The relative merits of using a high-sensitivity cardiac Troponin T assay compared to a nonhigh-sensitivity troponin T assay after noncardiac surgery.

INTRODUCTION: Troponin elevation after noncardiac surgery is associated with an elevated risk of 30-day mortality. Little is known about relative merit of using a high-sensitivity Troponin T (hsTnT), the fifth-generation assay, vs the nonhigh sensitivity Troponin T (non-hsTnT), the fourth-generation assay, in the noncardiac surgery setting. We aimed to identify whether hsTnT can identify additional patients at risk that would have gone undetected with non-hsTnT measurement. METHODS: The VISION Study included 40,004 noncardiac surgery patients with postoperative troponin measurements. Among them, 1,806 patients had both fourth-generation non-hsTnT and fifth-generation hsTnT concomitant measurements (4,451 paired results). We compared the absolute concentrations, the timing, and the impact of different thresholds on predicting 30-day major cardiovascular complications (composite of death, nonfatal cardiac arrest, coronary revascularization, and congestive heart failure). RESULTS: Based on the manufacturers' threshold of 14 ng/L, 580 (32.1%) patients had postoperative hsTnT concentrations greater than the threshold, while their non-hsTnT concentrations were below the manufacturer's threshold. These 580 patients had higher risk of major cardiovascular events (OR 2.33; CI 95% 1.04-5.23; P = .049) than patients with hsTnT concentrations below the manufacturer threshold. Among patients with myocardial injury after noncardiac surgery, only 50% would be detected by the fourth-generation non-hsTnT assay at 6 to 12 hours postoperative as compared to 85% with the fifth-generation hsTnT assay (P-value < .001). CONCLUSIONS: Within the first 3 postoperative days, fifth-generation hsTnT identified at least 1 in 3 patients with troponin elevation that would have gone undetected by fourth-generation non-hsTnT using published thresholds in this setting. Furthermore, fifth-generation hsTnT identified patients with an elevation earlier than fourth-generation non-hsTnT, indicating potential to improve postoperative risk stratification.

Canadian Internal Medicine Ultrasound (CIMUS) Expert Consensus Statement on the Use of Lung Ultrasound for the Assessment of Medical Inpatients With Known or Suspected Coronavirus Disease 2019.

OBJECTIVES: To develop a consensus statement on the use of lung ultrasound (LUS) in the assessment of symptomatic general medical inpatients with known or suspected coronavirus disease 2019 (COVID-19). METHODS: Our LUS expert panel consisted of 14 multidisciplinary international experts. Experts voted in 3 rounds on the strength of 26 recommendations as "strong," "weak," or "do not recommend." For recommendations that reached consensus for do not recommend, a fourth round was conducted to determine the strength of those recommendations, with 2 additional recommendations considered. RESULTS: Of the 26 recommendations, experts reached consensus on 6 in the first round, 13 in the second, and 7 in the third. Four recommendations were removed because of redundancy. In the fourth round, experts considered 4 recommendations that reached consensus for do not recommend and 2 additional scenarios; consensus was reached for 4 of these. Our final recommendations consist of 24 consensus statements; for 2 of these, the strength of the recommendations did not reach consensus. CONCLUSIONS: In symptomatic medical inpatients with known or suspected COVID-19, we recommend the use of LUS to: (1) support the diagnosis of pneumonitis but not diagnose COVID-19, (2) rule out concerning ultrasound features, (3) monitor patients with a change in the clinical status, and (4) avoid unnecessary additional imaging for patients whose pretest probability of an alternative or superimposed diagnosis is low. We do not recommend the use of LUS to guide admission and discharge decisions. We do not recommend routine serial LUS in patients without a change in their clinical condition.

Automatic multi-view pose estimation in focused cardiac ultrasound.

Focused cardiac ultrasound (FoCUS) is a valuable point-of-care method for evaluating cardiovascular structures and function, but its scope is limited by equipment and operator's experience, resulting in primarily qualitative 2D exams. This study presents a novel framework to automatically estimate the 3D spatial relationship between standard FoCUS views. The proposed framework uses a multi-view U-Net-like fully convolutional neural network to regress line-based heatmaps representing the most likely areas of intersection between input images. The lines that best fit the regressed heatmaps are then extracted, and a system of nonlinear equations based on the intersection between view triplets is created and solved to determine the relative 3D pose between all input images. The feasibility and accuracy of the proposed pipeline were validated using a novel realistic in silico FoCUS dataset, demonstrating promising results. Interestingly, as shown in preliminary experiments, the estimation of the 2D images' relative poses enables the application of 3D image analysis methods and paves the way for 3D quantitative assessments in FoCUS examinations.

The association of attentional foci and image interpretation accuracy in novices interpreting lung ultrasound images: an eye-tracking study.

It is unclear, where learners focus their attention when interpreting point-of-care ultrasound (POCUS) images. This study seeks to determine the relationship between attentional foci metrics with lung ultrasound (LUS) interpretation accuracy in novice medical learners. A convenience sample of 14 medical residents with minimal LUS training viewed 8 LUS cineloops, with their eye-tracking patterns recorded. Areas of interest (AOI) for each cineloop were mapped independently by two experts, and externally validated by a third expert. Primary outcome of interest was image interpretation accuracy, presented as a percentage. Eye tracking captured 10 of 14 participants (71%) who completed the study. Participants spent a mean total of 8 min 44 s ± standard deviation (SD) 3 min 8 s on the cineloops, with 1 min 14 s ± SD 34 s spent fixated in the AOI. Mean accuracy score was 54.0% ± SD 16.8%. In regression analyses, fixation duration within AOI was positively associated with accuracy [beta-coefficients 28.9 standardized error (SE) 6.42, P = 0.002). Total time spent viewing the videos was also significantly associated with accuracy (beta-coefficient 5.08, SE 0.59, P < 0.0001). For each additional minute spent fixating within the AOI, accuracy scores increased by 28.9%. For each additional minute spent viewing the video, accuracy scores increased only by 5.1%. Interpretation accuracy is strongly associated with time spent fixating within the AOI. Image interpretation training should consider targeting AOIs.

Bedside Ultrasound of Quadriceps to Predict Rehospitalization and Functional Decline in Hospitalized Elders.

OBJECTIVE: To evaluate the capacity of total anterior thigh thickness, quadriceps muscle thickness, and quadriceps contractile index, all measured by bedside ultrasound, to predict rehospitalization, functional decline, and death in elderly patients 3 months after hospital discharge. To evaluate intra and interobserver reproducibility of the dominant thigh evaluation method by point of care ultrasound. METHODS: Cohort study of patients aged 65 years or more admitted to a medium complexity unit in a teaching hospital in southern Brazil. Comprehensive geriatric assessment and ultrasound evaluation of the dominant thigh of each participant were performed. After 3 months of hospital discharge, telephone contact was made to evaluate the outcomes of rehospitalization or death and functional decline-assessed by the 100 points Barthel scale and defined as a decrease of five or more points. RESULTS: 100 participants were included. There was no statistically significant difference between intraobserver measurements in the GEE method analysis (p > 0.05), and the mean bias obtained in Bland-Altman plots was close to zero in all four analyses performed, suggesting good intra and interobserver agreement. There was a significant correlation between the echographic measurements (quadriceps thickness and contractile index) and gait speed, timed up and go, and handgrip tests. There was a significant association between contractile index (quadriceps thickness over total anterior thigh thickness multiplied by 100) lower than 60% and functional decline (relative risk 1.35; CI 95% 1.10-1.65; p = 0.003) as well as between the thickness of the quadriceps and rehospitalization or death, in both individuals with preserved walking capacity and in bedridden elders (relative risk 1.34; CI 95% 1.02-1.75; p = 0.04). CONCLUSION: The ultrasonographic method to evaluate thigh thickness was easily applicable and reproducible. The thickness of the quadriceps could predict rehospitalization or death, even in those patients without walking capacity-unable to perform gait speed and timed up and go tests. Additionally, the contractile index was associated with functional decline after 3 months of hospital discharge. This is a promising result, which highlights the bedside ultrasound of the quadriceps as a potential tool for the prognosis evaluation of bedridden hospitalized elderly patients.

Randomized study of effectiveness of computerized ultrasound simulators for an introductory course for residents in Brazil.

PURPOSE: This study aimed to assess the impact of ultrasound simulation (SonoSim) on educational outcomes of an introductory point-of-care ultrasound course compared to hands-on training with live models alone. METHODS: Fifty-three internal medicine residents without ultrasound experience were randomly assigned to control or experimental groups. They participated in an introductory point-of-care ultrasound course covering eight topics in eight sessions from June 23, 2014 until July 18, 2014. Both participated in lecture and hands-on training, but experimental group received an hour of computerized simulator training instead of a second hour of hands-on training. We assessed clinical knowledge and image acquisition with written multiple-choice and practical exams, respectively. Of the 53 enrolled, 40 participants (75.5%) completed the course and all testing. RESULTS: For the 30-item written exam, mean score of the experimental group was 23.1±3.4 (n=21) vs. 21.8±4.8 (n=19), (P>0 .05). For the practical exam, mean score for both groups was 8.7 out of 16 (P>0 .05). CONCLUSION: The substitution of eight hours of ultrasound simulation training for live model scanning in a 24 hour training course did not enhance performance on written and image acquisition tests in an introductory ultrasound course for residents. This result suggests that ultrasound simulation technology used as a substitute for live model training on an hour-for-hour basis, did not improve learning outcomes. Further investigation into simulation as a total replacement for live model training will provide a clearer picture of the efficacy of ultrasound simulators in medical education.