using tenfold cross-validation, their predictive performance was evaluated. Model performance was compared using the Brier score (BS), The area under the receiver operating characteristics curve (AUROC), The area and precision-recall curve (AUCPR), Hosmer-Lemeshow (H-L) goodness-of-fit test, precision, sensitivity, accuracy, F1-score, and Matthews correlation coefficient (MCC). The study included 2025 unique patients admitted to the hospital's ED, with a total percentage of hospital deaths at approximately 19%. In the training group and the validation group, 274 of 1476 (18.6%) and 152 of 728 (20.8%) patients died during hospitalization, respectively. According to the evaluation of the presented framework, EL models, particularly Bagging, predicted in-hospital mortality with the highest AUROC (0.839, CI (0.802-0.875)) and AUCPR = 0.64 comparable in terms of discrimination power with LR (AUROC (0.826, CI (0.787-0.864)) and AUCPR = 0.61). XGB achieved the highest precision (0.83), sensitivity (0.831), accuracy (0.842), F1-score (0.833), and the highest MCC (0.48). Additionally, the most accurate models in the unbalanced dataset belonged to RF with the lowest BS (0.128). Although all studied models overestimate mortality risk and have insufficient calibration (P > 0.05), stacking demonstrated relatively good agreement between predicted and actual mortality. EL models are not superior to LR in predicting in-hospital mortality in the ED. Both EL and LR models can be considered as screening tools to identify patients at risk of mortality.

Development and validation of prognostic models for small cell lung cancer patients with liver metastasis: a SEER population-based study

BACKGROUND

This study was to establish and validate prediction models to predict the cancer-specific survival (CSS) and overall survival (OS) of small-cell lung cancer (SCLC) patients with liver metastasis.

METHODS

In the retrospective cohort study, SCLC patients with liver metastasis between 2010 and 2015 were retrospectively retrieved from the Surveillance, Epidemiology, and End Results (SEER) database. Patients were randomly divided into the training group and testing group (3: 1 ratio). The Cox proportional hazards model was used to determine the predictive factors for CSS and OS in SCLC with liver metastasis. The prediction models were conducted based on the predictive factors. The performances of the prediction models were evaluated by concordance indexes (C-index), and calibration plots. The clinical value of the models was evaluated by decision curve analysis (DCA).

RESULTS

In total, 8,587 patients were included, with 154 patients experiencing CSS and 154 patients experiencing OS. The median follow-up was 3 months. Age, gender, marital status, N stage, lung metastases, multiple metastases surgery of metastatic site, chemotherapy, and radiotherapy were independent predictive factors for the CSS and OS of SCLC patients with liver metastasis. The prediction models presented good performances of CSS and OS among patients with liver metastasis, with the C-index for CSS being 0.724, whereas the C-index for OS was 0.732, in the training set. The calibration curve showed a high degree of consistency between the actual and predicted CSS and OS. DCA suggested that the prediction models provided greater net clinical benefit to these patients.

CONCLUSION

Our prediction models showed good predictive performance for the CSS and OS among SCLC patients with liver metastasis. Our developed nomograms may help clinicians predict CSS and OS in SCLC patients with liver metastasis.

Prognostic models and autoimmune liver diseases

Autoimmune liver diseases (AILDs) are complex diseases with unknown causes and immune-mediated pathophysiology. In primary biliary cholangitis (PBC) and autoimmune hepatitis (AIH) disease modifying drugs are available which improve patient quality and quantity of life. In primary sclerosing cholangitis (PSC) no medical therapy is available and the only accepted treatment is liver transplantation (LT). PBC, PSC and AIH possess features that describe the archetype of patients within each disorder. On the other hand, the classical disorders are not homogeneous, and patients within each diagnosis may present with a range of clinical, biochemical, serological, and histological findings. Singularly, they are considered rare diseases, but together, they account for approximately 20% of LTs in Europe and USA. Management of these patients is complex, as AILDs are relatively uncommon in clinical practice with challenges in developing expertise, disease presentation can be sneaky, clinical phenotypes and disease course are heterogeneous. Prognostic models are key tools for clinicians to assess patients' risk and to provide personalized care to patients. Aim of this review is to discuss challenges of the management of AILDs and how the available prognostic models can help. We will discuss the prognostic models developed in AILDs, with a special focus on the prognostic models that can support the clinical management of patients with AILDs: in PBC models based on ursodeoxycholic acid (UDCA) response and markers of liver fibrosis; in PSC several markers including biochemistry, disease stage and radiological semiquantitative markers; and finally in AIH, markers of disease stage and disease activity.

Humero-femoral index and diabetes risk in the US population- a case study

Background

The between-subject variability in diabetes risk persists in epidemiological studies, even after accounting for obesity. We investigated whether the humero-femoral index (HFI) was associated with prevalence of type 2 diabetes mellitus (T2DM) and assessed the incremental value of HFI as a marker of T2DM.

Methods

This population-based cross-sectional study used data from the National Health and Nutrition Examination Survey from 1999 to 2018. We assessed 42,088 adults aged ≥ 30 years. HFI was defined as the upper arm length/upper leg length ratio. The outcome included undiagnosed diabetes (based on 2-hour plasma glucose levels, fasting glucose and hemoglobin A1C) and history of diabetes (diagnosed diabetes or taking antidiabetic drugs).

Results

As compared with the bottom quartile, the prevalence ratio of T2DM was 1.28 (95% CI 1.19-1.38) in the second, 1.61 (95% CI 1.50-1.72) in the third, and 1.75 (95% CI 1.64-1.88) in the fourth quartile of HFI (P for trend < 0.0001). The positive association remained consistent within different patterns of BMI and WC in men but was rendered null in women. After adding HFI to the reference model (including WC only), the discrimination slopes increased by 60.0% in men and 51.1% in women.

Conclusion

Our findings suggest that HFI may be a key component in body structure contributing to the risk of T2DM. In men, the highest HFI was associated with elevated prevalence of T2DM, independent of BMI and WC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40200-023-01251-z.

Statistical primer: using prognostic models to predict the future: what cardiothoracic surgery can learn from Strictly Come Dancing

OBJECTIVES

Prognostic models are widely used across medicine and within cardiothoracic surgery, where predictive tools such as EuroSCORE are commonplace. Such models are a useful component of clinical assessment but may be misapplied. In this article, we demonstrate some of the major issues with risk scores by using the popular BBC television programme Strictly Come Dancing (known as Dancing with the Stars in many other countries) as an example.

METHODS

We generated a multivariable prognostic model using data from the then-completed 19 series of Strictly Come Dancing to predict prospectively the results of the 20th series.

RESULTS

The initial model based solely on demographic data was limited in its predictive value (0.25, 0.22; R2 and Spearman's rank correlation, respectively) but was substantially improved following the introduction of early judges' scores deemed representative of whether contestants could actually dance (0.40, 0.30). We then utilize our model to discuss the difficulties and pitfalls in using and interpreting prognostic models in cardiothoracic surgery and beyond, particularly where these do not adequately capture potentially important prognostic information.

CONCLUSION

Researchers and clinicians alike should use prognostic models cautiously and not extrapolate conclusions from demographic data alone.

Prediction of Mortality and Morbidity After Severe Traumatic Brain Injury: A Comparison Between Rotterdam and Richmond Computed Tomography Scan Scoring System

OBJECTIVE

Accurate prediction of the morbidity and mortality outcomes of traumatic brain injury patients is still challenging. In the present study, we aimed to compare the predictive value of the Richmond and Rotterdam scoring systems as two novel computed tomography-based predictive models.

METHODS

We retrospectively analyzed 1400 subjects who suffered from severe traumatic brain injury and were admitted to Emtiaz Hospital, a tertiary referral trauma center in Shiraz, south of Iran, from January 2018 to December 2019. We evaluated the 1-month results; considering two primary factors: mortality and morbidity. The patients' condition was the basis for this assessment. We conducted a logistic regression analysis to determine the association between scoring systems and outcomes. To determine the optimal threshold value, we utilized the receiver operating characteristic curve model.

RESULTS

The mean age of participants was 36.61 ± 17.58 years, respectively. Concerning predicting the mortality rate, the area under the curve (AUC) for the Rotterdam score was relatively low 0.64 (95% confidence interval: 0.60, 0.67), while the Richmond score had a higher AUC 0.74 (0.71-0.77), which demonstrated the superiority of this scoring system. Moreover, the Richmond score was more accurate for predicting 1-month morbidity with AUC: 0.71 (0.69, 0.74) versus 0.62 (0.59, 0.65).

CONCLUSIONS

The Richmond scoring system demonstrated more accurate predictions for the present outcomes. The simplicity and predictive value of the Richmond score make this system an ideal option for use in emergency settings and centers with high patient loads.

Development and validation of colorectal cancer risk prediction tools: A comparison of models

BACKGROUND

Identification of individuals at elevated risk can improve cancer screening programmes by permitting risk-adjusted screening intensities. Previous work introduced a prognostic model using sex, age and two preceding faecal haemoglobin concentrations to predict the risk of colorectal cancer (CRC) in the next screening round. Using data of 3 screening rounds, this model attained an area under the receiver-operating-characteristic curve (AUC) of 0.78 for predicting advanced neoplasia (AN). We validated this existing logistic regression (LR) model and attempted to improve it by applying a more flexible machine-learning approach.

METHODS

We trained an existing LR and a newly developed random forest (RF) model using updated data from 219,257 third-round participants of the Dutch CRC screening programme until 2018. For both models, we performed two separate out-of-sample validations using 1,137,599 third-round participants after 2018 and 192,793 fourth-round participants from 2020 onwards. We evaluated the AUC and relative risks of the predicted high-risk groups for the outcomes AN and CRC.

RESULTS

For third-round participants after 2018, the AUC for predicting AN was 0.77 (95% CI: 0.76-0.77) using LR and 0.77 (95% CI: 0.77-0.77) using RF. For fourth-round participants, the AUCs were 0.73 (95% CI: 0.72-0.74) and 0.73 (95% CI: 0.72-0.74) for the LR and RF models, respectively. For both models, the 5% with the highest predicted risk had a 7-fold risk of AN compared to average, whereas the lowest 80% had a risk below the population average for third-round participants.

CONCLUSION

The LR is a valid risk prediction method in stool-based screening programmes. Although predictive performance declined marginally, the LR model still effectively predicted risk in subsequent screening rounds. An RF did not improve CRC risk prediction compared to an LR, probably due to the limited number of available explanatory variables. The LR remains the preferred prediction tool because of its interpretability.

Prognostic models of in-hospital mortality of intensive care patients using neural representation of unstructured text: A systematic review and critical appraisal

OBJECTIVE

To review and critically appraise published and preprint reports of prognostic models of in-hospital mortality of patients in the intensive-care unit (ICU) based on neural representations (embeddings) of clinical notes.

METHODS

PubMed and arXiv were searched up to August 1, 2022. At least two reviewers independently selected the studies that developed a prognostic model of in-hospital mortality of intensive-care patients using free-text represented as embeddings and extracted data using the CHARMS checklist. Risk of bias was assessed using PROBAST. Reporting on the model was assessed with the TRIPOD guideline. To assess the machine learning components that were used in the models, we present a new descriptive framework based on different techniques to represent text and provide predictions from text. The study protocol was registered in the PROSPERO database (CRD42022354602).

RESULTS

Eighteen studies out of 2,825 were included. All studies used the publicly-available MIMIC dataset. Context-independent word embeddings are widely used. Model discrimination was provided by all studies (AUROC 0.75-0.96), but measures of calibration were scarce. Seven studies used both structural clinical variables and notes. Model discrimination improved when adding clinical notes to variables. None of the models was externally validated and often a simple train/test split was used for internal validation. Our critical appraisal demonstrated a high risk of bias in all studies and concerns regarding their applicability in clinical practice.

CONCLUSION

All studies used a neural architecture for prediction and were based on one publicly available dataset. Clinical notes were reported to improve predictive performance when used in addition to only clinical variables. Most studies had methodological, reporting, and applicability issues. We recommend reporting both model discrimination and calibration, using additional data sources, and using more robust evaluation strategies, including prospective and external validation. Finally, sharing data and code is encouraged to improve study reproducibility.

A systematic review with external validation

OBJECTIVES

To identify prognostic models which estimate the risk of critical COVID-19 in hospitalized patients and to assess their validation properties.

STUDY DESIGN AND SETTING

We conducted a systematic review in Medline (up to January 2021) of studies developing or updating a model that estimated the risk of critical COVID-19, defined as death, admission to intensive care unit, and/or use of mechanical ventilation during admission. Models were validated in two datasets with different backgrounds (HM [private Spanish hospital network], n = 1,753, and ICS [public Catalan health system], n = 1,104), by assessing discrimination (area under the curve [AUC]) and calibration (plots).

RESULTS

We validated 18 prognostic models. Discrimination was good in nine of them (AUCs ≥ 80%) and higher in those predicting mortality (AUCs 65%-87%) than those predicting intensive care unit admission or a composite outcome (AUCs 53%-78%). Calibration was poor in all models providing outcome's probabilities and good in four models providing a point-based score. These four models used mortality as outcome and included age, oxygen saturation, and C-reactive protein among their predictors.

CONCLUSION

The validity of models predicting critical COVID-19 by using only routinely collected predictors is variable. Four models showed good discrimination and calibration when externally validated and are recommended for their use.

Prediction of long-term mortality following hip fracture surgery: evaluation of three risk models

INTRODUCTION

Several prognostic models have been developed for mortality in hip fracture patients, but their accuracy for long-term prediction is unclear. This study evaluates the performance of three models assessing 30-day, 1-year and 8-year mortality after hip fracture surgery: the Nottingham Hip Fracture Score (NHFS), the model developed by Holt et al. and the Hip fracture Estimator of Mortality Amsterdam (HEMA).

MATERIALS AND METHODS

Patients admitted with a fractured hip between January 2012 and June 2013 were included in this retrospective cohort study. Relevant variables used by the three models were collected, as were mortality data. Predictive performance was assessed in terms of discrimination with the area under the receiver operating characteristic curve and calibration with the Hosmer-Lemeshow goodness-of-fit test. Clinical usefulness was evaluated by determining risk groups for each model, comparing differences in mortality using Kaplan-Meier curves, and by assessing positive and negative predictive values.

RESULTS

A total of 344 patients were included for analysis. Observed mortality rates were 6.1% after 30 days, 19.1% after 1 year and 68.6% after 8 years. The NHFS and the model by Holt et al. demonstrated good to excellent discrimination and adequate calibration for both short- and long-term mortality prediction, with similar clinical usefulness measures. The HEMA demonstrated inferior prediction of 30-day and 8-year mortality, with worse discriminative abilities and a significant lack of fit.

CONCLUSIONS

The NHFS and the model by Holt et al. allowed for accurate identification of low- and high-risk patients for both short- and long-term mortality after a fracture of the hip. The HEMA performed poorly. When considering predictive performance and ease of use, the NHFS seems most suitable for implementation in daily clinical practice.

Estimating the Effectiveness of Kidney Cancer Screening Within Lung Cancer Screening Programmes: A Validation in UK Biobank

In the absence of population-based screening, addition of screening for kidney cancer to lung cancer screening could provide an efficient and low-resource means to improve early detection. In this study, we used the UK Biobank cohort (n = 442 865) to determine the performance of the Yorkshire Lung Cancer Screening Trial (YLST) eligibility criteria for selecting individuals for kidney cancer screening. We measured the performance of two models widely used to determine eligibility for lung cancer screening (PLCO[m2012] and the Liverpool-Lung-Project-v2) and the performance of the combined YLST criteria. We found that the lung cancer models have discrimination (area under the receiver operating curve) between 0.60 and 0.68 for kidney cancer. In the UK, one in four cases (25%) of kidney cancer cases is expected to occur in those eligible for lung cancer screening, and one case of kidney cancer detected for every 200 people invited to lung cancer screening. These results suggest that adding kidney cancer screening to lung cancer screening would be an effective strategy to improve early detection rates of kidney cancer. However, most kidney cancers would not be picked up by this approach. This analysis does not address other important considerations about kidney cancer screening, such as overdiagnosis. PATIENT SUMMARY: It has been proposed that adding-on kidney cancer screening to lung cancer screening (both carried out by a computed tomography scan of the chest/abdomen) would be an easy and low-cost way of detecting cases of kidney cancer earlier, when these can be treated more easily. Lung cancer screening is usually targeted at people who are at a high risk (eg, older smokers); therefore, here we look at whether the same group of people are also at a high risk of kidney cancer. Our analysis shows that one in four people later diagnosed with kidney cancer are also at a high risk of lung cancer; hence, a combined screening programme could detect up to a quarter of kidney cancers.

Construction of a novel MPT-driven necrosis-related lncRNAs signature for prognosis prediction in laryngeal squamous cell carcinoma

Mitochondrial permeability transition (MPT)-driven necrosis, a type of programmed cell death, has recently gained much attention in a variety of tumor types. Few studies have been performed to explore the role of MPT-driven necrosis-related lncRNAs (MPTDNRlncRNAs) in laryngeal squamous cell carcinoma (LSCC). The purpose of our study is to screen MPTDNRlncRNAs with prognostic value and to explore their potential roles in LSCC. The RNA-sequencing data and the corresponding clinical data of LSCC patients were obtained from the TCGA database, and those MPT-driven necrosis-related genes were extracted from the Gene Set Enrichment Analysis (GSEA) database. We identified MPTDNRlncRNAs differentially expressed in LSCC. Also, we gained MPT-driven necrosis-related prognostic lncRNAs by univariate cox regression analysis. A novel MPTDNRlncRNAs signature was constructed by LASSO-COX. The accuracy and utility of the MPTDNRlncRNAs signature were evaluated via a variety of statistical methods. Multiple bioinformatics tools were used to explore the underlying difference in biological functions and signaling pathways between the different risk groups. The expressions levels of MPTDNRlncRNAs were analyzed using RT-qPCR in LSCC cell line. Finally, we identified A 5 MPTDNRlncRNAs signature in LSCC. Our prognostic model demonstrated an efficient ability to predict outcomes. The proportion difference of immune cells in the subgroups were significant, such as M0 macrophage and T follicular helper cells. The in vitro experiments suggested that our MPTDNRlncRNAs were significantly different. This 5 MPTDNRlncRNAs signature is a prognostic biomarker for LSCC. More importantly, the novel biologic prognostic model can be utilized for personalized immunotherapy in LSCC patients.

A study protocol of external validation of eight COVID-19 prognostic models for predicting mortality risk in older populations in a hospital, primary care, and nursing home setting

BACKGROUND

The COVID-19 pandemic has a large impact worldwide and is known to particularly affect the older population. This paper outlines the protocol for external validation of prognostic models predicting mortality risk after presentation with COVID-19 in the older population. These prognostic models were originally developed in an adult population and will be validated in an older population (≥ 70 years of age) in three healthcare settings: the hospital setting, the primary care setting, and the nursing home setting.

METHODS

Based on a living systematic review of COVID-19 prediction models, we identified eight prognostic models predicting the risk of mortality in adults with a COVID-19 infection (five COVID-19 specific models: GAL-COVID-19 mortality, 4C Mortality Score, NEWS2 + model, Xie model, and Wang clinical model and three pre-existing prognostic scores: APACHE-II, CURB65, SOFA). These eight models will be validated in six different cohorts of the Dutch older population (three hospital cohorts, two primary care cohorts, and a nursing home cohort). All prognostic models will be validated in a hospital setting while the GAL-COVID-19 mortality model will be validated in hospital, primary care, and nursing home settings. The study will include individuals ≥ 70 years of age with a highly suspected or PCR-confirmed COVID-19 infection from March 2020 to December 2020 (and up to December 2021 in a sensitivity analysis). The predictive performance will be evaluated in terms of discrimination, calibration, and decision curves for each of the prognostic models in each cohort individually. For prognostic models with indications of miscalibration, an intercept update will be performed after which predictive performance will be re-evaluated.

DISCUSSION

Insight into the performance of existing prognostic models in one of the most vulnerable populations clarifies the extent to which tailoring of COVID-19 prognostic models is needed when models are applied to the older population. Such insight will be important for possible future waves of the COVID-19 pandemic or future pandemics.

A model based on immune-related lncRNA pairs and its potential prognostic value in immunotherapy for melanoma

A model based on long non-coding RNA (lncRNA) pairs independent of expression quantification was constructed to evaluate prognosis melanoma and response to immunotherapy in melanoma. RNA sequencing data and clinical information were retrieved and downloaded from The Cancer Genome Atlas and the Genotype-Tissue Expression databases. We identified differentially expressed immune-related lncRNAs (DEirlncRNAs), matched them, and used least absolute shrinkage and selection operator and Cox regression to construct predictive models. The optimal cutoff value of the model was determined using a receiver operating characteristic curve and used to categorize melanoma cases into high-risk and low-risk groups. The predictive efficacy of the model with respect to prognosis was compared with that of clinical data and ESTIMATE (Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data). Then, we analyzed the correlations of risk score with clinical characteristics, immune cell invasion, anti-tumor, and tumor-promoting activities. Differences in survival, degree of immune cell infiltration, and intensity of anti-tumor and tumor-promoting activities were also evaluated in the high- and low-risk groups. A model based on 21 DEirlncRNA pairs was established. Compared with ESTIMATE score and clinical data, this model could better predict outcomes of melanoma patients. Follow-up analysis of the model's effectiveness showed that patients in the high-risk group had poorer prognosis and were less likely to benefit from immunotherapy compared with those in the low-risk group. Moreover, there were differences in tumor-infiltrating immune cells between the high-risk and low-risk groups. By pairing the DEirlncRNA, we constructed a model to evaluate the prognosis of cutaneous melanoma independent of a specific level of lncRNA expression.

Prognostication in myeloproliferative neoplasms, including mutational abnormalities

Increasing knowledge of the molecular features of myeloproliferative neoplasms (MPNs) is being combined with existing prognostic models based on clinical, laboratory, and cytogenetic information. Mutation-enhanced international prognostic systems (MIPSS) for polycythemia vera (PV) and essential thrombocythemia (ET) have improved prognostic assessments. In the case of overt primary myelofibrosis (PMF), the MIPSS70 and its later revisions (MIPSS70+ and MIPSS70+ version 2.0) effectively predicted the overall survival (OS) of patients. Because post-PV and post-ET myelofibrosis have different biological and clinical courses compared to overt PMF, the myelofibrosis secondary to PV and ET-prognostic model was developed. Although these molecular-inspired prognostic models need to be further validated in future studies, they are expected to improve the prognostic power in patients with MPNs in the molecular era. Efforts are being made to predict survival after the use of specific drugs or allogeneic hematopoietic stem cell transplantation. These treatment outcome prediction models enable the establishment of personalized treatment strategies, thereby improving the OS of patients with MPNs.

Long-term prediction of functional independence using adjusted and unadjusted single items of the functional independence measure (FIM) at discharge from rehabilitation

CONTEXT

Being able to survive in the long-term independently is of concern to patients with spinal cord injury (SCI), their relatives, and to those providing or planning health care, especially at rehabilitation discharge. Most previous studies have attempted to predict functional dependency in activities of daily living within one year after injury.

OBJECTIVES

(1) build 18 different predictive models, each model using one FIM (Functional Independence Measure) item, assessed at discharge, as independent predictor of total FIM score at chronic phase (3-6 years post-injury) (2) build three different predictive models, using in each model an item from a different FIM domain with the highest predictive power obtained in objective (1) to predict "good" functional independence at chronic phase and (3) adjust the 3 models from objective (2) with known confounding factors.

METHODS

This observational study included 461 patients admitted to rehabilitation between 2009 and 2019. We applied regression models to predict total FIM score and "good" functional independence (FIM motor score ≥ 65) reporting adjusted R², odds ratios, ROC-AUC (95% CI) tested using 10-fold cross-validation.

RESULTS

The top three predictors, each from a different FIM domain, were Toilet (adjusted R² = 0.53, Transfers domain), Toileting (adjusted R² = 0.46, Self-care domain), and Bowel (adjusted R² = 0.35, Sphincter control domain). These three items were also predictors of "good" functional independence (AUC: 0.84-0.87) and their predictive power increased (AUC: 0.88-0.93) when adjusted by age, paraplegia, time since injury, and length of stay.

CONCLUSIONS

Discharge FIM items accurately predict long-term functional independence.

Development and validation of echocardiography-based machine-learning models to predict mortality

BACKGROUND

Echocardiography (echo) based machine learning (ML) models may be useful in identifying patients at high-risk of all-cause mortality.

METHODS

We developed ML models (ResNet deep learning using echo videos and CatBoost gradient boosting using echo measurements) to predict 1-year, 3-year, and 5-year mortality. Models were trained on the Mackay dataset, Taiwan (6083 echos, 3626 patients) and validated in the Alberta HEART dataset, Canada (997 echos, 595 patients). We examined the performance of the models overall, and in subgroups (healthy controls, at risk of heart failure (HF), HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF)). We compared the models' performance to the MAGGIC risk score, and examined the correlation between the models' predicted probability of death and baseline quality of life as measured by the Kansas City Cardiomyopathy Questionnaire (KCCQ).

FINDINGS

Mortality rates at 1-, 3- and 5-years were 14.9%, 28.6%, and 42.5% in the Mackay cohort, and 3.0%, 10.3%, and 18.7%, in the Alberta HEART cohort. The ResNet and CatBoost models achieved area under the receiver-operating curve (AUROC) between 85% and 92% in internal validation. In external validation, the AUROCs for the ResNet (82%, 82%, and 78%) were significantly better than CatBoost (78%, 73%, and 75%), for 1-, 3- and 5-year mortality prediction respectively, with better or comparable performance to the MAGGIC score. ResNet models predicted higher probability of death in the HFpEF and HFrEF (30%-50%) subgroups than in controls and at risk patients (5%-20%). The predicted probabilities of death correlated with KCCQ scores (all p < 0.05).

INTERPRETATION

Echo-based ML models to predict mortality had good internal and external validity, were generalizable, correlated with patients' quality of life, and are comparable to an established HF risk score. These models can be leveraged for automated risk stratification at point-of-care.

FUNDING

Funding for Alberta HEART was provided by an Alberta Innovates - Health Solutions Interdisciplinary Team Grant no. AHFMRITG 200801018. P.K. holds a Canadian Institutes of Health Research (CIHR) Sex and Gender Science Chair and a Heart & Stroke Foundation Chair in Cardiovascular Research. A.V. and V.S. received funding from the Mitacs Globalink Research Internship.

Cuproptosis-related LncRNA signatures as a prognostic model for head and neck squamous cell carcinoma

Cuproptosis is a novel, distinct form of regulated cell death. However, little is known about the role of cuproptosis-related lncRNAs (CRlncRNAs) in head and neck squamous cell carcinoma (HNSCC). This study aimed to identify a CRlncRNAs signature, explore its prognostic value in HNSCC. RNA-seq data and relevant clinical data were downloaded from The Cancer Genome Atlas (TCGA) database, and cuproptosis-related genes were identified from a search of the relevant candidate-gene literature. Analysis of differentially expressed lncRNAs (DElncRNAs) was performed using the R package "edgeR". The intersection of the lncRNAs between DElncRNAs and CRlncRNAs was obtained using the R package "Venn Diagram". Univariate Cox regression was used to identify cuproptosis-related prognostic lncRNAs. LASSO-Cox method was used to narrow these cuproptosis-related prognostic lncRNAs and construct a prognostic model. Multiple statistical methods were used to evaluate the predictive ability of the model. Moreover, the relationships between the model and immune cell subpopulations, related functions and pathways and drug sensitivity were explored. Then, two risk groups were established according to the risk score calculated by the CRlncRNAs signature included three lncRNAs. In HNSCC patients, the risk score was a better predictor of survival than traditional clinicopathological features. In addition, significant differences in immune cells such as B cells, T cells and macrophages were observed between the two groups. Finally, the high-risk group had a lower IC50 for certain chemotherapeutic agents, such as cisplatin and cetuximab. This 3 CRlncRNAs signature is a powerful prognostic biomarker for predicting clinical outcomes and therapeutic responses in HNSCC patients.

The majority of 922 prediction models supporting breast cancer decision-making are at high risk of bias

OBJECTIVES

To systematically review the currently available prediction models that may support treatment decision-making in breast cancer.

STUDY DESIGN AND SETTING

Literature was systematically searched to identify studies reporting on development of prediction models aiming to support breast cancer treatment decision-making, published between January 2010 and December 2020. Quality and risk of bias were assessed using the Prediction model Risk Of Bias (ROB) Assessment Tool (PROBAST).

RESULTS

After screening 20,460 studies, 534 studies were included, reporting on 922 models. The 922 models predicted: mortality (n = 417 45%), recurrence (n = 217, 24%), lymph node involvement (n = 141, 15%), adverse events (n = 58, 6%), treatment response (n = 56, 6%), or other outcomes (n = 33, 4%). In total, 285 models (31%) lacked a complete description of the final model and could not be applied to new patients. Most models (n = 878, 95%) were considered to contain high ROB.

CONCLUSION

A substantial overlap in predictor variables and outcomes between the models was observed. Most models were not reported according to established reporting guidelines or showed methodological flaws during the development and/or validation of the model. Further development of prediction models with thorough quality and validity assessment is an essential first step for future clinical application.

Prognostic Scoring Systems in Primary Biliary Cholangitis: An Update

Primary biliary cholangitis (PBC) is a complex, chronic disease with a heterogeneous presentation, disease progression, and response to therapy. Several prognostic models based on disease stage and/or treatment response enhance risk stratification and therapeutic management. Recent work on disease modeling proposed early prediction of outcomes at PBC onset, yet this has not been implemented in clinical practice. Although early stratification of patients based on their individual risk of developing end-stage liver disease may prove cost-effective and actually become matter of medical deontology to timely offer the best therapeutic option, given the forthcoming availability of novel, disease-modifying drugs. This review outlines established and novel prognostic systems in PBC and provides some perspectives on the potential role of omics-derived biomarkers in developing reliable risk prediction models and promoting the implementation of personalized medicine in PBC.

Predicting mortality in pulmonary tuberculosis: A systematic review of prognostic models

BACKGROUND

Pulmonary tuberculosis is a highly prevalent disease in low-income countries; clinical prediction tools allow healthcare personnel to catalog patients with a higher risk of death in order to prioritize medical attention.

METHODOLOGY

We conducted a literature search on prognostic models aimed to predict mortality in patients diagnosed with pulmonary tuberculosis. We included prospective and retrospective studies where prognostic models predicting mortality were either developed or validated in patients diagnosed with pulmonary tuberculosis. Three reviewers independently assessed the quality of the included studies using the PROBAST tool (Prediction model study Risk of Bias Assessment Tool). A narrative review of the characteristics of each model was conducted.

RESULTS

Six articles (n = 3553 patients) containing six prediction models were included in the review. Most studies (5 out of 6) were retrospective cohorts, only one study was a prospective case-control study. All the studies had a high risk of bias according to the PROBAST tool in the overall assessment. Regarding the applicability of the prediction models, three studies had a low concern of applicability, two high concern and one unclear concern. Five studies developed new prediction rules. In general, the presented models had a good discriminatory ability, with areas under the curve fluctuating between 0.65 up to 0.91.

CONCLUSION

None of the prognostic models included in the review accurately predict mortality in patients with pulmonary tuberculosis, due to great heterogeneity in the population and a high risk of bias.

Prognostic models for outcome prediction in patients with advanced hepatocellular carcinoma treated by systemic therapy: a systematic review and critical appraisal

Objective

To describe and analyze the predictive models of the prognosis of patients with hepatocellular carcinoma (HCC) undergoing systemic treatment.

Design

Systematic review.

Data sources

PubMed and Embase until December 2020 and manually searched references from eligible articles.

Eligibility criteria for study selection

The development, validation, or updating of prognostic models of patients with HCC after systemic treatment.

Results

The systematic search yielded 42 eligible articles: 28 articles described the development of 28 prognostic models of patients with HCC treated with systemic therapy, and 14 articles described the external validation of 32 existing prognostic models of patients with HCC undergoing systemic treatment. Among the 28 prognostic models, six were developed based on genes, of which five were expressed in full equations; the other 22 prognostic models were developed based on common clinical factors. Of the 28 prognostic models, 11 were validated both internally and externally, nine were validated only internally, two were validated only externally, and the remaining six models did not undergo any type of validation. Among the 28 prognostic models, the most common systemic treatment was sorafenib (n = 19); the most prevalent endpoint was overall survival (n = 28); and the most commonly used predictors were alpha-fetoprotein (n = 15), bilirubin (n = 8), albumin (n = 8), Child–Pugh score (n = 8), extrahepatic metastasis (n = 7), and tumor size (n = 7). Further, among 32 externally validated prognostic models, 12 were externally validated > 3 times.

Conclusions

This study describes and analyzes the prognostic models developed and validated for patients with HCC who have undergone systemic treatment. The results show that there are some methodological flaws in the model development process, and that external validation is rarely performed. Future research should focus on validating and updating existing models, and evaluating the effects of these models in clinical practice.

Systematic review registration

PROSPERO CRD42020200187.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-022-09841-5.

Which model is superior in predicting ICU survival: artificial intelligence versus conventional approaches

Background

A disease severity classification system is widely used to predict the survival of patients admitted to the intensive care unit with different diagnoses. In the present study, conventional severity classification systems were compared with artificial intelligence predictive models (Artificial Neural Network and Decision Tree) in terms of the prediction of the survival rate of the patients admitted to the intensive care unit.

Methods

This retrospective cohort study was performed on the data of the patients admitted to the ICU of Ghaemshahr’s Razi Teaching Care Center from March 20th, 2017, to September 22nd, 2019. The required data for calculating conventional severity classification models (SOFA, SAPS II, APACHE II, and APACHE IV) were collected from the patients’ medical records. Subsequently, the score of each model was calculated. Artificial intelligence predictive models (Artificial Neural Network and Decision Tree) were developed in the next step. Lastly, the performance of each model in predicting the survival of the patients admitted to the intensive care unit was evaluated using the criteria of sensitivity, specificity, accuracy, F-measure, and area under the ROC curve. Also, each model was validated externally. The R program, version 4.1, was used to create the artificial intelligence models, and SPSS Statistics Software, version 21, was utilized to perform statistical analysis.

Results

The area under the ROC curve of SOFA, SAPS II, APACHE II, APACHE IV, multilayer perceptron artificial neural network, and CART decision tree were 76.0, 77.1, 80.3, 78.5, 84.1, and 80.0, respectively.

Conclusion

The results showed that although the APACHE II model had better results than other conventional models in predicting the survival rate of the patients admitted to the intensive care unit, the other conventional models provided acceptable results too. Moreover, the findings showed that the artificial neural network model had the best performance among all the studied models, indicating the discrimination power of this model in predicting patient survival compared to the other models.

ADC textural features in patients with single brain metastases improve clinical risk models

AIMS

In this retrospective study we performed a quantitative textural analysis of apparant diffusion coefficient (ADC) images derived from diffusion weighted MRI (DW-MRI) of single brain metastases (BM) patients from different primary tumors and tested whether these imaging parameters may improve established clinical risk models.

METHODS

We identified 87 patients with single BM who had a DW-MRI at initial diagnosis. Applying image segmentation, volumes of contrast-enhanced lesions in T1 sequences, hyperintense T2 lesions (peritumoral border zone (T2PZ)) and tumor-free gray and white matter compartment (GMWMC) were generated and registered to corresponding ADC maps. ADC textural parameters were generated and a linear backward regression model was applied selecting imaging features in association with survival. A cox proportional hazard model with backward regression was fitted for the clinical prognostic models (diagnosis-specific graded prognostic assessment score (DS-GPA) and the recursive partitioning analysis (RPA)) including these imaging features.

RESULTS

Thirty ADC textural parameters were generated and linear backward regression identified eight independent imaging parameters which in combination predicted survival. Five ADC texture features derived from T2PZ, the volume of the T2PZ, the normalized mean ADC of the GMWMC as well as the mean ADC slope of T2PZ. A cox backward regression including the DS-GPA, RPA and these eight parameters identified two MRI features which improved the two risk scores (HR = 1.14 [1.05;1.24] for normalized mean ADC GMWMC and HR = 0.87 [0.77;0.97]) for ADC 3D kurtosis of the T2PZ.) CONCLUSIONS: Textural analysis of ADC maps in patients with single brain metastases improved established clinical risk models. These findings may aid to better understand the pathogenesis of BM and may allow selection of patients for new treatment options.

Pathological and genetic markers improve recurrence prognostication with the University of California Los Angeles Integrated Staging System for patients with clear cell renal cell carcinoma

PURPOSE

To elucidate which patients with clear cell renal cell carcinoma have the highest risk for disease relapse after curative nephrectomy is challenging but is acutely relevant in the era of approved adjuvant therapies. Pathological and genetic markers were used to improve the University of California Los Angeles Integrated Staging System (UISS) for the risk stratification and prognostication of recurrence free survival (RFS).

PATIENTS AND METHODS

Necrosis, sarcomatoid features, Rhabdoid features, chromosomal loss 9p, combined chromosomal loss 3p14q and microvascular invasion (MVI) were tested in univariable and multivariable analyses for their ability to improve the discriminatory ability of the UISS.

RESULTS

In the development cohort, during the median follow-up time of 43.4 months (±SD 54.1 months), 50/240 (21%) patients developed disease recurrence. MVI (HR: 2.22; p = 0.013) and the combined loss of chromosome 3p/14q (HR: 2.89; p = 0.004) demonstrated independent association with RFS and were used to improve the assignment to the UISS risk category. In the current UISS high-risk group, only 7/50 (14%) recurrence cases were correctly identified; while in the improved system, 23/50 (45%) were correctly prognosticated. The concordance index meaningfully improved from 0.55 to 0.68 to distinguish patients at intermediate risk versus high risk. Internal validation demonstrated a robust prognostication of RFS. In the external validation cohort, there was no case with disease recurrence in the low-risk group, and the mean RFS times were 13.2 (±1.8) and 8.2 (±0.8) years in the intermediate and high-risk groups, respectively.

CONCLUSIONS

Adding MVI and combined chromosomal loss3p/14q to the UISS improves the ability to define the patient group with clear cell renal cell carcinomawho are at the highest risk for disease relapse after surgical treatment.

A robust and parsimonious machine learning method to predict ICU admission of COVID-19 patients

In this article, we discuss the development of prognostic machine learning (ML) models for COVID-19 progression, by focusing on the task of predicting ICU admission within (any of) the next 5 days. On the basis of 6,625 complete blood count (CBC) tests from 1,004 patients, of which 18% were admitted to intensive care unit (ICU), we created four ML models, by adopting a robust development procedure which was designed to minimize risks of bias and over-fitting, according to reference guidelines. The best model, a support vector machine, had an AUC of .85, a Brier score of .14, and a standardized net benefit of .69: these scores indicate that the model performed well over a variety of prediction criteria. We also conducted an interpretability study to back up our findings, showing that the data on which the developed model is based is consistent with the current medical literature. This also demonstrates that CBC data and ML methods can be used to predict COVID-19 patients' ICU admission at a relatively low cost: in particular, since CBC data can be quickly obtained by means of routine blood exams, our models could be used in resource-constrained settings and provide health practitioners with rapid and reliable indications.

Modification of the TRISS: simple and practical mortality prediction after trauma in an all-inclusive registry

PURPOSE

Numerous studies have modified the Trauma Injury and Severity Score (TRISS) to improve its predictive accuracy for specific trauma populations. The aim of this study was to develop and validate a simple and practical prediction model that accurately predicts mortality for all acute trauma admissions.

METHODS

This retrospective study used Dutch National Trauma Registry data recorded between 2015 and 2018. New models were developed based on nonlinear transformations of TRISS variables (age, systolic blood pressure (SBP), Glasgow Coma Score (GCS) and Injury Severity Score (ISS)), the New Injury Severity Score (NISS), the sex-age interaction, the best motor response (BMR) and the American Society of Anesthesiologists (ASA) physical status classification. The models were validated in 2018 data and for specific patient subgroups. The models' performance was assessed based on discrimination (areas under the curve (AUCs)) and by calibration plots. Multiple imputation was applied to account for missing values.

RESULTS

The mortality rates in the development and validation datasets were 2.3% (5709/245363) and 2.5% (1959/77343), respectively. A model with sex, ASA class, and nonlinear transformations of age, SBP, the ISS and the BMR showed significantly better discrimination than the TRISS (AUC 0.915 vs. 0.861). This model was well calibrated and demonstrated good discrimination in different subsets of patients, including isolated hip fractures patients (AUC: 0.796), elderly (AUC: 0.835), less severely injured (ISS16) (AUC: 878), severely injured (ISS ≥ 16) (AUC: 0.889), traumatic brain injury (AUC: 0.910). Moreover, discrimination for patients admitted to the intensive care (AUC: s0.846), and for both non-major and major trauma center patients was excellent, with AUCs of 0.940 and 0.895, respectively.

CONCLUSION

This study presents a simple and practical mortality prediction model that performed well for important subgroups of patients as well as for the heterogeneous population of all acute trauma admissions in the Netherlands. Because this model includes widely available predictors, it can also be used for international evaluations of trauma care within institutions and trauma systems.

Diagnostic Evidence GAuge of Single cells (DEGAS): a flexible deep transfer learning framework for prioritizing cells in relation to disease

We propose DEGAS (Diagnostic Evidence GAuge of Single cells), a novel deep transfer learning framework, to transfer disease information from patients to cells. We call such transferrable information "impressions," which allow individual cells to be associated with disease attributes like diagnosis, prognosis, and response to therapy. Using simulated data and ten diverse single-cell and patient bulk tissue transcriptomic datasets from glioblastoma multiforme (GBM), Alzheimer's disease (AD), and multiple myeloma (MM), we demonstrate the feasibility, flexibility, and broad applications of the DEGAS framework. DEGAS analysis on myeloma single-cell transcriptomics identified PHF19high myeloma cells associated with progression. Availability: https://github.com/tsteelejohnson91/DEGAS .

Predicting SARS-CoV-2 infection duration at hospital admission:a deep learning solution

COVID-19 cases are increasing around the globe with almost 5 million of deaths. We propose here a deep learning model capable of predicting the duration of the infection by means of information available at hospital admission. A total of 222 patients were enrolled in our observational study. Anagraphical and anamnestic data, COVID-19 signs and symptoms, COVID-19 therapy, hematochemical test results, and prior therapies administered to patients are used as predictors. A set of 55 features, all of which can be taken in the first hours of the patient’s hospitalization, was considered. Different solutions were compared achieving the best performance with a sequential convolutional neural network-based model merged in an ensemble with two different meta-learners linked in cascade. We obtained a median absolute error of 2.7 days (IQR = 3.0) in predicting the duration of the infection; the error was equally distributed in the infection duration range. This tool could preemptively give an outlook of the COVID-19 patients’ expected path and the associated hospitalization effort. The proposed solution could be viable in tackling the huge burden and the logistics complexity of hospitals or rehabilitation centers during the pandemic waves.

External validation of prognostic models predicting outcome after chronic subdural hematoma

BACKGROUND

Several prognostic models for outcomes after chronic subdural hematoma (CSDH) treatment have been published in recent years. However, these models are not sufficiently validated for use in daily clinical practice. We aimed to assess the performance of existing prediction models for outcomes in patients diagnosed with CSDH.

METHODS

We systematically searched relevant literature databases up to February 2021 to identify prognostic models for outcome prediction in patients diagnosed with CSDH. For the external validation of prognostic models, we used a retrospective database, containing data of 2384 patients from three Dutch regions. Prognostic models were included if they predicted either mortality, hematoma recurrence, functional outcome, or quality of life. Models were excluded when predictors were absent in our database or available for < 150 patients in our database. We assessed calibration, and discrimination (quantified by the concordance index C) of the included prognostic models in our retrospective database.

RESULTS

We identified 1680 original publications of which 1656 were excluded based on title or abstract, mostly because they did not concern CSDH or did not define a prognostic model. Out of 18 identified models, three could be externally validated in our retrospective database: a model for 30-day mortality in 1656 patients, a model for 2 months, and another for 3-month hematoma recurrence both in 1733 patients. The models overestimated the proportion of patients with these outcomes by 11% (15% predicted vs. 4% observed), 1% (10% vs. 9%), and 2% (11% vs. 9%), respectively. Their discriminative ability was poor to modest (C of 0.70 [0.63-0.77]; 0.46 [0.35-0.56]; 0.59 [0.51-0.66], respectively).

CONCLUSIONS

None of the examined models showed good predictive performance for outcomes after CSDH treatment in our dataset. This study confirms the difficulty in predicting outcomes after CSDH and emphasizes the heterogeneity of CSDH patients. The importance of developing high-quality models by using unified predictors and relevant outcome measures and appropriate modeling strategies is warranted.

GRADE concept paper 2: Concepts for judging certainty on the calibration of prognostic models in a body of validation studies

``In this paper, we highlight key concepts...'' is background.The sentence ``IN this paper, we highlight key concepts in evaluating the certainty of evidence regarding the calibration of prognostic models'' is methods. The rest is results and conclusion. Brognostic models combine several prognostic factors to provide an estimate of the likelihood (or risk) of future events in individual patients, conditional on their prognostic factor values. A fundamental part of evaluating prognostic models is undertaking studies to determine whether their predictive performance, such as calibration and discrimination, is reproduced across settings. Systematic reviews and meta-analyses of studies evaluating prognostic models' performance are a necessary step for selection of models for clinical practice and for testing the underlying assumption that their use will improve outcomes, including patient's reassurance and optimal future planning. In this paper, we highlight key concepts in evaluating the certainty of evidence regarding the calibration of prognostic models. Four concepts are key to evaluating the certainty of evidence on prognostic models' performance regarding calibration. The first concept is that the inference regarding calibration may take 1 of 2 forms: deciding whether 1 is rating certainty that a model's performance is satisfactory or, instead, unsatisfactory, in either case defining the threshold for satisfactory (or unsatisfactory) model performance. Second, inconsistency is the critical GRADE domain to deciding whether we are rating certainty in the model performance being satisfactory or unsatisfactory. Third, depending on whether 1 is rating certainty in satisfactory or unsatisfactory performance, different patterns of inconsistency of results across studies will inform ratings of certainty of evidence. Fourth, exploring the distribution of point estimates of observed to expected ratio across individual studies, and its determinants, will bear on the need for and direction of future research.

Genome-wide analysis of methylation CpG sites in gene promoters identified four pairs of CpGs-mRNAs associated with lung adenocarcinoma prognosis

BACKGROUND

Activation of oncogenes through promoter hypomethylation and silencing of tumor suppressor genes induced by promoter hypermethylation played essential roles in the progression of lung adenocarcinoma (LUAD). This study aimed to identify the LUAD prognostic CpG sites and the regulated genes which contributed to LUAD progression.

METHODS

Methylation profiles from TCGA and GSE60645 were used to screen the differentially methylated CpGs. Then, the Log-rank test was adopted to identify LUAD prognosis-associated CpGs. Differential gene expression and survival analyses were further performed to suggest the roles of methylation-driven genes in LUAD prognosis. Finally, models and nomograms were constructed to predict the prognosis of LUAD.

RESULTS

A total of 1891 CpGs at gene promoters were differentially methylated. Among them, 54 CpGs were significantly associated with LUAD prognosis. Nine of them showed significant correlations with the expression of four genes (CCDC181, CFTR, PPP1R16B, MYEOV). CCDC181, CFTR and PPP1R16B were aberrantly down-regulated in LUAD, while MYEOV was up-regulated. All of them were significantly associated with LUAD prognosis. The LASSO regression analysis indicated that tumor stages, cg09181792, cg16998150, cg22779330 and PPP1R16B were promising prognostic factors. The AUC (area under the curve) of the model containing the clinical predictors was 0.643. The combination of CpGs and PPP1R16B with clinical variables significantly improved the predictive efficiency with an AUC of 0.714 (P = 0.036).

CONCLUSION

This study identified four pairs of promoter CpGs and genes that were significantly associated with LUAD prognosis. The integration of CpGs methylation and gene expression showed better predictive ability for LUAD prognosis.

Prognostic models for mortality after cardiac surgery in patients with infective endocarditis: a systematic review and aggregation of prediction models

BACKGROUND

There are several prognostic models to estimate the risk of mortality after surgery for active infective endocarditis (IE). However, these models incorporate different predictors and their performance is uncertain.

OBJECTIVE

We systematically reviewed and critically appraised all available prediction models of postoperative mortality in patients undergoing surgery for IE, and aggregated them into a meta-model.

DATA SOURCES

We searched Medline and EMBASE databases from inception to June 2020.

STUDY ELIGIBILITY CRITERIA

We included studies that developed or updated a prognostic model of postoperative mortality in patient with IE.

METHODS

We assessed the risk of bias of the models using PROBAST (Prediction model Risk Of Bias ASsessment Tool) and we aggregated them into an aggregate meta-model based on stacked regressions and optimized it for a nationwide registry of IE patients. The meta-model performance was assessed using bootstrap validation methods and adjusted for optimism.

RESULTS

We identified 11 prognostic models for postoperative mortality. Eight models had a high risk of bias. The meta-model included weighted predictors from the remaining three models (EndoSCORE, specific ES-I and specific ES-II), which were not rated as high risk of bias and provided full model equations. Additionally, two variables (age and infectious agent) that had been modelled differently across studies, were estimated based on the nationwide registry. The performance of the meta-model was better than the original three models, with the corresponding performance measures: C-statistics 0.79 (95% CI 0.76-0.82), calibration slope 0.98 (95% CI 0.86-1.13) and calibration-in-the-large -0.05 (95% CI -0.20 to 0.11).

CONCLUSIONS

The meta-model outperformed published models and showed a robust predictive capacity for predicting the individualized risk of postoperative mortality in patients with IE.

PROTOCOL REGISTRATION

PROSPERO (registration number CRD42020192602).

Predicting prognosis in COVID-19 patients using machine learning and readily available clinical data

Rationale

Prognostic tools for aiding in the treatment of hospitalized COVID-19 patients could help improve outcome by identifying patients at higher or lower risk of severe disease. The study objective was to develop models to stratify patients by risk of severe outcomes during COVID-19 hospitalization using readily available information at hospital admission.

Methods

Hierarchical ensemble classification models were trained on a set of 229 patients hospitalized with COVID-19 to predict severe outcomes, including ICU admission, development of acute respiratory distress syndrome, or intubation, using easily attainable attributes including basic patient characteristics, vital signs at admission, and basic lab results collected at time of presentation. Each test stratifies patients into groups of increasing risk. An additional cohort of 330 patients was used for blinded, independent validation. Shapley value analysis evaluated which attributes contributed most to the models’ predictions of risk.

Main results

Test performance was assessed using precision (positive predictive value) and recall (sensitivity) of the final risk groups. All test cut-offs were fixed prior to blinded validation. In development and validation, the tests achieved precision in the lowest risk groups near or above 0.9. The proportion of patients with severe outcomes significantly increased across increasing risk groups. While the importance of attributes varied by test and patient, C-reactive protein, lactate dehydrogenase, and D-dimer were often found to be important in the assignment of risk.

Conclusions

Risk of severe outcomes for patients hospitalized with COVID-19 infection can be assessed using machine learning-based models based on attributes routinely collected at hospital admission.

Development and internal validation of prognostic models for recovery in patients with non-specific neck pain presenting in primary care

OBJECTIVES

Development and internal validation of prognostic models for post-treatment and 1-year recovery in patients with neck pain in primary care.

DESIGN

Prospective cohort study.

SETTING

Primary care manual therapy practices.

PARTICIPANTS

Patients with non-specific neck pain of any duration (n=1193).

INTERVENTION

Usual care manual therapy.

OUTCOME MEASURES

Recovery defined in terms of pain intensity, disability, and global perceived improvement directly post-treatment and at 1-year follow-up.

RESULTS

All post-treatment models exhibited acceptable discriminative performance after derivation (AUC≥0.7). The developed post-treatment disability model exhibited the best overall performance (R²=0.24; IQR, 0.22-0.26), discrimination (AUC=0.75; 95% CI, 0.63-0.84), and calibration (slope 0.92; IQR, 0.91-0.93). After internal validation and penalization, this model retained acceptable discriminative performance (AUC=0.74). The five other models, including those predicting 1-year recovery, did not reach acceptable discriminative performance after internal validation. Baseline pain duration, disability, and pain intensity were consistent predictors across models.

CONCLUSION

A post-treatment prognostic model for disability was successfully developed and internally validated. This model has potential to inform primary care clinicians about a patient's individual prognosis after treatment, but external validation is required before clinical use can be recommended.

Data-driven prediction of decannulation probability and timing in patients with severe acquired brain injury

BACKGROUND AND OBJECTIVES

From a rehabilitation perspective, removal of tracheostomy in patients with severe acquired brain injuries (sABI) is a crucial step. Predictive parameters for a successful decannulation are currently still a focus of the research for sABI patients, especially for those presenting a disorder of consciousness. For this reason, we adopted a data-driven approach predicting decannulation probability and timing using ensemble learning models in patients in intensive rehabilitation units.

METHODS

327 patients, 186 of which were successfully decannulated during their intensive rehabilitative stay, were recruited in a non-concurrent retrospective study. Decannulation probability and timing were predicted using data available within one week from admission at the rehabilitation unit. Two predictive models were trained and cross-validated independently, with the first being an ensemble of a Support Vector Machine and Random Forests and the second an Adaptive Boosting with a Support Vector Regression as weak learner. Confusion matrix, accuracy and AUC were considered as evaluation metrics for the classifier and median absolute error was considered for the regressor. To quantify the advantages in the clinical practice of using the latter prediction, we compared timing estimation with a timing guess (median) calculated on available data. The comparison was based on a Wilcoxon signed rank test.

RESULTS

Decannulation probability was successfully predicted with an accuracy of 84.8% (AUC = 0.85) and timing with a median absolute error of 25.7 days [IQR = 25.6]. This resulted in a significant improvement with respect to the weaning time guess (p<0.05) with an effect size of 71.7%. Furthermore, dichotomizing the regression prediction with a threshold (3 months from the event), resulted in a prediction accuracy of 77.5% (AUC = 0.82) on the test set.

DISCUSSIONS

A model capable of providing a prediction on decannulation probability and timing was developed and cross-validated, built on data taken at admission to the intensive rehabilitation unit. Translated in clinical practice, this information can support the clinical decision process and provide a mean to improve both in-hospital and domiciliary care organization.

Follicular Lymphoma: Refining Prognostic Models and Impact of Pod-24 in Clinical Outcomes

Follicular lymphoma (FL) is the most common indolent lymphoma, accounting for 20%-25% of all non-Hodgkin's lymphomas (NHLs). It is a malignancy with variable biologic presentation and heterogeneous clinical outcomes. Several models incorporating clinical laboratory variables and molecular biomarkers are able to predict its prognosis, allowing to stratify patients into different risk groups. However, these prognostic scores should not be used to indicate first-line treatment or risk-adapted therapeutic recommendations. Over the past 5 years, progression of disease within 24 months (POD-24) of first-line chemo-immunotherapy has emerged as a robust adverse prognostic factor, capable of assessing overall survival and identifying high-risk patients with indication for more aggressive therapeutic approaches, such as consolidation based in autologous stem cell transplantation. It should be reinforced that POD-24 is not a baseline measurement, it is based on a post-treatment strategy, and is usually applied to patients with a high tumor burden. The identification of newly diagnosed patients at high risk for disease progression, particularly those with low tumor volume is still a challenge in the context of FL. Therefore, the primary purpose of this review is to provide an overview of the main prognostic models validated to date for FL. Moreover, using these scores, which incorporate clinical and genetic variables, we aim to identify individuals with newly diagnosed FL, advanced disease, and low tumor burden with a high probability of progression or relapse within 24 months of first treatment. Thus, a decision regarding risk-adapted induction therapy could be better stablished for these subset of patients.

Machine learning identifies ICU outcome predictors in a multicenter COVID-19 cohort

BACKGROUND

Intensive Care Resources are heavily utilized during the COVID-19 pandemic. However, risk stratification and prediction of SARS-CoV-2 patient clinical outcomes upon ICU admission remain inadequate. This study aimed to develop a machine learning model, based on retrospective & prospective clinical data, to stratify patient risk and predict ICU survival and outcomes.

METHODS

A Germany-wide electronic registry was established to pseudonymously collect admission, therapeutic and discharge information of SARS-CoV-2 ICU patients retrospectively and prospectively. Machine learning approaches were evaluated for the accuracy and interpretability of predictions. The Explainable Boosting Machine approach was selected as the most suitable method. Individual, non-linear shape functions for predictive parameters and parameter interactions are reported.

RESULTS

1039 patients were included in the Explainable Boosting Machine model, 596 patients retrospectively collected, and 443 patients prospectively collected. The model for prediction of general ICU outcome was shown to be more reliable to predict "survival". Age, inflammatory and thrombotic activity, and severity of ARDS at ICU admission were shown to be predictive of ICU survival. Patients' age, pulmonary dysfunction and transfer from an external institution were predictors for ECMO therapy. The interaction of patient age with D-dimer levels on admission and creatinine levels with SOFA score without GCS were predictors for renal replacement therapy.

CONCLUSIONS

Using Explainable Boosting Machine analysis, we confirmed and weighed previously reported and identified novel predictors for outcome in critically ill COVID-19 patients. Using this strategy, predictive modeling of COVID-19 ICU patient outcomes can be performed overcoming the limitations of linear regression models. Trial registration "ClinicalTrials" (clinicaltrials.gov) under NCT04455451.

Predicting critical illness on initial diagnosis of COVID-19 based on easily obtained clinical variables: development and validation of the PRIORITY model

OBJECTIVES

We aimed to develop and validate a prediction model, based on clinical history and examination findings on initial diagnosis of coronavirus disease 2019 (COVID-19), to identify patients at risk of critical outcomes.

METHODS

We used data from the SEMI-COVID-19 Registry, a cohort of consecutive patients hospitalized for COVID-19 from 132 centres in Spain (23rd March to 21st May 2020). For the development cohort, tertiary referral hospitals were selected, while the validation cohort included smaller hospitals. The primary outcome was a composite of in-hospital death, mechanical ventilation, or admission to intensive care unit. Clinical signs and symptoms, demographics, and medical history ascertained at presentation were screened using least absolute shrinkage and selection operator, and logistic regression was used to construct the predictive model.

RESULTS

There were 10 433 patients, 7850 in the development cohort (primary outcome 25.1%, 1967/7850) and 2583 in the validation cohort (outcome 27.0%, 698/2583). The PRIORITY model included: age, cardiovascular disease, chronic kidney disease, dyspnoea, tachypnoea, confusion, systolic blood pressure, and SpO₂ ≤93% or oxygen requirement. The model showed high discrimination for critical illness in both the development (C-statistic 0.823; 95% confidence interval (CI) 0.813, 0.834) and validation (C-statistic 0.794; 95%CI 0.775, 0.813) cohorts. A freely available web-based calculator was developed based on this model (https://www.evidencio.com/models/show/2344).

CONCLUSIONS

The PRIORITY model, based on easily obtained clinical information, had good discrimination and generalizability for identifying COVID-19 patients at risk of critical outcomes.

Spine growth modulation with titanium implant: comparisons to observation and bracing in early adolescent idiopathic scoliosis

Outcomes of a pilot study of spine growth modulation (GM) were compared to those of untreated and braced patients from a concurrent bracing effectiveness trial (BrAIST). The purpose of this study was to determine probabilities of progression (PP) to fusion indications (≥45°) in a cohort of subjects who underwent GM surgery, and to compare GM outcomes to those of matched BrAIST subjects. Secondary analyses were conducted comparing two prospective longitudinal studies. In one, a vertebral GM system was implanted in 6 highly skeletally immature AIS patients. The control group provided by BrAIST was comprised of a subset of untreated or braced subjects that fit the eligibility criteria of the GM study. GM outcomes were compared to predictions from two prognostic logistic regression models derived from BrAIST to estimate risk of curve progression to ≥45°. If the GM patients were untreated, PPs ranged from 68-98%. If braced for 18 hours/day, progression was expected in two of six, one with a PP of 71%. This latter patient not only did not progress, his curve decreased >20°. In the matched cohort, two were untreated and quickly progressed, whereas two were braced and did not progress. Therefore, the bracing models and matched cohort confirmed the initial assumption that all GM patients were at high risk if untreated. They also supported the probable benefit of the GM system, as 3 of 6 benefited from GM relative to predictions for untreated patients, and one of 6 benefited compared to predictions for highly compliant braced patients.

Internal validation and comparison of the prognostic performance of models based on six emergency scoring systems to predict in-hospital mortality in the emergency department

Background

Medical scoring systems are potentially useful to make optimal use of available resources. A variety of models have been developed for illness measurement and stratification of patients in Emergency Departments (EDs). This study was aimed to compare the predictive performance of the following six scoring systems: Simple Clinical Score (SCS), Worthing physiological Score (WPS), Rapid Acute Physiology Score (RAPS), Rapid Emergency Medicine Score (REMS), Modified Early Warning Score (MEWS), and Routine Laboratory Data (RLD) to predict in-hospital mortality.

Methods

A prospective single-center observational study was conducted from March 2016 to March 2017 in Edalatian ED in Emam Reza Hospital, located in the northeast of Iran. All variables needed to calculate the models were recorded at the time of admission and logistic regression was used to develop the models’ prediction probabilities. The Area Under the Curve for Receiver Operating Characteristic (AUC-ROC) and Precision-Recall curves (AUC-PR), Brier Score (BS), and calibration plots were used to assess the models’ performance. Internal validation was obtained by 1000 bootstrap samples. Pairwise comparison of AUC-ROC was based on the DeLong test.

Results

A total of 2205 patients participated in this study with a mean age of 61.8 ± 18.5 years. About 19% of the patients died in the hospital. Approximately 53% of the participants were male. The discrimination ability of SCS, WPS, RAPS, REMS, MEWS, and RLD methods were 0.714, 0.727, 0.661, 0.678, 0.698, and 0.656, respectively. Additionally, the AUC-PR of SCS, WPS, RAPS, REMS, EWS, and RLD were 0.39, 0.42, 0.35, 0.34, 0.36, and 0.33 respectively. Moreover, BS was 0.1459 for SCS, 0.1713 for WPS, 0.0908 for RAPS, 0.1044 for REMS, 0.1158 for MEWS, and 0.073 for RLD. Results of pairwise comparison which was performed for all models revealed that there was no significant difference between the SCS and WPS. The calibration plots demonstrated a relatively good concordance between the actual and predicted probability of non-survival for the SCS and WPS models.

Conclusion

Both SCS and WPS demonstrated fair discrimination and good calibration, which were superior to the other models. Further recalibration is however still required to improve the predictive performance of all available models and their use in clinical practice is still unwarranted.

Assessment of acute heart failure prognosis: the promising role of prognostic models and biomarkers

Numerous models and biomarkers have been proposed to estimate prognosis and improve decision-making in patients with acute heart failure (AHF). The present literature review provides a critical appraisal of externally validated prognostic models in AHF, combining clinical data and biomarkers. We perform a literature review of clinical studies, using the following terms: "acute heart failure," "acute decompensated heart failure," "prognostic models," "risk scores," "mortality," "death," "hospitalization," "admission," and "biomarkers." We searched MEDLINE and EMBASE databases from 1990 to 2020 for studies documenting prognostic models in AHF. External validation of each prognostic model to another AHF cohort, containing at least one biomarker, was prerequisites for study selection. Among 358 initially screened studies, 9 of them fulfilled all searching criteria. The majority of prognostic models were simplified, including a narrow number of variables (up to 10), with good performance regarding calibration and discrimination (c-statistics > 0.65). Unfortunately, the derived and validated cohorts showed a wide variety in patients' characteristics (e.g., cause of AHF and therapy). Moreover, the prognostic models used various time-points and a plethora of combinations of variables determining different cut-off values. Although the application of valid prognostic models in AHF population is quite promising, a precise methodological approach should be set for the derivation and validation of prognostic models in AHF with unified characteristics to establish an effective performance in clinical practice.

Validation of the geriatric trauma outcome scores in predicting outcomes of elderly trauma patients

BACKGROUND

Using three patient characteristics, including age, Injury Severity Score (ISS) and transfusion within 24 h of admission (yes vs. no), the Geriatric Trauma Outcome Score (GTOS) and Geriatric Trauma Outcome Score II (GTOS II) have been developed to predict mortality and unfavourable discharge (to a nursing home or hospice facility), of those who were ≥65 years old, respectively.

OBJECTIVES

This study aimed to validate the GTOS and GTOS II models. For the nested-cohort requiring intensive care, we compared the GTOS scores with two ICU prognostic scores - the Acute Physiology and Chronic Health Evaluation (APACHE) III and Australian and New Zealand Risk of Death (ANZROD).

METHODS

All elderly trauma patients admitted to the State Trauma Unit between 2009 and 2019 were included. The discrimination ability and calibration of the GTOS and GTOS II scores were assessed by the area under the receiver-operating-characteristic (AUROC) curve and a calibration plot, respectively.

RESULTS

Of the 57,473 trauma admissions during the study period, 15,034 (26.2%) were ≥65 years-old. The median age and ISS of the cohort were 80 (interquartile range [IQR] 72-87) and 6 (IQR 2-9), respectively; and the average observed mortality was 4.3%. The ability of the GTOS to predict mortality was good (AUROC 0.838, 95% confidence interval [CI] 0.821-0.855), and better than either age (AUROC 0.603, 95%CI 0.581-0.624) or ISS (AUROC 0.799, 95%CI 0.779-0.819) alone. The GTOS II's ability to predict unfavourable discharge was satisfactory (AUROC 0.707, 95%CI 0.696-0.719) but no better than age alone. Both GTOS and GTOS II scores over-estimated risks of the adverse outcome when the predicted risks were high. The GTOS score (AUROC 0.683, 95%CI 0.591-0.775) was also inferior to the APACHE III (AUROC 0.783, 95%CI 0.699-0.867) or ANZROD (AUROC 0.788, 95%CI 0.705-0.870) in predicting mortality for those requiring intensive care.

CONCLUSIONS

The GTOS scores had a good ability to discriminate between survivors and non-survivors in the elderly trauma patients, but GTOS II scores were no better than age alone in predicting unfavourable discharge. Both GTOS and GTOS II scores were not well-calibrated when the predicted risks of adverse outcome were high.

Challenges and solutions in prognostic prediction models in spinal disorders

Methodological shortcomings in prognostic modeling for patients with spinal disorders are highly common. This general commentary discusses methodological challenges related to the specific nature of this field. Five specific methodological challenges in prognostic modeling for patients with spinal disorders are presented with their potential solutions, as related to the choice of study participants, purpose of studies, limitations in measurements of outcomes and predictors, complexity of recovery predictions, and confusion of prognosis and treatment response. Large studies specifically designed for prognostic model research are needed, using standard baseline measurement sets, clearly describing participants' recruitment and accounting and correcting for measurement limitations.