Interaction of Cigarette Smoking and Polygenic Risk Score on Reduced Lung Function

IMPORTANCE

The risk of airflow limitation and chronic obstructive pulmonary disease (COPD) is influenced by combinations of cigarette smoking and genetic susceptibility, yet it remains unclear whether gene-by-smoking interactions are associated with quantitative measures of lung function.

OBJECTIVE

To assess the interaction of cigarette smoking and polygenic risk score in association with reduced lung function.

DESIGN, SETTING, AND PARTICIPANTS

This UK Biobank cohort study included UK citizens of European ancestry aged 40 to 69 years with genetic and spirometry data passing quality control metrics. Data was analyzed from July 2020 to March 2021.

EXPOSURES

PRS of combined forced expiratory volume in 1 second (FEV1) and percent of forced vital capacity exhaled in the first second (FEV1/FVC), self-reported pack-years of smoking, ever- vs never-smoking status, and current- vs former- or never-smoking status.

MAIN OUTCOMES AND MEASURES

FEV1/FVC was the primary outcome. Models were used to test for interactions with models, including the main effects of PRS, different smoking variables, and their cross-product terms. The association between pack-years of smoking and FEV1/FVC were compared for those in the highest vs lowest decile of estimated genetic risk for low lung function.

RESULTS

We included 319 730 individuals, of whom 24 915 (8%) had moderate-to-severe COPD cases, and 44.4% were men. Participants had a mean (SD) age 56.5 of (8.02) years. The PRS and pack-years were significantly associated with lower FEV1/FVC (PRS: β, -0.03; 95% CI, -0.031 to -0.03; pack-years: β, -0.0064; 95% CI, -0.0064 to -0.0063) and the interaction term (β, -0.0028; 95% CI, -0.0029 to -0.0026). A stepwise increment in estimated effect sizes for these interaction terms was observed per 10 pack-years of smoking exposure. The interaction of PRS with 11 to 20, 31 to 40, and more than 50 pack-years categories were β (interaction) -0.0038 (95% CI, -0.0046 to -0.0031); -0.013 (95% CI, -0.014 to -0.012); and -0.017 (95% CI, -0.019 to -0.016), respectively. There was evidence of significant interaction between PRS with ever- or never- smoking status (β, interaction; -0.0064; 95% CI, -0.0068 to -0.0060) and current or not-current smoking (β, interaction; -0.0091; 95% CI, -0.0097 to -0.0084). For any given level of pack-years of smoking exposure, FEV1/FVC was significantly lower for individuals in the tenth decile (ie, highest risk) than the first decile (ie, lowest risk) of genetic risk. For every 20 pack-years of smoking, those in the tenth decile compared with the first decile of genetic risk showed nearly a 2-fold reduction in FEV1/FVC.

CONCLUSIONS AND RELEVANCE

COPD is characterized by diminished lung function, and our analyses suggest there is substantial interaction between genome-wide PRS and smoking exposures. While smoking was associated with decreased lung function across all genetic risk categories, the associations were strongest in individuals with higher estimated genetic risk.

Clinical and inflammatory features based machine learning model for fatal risk prediction of hospitalized COVID-19 patients: results from a retrospective cohort study

OBJECTIVES

To appraise effective predictors for COVID-19 mortality in a retrospective cohort study.

METHODS

A total of 1270 COVID-19 patients, including 984 admitted in Sino French New City Branch (training and internal validation sets randomly split at 7:3 ratio) and 286 admitted in Optical Valley Branch (external validation set) of Wuhan Tongji hospital, were included in this study. Forty-eight clinical and laboratory features were screened with LASSO method. Further multi-tree extreme gradient boosting (XGBoost) machine learning-based model was used to rank importance of features selected from LASSO and subsequently constructed death risk prediction model with simple-tree XGBoost model. Performances of models were evaluated by AUC, prediction accuracy, precision, and F1 scores.

RESULTS

Six features, including disease severity, age, levels of high-sensitivity C-reactive protein (hs-CRP), lactate dehydrogenase (LDH), ferritin, and interleukin-10 (IL-10), were selected as predictors for COVID-19 mortality. Simple-tree XGBoost model conducted by these features can predict death risk accurately with >90% precision and >85% sensitivity, as well as F1 scores >0.90 in training and validation sets.

CONCLUSION

We proposed the disease severity, age, serum levels of hs-CRP, LDH, ferritin, and IL-10 as significant predictors for death risk of COVID-19, which may help to identify the high-risk COVID-19 cases. KEY MESSAGES A machine learning method is used to build death risk model for COVID-19 patients. Disease severity, age, hs-CRP, LDH, ferritin, and IL-10 are death risk factors. These findings may help to identify the high-risk COVID-19 cases.

Performance of a Prostate-Specific Membrane Antigen Positron Emission Tomography/Computed Tomography-Derived Risk-Stratification Tool for High-risk and Very High-risk Prostate Cancer

IMPORTANCE

Prostate-specific membrane antigen (PSMA) positron emission tomography/computed tomography (PET/CT) can detect low-volume, nonlocalized (ie, regional or metastatic) prostate cancer that was occult on conventional imaging. However, the long-term clinical implications of PSMA PET/CT upstaging remain unclear.

OBJECTIVES

To evaluate the prognostic significance of a nomogram that models an individual's risk of nonlocalized upstaging on PSMA PET/CT and to compare its performance with existing risk-stratification tools.

DESIGN, SETTING, AND PARTICIPANTS

This cohort study included patients diagnosed with high-risk or very high-risk prostate cancer (ie, prostate-specific antigen [PSA] level >20 ng/mL, Gleason score 8-10, and/or clinical stage T3-T4, without evidence of nodal or metastatic disease by conventional workup) from April 1995 to August 2018. This multinational study was conducted at 15 centers. Data were analyzed from December 2020 to March 2021.

EXPOSURES

Curative-intent radical prostatectomy (RP), external beam radiotherapy (EBRT), or EBRT plus brachytherapy (BT), with or without androgen deprivation therapy.

MAIN OUTCOMES AND MEASURES

PSMA upstage probability was calculated from a nomogram using the biopsy Gleason score, percentage positive systematic biopsy cores, clinical T category, and PSA level. Biochemical recurrence (BCR), distant metastasis (DM), prostate cancer-specific mortality (PCSM), and overall survival (OS) were analyzed using Fine-Gray and Cox regressions. Model performance was quantified with the concordance (C) index.

RESULTS

Of 5275 patients, the median (IQR) age was 66 (60-72) years; 2883 (55%) were treated with RP, 1669 (32%) with EBRT, and 723 (14%) with EBRT plus BT; median (IQR) PSA level was 10.5 (5.9-23.2) ng/mL; 3987 (76%) had Gleason grade 8 to 10 disease; and 750 (14%) had stage T3 to T4 disease. Median (IQR) follow-up was 5.1 (3.1-7.9) years; 1221 (23%) were followed up for at least 8 years. Overall, 1895 (36%) had BCR, 851 (16%) developed DM, and 242 (5%) died of prostate cancer. PSMA upstage probability was significantly prognostic of all clinical end points, with 8-year C indices of 0.63 (95% CI, 0.61-0.65) for BCR, 0.69 (95% CI, 0.66-0.71) for DM, 0.71 (95% CI, 0.67-0.75) for PCSM, and 0.60 (95% CI, 0.57-0.62) for PCSM (P < .001). The PSMA nomogram outperformed existing risk-stratification tools, except for similar performance to Staging Collaboration for Cancer of the Prostate (STAR-CAP) for PCSM (eg, DM: PSMA, 0.69 [95% CI, 0.66-0.71] vs STAR-CAP, 0.65 [95% CI, 0.62-0.68]; P < .001; Memorial Sloan Kettering Cancer Center nomogram, 0.57 [95% CI, 0.54-0.60]; P < .001; Cancer of the Prostate Risk Assessment groups, 0.53 [95% CI, 0.51-0.56]; P < .001). Results were validated in secondary cohorts from the Surveillance, Epidemiology, and End Results database and the National Cancer Database.

CONCLUSIONS AND RELEVANCE

These findings suggest that PSMA upstage probability is associated with long-term, clinically meaningful end points. Furthermore, PSMA upstaging had superior risk discrimination compared with existing tools. Formerly occult, PSMA PET/CT-detectable nonlocalized disease may be the main driver of outcomes in high-risk patients.

Development and external validation of a prediction model for tube feeding dependency for at least four weeks during chemoradiotherapy for head and neck cancer

BACKGROUND & AIMS

Patients who receive chemoradiotherapy or bioradiotherapy (CRT/BRT) for locally advanced head and neck squamous cell carcinoma (LAHNSCC) often experience high toxicity rates interfering with oral intake, causing tube feeding (TF) dependency. International guidelines recommend gastrostomy insertion when the expected use of TF exceeds 4 weeks. We aimed to develop and externally validate a prediction model to identify patients who need TF ≥ 4 weeks and would benefit from prophylactic gastrostomy insertion.

METHODS

A retrospective multicenter cohort study was performed in four tertiary head and neck cancer centers in the Netherlands. The prediction model was developed using data from University Medical Center Utrecht and the Netherlands Cancer Institute and externally validated using data from Maastricht University Medical Center and Radboud University Medical Center. The primary endpoint was TF dependency ≥4 weeks initiated during CRT/BRT or within 30 days after CRT/BRT completion. Potential predictors were extracted from electronic health records and radiotherapy dose-volume parameters were calculated.

RESULTS

The developmental and validation cohort included 409 and 334 patients respectively. Multivariable analysis showed predictive value for pretreatment weight change, texture modified diet at baseline, ECOG performance status, tumor site, N classification, mean radiation dose to the contralateral parotid gland and oral cavity. The area under the receiver operating characteristics curve for this model was 0.73 and after external validation 0.62. Positive and negative predictive value for a risk of 90% or higher for TF dependency ≥4 weeks were 81.8% and 42.3% respectively.

CONCLUSIONS

We developed and externally validated a prediction model to estimate TF-dependency ≥4 weeks in LAHNSCC patients treated with CRT/BRT. This model can be used to guide personalized decision-making on prophylactic gastrostomy insertion in clinical practice.

Nomogram to predict contralateral breast cancer risk in breast cancer survivors: A SEER-based study

The main purpose of this study was to build a prediction model for patients with contralateral breast cancer (CBC) using competing risks methodology. The aim is to help clinicians predict the probability of CBC in breast cancer (BC) survivors.We reviewed data from the Surveillance, Epidemiology, and End Results database of 434,065 patients with BC. Eligible patients were used to quantify the association between the development of CBC and multiple characteristics of BC patients using competing risk models. A nomogram was also created to facilitate clinical visualization and analysis. Finally, the stability of the model was verified using concordance index and calibration plots, and decision curve analysis was used to evaluate the clinical utility of the model by calculating the net benefit.Four hundred thirty-four thousand sixty-five patients were identified, of whom 6944 (1.6%) developed CBC in the 10 years follow-up. The 10-year cumulative risk of developing CBC was 2.69%. According to a multivariate competing risk model, older patients with invasive lobular carcinoma who had undergone unilateral BC surgery, and whose tumor was better differentiated, of smaller size and ER-negative/PR-positive, had a higher risk of CBC. The calibration plots illustrated an acceptable correlation between the prediction by nomogram and actual observation, as the calibration curve was closed to the 45° diagonal line. The concordance index for the nomogram was 0.65, which indicated it was well calibrated for individual risk of CBC. Decision curve analysis produced a wide range of risk thresholds under which the model we built would yield a net benefit.BC survivors remain at high risk of developing CBC. Patients with CBC have a worse clinical prognosis compared to those with unilateral BC. We built a predictive model for the risk of developing CBC based on a large data cohort to help clinicians identify patients at high risk, which can then help them plan individualized surveillance and treatment.

Cost-effectiveness analysis of a model of first-trimester prediction and prevention of preterm pre-eclampsia compared with usual care

OBJECTIVES

Pre-eclampsia (PE) causes substantial maternal and neonatal mortality and morbidity. In addition to the personal impact on women, children and their families, PE has a significant economic impact on our society. Recent research suggests that a first-trimester multivariate model is highly predictive of preterm (< 37 weeks' gestation) PE and can be combined successfully with targeted prophylaxis (low-dose aspirin), resulting in an 80% reduction in prevalence of disease. The aim of this study was to examine the potential health outcomes and cost implications following introduction of first-trimester prediction and prevention of preterm PE within a public healthcare setting, compared with usual care, and to conduct a cost-effectiveness analysis to inform health-service decisions regarding implementation of such a program.

METHODS

A decision-analytic model was used to compare usual care with the proposed first-trimester screening intervention within the obstetric population (n = 6822) attending two public hospitals within a metropolitan district health service in New South Wales, Australia, between January 2015 and December 2016. The model, applied from early pregnancy, included exposure to a variety of healthcare professionals and addressed type of risk assessment (usual care or first-trimester screening) and use of (compliance with) low-dose aspirin prescribed prophylactically for prevention of PE. All pathways culminated in six possible health outcomes, ranging from no PE to maternal death. Results were presented as the number of cases of PE gained/avoided and the incremental increase/decrease in economic costs arising from the intervention compared with usual care. Significant assumptions were tested in sensitivity/uncertainty analyses.

RESULTS

The intervention produced, across all gestational ages, 31 fewer cases of PE and reduced aggregate economic health-service costs by 1 431 186 Australian dollars over the 2-year period. None of the tested iterations of uncertainty analyses reported additional cases of PE or higher economic costs. The new intervention based on first-trimester screening dominated usual care.

CONCLUSION

This cost-effectiveness analysis demonstrated a reduction in prevalence of preterm PE and substantial cost savings associated with a population-based program of first-trimester prediction and prevention of PE, and supports implementation of such a policy. © 2020 International Society of Ultrasound in Obstetrics and Gynecology.

Risk of bias in studies on prediction models developed using supervised machine learning techniques: systematic review

OBJECTIVE

To assess the methodological quality of studies on prediction models developed using machine learning techniques across all medical specialties.

DESIGN

Systematic review.

DATA SOURCES

PubMed from 1 January 2018 to 31 December 2019.

ELIGIBILITY CRITERIA

Articles reporting on the development, with or without external validation, of a multivariable prediction model (diagnostic or prognostic) developed using supervised machine learning for individualised predictions. No restrictions applied for study design, data source, or predicted patient related health outcomes.

REVIEW METHODS

Methodological quality of the studies was determined and risk of bias evaluated using the prediction risk of bias assessment tool (PROBAST). This tool contains 21 signalling questions tailored to identify potential biases in four domains. Risk of bias was measured for each domain (participants, predictors, outcome, and analysis) and each study (overall).

RESULTS

152 studies were included: 58 (38%) included a diagnostic prediction model and 94 (62%) a prognostic prediction model. PROBAST was applied to 152 developed models and 19 external validations. Of these 171 analyses, 148 (87%, 95% confidence interval 81% to 91%) were rated at high risk of bias. The analysis domain was most frequently rated at high risk of bias. Of the 152 models, 85 (56%, 48% to 64%) were developed with an inadequate number of events per candidate predictor, 62 handled missing data inadequately (41%, 33% to 49%), and 59 assessed overfitting improperly (39%, 31% to 47%). Most models used appropriate data sources to develop (73%, 66% to 79%) and externally validate the machine learning based prediction models (74%, 51% to 88%). Information about blinding of outcome and blinding of predictors was, however, absent in 60 (40%, 32% to 47%) and 79 (52%, 44% to 60%) of the developed models, respectively.

CONCLUSION

Most studies on machine learning based prediction models show poor methodological quality and are at high risk of bias. Factors contributing to risk of bias include small study size, poor handling of missing data, and failure to deal with overfitting. Efforts to improve the design, conduct, reporting, and validation of such studies are necessary to boost the application of machine learning based prediction models in clinical practice.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42019161764.

External validation of the RISC, RISC-Malawi, and PERCH clinical prediction rules to identify risk of death in children hospitalized with pneumonia

BACKGROUND

Existing scores to identify children at risk of hospitalized pneumonia-related mortality lack broad external validation. Our objective was to externally validate three such risk scores.

METHODS

We applied the Respiratory Index of Severity in Children (RISC) for HIV-negative children, the RISC-Malawi, and the Pneumonia Etiology Research for Child Health (PERCH) scores to hospitalized children in the Pneumonia REsearch Partnerships to Assess WHO REcommendations (PREPARE) data set. The PREPARE data set includes pooled data from 41 studies on pediatric pneumonia from across the world. We calculated test characteristics and the area under the curve (AUC) for each of these clinical prediction rules.

RESULTS

The RISC score for HIV-negative children was applied to 3574 children 0-24 months and demonstrated poor discriminatory ability (AUC = 0.66, 95% confidence interval (CI) = 0.58-0.73) in the identification of children at risk of hospitalized pneumonia-related mortality. The RISC-Malawi score had fair discriminatory value (AUC = 0.75, 95% CI = 0.74-0.77) among 17 864 children 2-59 months. The PERCH score was applied to 732 children 1-59 months and also demonstrated poor discriminatory value (AUC = 0.55, 95% CI = 0.37-0.73).

CONCLUSIONS

In a large external application of the RISC, RISC-Malawi, and PERCH scores, a substantial number of children were misclassified for their risk of hospitalized pneumonia-related mortality. Although pneumonia risk scores have performed well among the cohorts in which they were derived, their performance diminished when externally applied. A generalizable risk assessment tool with higher sensitivity and specificity to identify children at risk of hospitalized pneumonia-related mortality may be needed. Such a generalizable risk assessment tool would need context-specific validation prior to implementation in that setting.

A Fetal Risk Stratification Pathway for Neonatal Aortic Coarctation Reduces Medical Exposure

OBJECTIVE

To test the hypothesis that a fetal stratification pathway will effectively discriminate between infants at different levels of risk for surgical coarctation and reduce unnecessary medicalization.

STUDY DESIGN

We performed a pre-post nonrandomized study in which we prospectively assigned fetuses with prenatal concern for coarctation to 1 of 3 risk categories and implemented a clinical pathway for postnatal management. Postnatal clinical outcomes were compared with those in a historical control group that were not triaged based on the pathway.

RESULTS

The study cohort comprised 109 fetuses, including 57 treated along the fetal coarctation pathway and 52 historical controls. Among mild-risk fetuses, 3% underwent surgical coarctation repair (0% of those without additional heart defects), compared with 27% of moderate-risk and 63% of high-risk fetuses. The combined fetal aortic, mitral, and isthmus z-score best discriminated which infants underwent surgery (area under the curve = 0.78; 95% CI, 0.66-0.91). Compared with historical controls, infants triaged according to the fetal coarctation pathway had fewer delivery location changes (76% vs 55%; P = .025) and less umbilical venous catheter placement (74% vs 51%; P = .046). Trends toward shorter intensive care unit stay, hospital stay, and time to enteral feeding did not reach statistical significance.

CONCLUSIONS

A stratified risk-assignment pathway effectively identifies a group of fetuses with a low rate of surgical coarctation and reduces unnecessary medicalization in infants who do not undergo aortic surgery. Incorporation of novel measurements or imaging techniques may improve the specificity of high-risk criteria.

D-Dimer Testing for the Exclusion of Pulmonary Embolism Among Hospitalized Patients With COVID-19

This prognostic study evaluates the use of plasma D-dimer concentrations to rule out pulmonary embolism among patients hospitalized with COVID-19.

A Continuously Benchmarked and Crowdsourced Challenge for Rapid Development and Evaluation of Models to Predict COVID-19 Diagnosis and Hospitalization

Importance

Machine learning could be used to predict the likelihood of diagnosis and severity of illness. Lack of COVID-19 patient data has hindered the data science community in developing models to aid in the response to the pandemic.

Objectives

To describe the rapid development and evaluation of clinical algorithms to predict COVID-19 diagnosis and hospitalization using patient data by citizen scientists, provide an unbiased assessment of model performance, and benchmark model performance on subgroups.

Design, Setting, and Participants

This diagnostic and prognostic study operated a continuous, crowdsourced challenge using a model-to-data approach to securely enable the use of regularly updated COVID-19 patient data from the University of Washington by participants from May 6 to December 23, 2020. A postchallenge analysis was conducted from December 24, 2020, to April 7, 2021, to assess the generalizability of models on the cumulative data set as well as subgroups stratified by age, sex, race, and time of COVID-19 test. By December 23, 2020, this challenge engaged 482 participants from 90 teams and 7 countries.

Main Outcomes and Measures

Machine learning algorithms used patient data and output a score that represented the probability of patients receiving a positive COVID-19 test result or being hospitalized within 21 days after receiving a positive COVID-19 test result. Algorithms were evaluated using area under the receiver operating characteristic curve (AUROC) and area under the precision recall curve (AUPRC) scores. Ensemble models aggregating models from the top challenge teams were developed and evaluated.

Results

In the analysis using the cumulative data set, the best performance for COVID-19 diagnosis prediction was an AUROC of 0.776 (95% CI, 0.775-0.777) and an AUPRC of 0.297, and for hospitalization prediction, an AUROC of 0.796 (95% CI, 0.794-0.798) and an AUPRC of 0.188. Analysis on top models submitting to the challenge showed consistently better model performance on the female group than the male group. Among all age groups, the best performance was obtained for the 25- to 49-year age group, and the worst performance was obtained for the group aged 17 years or younger.

Conclusions and Relevance

In this diagnostic and prognostic study, models submitted by citizen scientists achieved high performance for the prediction of COVID-19 testing and hospitalization outcomes. Evaluation of challenge models on demographic subgroups and prospective data revealed performance discrepancies, providing insights into the potential bias and limitations in the models.

Clinical Prediction Models in Sports Medicine: A Guide for Clinicians and Researchers

SYNOPSIS

Participating in sport carries inherent risk of injury. Clinicians execute high-level clinical reasoning and decision making to support athletes to achieve the best outcomes. Accurately diagnosing a problem, estimating prognosis, or selecting the most suitable intervention for each athlete is challenging. Clinical prediction models are tools to assist clinicians in estimating the risk or probability of a health outcome for an individual by using data from multiple predictors. Although common in general medical literature, clinical prediction models are rare in sports medicine. The purpose of this article was to (1) describe the steps required to develop and validate (ie, evaluate) a clinical prediction model for clinical researchers, and (2) help sports medicine clinicians understand and interpret clinical prediction model studies. Using a case study to illustrate how to implement clinical prediction models in practice, we address the following issues in developing and validating a clinical prediction model: study design and data, sample size, missing data, selecting predictors, handling continuous predictors, model fitting, internal and external validation, performance measures, reporting, and model presentation. Our work builds on initiatives to improve diagnostic and prognostic clinical research, including the PROGnosis RESearch Strategy (PROGRESS) series of papers and textbook and the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement. J Orthop Sports Phys Ther 2021;51(10):517-525. doi:10.2519/jospt.2021.10697.

Performance of prediction models for nephropathy in people with type 2 diabetes: systematic review and external validation study

OBJECTIVES

To identify and assess the quality and accuracy of prognostic models for nephropathy and to validate these models in external cohorts of people with type 2 diabetes.

DESIGN

Systematic review and external validation.

DATA SOURCES

PubMed and Embase.

ELIGIBILITY CRITERIA

Studies describing the development of a model to predict the risk of nephropathy, applicable to people with type 2 diabetes.

METHODS

Screening, data extraction, and risk of bias assessment were done in duplicate. Eligible models were externally validated in the Hoorn Diabetes Care System (DCS) cohort (n=11 450) for the same outcomes for which they were developed. Risks of nephropathy were calculated and compared with observed risk over 2, 5, and 10 years of follow-up. Model performance was assessed based on intercept adjusted calibration and discrimination (Harrell's C statistic).

RESULTS

41 studies included in the systematic review reported 64 models, 46 of which were developed in a population with diabetes and 18 in the general population including diabetes as a predictor. The predicted outcomes included albuminuria, diabetic kidney disease, chronic kidney disease (general population), and end stage renal disease. The reported apparent discrimination of the 46 models varied considerably across the different predicted outcomes, from 0.60 (95% confidence interval 0.56 to 0.64) to 0.99 (not available) for the models developed in a diabetes population and from 0.59 (not available) to 0.96 (0.95 to 0.97) for the models developed in the general population. Calibration was reported in 31 of the 41 studies, and the models were generally well calibrated. 21 of the 64 retrieved models were externally validated in the Hoorn DCS cohort for predicting risk of albuminuria, diabetic kidney disease, and chronic kidney disease, with considerable variation in performance across prediction horizons and models. For all three outcomes, however, at least two models had C statistics >0.8, indicating excellent discrimination. In a secondary external validation in GoDARTS (Genetics of Diabetes Audit and Research in Tayside Scotland), models developed for diabetic kidney disease outperformed those for chronic kidney disease. Models were generally well calibrated across all three prediction horizons.

CONCLUSIONS

This study identified multiple prediction models to predict albuminuria, diabetic kidney disease, chronic kidney disease, and end stage renal disease. In the external validation, discrimination and calibration for albuminuria, diabetic kidney disease, and chronic kidney disease varied considerably across prediction horizons and models. For each outcome, however, specific models showed good discrimination and calibration across the three prediction horizons, with clinically accessible predictors, making them applicable in a clinical setting.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42020192831.

Comparison of the 2019 European Alliance of Associations for Rheumatology/American College of Rheumatology Systemic Lupus Erythematosus Classification Criteria With Two Sets of Earlier Systemic Lupus Erythematosus Classification Criteria

OBJECTIVE

The Systemic Lupus International Collaborating Clinics (SLICC) 2012 systemic lupus erythematosus (SLE) classification criteria and the revised American College of Rheumatology (ACR) 1997 criteria are list based, counting each SLE manifestation equally. We derived a classification rule based on giving variable weights to the SLICC criteria and compared its performance to the revised ACR 1997, the unweighted SLICC 2012, and the newly reported European Alliance of Associations for Rheumatology (EULAR)/ACR 2019 criteria sets.

METHODS

The physician-rated patient scenarios used to develop the SLICC 2012 classification criteria were reemployed to devise a new weighted classification rule using multiple linear regression. The performance of the rule was evaluated on an independent set of expert-diagnosed patient scenarios and compared to the performance of the previously reported classification rules.

RESULTS

The weighted SLICC criteria and the EULAR/ACR 2019 criteria had less sensitivity but better specificity compared to the list-based revised ACR 1997 and SLICC 2012 classification criteria. There were no statistically significant differences between any pair of rules with respect to overall agreement with the physician diagnosis.

CONCLUSION

The 2 new weighted classification rules did not perform better than the existing list-based rules in terms of overall agreement on a data set originally generated to assess the SLICC criteria. Given the added complexity of summing weights, researchers may prefer the unweighted SLICC criteria. However, the performance of a classification rule will always depend on the populations from which the cases and non-cases are derived and whether the goal is to prioritize sensitivity or specificity.

Optimizing imaging in the pediatric trauma patient, part 1: head and neck trauma

Trauma is the leading cause of death in the pediatric population and is among the most common reasons for ED visits by children. Imaging is an important tool for the diagnosis and management of pediatric trauma, but there are risks associated with exposure to ionizing radiation. In pediatric head and neck injuries, clinical findings and clinical decision tools can help inform selection of the most appropriate imaging modalities for the trauma patient, while also reducing unnecessary radiation exposure. This supplement reviews evidence-based recommendations for imaging decisions and interpretations in skull fractures, traumatic brain injuries, abusive head trauma, cervical spine injuries, and facial bone fractures. Examples demonstrating imaging modalities and specific findings for the types of injuries are also provided.

Prediction of encephalopathy in perinatal asphyxia score: reaching the unreached

OBJECTIVES

Neonates who develop moderate to severe encephalopathy following perinatal asphyxia will benefit from therapeutic hypothermia. Current National Institute of Child Health and Human Development (NICHD) criteria for identifying encephalopathic neonates needing therapeutic hypothermia has high specificity. This results in correctly identifying neonates who have already developed moderate to severe encephalopathy but miss out many potential beneficiaries who progress to develop moderate to severe encephalopathy later. The need is therefore not just to diagnose encephalopathy, but to predict development of encephalopathy and extend the therapeutic benefit for all eligible neonates. The primary objective of the study was to develop and validate the statistical model for prediction of moderate to severe encephalopathy following perinatal asphyxia and compare with current NICHD criteria.

METHODS

The study was designed as prospective observational study. It was carried out in a single center Level 3 perinatal unit in India. Neonates>35 weeks of gestation and requiring resuscitation at birth were included. Levels of resuscitation and blood gas lactate were used to determine the pre-test probability, Thompson score between 3 and 5 h of life was used to determine post-test probability of developing encephalopathy. Primary outcome measure: Validation of Prediction of Encephalopathy in Perinatal Asphyxia (PEPA) score by Holdout method.

RESULTS

A total of 55 babies were included in the study. The PEPA score was validated by Holdout method where the fitted receiver-operating characteristic (ROC) area for the training and test sample were comparable (p=0.758). The sensitivity and specificity of various PEPA scores for prediction of encephalopathy ranged between 74 and 100% in contrast to NICHD criteria which was 42%. PEPA score of 30 had a best combination of sensitivity and specificity of 95 and 89% respectively.

CONCLUSIONS

PEPA score has a higher sensitivity than NICHD criteria for prediction of Encephalopathy in asphyxiated neonates.

Which older emergency patients are at risk of intracranial bleeding after a fall? A protocol to derive a clinical decision rule for the emergency department

INTRODUCTION

Falling on level ground is now the most common cause of traumatic intracranial bleeding worldwide. Older adults frequently present to the emergency department (ED) after falling. It can be challenging for clinicians to determine who requires brain imaging to rule out traumatic intracranial bleeding, and often head injury decision rules do not apply to older adults who fall. The goal of our study is to derive a clinical decision rule, which will identify older adults who present to the ED after a fall who do not have clinically important intracranial bleeding.

METHODS AND ANALYSIS

This is a prospective cohort study enrolling patients aged 65 years or older, who present to the ED of 11 hospitals in Canada and the USA within 48 hours of having a fall. Patients are included if they fall on level ground, off a chair, toilet seat or out of bed. The primary outcome is the diagnosis of clinically important intracranial bleeding within 42 days of the index ED visit. An independent adjudication committee will determine the primary outcome, blinded to all other data. We are collecting data on 17 potential predictor variables. The treating physician completes a study data form at the time of initial assessment, prior to brain imaging. Data extraction is supplemented by an independent, structured electronic medical record review. We will perform binary recursive partitioning using Classification and Regression Trees to derive a clinical decision rule.

ETHICS AND DISSEMINATION

The study was initially approved by the Hamilton Integrated Research Ethics Committee and subsequently approved by the research ethics boards governing all participating sites. We will disseminate our results by journal publication, presentation at international meetings and social media.

TRIAL REGISTRATION NUMBER

NCT03745755.

Sex Differences in Outcomes Following Left Atrial Appendage Closure

OBJECTIVE

To evaluate the effects of female sex on in-hospital outcomes and to provide estimates for sex-specific prediction models of adverse outcomes following left atrial appendage closure (LAAC).

PATIENTS AND METHODS

Cohort-based observational study querying the National Inpatient Sample database between October 1, 2015, and December 31, 2017. Demographics, baseline characteristics, and comorbidities were assessed with the Charlson Comorbidity Index (CCI), Elixhauser Comorbidity Index score (ECS), and CHA₂DS₂-VASc score. The primary outcome was in-hospital major adverse events (MAEs) defined as the composite of bleeding, vascular, cardiac complications, post-procedural stroke, and acute kidney injury. The associations of the CCI, ECS, and CHA₂DS₂-VASc score with in-hospital MAE were examined using logistic regression models for women and men, respectively.

RESULTS

A total of 3294 hospitalizations were identified, of which 1313 (40%) involved women and 1981 (60%) involved men. Women were older (76.3±7.7 vs 75.2±8.4 years, P<.001), had a higher CHA₂DS₂-VASc score (4.9±1.4 vs 3.9±1.4, P<.001) but showed lower CCI and ECS compared with men (2.1±1.9 vs 2.3±1.9, P=.01; and 9.3±5.9 vs 9.9±5.7, P=.002, respectively). The primary composite outcome occurred in 4.6% of patients and was higher in women compared with men (women 5.6% vs men 4.0%, P=.04), and this was mainly driven by the occurrence of cardiac complications (2.4% vs 1.2%, P=.01). In women, older age, higher median income, and higher CCI (adjusted odds ratio [aOR], 1.32; 95% confidence interval [CI], 1.21 to 1.44; P<.001), ECS (aOR, 1.04; 95% CI, 1.02 to 1.07; P=.002), and CHA₂DS₂-VASc score (aOR, 1.24; 95% CI, 1.10 to 1.39; P<.001) were associated with increased risk of in-hospital MAE. In men, non-White race/ethnicity, lower median income, and higher ECS (aOR, 1.06; 95% CI, 1.04 to 1.09; P<.001) were associated with increased risk of in-hospital MAE.

CONCLUSION

Women had higher rates of in-hospital adverse events following LAAC than men did. Women with older age and higher median income, CCI, ECS, and CHA₂DS₂-VASc scores were associated with in-hospital adverse events, whereas men with non-White race/ethnicity, lower median income, and higher ECS were more likely to experience adverse events. Further research is warranted to identify sex-specific, racial/ethnic, and socioeconomic pathways during the patient selection process to minimize complications in patients undergoing LAAC.

A Tale of Two Cities: Applying the Boston Syncope Criteria to Jerusalem

BACKGROUND

Syncope is a common reason for emergency department (ED) visits; however, the decision to admit or discharge patients after a syncopal episode remains challenging for emergency physicians. Decision rules such as the Boston Syncope Criteria have been developed in an attempt to aid clinicians in identifying high-risk patients as well as those who can be safely discharged, but applying these rules to different populations remains unclear.

OBJECTIVES

To determine whether the Boston Syncope Criteria are valid for emergency department patients in Israel.

METHODS

This retrospective cohort convenience sample included patients who visited a tertiary care hospital in Jerusalem from August 2018 to July 2019 with a primary diagnosis of syncope. Thirty-day follow-up was performed using a national health system database. The Boston Syncope Criteria were retrospectively applied to each patient to determine whether they were at high risk for an adverse outcome or critical intervention, versus low risk and could be discharged.

RESULTS

A total of 198 patients fulfilled the inclusion criteria and completed follow-up. Of these, 21 patients had either an adverse outcome or critical intervention. The rule detected 20/21 with a sensitivity of 95%, a specificity of 66%, and a negative predictive value of 99.

CONCLUSIONS

The Boston Syncope Criteria may be useful for physicians in other locations throughout the world to discharge low-risk syncope patients as well as identify those at risk of complications.

Validation of a Clinical Prediction Rule to Predict Asymptomatic Chlamydia and Gonorrhea Infections Among Internet-Based Testers

BACKGROUND

Clinical prediction rules (CPRs) can be used in sexually transmitted infection (STI) testing environments to prioritize individuals at the highest risk of infection and optimize resource allocation. We previously derived a CPR to predict asymptomatic chlamydia and/or gonorrhea (CT/NG) infection among women and heterosexual men at in-person STI clinics based on 5 predictors. Population differences between clinic-based and Internet-based testers may limit the tool's application across settings. The primary objective of this study was to assess the validity, sensitivity, and overall performance of this CPR within an Internet-based testing environment (GetCheckedOnline.com).

METHODS

We analyzed GetCheckedOnline online risk assessment and laboratory data from October 2015 to June 2019. We compared the STI clinic population used for CPR derivation (data previously published) and the GetCheckedOnline validation population using χ2 tests. Calibration and discrimination were assessed using the Hosmer-Lemeshow goodness-of-fit test and the area under the receiver operating curve, respectively. Sensitivity and the fraction of total screening tests offered were quantified for CPR-predicted risk scores.

RESULTS

Asymptomatic CT/NG infection prevalence in the GetCheckedOnline population (n = 5478) was higher than in the STI clinic population (n = 10,437; 2.4% vs. 1.8%, P = 0.007). When applied to GetCheckedOnline, the CPR had reasonable calibration (Hosmer-Lemeshow, P = 0.90) and discrimination (area under the receiver operating characteristic, 0.64). By screening only individuals with total risk scores ≥4, we would detect 97% of infections and reduce screening by 14%.

CONCLUSIONS

The application of an existing CPR to detect asymptomatic CT/NG infection is valid within an Internet-based STI testing environment. Clinical prediction rules applied online can reduce unnecessary STI testing and optimize resource allocation within publicly funded health systems.