Explore the vast range of titles available.

Clinical prognostic models can assist patient care decisions. However, their performance can drift over time and location, necessitating model monitoring and updating. Despite rapid and significant changes during the pandemic, prognostic models for COVID-19 patients do not currently account for these drifts. We develop a framework for continuously monitoring and updating prognostic models and apply it to predict 28-day survival in COVID-19 patients. We use demographic, laboratory, and clinical data from electronic health records of 34912 hospitalized COVID-19 patients from March 2020 until May 2022 and compare three modeling methods. Model calibration performance drift is immediately detected with minor fluctuations in discrimination. The overall calibration on the prospective validation cohort is significantly improved when comparing the dynamically updated models against their static counterparts. Our findings suggest that, using this framework, models remain accurate and well-calibrated across various waves, variants, race and sex and yield positive net-benefits. Despite rapid and significant changes during the pandemic, prognostic models for COVID-19 patients do not currently account for data drifts. Here, the authors develop a framework for continuously monitoring and updating prognostic models and applied it to predict 28-day survival in COVID-19 patients.

Background The number of cases from the coronavirus disease 2019 (COVID-19) global pandemic has overwhelmed existing medical facilities and forced clinicians, patients, and families to make pivotal decisions with limited time and information. Main body While machine learning (ML) methods have been previously used to augment clinical decisions, there is now a demand for “Emergency ML.” Throughout the patient care pathway, there are opportunities for ML-supported decisions based on collected vitals, laboratory results, medication orders, and comorbidities. With rapidly growing datasets, there also remain important considerations when developing and validating ML models. Conclusion This perspective highlights the utility of evidence-based prediction tools in a number of clinical settings, and how similar models can be deployed during the COVID-19 pandemic to guide hospital frontlines and healthcare administrators to make informed decisions about patient care and managing hospital volume.

Antithrombotic guidance statements for hospitalized patients with coronavirus disease 2019 (COVID‐19) suggest a universal thromboprophylactic strategy with potential to escalate doses in high‐risk patients. To date, no clear approach exists to discriminate patients at high risk for venous thromboembolism (VTE). The objective of this study is to externally validate the IMPROVE‐DD risk assessment model (RAM) for VTE in a large cohort of hospitalized patients with COVID‐19 within a multihospital health system. This retrospective cohort study evaluated the IMPROVE‐DD RAM on adult inpatients with COVID‐19 hospitalized between March 1, 2020, and April 27, 2020. Diagnosis of VTE was defined by new acute deep venous thrombosis or pulmonary embolism by Radiology Department imaging or point‐of‐care ultrasound. The receiver operating characteristic (ROC) curve was plotted and area under the curve (AUC) calculated. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated using standard methods. A total of 9407 patients were included, with a VTE prevalence of 2.9%. The VTE rate was 0.4% for IMPROVE‐DD score 0‐1 (low risk), 1.3% for score 2‐3 (moderate risk), and 5.3% for score ≥ 4 (high risk). Approximately 45% of the total population scored high VTE risk, while 21% scored low VTE risk. IMPROVE‐DD discrimination of low versus medium/high risk showed sensitivity of 0.971, specificity of 0.218, PPV of 0.036, and NPV of 0.996. ROC AUC was 0.702. The IMPROVE‐DD VTE RAM demonstrated very good discrimination to identify hospitalized patients with COVID‐19 as low, moderate, and high VTE risk in this large external validation study with potential to individualize thromboprophylactic strategies.

Chest radiographs (CXRs) are the most widely available radiographic imaging modality used to detect respiratory diseases that result in lung opacities. CXR reports often use non-standardized language that result in subjective, qualitative, and non-reproducible opacity estimates. Our goal was to develop a robust deep transfer learning framework and adapt it to estimate the degree of lung opacity from CXRs. Following CXR data selection based on exclusion criteria, segmentation schemes were used for ROI (Region Of Interest) extraction, and all combinations of segmentation, data balancing, and classification methods were tested to pick the top performing models. Multifold cross validation was used to determine the best model from the initial selected top models, based on appropriate performance metrics, as well as a novel Macro-Averaged Heatmap Concordance Score (MA HCS). Performance of the best model is compared against that of expert physician annotators, and heatmaps were produced. Finally, model performance sensitivity analysis across patient populations of interest was performed. The proposed framework was adapted to the specific use case of estimation of degree of CXR lung opacity using ordinal multiclass classification. Acquired between March 24, 2020, and May 22, 2020, 38,365 prospectively annotated CXRs from 17,418 patients were used. We tested three neural network architectures (ResNet-50, VGG-16, and ChexNet), three segmentation schemes (no segmentation, lung segmentation, and lateral segmentation based on spine detection), and three data balancing strategies (undersampling, double-stage sampling, and synthetic minority oversampling) using 38,079 CXR images for training, and validation with 286 images as the out-of-the-box dataset that underwent expert radiologist adjudication. Based on the results of these experiments, the ResNet-50 model with undersampling and no ROI segmentation is recommended for lung opacity classification, based on optimal values for the MAE metric and HCS (Heatmap Concordance Score). The degree of agreement between the opacity scores predicted by this model with respect to the two sets of radiologist scores (OR or Original Reader and OOBTR or Out Of Box Reader) in terms of performance metrics is superior to the inter-radiologist opacity score agreement.

Introduction: Limited information exists on patients with suspected coronavirus disease 2019 (COVID-19) who return to the emergency department (ED) during the first wave. In this study we aimed to identify predictors of ED return within 72 hours for patients with suspected COVID-19. Methods: Incorporating data from 14 EDs within an integrated healthcare network in the New York metropolitan region from March 2–April 27, 2020, we analyzed this data on predictors for a return ED visit—including demographics, comorbidities, vital signs, and laboratory results. Results: In total, 18,599 patients were included in the study. The median age was 46 years old [interquartile range 34-58]), 50.74% were female, and 49.26% were male. Overall, 532 (2.86%) returned to the ED within 72 hours, and 95.49% were admitted at the return visit. Of those tested for COVID-19, 59.24% (4704/ 7941) tested positive. Patients with chief complaints of “fever” or “flu” or a history of diabetes or renal disease were more likely to return at 72 hours. Risk of return increased with persistently abnormal temperature (odds ratio [OR] 2.43, 95% CI 1.8-3.2), respiratory rate (2.17, 95% CI 1.6-3.0), and chest radiograph (OR 2.54, 95% CI 2.0-3.2). Abnormally high neutrophil counts, low platelet counts, high bicarbonate values, and high aspartate aminotransferase levels were associated with a higher rate of return. Risk of return decreased when discharged on antibiotics (OR 0.12, 95% CI 0.0-0.3) or corticosteroids (OR 0.12, 95% CI 0.0-0.9). Conclusion: The low overall return rate of patients during the first COVID-19 wave indicates that physicians’ clinical decision-making successfully identified those acceptable for discharge.

A 57–year–old male former smoker presented to the Emergency Department (ED) with blurry vision, headache, and generalized weakness. He was hypoxic on room air and ECG showed sinus tachycardia. A CT pulmonary angiogram was ordered in the ED and revealed no pulmonary embolism but incidentally noted a likely significant stenosis in the proximal LAD. Subsequent cardiac catheterization revealed a 90% stenotic lesion with percutaneous coronary intervention leading to symptom resolution. Unlike coronary CTA, CTPA is performed with non–ECG gated helical scanning and generates motion artifacts associated with myocardial contraction. However, the timing of vascular contrast opacification during CTPA often allows for at least partial evaluation of the coronary arteries, especially ostial and proximal segments. Physicians ordering and evaluating noncardiac–focused CT chest studies, particularly in the ED, should remain cognizant of the radiographic appearance of underlying CAD, particularly life–threatening incidental coronary pathologies to allow for timely management and intervention.

Despite the increasing use of Coronary Artery Calcium (CAC) scoring for cardiovascular risk stratification in asymptomatic patients, the gender differences in CAC among symptomatic patients have not been well evaluated. We analyzed patients presenting to the emergency department (ED) with chest pain suggesting possible coronary artery disease (CAD) who received coronary computed tomography angiography (CCTA). Ordinal logistic regression was used to determine the odds ratio for the association of traditional cardiovascular risk factors and CAC. Patients with a CAC score ≥ 100 were followed for cardiovascular events or changes in medical management. Our cohort included 542 individuals (263 male, 279 female). Ordinal logistic regression model showed that among traditional cardiovascular risk factors, male sex had the highest odds ratio (OR) of 3.04 (p < 0.001, 95% CI [2.01, 4.59]) for the presence of CAC. Also, males had more diffuse distribution of coronary atherosclerosis (p=0.01). Subgroup analysis revealed that obesity was a bigger risk factor in male patients (OR 2.16), while smoking showed the greatest effect (OR 4.27) on CAC in women. Of patients who had CAC > 100 with an average follow-up of 346 days, there was an increase in both aspirin and statin use, yet significant sex differences were observed especially in patients with non-obstructive lesions on CCTA. Among male patients with non-obstructive lesions, 68.2% were on aspirin and 86.4% were on statin therapy after the CCTA compared to 27.3% and 45.5% respectively in their female counterparts. In conclusion, sex not only is the most powerful predictor for higher CAC among traditional cardiovascular risk factors in symptomatic patients but also influences the contribution of various traditional risk factors to elevated CAC. Furthermore, the discovery of CAD led to the initiation of medical therapy in male patients more frequently than in female patients, even after adjusting for the degree of luminal stenosis detected on coronary CT angiography.

ObjectivesAs the relationship between CT scout landmarks and chest CT boundaries is not known, the selected scan length is often greater than necessary for the CT scan, resulting in increased radiation dose to the neck and upper abdomen. The purpose of this study is to establish the relationship between CT scout landmarks with the superior and inferior boundaries of the lungs on chest CT.MethodsRetrospective comparison of the location of the top of the first rib on frontal scout and the most inferior costophrenic angle on lateral scout to the chest CT slice just above and below the lungs. The percent of scans that would exclude part of the lung based on CT initiated at several distances above or below these landmarks was calculated.ResultsThere was 2.7 times greater variability between scout landmarks and lung boundaries inferiorly than superiorly on chest CT (p < 0.001). Initiating CT at the top of the first rib on scout did not exclude any lung on CT. Initiating CT 0, 1, 2, 3, and 4 cm inferior to the CPA on lateral scout excluded part of the lung in 45.7%, 12.9%, 4.3%, 1.9%, and 0.8% of CTs.ConclusionsChest CT to include the lungs should be performed from the top of the first rib to 3 or 4 cm below the costophrenic angle on lateral topogram.Key Points• There is a greater motion at the inferior lung than at the superior lung.• Chest CT acquisition from the top of the first rib on scout would not exclude the lung.• Chest CT acquisition from CPA on lateral scout would exclude the lung 46% of time.

Traditionally, cholecystectomy for cholecystitis is performed within 3 days of the onset of symptoms or after 5 weeks, allowing for resolution of the inflammatory response. This study reviewed the outcomes of cholecystectomy performed for patients with gallstone disease in the acute (n = 45), intermediate (n = 55), and delayed (n = 102) periods after the onset of symptoms. The medical records of 202 patients who underwent laparoscopic cholecystectomy at a large municipal hospital were reviewed retrospectively. The primary outcomes studied were length of hospital stay, conversion to open cholecystectomy, and complications. There was no significant difference in the conversion rate (acute [18%] vs intermediate [20%] vs delayed [11%]) or complication rate (acute [16%] vs intermediate [9%] vs delayed [7%]) among the 3 groups. The delayed group had a significantly shorter length of hospital stay than the intermediate or acute group (3.1 ± 3.8 vs 4.3 ± 3.8 vs 1.7 ± 2.1, respectively, P < .001). Patients who present with acute symptoms of cholecystitis should undergo surgery during the same admission, regardless of the duration of symptoms.

Fluoroscopic imaging guidance is frequently used in performing spinal interventional techniques. Reference level standards are a quality improvement tool to help reduce radiation dose and serve as benchmarks for physicians and their technologists to achieve reasonable radiation exposure while performing fluoroscopically-guided spinal procedures. There are limited data describing radiation dose for musculoskeletal injections - in particular, spinal injections without any published reference standards. The purpose of this study is to perform a practice audit of radiation doses of fluoroscopically-guided spinal injections to establish preliminary reference levels as a quality improvement tool for potential use in future radiation reduction measures. Retrospective, observational study. An academic-based subspecialty, high volume pain medicine practice. A retrospective analysis of 6,234 spinal injections of 9 different types performed by experienced practitioners between January and December 2012 was conducted under an institutional review board's approval with HIPAA compliance and waiver of informed consent. Cumulative radiation dose (in mGy) and exposure time (in seconds) distributions (percentiles) as displayed on the C-arm were calculated per injection for each type of fluoroscopically-guided spinal injection. Confidence intervals for the dose distributions were determined by using bootstrap resampling and were used to determine preliminary reference levels. Proposed preliminary reference levels of cumulative radiation dose (in mGy) and exposure time (in seconds) for fluoroscopically-guided spinal interventional procedures are provided for lumbar transforaminal (13 mGy, 30 s), cervical transforaminal (6 mGy, 49 s), caudal epidural (12 mGy, 23s), cervical facet injection (3 mGy, 36 s), lumbar facet injection (9 mGy, 20s), interlaminar (13mGy, 39s), lumbar radiofrequency denervation (7 mGy, 17s), lumbar sympathetic block (21 mGy, 39s), cervical medial branch block (2 mGy, 25 s), lumbar medial branch block (4 mGy, 12s) and sacroiliac joint injections (18 mGy, 37s). Study performed at a single subspecialty institution using only one type of C-arm which limits generalizability. Radiation doses and preliminary reference levels of fluoroscopically-guided interventional spine procedures performed by experienced practitioners are made available without correction for body habitus or field of view, magnification or subtraction techniques or continuous vs pulsed mode. A registry of radiation-dose data for fluoroscopically-guided interventional spine procedures would be the next step to refine this data. Spinal procedures, radiation dose, patient safety.