Scoring Algorithms Research Articles

Antimicrobial Stewardship Strategies Including Point-of-Care Testing (POCT) for Pediatric Patients with Upper-Respiratory-Tract Infections in Primary Care: A Systematic Review of Economic Evaluations.

Upper-respiratory-tract infections (URTIs) are among the main causes of antibiotic prescriptions in pediatric patients. Over one-third of all antibiotic prescriptions for URTIs in children are estimated to be inappropriate, as the majority of URTIs are caused by viral agents. Several strategies, including clinical scoring algorithms and different point-of-care tests (POCTs) have been developed to help discriminate bacterial from viral URTIs in the outpatient clinical setting. A systematic review of the literature was conducted following PRISMA guidelines with the objective of summarizing evidence from health–economic evaluations on the use of POCT for URTIs in pediatric outpatients. A total of 3375 records identified from four databases and other sources were screened, of which 8 met the inclusion criteria. Four studies were classified as being of high reporting quality, and three were of medium quality. Five out of eight studies concluded in favor of strategies that included POCTs, with an additional study finding several POCTs to be cost-effective compared to usual care but over an acceptable WTP threshold. This review found POCT could be a valuable tool for antimicrobial stewardship strategies targeted towards childhood URTIs in primary care.

Fisher Scoring with Condition-Based Ensemble Supervised Learning Classification Technique for Prediction in PFZ

Potential fishing zone (PFZ) alerts are critical in anticipating fishing places. Earlier PFZ predictions are based on NOAA’s advanced very high-resolution radiometer (AVHRR). To a significant degree, the expansion of the fishing industry may be attributed to the influence of research on fish growers, fishermen, fisheries planners, and managers. Artificial intelligence (AI) technologies are increasingly being used to improve the sustainability of smart fishing. While sustainability is frequently touted to be the intended consequence of AI applications, there is no data currently on how AI contributes to sustainable fishing. The purpose of this paper is to perform a feature selection using the fisher’s score (FS) technique to find the optimal features for final classification. Normalization is used as a preprocessing technique to remove missing and irrelevant data. Here, the collected features, financial derivatives, and geometrical features are used, which leads to poor classification accuracy for predicting the PFZ. Therefore, to improve the accuracy of the condition-based ensemble machine learning and deep learning classification technique (CECT), FS is used and provides the minimum number of features for classification. The experiment is carried out on collected data and tested with existing techniques in terms of accuracy, sensitivity, specificity, and F-measure. The simulation results proved that the proposed technique achieved 96.11% accuracy and 96% specificity compared to the FS technique.

Test Score Algorithms for Budgeted Stochastic Utility Maximization

Motivated by recent developments in designing algorithms based on individual item scores for solving utility maximization problems, we study the framework of using test scores, defined as a statistic of observed individual item performance data, for solving the budgeted stochastic utility maximization problem. We extend an existing scoring mechanism, namely, the replication test scores, to incorporate heterogeneous item costs as well as item values. We show that a natural greedy algorithm that selects items solely based on their replication test scores outputs solutions within a constant factor of the optimum for the class of functions satisfying an extended diminishing returns property. Our algorithms and approximation guarantees assume that test scores are noisy estimates of certain expected values with respect to marginal distributions of individual item values, thus making our algorithms practical and extending previous work that assumes noiseless estimates. Moreover, we show how our algorithm can be adapted to the setting in which items arrive in a streaming fashion while maintaining the same approximation guarantee. We present numerical results, using synthetic data and data sets from the Academia.StackExchange Q&A forum, which show that our test score algorithm can achieve competitiveness and in some cases better performance than a benchmark algorithm that requires access to a value oracle to evaluate function values. Funding: This research was supported, in part, by the Institute for Basic Science [Grants IBS-R029-C1, IBS-R029-Y2].

Evaluation of Orbital Lymphoproliferative and Inflammatory Disorders by Gene Expression Analysis.

Non-specific orbital inflammation (NSOI) and IgG4-related orbital disease (IgG4-ROD) are often challenging to differentiate. Furthermore, it is still uncertain how chronic inflammation, such as IgG4-ROD, can lead to mucosa-associated lymphoid tissue (MALT) lymphoma. Therefore, we aimed to evaluate the diagnostic value of gene expression analysis to differentiate orbital autoimmune diseases and elucidate genetic overlaps. First, we established a database of NSOI, relapsing NSOI, IgG4-ROD and MALT lymphoma patients of our orbital center (2000–2019). In a consensus process, three typical patients of the above mentioned three groups (mean age 56.4 ± 17 years) at similar locations were selected. Afterwards, RNA was isolated using the RNeasy FFPE kit (Qiagen) from archived paraffin-embedded tissues. The RNA of these 12 patients were then subjected to gene expression analysis (NanoString nCounter®), including a total of 1364 target genes. The most significantly upregulated and downregulated genes were used for a machine learning algorithm to distinguish entities. This was possible with a high probability (p < 0.0001). Interestingly, gene expression patterns showed a characteristic overlap of lymphoma with IgG4-ROD and NSOI. In contrast, IgG4-ROD shared only altered expression of one gene regarding NSOI. To validate our potential biomarker genes, we isolated the RNA of a further 48 patients (24 NSOI, 11 IgG4-ROD, 13 lymphoma patients). Then, gene expression pattern analysis of the 35 identified target genes was performed using a custom-designed CodeSet to assess the prediction accuracy of the multi-parameter scoring algorithms. They showed high accuracy and good performance (AUC ROC: IgG4-ROD 0.81, MALT 0.82, NSOI 0.67). To conclude, genetic expression analysis has the potential for faster and more secure differentiation between NSOI and IgG4-ROD. MALT-lymphoma and IgG4-ROD showed more genetic similarities, which points towards progression to lymphoma.

Comparative Effectiveness of Ustekinumab and Anti-TNF Agent as First-Line Biological Therapy in Luminal Crohn's Disease: A Retrospective Study From 2 Referral Centers.

Real-life data on the efficacy of ustekinumab as first-line therapy for the treatment of luminal Crohn's disease (CD) compared with anti-tumor necrosis factor (anti-TNF) agents are lacking. We compared the clinical response rates at 3 months in 2 cohorts of biologic-naïve patients treated by ustekinumab and anti-TNF agents. Biologic-naïve patients starting either ustekinumab or an anti-TNF agent for luminal CD between 2016 and 2019 in 2 tertiary centers were retrospectively included. The primary endpoint was clinical response at 3 months, defined as a Harvey-Bradshaw Index <4 or a 3-point drop in the score without steroids, need for CD-related surgery, or treatment discontinuation owing to failure or intolerance. Patients treated with ustekinumab were matched to patients receiving anti-TNF agents by a propensity score algorithm. We included 156 patients starting anti-TNF agents (95 adalimumab and 61 infliximab) and 50 ustekinumab. After matching, clinical response rates at 3 months were 64% and 86% in the ustekinumab and anti-TNF groups, respectively (P = .01). At 12 months, in multivariate analysis adjusted for disease duration, location, concomitant immunosuppressant and steroids, and symptoms, clinical remission was independently associated with the biological therapy received (odds ratio, 2.6 for anti-TNF agent vs ustekinumab; P = .02). With a median follow-up duration of 40 (interquartile range, 23-52) months, no difference was observed in terms of time to drug withdrawal (P = .29) or safety. This retrospective real-world data suggest that an anti-TNF agent as a first-line biological therapy is associated with higher rates of response at 3 months than ustekinumab in patients with CD.

Impact of laparoscopic surgical algorithm to triage timing of tumor reductive surgery on overall survival in newly diagnosed advanced ovarian cancer (548)

<b>Objectives:</b> To estimate the effects of implementing a laparoscopic scoring algorithm to triage timing of tumor reductive surgery (TRS) in patients with newly diagnosed advanced-stage ovarian cancer on overall survival. <b>Methods:</b> In April 2013, we initiated a prospective quality improvement project utilizing laparoscopic scoring assessment to determine primary resectability in patients with newly diagnosed advanced- stage ovarian cancer in order to improve complete gross resection rates (R0). Current data were analyzed through March 2021. Patients who were medically inoperable or had distant metastatic disease did not undergo laparoscopic assessment and received neoadjuvant chemotherapy (NACT). A validated scoring method was used, and patients with predictive index value (PIV) > 10 received NACT, and those with PIV < 10 were offered primary TRS (algorithm updated from PIV <8 on March 1, 2018). Univariate and multivariate analysis was performed for effects of overall survival (OS). <b>Results:</b> A total of 1,124 patients were included in our analysis; the median follow-up was 28 months (1-96 months). Three hundred forty-five patients (31%) underwent laparoscopic scoring assessment, of which 206 patients (59%) had PIV <10, 112 patients (32%) had PIV > 10, 10 patients (3%) were unscorable, and 17 patients (5%) had an outside laparoscopic assessment with score unknown. Six hundred sixty-four (59%) patients were triaged directly to NACT without laparoscopic assessment. Patients triaged to primary TRS had a lower median CA-125 (404 vs 647 U/mL, p<0.001) and lower median platelet count at diagnosis (312 vs 365, p<0.001) compared to those triaged to NACT by laparoscopy. There was no difference in complete gross resection rates (R0) between those patients triaged to primary TRS versus NACT (84% vs 75%, p=0.32); however, 13% of patients triaged to NACT (p<0.001) were not candidates for interval TRS due to either stable disease (4%) or death (6%). Patients who underwent primary TRS had a significantly longer OS (85.4 months) compared to those who underwent NACT following laparoscopy (36.6 months) or were triaged to NACT and not candidates for laparoscopic assessment (38.5 months, p<0.001). Age (p<0.001), advanced stage (p<0.001), Charlson comorbidity index (p<0.001), platelet count at diagnosis (p=0.001), germline <i>BRCA</i> mutation status (p<0.001), and complete gross resection at TRS (<0.001) were factors significantly associated with OS in our patient population. <b>Conclusions:</b> Implementation of laparoscopic scoring triage in patients with newly diagnosed advanced-stage ovarian patients resulted in high complete gross resection rates at TRS and appropriately triaged select patients to primary TRS who had significantly longer OS. Laparoscopic scoring assessment is a valuable tool to incorporate into surgical practice in advanced ovarian cancer.

Comparison of three validated PD-L1 immunohistochemical assays in urothelial carcinoma of the bladder: interchangeability and issues related to patient selection.

Different programmed cell death-ligand 1 (PD-L1) assays and scoring algorithms are being used in the evaluation of PD-L1 expression for the selection of patients for immunotherapy in specific settings of advanced urothelial carcinoma (UC). In this paper, we sought to investigate three approved assays (Ventana SP142 and SP263, and Dako 22C3) in UC with emphasis on implications for patient selection for atezolizumab/pembrolizumab as the first line of treatment. Tumors from 124 patients with invasive UC of the bladder were analyzed using tissue microarrays (TMA). Serial sections were stained with SP263 and SP142 on Ventana Benchmark Ultra and with 22C3 on Dako Autostainer Link 48. Stains were evaluated independently by two observers and scored using the combined positive score (CPS) and tumor infiltrating immune cells (IC) algorithms. Differences in proportions (DP), overall percent agreement (OPA), positive percent agreement (PPA), negative percent agreement (NPA), and Cohen κ were calculated for all comparable cases. Good overall concordance in analytic performance was observed for 22C3 and SP263 with both scoring algorithms; specifically, the highest OPA was observed between 22C3 and SP263 (89.6%) when using CPS. On the other hand, SP142 consistently showed lower positivity rates with high differences in proportions (DP) compared with 22C3 and SP263 with both CPS and IC, and with a low PPA, especially when using the CPS algorithm. In conclusion, 22C3 and SP263 assays show comparable analytical performance while SP142 shows divergent staining results, with important implications for the selection of patients for both pembrolizumab and atezolizumab.

Comparison of accuracy, sensitivity and specifity of Bahrudin score vs Siriraj score vs Gajah Mada algorithm in diagnosing type of stroke

IntroductionRapid diagnosis is crucial for stroke patients since it is an emergency that may result in morbidity and mortality. The gold standard, which is a CT scan of the brain is not always feasible, hence, Siriraj and Bahrudin Score, as well as Gajah Mada Algorithm are likely to be alternatives. ObjectivesThis study aims to determine the specificity and sensitivity of the Siriraj and Bahrudin Score as well as Gajah Mada Algorithm. MethodA cross-sectional study was undertaken at five network hospitals of the Medical Faculty University in Muhammadiyah Malang. It involves a sample of 304 medical records used to determine the sensitivity, specificity, accuracy, and receiver operating characteristic (ROC) curve with output area under the curve (AUC). ResultThe result showed that the sensitivity and specificity of Bahrudin and Siriraj Acore, as well as Gajah Mada Algorithm for determining infarct stroke was 91.3% vs 89.7% vs 61.2% and 67.7% vs 69.4% vs 77.4%, respectively. For determining hemorrhagic stroke the values were 67.7% vs 69.4% vs 77.4% and 91.3% vs 89.7% vs 61.2%, respectively. Furthermore, the area under the curve of Bahrudin and Siriraj Scores was better than Gajah Mada Algorithm. ConclusionIn conclusion, Bahrudin and Siriraj Scores have good accuracy, sensitivity, and specificity in diagnosing stroke than Gajah Mada Algorithm.

Estimation of ischemic core in acute ischemic stroke with CT angiography and non-contrast CT: Attenuation changes in ASPECTS regions vs. automated ASPECTS scoring.

PurposeReperfusion therapies for acute ischemic stroke due to large-vessel occlusion (AIS-LVO) are highly time-dependent, and large infarction is related to poor outcomes and risk of symptomatic hemorrhage. It is of significance to investigate and optimize the screening means and selection criteria for reperfusion therapies to identify more appropriate patients with better outcomes. This study aimed to compare the performance of attenuation changes vs. automated Alberta Stroke Program Early CT Score (ASPECTS) and using CT angiography (CTA) source images vs. non-contrast CT (NCCT) in distinguishing the infarction extent of ischemic core volumes ≥ 70 ml within different time windows.MethodsA total of 73 patients with AIS-LVO who received multimodal CT were analyzed. The automated software was used to calculate ASPECTS. Attenuation change was defined as the sum of products of relative Hounsfield unit (rHU) values times weighting factors of all 10 ASPECTS regions. rHU value of each region was the HU of the ischemic side over that of the contralateral. The corresponding weighting factors were the regression coefficients derived from a multivariable linear regression model which was used to correlate regional rHU with ischemic core volumes, because each region in the ASPECTS template is weighted disproportionally in the ASPECTS system. Automated ASPECTS and attenuation changes were both calculated using CTA and NCCT, respectively.ResultsAttenuation changes were correlated with ischemic core volumes within different time windows (Rho ranging from 0.439 to 0.637). In classification of the ischemic core ≥ 70 ml, the performances of attenuation changes were comparable with ASPECTS (area under the curve [AUC] ranging from 0.799 to 0.891), with DeLong’s test (P = 0.079, P = 0.373); using CTA (AUC = 0.842) was not different from NCCT (AUC = 0.838).ConclusionAttenuation changes in ASPECTS regions were correlated with ischemic core volumes. In the classification of infarction volumes, attenuation changes had a high diagnostic ability comparable with automated ASPECTS. Measurement of attenuation changes is not involved in complicated scoring algorithms. This measurement can be used as an available, rapid, reliable, and accurate means to evaluate infarction extent within different time windows. The usefulness of infarction volumes measured by attenuation changes to identify more appropriate patients for reperfusion therapies can be validated in future clinical trials.

Admission chest CT findings and risk assessment for stroke-associated pneumonia.

IntroductionStroke-associated pneumonia (SAP) is a significant cause of morbidity and mortality after stroke. Various factors, including dysphagia and stroke severity, are closely related to SAP risk; however, the contribution of the baseline pulmonary parenchymal status to this interplay is an understudied field. Herein, we evaluated the prognostic performance of admission chest computed tomography (CT) findings in predicting SAP.MethodsWe evaluated admission chest CT images, acquired as part of a COVID-19-related institutional policy, in a consecutive series of acute ischemic stroke patients. The pulmonary opacity load at baseline was quantified using automated volumetry and visual scoring algorithms. The relationship between pulmonary opacities with risk of pneumonia within 7 days of symptom onset (i.e., SAP) was evaluated by bivariate and multivariate analyses.ResultsTwenty-three percent of patients in our cohort (n = 100) were diagnosed with SAP. Patients with SAP were more likely to have atrial fibrillation, COPD, severe neurological deficits, and dysphagia. The visual opacity score on chest CT was significantly higher among patients who developed SAP (p = 0.014), while no such relationship was observed in terms of absolute or relative opacity volume. In multivariate analyses, admission stroke severity, presence of dysphagia and a visual opacity score of ≥ 3 (OR 6.37, 95% CI 1.61–25.16; p = 0.008) remained significantly associated with SAP risk.ConclusionsPulmonary opacity burden, as evaluated on admission chest CT, is significantly associated with development of pneumonia within initial days of stroke. This association is independent of other well-known predisposing factors for SAP, including age, stroke severity, and presence of dysphagia.Supplementary InformationThe online version contains supplementary material available at 10.1007/s13760-022-02043-7.

Associations of specific types of physical activities with 10-year risk of cardiovascular disease among adults: Data from the national health and nutrition examination survey 1999–2006

BackgroundPhysical activity plays a key role in the prevention of cardiovascular disease (CVD). However, previous studies focused predominantly on the associations of the total amount of physical activity with CVD. There were few evidences on the associations of specific sport disciplines with CVD. Furthermore, little was known on the interactions between the different types of sports on CVD risk. Therefore, this study aimed to examine the independent associations of specific types of physical activities with the 10-year risk of CVD, and further evaluate the interactions between specific types of physical activities on the 10-year risk of CVD in US adults.MethodsThis study used the data of the National Health and Nutrition Examination Survey (NHANES) 1999-2006. Participants aged ≥ 30 years and with free of CVD were eligible. The physical activity questionnaire is used to collect general information on leisure-time activities in the past 30 days, including the frequency, duration, and intensity of participation in each activity. The exposures of interest included cycling, swimming, aerobics, running, American Football, basketball, and racquet sports. The Framingham risk score algorithm was used to assess 10-year CVD risk based on age, high density lipoprotein, total cholesterol, systolic blood pressure, smoking status, and diabetes. A higher total score reflects a greater risk of CVD.ResultsThis study included 10829 participants. Compared to no participation, participation in cycling (β = −0.890, 95% CI:−1.278,−0.502, P < 0.001), running (β = −1.466, 95% CI:−1.837,−1.095, P < 0.001), American Football (β = −2.934, 95% CI:−3.750,−2.119, P < 0.001), basketball (β = −1.968, 95% CI:−2.645,−1.291, P < 0.001), and aerobics (β = −0.980, 95% CI:−1.352,−0.608, P < 0.001) was associated with a lower CVD risk. Furthermore, cycling was antagonistic with basketball and racquet sports in the associations with CVD risk. An antagonistic action between swimming and aerobics was also observed. Nevertheless, running was synergistic with cycling, aerobics, and racquet sports in the associations with CVD risk.ConclusionsThere were inverse associations of specific types of physical activities with CVD risk. Furthermore, there might be synergistic and antagonistic associations of multiple types of physical activities with CVD risk.

Lipid Fractions and Severity of Coronary Artery Lesions in Acute Myocardial Infarction: A Retrospective Single-Center Study

The relationship between different lipid fractions and the occurrence of atherosclerotic coronary artery disease(CAD) is a well-known fact. However, knowledge of the association of different lipid fractions and the severity and extension of coronary lesions has been rarely reported in clinical studies and is often contradictory. This study aimed to evaluate the association between different lipid fractions and the severity and extent of the coronary atherosclerotic lesion in patients with ST elevated acute myocardial infarction (STEMI). We analyzed data from 50 consecutive patients with STEMI who underwent coronary angiography during hospitalization at the Clinical Hospital in Tetovo. The severity and extent of CAD were defined using the Syntax score (SXscore) algorithm stratified according to tertiles. Laboratory analysis for lipid fractions was performed. It was noticed a statistically significant association between SXscore tertiles and the mean of total cholesterol (TC) to high-density lipoprotein cholesterol (HDL) ratio (TC/HDL), (p=0.008). Correlation test showed week but significant positive correlation between SXscore tertiles and TC/HDL ratio (r=0.339, p=0.016) and Non-HDL (r=0.309, p=0.034). In regression analysis, reduced HDL and increased TC values statistically significantly predicted SXscore (p&lt;0.05). Among all lipid variables examined, the TC/HDL ratio appeared as the most powerful indicator of severity and extension of the coronary lesion in patients with STEMI. The TC/HDL ratio should be used as an easy, non-invasive, and inexpensive method to measure the severity of CAD to determine the diagnosis strategy and treatment of patients with atherosclerotic coronary diseases at risk of acute infarction.

Statement of the Spanish Interdisciplinary Vascular Prevention Committee on the updated European Guidelines on Cardiovascular Disease Prevention

We present the Spanish adaptation of the 2021 European Guidelines on Cardiovascular Disease (CVD) prevention in clinical practice. The current guidelines besides the individual approach greatly emphasize on the importance of population level approaches to the prevention of cardiovascular diseases. Systematic global CVD risk assessment is recommended in individuals with any major vascular risk factor. Regarding LDL-Cholesterol, blood pressure, and glycemic control in patients with diabetes mellitus, goals and targets remain as recommended in previous guidelines. However, it is proposed a new, stepwise approach (Step 1 and 2) to treatment intensification as a tool to help physicians and patients pursue these targets in a way that fits patient profile. After Step 1, considering proceeding to the intensified goals of Step 2 is mandatory, and this intensification will be based on 10-year CVD risk, lifetime CVD risk and treatment benefit, comorbidities and patient preferences.The updated SCORE algorithm—SCORE2, SCORE-OP— is recommended in these guidelines, which estimates an individual’s 10-year risk of fatal and non-fatal CVD events (myocardial infarction, stroke) in healthy men and women aged 40–89 years. Another new and important recommendation is the use of different categories of risk according different age groups (<50, 50−69, ≥70 years).Different flow charts of CVD risk and risk factor treatment in apparently healthy persons, in patients with established atherosclerotic CVD, and in diabetic patients are recommended. Patients with chronic kidney disease are considered high risk or very high-risk patients according to the levels of glomerular filtration rate and albumin-to-creatinine ratio. New lifestyle recommendations adapted to the ones published by the Spanish Ministry of Health as well as recommendations focused on the management of lipids, blood pressure, diabetes and chronic renal failure are included.

408 Evaluation Of A Deep Learning Based Fully Automated Coronary Calcium Scoring Algorithm, CardIQ Suite Software

Introduction: CardIQ Suite is a new generation of cardiac CT imaging post-processing software. Automatic calcium scoring is enabled by the present design, although manual scoring and editing capabilities remain accessible. The purpose of this study is to compare an fully automated calcium scoring algorithm to expert readers using the SmartScore predicate device. Primary objectives, 1) Compare calcium scores obtained by manual and automated algorithm in the assigned CAC-DRS category. 2) Evaluate the efficiency of the automated scoring approach in comparison to the manual scoring method.

Comentario del CEIPV a las nuevas Guías Europeas de Prevención Cardiovascular 2021

We present the Spanish adaptation of the 2021 European Guidelines on Cardiovascular Disease Prevention in Clinical Practice. The current guidelines besides the individual approach greatly emphasize on the importance of population level approaches to the prevention of cardiovascular diseases. Systematic global cardiovascular disease risk assessment is recommended in individuals with any major vascular risk factor. Regarding LDL-cholesterol, blood pressure, and glycemic control in patients with diabetes mellitus, goals and targets remain as recommended in previous guidelines. However, it is proposed a new, stepwise approach (steps 1 and 2) to treatment intensification as a tool to help physicians and patients pursue these targets in a way that fits patient profile. After step 1, considering proceeding to the intensified goals of step 2 is mandatory, and this intensification will be based on 10-year cardiovascular disease risk, lifetime cardiovascular disease risk and treatment benefit, comorbidities and patient preferences.The updated SCORE algorithm ?SCORE2, SCORE2-OP? is recommended in these guidelines, which estimates an individual's 10-year risk of fatal and non-fatal cardiovascular disease events (myocardial infarction, stroke) in healthy men and women aged 40-89 years. Another new and important recommendation is the use of different categories of risk according to different age groups (<50, 50-69, ≥70 years).Different flow charts of cardiovascular disease risk and risk factor treatment in apparently healthy persons, in patients with established atherosclerotic cardiovascular disease, and in diabetic patients are recommended. Patients with chronic kidney disease are considered high risk or very high-risk patients according to the levels of glomerular filtration rate and albumin-to-creatinine ratio. New lifestyle recommendations adapted to the ones published by the Spanish Ministry of Health as well as recommendations focused on the management of lipids, blood pressure, diabetes and chronic renal failure are included.

Scaled Vecchia Approximation for Fast Computer-Model Emulation

Many scientific phenomena are studied using computer experiments consisting of multiple runs of a computer model while varying the input settings. Gaussian processes (GPs) are a popular tool for the analysis of computer experiments, enabling interpolation between input settings, but direct GP inference is computationally infeasible for large datasets. We adapt and extend a powerful class of GP methods from spatial statistics to enable the scalable analysis and emulation of large computer experiments. Specifically, we apply Vecchia’s ordered conditional approximation in a transformed input space, with each input scaled according to how strongly it relates to the computer-model response. The scaling is learned from the data by estimating parameters in the GP covariance function using Fisher scoring. Our methods are highly scalable, enabling estimation, joint prediction, and simulation in near-linear time in the number of model runs. In several numerical examples, our approach substantially outperformed existing methods.

Performance of a multisensor smart ring to evaluate sleep: in-lab and home-based evaluation of generalized and personalized algorithms.

Wearable sleep technology has rapidly expanded across the consumer market due to advances in technology and increased interest in personalized sleep assessment to improve health and mental performance. We tested the performance of a novel device, the Happy Ring, alongside other commercial wearables (Actiwatch 2, Fitbit Charge 4, Whoop 3.0, Oura Ring V2), against in-lab polysomnography (PSG) and at-home electroencephalography (EEG)-derived sleep monitoring device, the Dreem 2 Headband. Thirty-six healthy adults with no diagnosed sleep disorders and no recent use of medications or substances known to affect sleep patterns were assessed across 77 nights. Subjects participated in a single night of in-lab PSG and two nights of at-home data collection. The Happy Ring includes sensors for skin conductance, movement, heart rate, and skin temperature. The Happy Ring utilized two machine-learning derived scoring algorithms: a "generalized" algorithm that applied broadly to all users, and a "personalized" algorithm that adapted to individual subjects' data. Epoch-by-epoch analyses compared the wearable devices to in-lab PSG and to at-home EEG Headband. Compared to in-lab PSG, the "generalized" and "personalized" algorithms demonstrated good sensitivity (94% and 93%, respectively) and specificity (70% and 83%, respectively). The Happy Personalized model demonstrated a lower bias and more narrow limits of agreement across Bland-Altman measures. The Happy Ring performed well at home and in the lab, especially regarding sleep/wake detection. The personalized algorithm demonstrated improved detection accuracy over the generalized approach and other devices, suggesting that adaptable, dynamic algorithms can enhance sleep detection accuracy.

The impact of pediatric early warning score and rapid response algorithm training and implementation on interprofessional collaboration in a resource-limited setting.

Improved teamwork and communication have been associated with improved quality of care. Early Warning Scores (EWS) and rapid response algorithms are a way of identifying deteriorating patients and providing a common framework for communication and response between physicians and nurses. The impact of EWS implementation on interprofessional collaboration (IPC) has been minimally studied, especially in resource-limited settings. The study took place in the Pediatric Department of the main academic referral hospital in Rwanda between April 2019 and January 2020. Pediatric nurses and residents were trained on the use of the Pediatric Warning Score for Resource-Limited Settings (PEWS-RL) and a rapid response algorithm. Training included vital sign collection, PEWS-RL calculation, IPC and rapid response algorithm implementation. Prior to training, participants completed surveys on IPC with Likert scale responses (from "strongly disagree" to "strongly agree"). Follow-up surveys were then administered nine months later and also included an open-response question on the impact of the PEWS-RL implementation on IPC. Sixty-five (96%) nurses were trained and completed the pre-survey and thirty-seven (54%) of the trained nurses completed the post-survey. Twenty-two (59%) pediatric residents were trained in the workshop and completed the pre-survey and twenty-four physicians (4 pediatricians (40%) and 20 pediatric residents (53%)) completed the post-implementation survey. There was a statistically significant increase in the percent of nurses indicating strong agreement across all domains of communication and collaboration from the pre- to the post-survey. Although the percent of physicians indicating strong agreement increased in the post-survey for all items, only the "share information" item was statistically significant. Training and implementation of a PEWS-RL and a rapid response algorithm at a tertiary hospital in Rwanda resulted in significant improvement of nurse and physician ratings of IPC nine months later.

Abstract 1272: Confirmatory study of the IO Score, a tumor immune microenvironment (TIME) classifier, demonstrates efficacy in a real-world cohort of metastatic urothelial cancer (mUC) patients treated with immune checkpoint inhibitors (ICIs)

Abstract Background: Prognosis for patients with mUC is poor with fewer than 10% surviving two years after diagnosis. As the majority of mUC patients do not survive to a second line of therapy, optimization of first line therapy is critical, especially with recently approved non-ICI therapeutics. The IO Score, a 27-gene TIME classifier, has previously demonstrated a strong association with response to ICI therapy in mUC (IMvigor210; n= 348), NSCLC, renal cancer and a randomized TNBC trial (NeoTRIPaPDL1; n= 258). In a real-world study of ICI therapy in mUC pts (UNC-108), clinical benefit was seen in less than 30% of patients (Rose, 2021). The IO Score was evaluated for efficacy to ICI in this cohort to confirm findings from IMvigor210 where the IO Score was both independent and complementary to TMB. Methods: The previously defined IO score algorithm was applied to available RNAseq data from 77 patients in the UNC-108 study. Using a six-week front censor for OS, Cox proportional hazards for OS, and PFS, the IO Score, TMB, and TMB and/or IO Score positivity were calculated. Responses (CR/PR/SD for &amp;gt;6 months) were used to calculate both the ORR and OR. PD-L1 staining was unavailable. Results: Conclusions: These data confirm the association of the IO Score with response to ICI treatment in a second, independent, mUC cohort using the same algorithm and threshold previously established and validated in other tumor types. The IO Score provided independent biomarker value and identified more responders than TMB alone. The IO Score provides unique information on both the inflammatory and cancer associated fibroblast biologic signatures of response to ICI’s and can be complemented by other ICI associated biomarkers. Thus, this tool can inform therapeutic decisions and has the potential to improve outcome of mUC patients. Measure IO Score Positive TMB (10mut/MB) IO Score and/or TMB Biomarker Positive 28 (36.4%) 19 (24.7%) 35 (45.5%) OS HR = 0.39 95%CI 0.21-0.75; p=0.004 HR = 0.32 95%CI 0.14-0.72; p=0.006 HR = 0.34 95%CI 0.19-0.67; p&amp;lt;0.001 PFS HR = 0.45 95%CI 0.26-0.77; p=0.004 HR = 0.30 95%CI 0.16-0.59; p&amp;lt;0.001 HR = 0.36 95%CI 0.21-0.62; p&amp;lt;0.001 Objective response Rate ORR = 50% (n=14) ORR = 58% (n=11) ORR = 49% (n=17) Odds Ratio 7.17 95%CI 2.3-22.2; p&amp;lt;0.001 7.49 95%CI 2.4-23.8; p&amp;lt;0.001 12.3 95%CI 3.19-47.3; p&amp;lt;0.001 Citation Format: Tyler Nielsen, Robert Seitz, Matthew Varga, Catherine Cronister, David Hout, Brock Schweitzer, Douglas Ross, Kim McGregor. Confirmatory study of the IO Score, a tumor immune microenvironment (TIME) classifier, demonstrates efficacy in a real-world cohort of metastatic urothelial cancer (mUC) patients treated with immune checkpoint inhibitors (ICIs) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1272.

Accurate and generalizable quantitative scoring of liver steatosis from ultrasound images via scalable deep learning.

BACKGROUNDHepatic steatosis is a major cause of chronic liver disease. Two-dimensional (2D) ultrasound is the most widely used non-invasive tool for screening and monitoring, but associated diagnoses are highly subjective.AIMTo develop a scalable deep learning (DL) algorithm for quantitative scoring of liver steatosis from 2D ultrasound images.METHODSUsing multi-view ultrasound data from 3310 patients, 19513 studies, and 228075 images from a retrospective cohort of patients received elastography, we trained a DL algorithm to diagnose steatosis stages (healthy, mild, moderate, or severe) from clinical ultrasound diagnoses. Performance was validated on two multi-scanner unblinded and blinded (initially to DL developer) histology-proven cohorts (147 and 112 patients) with histopathology fatty cell percentage diagnoses and a subset with FibroScan diagnoses. We also quantified reliability across scanners and viewpoints. Results were evaluated using Bland-Altman and receiver operating characteristic (ROC) analysis.RESULTSThe DL algorithm demonstrated repeatable measurements with a moderate number of images (three for each viewpoint) and high agreement across three premium ultrasound scanners. High diagnostic performance was observed across all viewpoints: Areas under the curve of the ROC to classify mild, moderate, and severe steatosis grades were 0.85, 0.91, and 0.93, respectively. The DL algorithm outperformed or performed at least comparably to FibroScan control attenuation parameter (CAP) with statistically significant improvements for all levels on the unblinded histology-proven cohort and for “= severe” steatosis on the blinded histology-proven cohort.CONCLUSIONThe DL algorithm provides a reliable quantitative steatosis assessment across view and scanners on two multi-scanner cohorts. Diagnostic performance was high with comparable or better performance than the CAP.