Higher Area Under The Curve Research Articles

Performance of Deauville Criteria in [18F]FDG-PET/CT Diagnostics of Giant Cell Arteritis.

In this retrospective study, PET/CT data from 59 patients with suspected giant cell arteritis (GCA) were reviewed using the Deauville criteria to determine an optimal cut-off between PET positivity and negativity. Seventeen standardised vascular regions were analysed per patient by three investigators blinded to clinical information. Statistical analysis included ROC curves with areas under the curve (AUC), Cohen's and Fleiss' kappa (κ) to calculate sensitivity, specificity, accuracy, and agreement. According to final clinician's diagnosis and the revised 2017 ACR criteria GCA was confirmed in 29 of 59 (49.2 %) patients. With a diagnostic cut-off ≥ 4 (highest tracer uptake of a vessel wall exceeds liver uptake) for PET positivity, all investigators achieved high accuracy (range, 89.8-93.2%) and AUC (range, 0.94-0.97). Sensitivity and specificity ranged from 89.7-96.6% and 83.3-96.7%, respectively. Agreement between the three investigators suggested 'almost perfect agreement' (Fleiss' κ = 0.84) A Deauville score of ≥4 as threshold for PET positivity yielded excellent results with high accuracy and almost perfect inter-rater agreement, suggesting a standardized, reproducible, and reliable score in diagnosing GCA. However, the small sample size and reference standard could lead to biases. Therefore, verification in a multicentre study with a larger patient cohort and prospective setting is needed.

The Influence of Systolic Blood Pressure at the Time of Extubation on the Development of Postoperative Spinal Epidural Hematoma.

The most common cause of neurological complications after a biportal endoscopic spine surgery (BESS) is postoperative spinal epidural hematomas (POSEH). The objective of this study was to determine the influence of systolic blood pressure at extubation (e-SBP) on POSEH. A total of 352 patients who underwent single-level decompression surgery including laminectomy and/or discectomy with BESS under the diagnosis of spinal stenosis and herniated nucleus pulposus between August 1, 2018, and June 30, 2021, were reviewed retrospectively. The patients were divided into two, a POSEH group and a normal group without POSEH (no neurological complication). The e-SBP, demographic factors, and the preoperative and intraoperative factors suspected to influence the POSEH were analyzed. The e-SBP was converted to a categorical variable by the threshold level that was decided by maximum area under the curve (AUC) in receiver operating characteristic (ROC) curve analysis. Antiplatelet drugs (APDs) were taken in 21 patients (6.0%), discontinued in 24 patients (6.8%), and not taken in 307 patients (87.2%). Tranexamic acid (TXA) was used in 292 patients (83.0%) in the perioperative period. Of the 352 patients, 18 patients (5.1%) underwent revision surgery for the removal of POSEH. The POSEH and normal groups were homogenous in age, sex, diagnosis, operation segments, operation time, and lab findings that were related to blood clotting, whereas there were differences in e-SBP (163.7 ± 15.7 mmHg in POSEH group and 154.1 ± 18.3 mmHg in normal group), APD (4 takers, 2 stoppers, 12 non-takers in POSEH group and 16 takers, 22 stoppers, 296 non-takers in normal group), and TXA (12 use, 6 not use in POSEH group and 280 use, 54 not use in normal group) in single variable analysis. The highest AUC in the ROC curve analysis was 0.652 for 170 mmHg e-SBP (p < 0.05). There were 94 patients in the high e-SBP group (≥ 170 mmHg) and 258 patients in the low e-SBP group. In multivariable logistic regression analysis, only high e-SBP was a significant risk factor for POSEH (p = 0.013; odds ratio, 3.434). High e-SBP (≥ 170 mmHg) can influence the development of POSEH in biportal endoscopic spine surgery.

The role of antral histopathology in diagnosing paediatric celiac disease.

The aim of this study was to evaluate the diagnostic potential of gastric antrum histology in children suspected of having celiac disease (CD). The present retrospective study was performed on 224 patients who were suspected of having CD and had several duodenal and one gastric antrum biopsies. They were divided into 2 groups based on the definite diagnosis of CD. The statistical analysis was performed using SPSS version 22 software. Receiver operating characteristic (ROC) curves were drown and the area under the curves (AUCs) was calculated. Article authorized by the Editor-in-Chief. Based on MARSH criteria, out of 224 patients, 124 were diagnosed as definite CD and 100 patients comprised the non-celiac group. The AUC for the mean of all pathological tests was estimated to be 0.90 (p < 0.001). The pooled AUC for the combination of 3 pathologic findings with the highest AUCs (cell, crypt, and gland size) was estimated to be 0.89 (p < 0.001). We observed that the histological changes we found in the gastric antrum were identical to those found in the duodenum of paediatric CD patients. Because providing abiopsy from the gastric antrum is easier than getting multiple biopsies from the duodenum, we suggest using the criteria mentioned in this study in other studies with larger sample sizes.

Android IoT Lifelog System and Its Application to Motion Inference

In social science, health care, digital therapeutics, etc., smartphone data have played important roles to infer users’ daily lives. However, smartphone data collection systems could not be used effectively and widely because they did not exploit any Internet of Things (IoT) standards (e.g., oneM2M) and class labeling methods for machine learning (ML) services. Therefore, in this paper, we propose a novel Android IoT lifelog system complying with oneM2M standards to collect various lifelog data in smartphones and provide two manual and automated class labeling methods for inference of users’ daily lives. The proposed system consists of an Android IoT client application, an oneM2M-compliant IoT server, and an ML server whose high-level functional architecture was carefully designed to be open, accessible, and internationally recognized in accordance with the oneM2M standards. In particular, we explain implementation details of activity diagrams for the Android IoT client application, the primary component of the proposed system. Experimental results verified that this application could work with the oneM2M-compliant IoT server normally and provide corresponding class labels properly. As an application of the proposed system, we also propose motion inference based on three multi-class ML classifiers (i.e., k nearest neighbors, Naive Bayes, and support vector machine) which were created by using only motion and location data (i.e., acceleration force, gyroscope rate of rotation, and speed) and motion class labels (i.e., driving, cycling, running, walking, and stilling). When compared with confusion matrices of the ML classifiers, the k nearest neighbors classifier outperformed the other two overall. Furthermore, we evaluated its output quality by analyzing the receiver operating characteristic (ROC) curves with area under the curve (AUC) values. The AUC values of the ROC curves for all motion classes were more than 0.9, and the macro-average and micro-average ROC curves achieved very high AUC values of 0.96 and 0.99, respectively.

Impact of Emergency Department Crowding on Discharged Patient Experience.

While emergency department (ED) crowding has deleterious effects on patient care outcomes and operational efficiency, impacts on the experience for patients discharged from the ED are unknown. We aimed to study how patient-reported experience is affected by ED crowding to characterize which factors most impact discharged patient experience. This institutional review board-exempt, retrospective, cohort study included all discharged adult ED patients July 1, 2020-June 30, 2021 with at least some response data to the the National Research Corporation Health survey, sent to most patients discharged from our large, academic medical center ED. Our query yielded 9,401 unique encounters for 9,221 patients. Based on responses to the summary question of whether the patient was likely to recommend our ED, patients were categorized as "detractors" (scores 0-6) or "non-detractors" (scores 7-10). We assessed the relationship between census and patient experience by 1) computing percentage of detractors within each care area and assessing for differences in census and boarder burden between detractors and non-detractors, and 2) multivariable logistic regression assessing the relationship between likelihood of being a detractor in terms of the ED census and the patient's last ED care area. A second logistic regression controlled for additional patient- and encounter-specific covariates. Survey response rate was 24.8%. Overall, 13.9% of responders were detractors. There was a significant difference in the average overall ED census for detractors (average 3.70 more patients physically present at the time of arrival, 95% CI 2.33-5.07). In unadjusted multivariable analyses, three lower acuity ED care areas showed statistically significant differences of detractor likelihood with changes in patient census. The overall area under the curve (AUC) for the unadjusted model was 0.594 (CI 0.577-0.610). The adjusted model had higher AUC (0.673, CI 0.657-0.690]; P<0.001), with the same three care areas having significant differences in detractor likelihood based on patient census changes. Length of stay (OR 1.71, CI 1.50-1.95), leaving against medical advice/without being seen (OR 5.15, CI 3.84-6.89), and the number of ED care areas a patient visited (OR 1.16, CI 1.01-1.33) was associated with an increase in detractor likelihood. Patients arriving to a crowded ED and ultimately discharged are more likely to have negative patient experience. Future studies should characterize which variables most impact patient experience of discharged ED patients.

Association of local solid mechanical, hemodynamic and morphological characteristics with ruptured intracranial aneurysm.

The rupture of intracranial aneurysms (IAs) is a complicated phenomenon of which the mechanism is not fully understood. The purpose of this study is to associate local solid mechanical, hemodynamic, and morphological characteristics with rupture regions through statistical means, in an attempt to identify the parameters that are indicative of rupture propensity for IAs. Twenty patient-specific ruptured IA models were reconstructed from digital subtraction angiography (DSA), and applied in the analysis of wall tension, wall shear stress (WSS) and curvature. The precise rupture locations were marked out through intraoperative videos. Pearson correlation analysis was employed to investigate the correlations of these three parameters with patient characteristics and global geometric features. Univariate and multivariate logistic regression analysis were further performed on wall tension, WSS and curvature with regards to rupture and nonrupture regions. Receiver operating characteristic (ROC) analysis defining area under the curve (AUC) was performed on these three parameters. The univariate model of wall tension (AUC, 0.9750), WSS (AUC, 0.9300), curvature (0.8150) and their combined multivariate model (AUC, 0.9875) all present high AUC values. The wall tension, WSS and curvature are acceptable parameters relating to rupture regions. The rupture odd is more sensitive to the wall tension and WSS than curvature. Each logistic model is capable in discriminating ruptures from nonrupture regions, while the multivariate model is the most efficient.

Utility of the Total Thrombus-Formation Analysis System as a Tool for Evaluating Thrombogenicity and Monitoring Antithrombotic Therapy in Pediatric Fontan Patients.

There is no consensus regarding thromboprophylaxis after Fontan procedure, and novel tools to assess thrombogenicity are needed to establish optimal thromboprophylaxis. The Total Thrombus-formation Analysis System (T-TAS) was developed for the quantitative analysis of thrombus formation using microchips with thrombogenic surfaces. This prospective study evaluated the utility of T-TAS in the assessment of thrombogenicity in pediatric Fontan patients. The participants included 20 consecutive Fontan patients who underwent cardiac catheterization and 30 healthy controls. Blood samples collected without and with antithrombotic therapy (aspirin or aspirin and warfarin) were used for T-TAS to compute the area under the curve (AUC) in the atheroma (AR10-AUC30) and platelet (PL18-AUC10) chips. A higher AUC indicates higher thrombogenicity. T-TAS values showed that patients in the Fontan group without antithrombotic therapy had lower thrombogenicity than those in the control group [PL18-AUC10, median (interquartile range) 356 (313-394) vs. 408 (392-424); AR10-AUC30, median (interquartile range) 1270 (1178-1351) vs. 1382 (1338-1421)]. Aspirin and warfarin therapies significantly decreased PL18-AUC10 and AR10-AUC30, respectively, compared with those of patients without antithrombotic therapy (P < 0.001 for each comparison). Subgroup analysis divided by low (< 9mmHg) or high (≥ 9mmHg) central venous pressure (CVP) showed that CVP affects the reduction in AR10-AUC30 with antithrombotic therapy. T-TAS may be a useful tool for monitoring thrombogenicity and antithrombotic therapy in Fontan patients.

SCREENING OF US VETERANS AT HIGH RISK FOR OUTCOMES FOLLOWING COVID-19 BY FRAILTY STATUS

Abstract Frailty increases risk of adverse outcomes in the presence of stressors such as COVID-19 infection. We examined the association between different levels of frailty and outcomes following COVID-19 infection. This is a retrospective cohort study of US Veterans aged ≥50 years, active Veterans Health Administration (VHA) care users, and tested positive for COVID-19 between 02/15/2020 and 09/30/2021. VHA frailty index (VA-FI) was calculated from one year prior to the COVID-19 testing and divided into three groups: robust (≤0.1), prefrail (0.1-0.2), and frail (&amp;gt;0.2). The risk of hospitalization, ICU admission, ventilator use, and in-hospital mortality were calculated using logistic regression adjusted by age, sex, body-mass-index, and race. The performance of VA-FI in predicting outcomes was compared to age or Charlson comorbidity index (CCI) using area under the curve (AUC). Of 204,426 COVID-19 positive Veterans, 32,965 were hospitalized (age: 71.4±10.4 years, BMI: 29.5±7.1 kg/m2). We observed higher odds of hospitalization (frail, adjusted odds ratio (aOR)=8.64; prefrail, aOR=2.57), ICU admission (frail, aOR=1.58; prefrail, aOR=1.32), ventilator use (frail, aOR=1.97; prefrail, aOR=1.57), and mortality (frail aOR=2.15; prefrail, aOR=1.55) in frail and prefrail compared to robust Veterans. We observed that VA-FI had higher AUC in predicting hospitalization (AUC 0.75) independent of age (0.59) and CCI (0.63). Veterans with COVID-19 who were frail and prefrail had a higher risk of hospitalization, ICU admission, ventilator use, mortality compared to robust. VA-FI may be a useful tool at the time of COVID-19 diagnosis to triage patients at risk for adverse outcomes.

Development of a Diagnostic Model Focusing on Esophageal Dysmotility in Patients with Systemic Sclerosis.

Esophageal dysmotility is a common and neglected complication of systemic sclerosis (SSc) associated with poor prognosis, while the assessment remains a challenge. We aimed to develop a diagnostic model for esophageal dysmotility in SSc patients that provides individualized risk estimates. Seventy-five SSc patients who underwent high-resolution manometry (HRM) were included in the study. Esophageal widest diameter (WED) was measured on a chest CT scan. Esophageal parameters between patients with and without esophageal dysmotility were compared. Multivariate logistic regression analysis and Least Absolute Shrinkage and Selection Operator (LASSO) regression were used to fit the model. The diagnostic model was evaluated by discrimination and calibration. Internal validation was estimated using the enhanced bootstrap method with 1000 repetitions. Sixty-one systemic sclerosis patients (81.3%) were diagnosed with esophageal dysmotility according to the Chicago Classification v 3.0. The diagnostic model for evaluating the probability of esophageal dysmotility integrated clinical and imaging features, including disease duration, ILD, and WED. The model displayed good discrimination with an area under the curve (AUC) of 0.923 (95% CI: 0.837-1.000), a Brier score of 0.083, and good calibration. A high AUC value of 0.911 could still be achieved in the internal validation. The diagnostic model, which combines the disease duration, ILD, and imaging feature (WED), is an effective and noninvasive method for predicting esophageal dysmotility in SSc patients.

Association between initial presenting level of consciousness and patient acuity – A potential application for secondary triage in emergency ambulance calls

IntroductionConscious state assessment is important for the triage of emergency patients. In this study, we measured the association between ambulance patients’ conscious state and high versus lower acuity, with a view to informing telephone triage assessment of conscious state. MethodsData were analysed from one year of emergency ambulance incidents in Perth, Western Australia. Patient conscious state at the time of paramedic arrival was compared to acuity (based on paramedic assessment and management). We determined the proportion of high-acuity patients across six levels of consciousness (Alert, Confused, Drowsy, Voice Response, Pain Response, Unresponsive) overall, and within individual protocols of the Medical Priority Dispatch System (MPDS). ResultsThe proportion of high acuity patients increased with each step across the consciousness scale. Applying conscious state as a binary predictor of acuity, the largest increases occurred moving the threshold from Alert to Confused (22.0–48.6% high acuity) and Drowsy to Voice Response (61.9–89.5% high acuity). The Area Under the Curve (AUC) of the Receiver Operating Characteristic was 0.65. Within individual protocols, the highest AUC was in Cardiac Arrest (0.89), Overdose/Poisoning (0.81), Unknown Problem (0.76), Diabetic Problem, (0.74) and Convulsions/Fitting (0.73); and lowest in Heart problems (0.55), Abdominal Pain (0.55), Breathing Problems (0.55), Back Pain (0.53), and Chest Pain (0.52). ConclusionBased on these proportions of high acuity patients, it is reasonable to consider patients with any altered conscious state a high priority. The value of conscious state assessment for predicting acuity varies markedly between MPDS protocols. These findings could help inform secondary triage of ambulance patients during the emergency call.

Optimal method for detecting cervical lymph node metastasis from papillary thyroid cancer.

Papillary thyroid cancer (PTC) is often accompanied by cervical lymph node metastases (LNM). This study compared the diagnostic performance of fine needle aspiration (FNA), the combination of FNA and thyroglobulin measurement in the washout (FNA-Tg), and core needle biopsy(CNB) in order to determine the optimal method in detecting cervical LNM from PTC. PTC patients with suspicious cervical lymph nodes, who underwent ultrasonography-guided FNA, or CNB before surgery were retrospectively reviewed between January 2021 and April 2022. Patients' demographics, lymph node size, and results of FNA, FNA-Tg, CNB and surgical pathology were collected. Sensitivitity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy and the area under the curves (AUC) of receiver operating characteristic were obtained. FNA-Tg had a higher AUC with a threshold of 55 ng/mL than that with a threshold of 1 ng/mL (0.782 vs. 0.636, P = 0.005). The sensitivity of the combination of FNA and FNA-Tg was higher than that of FNA alone (91.3% vs. 81.0%, P = 0.01), but it was not different from that of CNB (91.3% vs. 88.0%, P > 0.05). No significant difference was found in specificity, PPV, NPV, accuracy and AUC among FNA alone, CNB alone, and the combination of FNA and FNA-Tg. FNA-Tg is useful to improve the sensitivity of FNA. The diagnostic performance of the combination of FNA and FNA-Tg is better than that of FNA or FNA-Tg alone, but is not different from that of CNB in detecting LNM from PTC.

Use of machine learning to diagnose somatic symptom disorder: Are the biomarkers beneficial for the diagnosis?

Objectives We used machine learning to incorporate three types of biomarkers (respiratory sinus arrhythmia, RSA; skin conductance, SC; finger temperature, FT) for examining the performance of diagnosing somatic symptom disorder (SSD). Methods We recruited 97 SSD subjects and 96 controls without psychiatric history or somatic distress. The values of RSA, SC and FT were recorded in three situations (resting state, under a cognitive task and under paced breathing) and compared for the two populations. We used machine learning to combine the biological signals and then applied receiver operating characteristic curve analysis to examine the performance of diagnosing SSD regarding the distinct indicators and situations. Subgroup analysis for subjects without depression/anxiety was also conducted. Results FT was significantly different between SSD patients and controls, especially in the resting state and under paced breathing. However, the biomarkers (0.75–0.76) did not reveal an area under the curve (AUC) comparable with the psychological questionnaires (0.86). Combining the biological and psychological indicators gave a high AUC (0.86–0.92). When excluding individuals with depression/anxiety, combining three biomarkers (0.79–0.83) and adopting psychological questionnaires (0.78) revealed a similar AUC. Conclusions The performance of RSA/SC/FT was unsatisfactory for diagnosing SSD but became comparable when excluding comorbid depression/anxiety.

Diagnostic serum biomarkers associated with ankylosing spondylitis.

Ankylosing spondylitis (AS) is an autoimmune rheumatic disease that mostly affects the axial skeleton. This study aimed to investigate reliable diagnostic serum biomarkers for AS. Serum samples were collected from 20 AS patients and 20 healthy controls (HCs) and analyzed using ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). The differential metabolites between the AS patients and HCs were profiled using univariate and multivariate statistical analyses. Pathway analysis and a heat map were also conducted. Random forest (RF) analysis and the least absolute shrinkage and selection operator (LASSO) were used to establish predictive and diagnostic models. After controlling the variable importance in the projection (VIP) value > 1 and false discovery rate (FDR) < 0.05, a total of 61 differential metabolites were identified from 995 metabolites, which exhibited significant differences in the pathway analysis and heat map between the AS patients and HCs. RF as a predictive model also identified differential metabolites with 95% predictive accuracy and a high area under the curve (AUC) of 1. A diagnostic model comprising nine metabolites (cysteinylglycine disulfide, choline, N6, N6, N6-trimethyllysine, histidine, sphingosine, fibrinopeptide A, glycerol 3-phosphate, 1-linoleoyl-GPA (18:2), and fibrinopeptide A (3-16)) was generated using LASSO regression, capable of distinguishing HCs from AS with a high AUC of 1. Our results indicated that the UPLC-MS/MS analysis method is a powerful tool for identifying AS metabolite profiles. We developed a nine-metabolites-based model serving as a diagnostic tool to separate AS patients from HCs, and the identified diagnostic biomarkers appeared to have a diagnostic value for AS.

Validation of a fully automated lab developed test for plasma phospho‐tau 181 levels for Alzheimer’s disease diagnosis

AbstractBackgroundThe potential usefulness of plasma phosphorylated tau (pTau) measurements for identifying individuals likely to be on the Alzheimer’s disease (AD) continuum is evident from recent literature. Here we report the validation of an immunoassay for measurement of pTau181 in human plasma that will aid in the evaluation of Alzheimer’s pathology in cases presenting with mild cognitive impairment.MethodThe Simoa pTau‐181 assay is a fully automated 2‐step digital sandwich immunoassay that utilizes paramagnetic beads with capture antibodies and biotinylated detector antibodies. Digital detection is achieved using Simoa discs via SβG (Streptavidin‐ß‐galactosidase) and RGP (Resorufin‐β‐D‐galactopyranoside) amplification of specific pTau181 immune complexes. Harmonization to the Simoa pTau‐181 method described by Karikari et al (Mol Psychiatry. 2021;26(2):429‐442) using 80 samples, establishes commutability with published results from ADNI cohorts. Analytical validation on HD‐X instrument followed CLSI guidelines. Thirty‐four AD cases from the Amsterdam Dementia Cohort (age 66 ± 8 years) diagnosed according to the NIA‐AA diagnostic AD criteria and had a CSF biomarker‐confirmed AD diagnosis, 36 age‐ and sex‐matched (age 66 ± 8 years) cognitively healthy controls from the Dutch Brain Research Registry were included for clinical validation of this LDT.ResultThe limits of detection (LoD) and quantification (LoQ) were 0.056 pg/mL and 0.276, respectively. Repeatability and reproducibility across 3 reagent lots using samples across the measuring range were 4.25% (95%CI:2.84‐5.65), 9.42% (95% CI: 4.46‐14.37) respectively. Linearity was demonstrated per CLSI EP06‐A, as ≤15% deviation at all dilutions. Cross reactivity against full length unphosphorylated‐Tau441 and synthetic peptides for pTau231 and pTau217 epitopes was below LoQ. ROC analysis resulted in a high AUC (area under the curve) of 0.897 (95% CI 0.825 to 0.969). Applying the cut‐offs described in Karikari1 (17.7 pg/mL) for this cohort resulted in a sensitivity of 94%, specificity of 72%.ConclusionThis test demonstrated acceptable analytical performance for CLIA laboratory implementation. High AUC of 0.897 to discriminate AD from non‐AD cases helps both rule‐in or rule‐out scenarios, and further allows for setting of cut‐off values that not only take the pretest probability, impact on screening algorithm, treatment protocols into consideration but also the costs incurred by misdiagnosis.

Pseudocholinesterase as a Biomarker for Untreated Wilson's Disease.

The aim of this study was to demonstrate that pseudocholinesterase (CHE) serum level is a useful diagnostic biomarker for untreated Wilson's disease (WD). Between 2013 and 2019, about 75 patients were referred to the outpatient department of the University of Düsseldorf with suspected Wilson's disease. In 31 patients with suspected Wilson's disease (WD-SUS-group), WD was excluded by means of investigations other than analysis of blood and urine. A total of 27 parameters of blood and urine in these 31 patients were compared to those of 20 de novo patients with manifest WD (WD-DEF-group), which parameter showed the highest significance level of difference between the WD-DEF-group and the WD-SUS-group. Thereafter, receiver operating characteristics (ROC-curves) were analyzed to evaluate which parameter showed the largest area under the curve (AUC) to detect WD. Finally, a logistic regression analysis was performed to analyze which combination of parameters allowed the best classification of the 51 patients either into the WD-DEF-group or into the WD-SUS-group. CHE showed the highest significance level for a difference between the WD-DEF- and WD-SUS-group, had the highest AUC, and, in combination with ceruloplasmin, allowed 100% correct classification. Without CHE, no other combination of parameters reached this level of correct classification. After the initiation of treatment, which regularly results in an improvement in CHE, the high diagnostic accuracy of this biomarker was lost. Cholinesterase turns out to be an excellent biomarker for differentiation between untreated de novo patients with manifest WD and heterozygotic gene carriers.

Utility of biomarkers and temporal artery biopsy length for investigating giant cell arteritis in Western Australia.

To explore demographic characteristics, biopsy length, and blood biomarker performance in an Australian cohort of patients who have undergone temporal artery biopsy (TAB) for giant cell arteritis (GCA). We extracted data on biopsies performed for GCA between January 2016 and December 2020 at public hospitals in Perth. Sensitivity, specificity, and area under the curve (AUC) were calculated for blood results. We evaluated the proportion of biopsies with post-fixation length less than 15 mm and explored several length associations. We retrospectively reviewed biopsies of 360 patients (65.8% female, mean age 72.1 years). Biopsy-positive patients were older (6.0 years, P < 0.01), and had higher C-reactive protein (CRP) (44.5mg/L, P < 0.01), erythrocyte sedimentation rate (ESR) (18.9 mm/h, P < 0.01), and platelets (86.8 × 103 /μL, P < 0.01) compared with biopsy-negative patients. CRP and platelets had the highest AUCs at 0.76 and 0.71, respectively. Sensitivities for CRP and ESR were 96.2% and 91.5%, respectively. Specificities were comparatively low at 41.3% for CRP and 37.4% for ESR. The proportion of biopsies with sub-optimal length was 55.9% and this varied significantly by site (P < 0.01). Smaller sites performed worse, with a sub-optimal biopsy rate of 87% amongst the three smallest sites. ESR and CRP are helpful preliminary investigations, especially in identifying low-risk patients, but their specificity is limited. Smaller centers had a higher proportion of biopsies with sub-optimal length. Considering the importance of biopsy length for TAB diagnostic value, reviewing biopsy data may assist services in developing improvement strategies.

A Computational Model of Blood D-Dimer, Cystatin C, and CRP Levels Predicts the Risk of Intracranial Aneurysms and their Rupture.

The aim of this study is to construct a computational model of blood D-dimer, cystatin C, and CRP levels and to predict the risk of intracranial aneurysms and their rupture. A total of 69 intracranial aneurysms patients were selected as the case group, including 28 cases in the ruptured group and 41 cases in the unruptured group. Another 64 non-intracranial aneurysm patients were selected as the control group. The detection results of serum D-dimer, cystatin C, and CRP were collected. The logistic regression computational model was used to analyze the occurrence and risk factors of intracranial aneurysms. The receiver operating curves (ROC) of serum D-dimer, cystatin C, and C reactive protein (CRP) levels for predicting intracranial aneurysms and their rupture were drawn, and the area under the curve (AUC), sensitivity, and specificity were calculated. The serum levels of D-dimer, cystatin C, and CRP in patients with intracranial aneurysms were significantly higher than those in the control group and the differences were statistically significant (P < 0.05). The serum levels of D-dimer, cystatin C, and CRP in patients with ruptured intracranial aneurysms were higher than those in patients with unruptured intracranial aneurysms, and the differences were also statistically significant (P < 0.05). The combined detection of serum D-dimer, cystatin C, and CRP levels has a higher AUC (0.9014) for predicting intracranial aneurysms and higher AUC (0.9412) for predicting ruptured intracranial aneurysms than D-dimer (0.7118 and 0.8750, respectively), cystatin C (0.6489 and 0.6180, respectively), and CRP (0.7764 and 0.6551, respectively) independent detection; the combined detection had a sensitivity of 93.75% and 87.80 for predicting the occurrence and rupture of intracranial aneurysms, and the specificity was 68.12% and 92.86%, respectively. The combined detection of serum D-dimer, cystatin C, and CRP levels is a very valuable indicator for predicting the occurrence and rupture of intracranial aneurysms, and combined detection can provide scientific evidence-based guidance for clinical prediction of the occurrence and rupture of intracranial aneurysms.

Machine Learning Approaches Incorporating High-Dimensional Longitudinal Data Improve the Prediction of Survival after Allogeneic Hematopoietic Cell Transplantation

Introduction: The European Society for Blood and Marrow Transplantation (EBMT) risk score and the Hematopoietic Cell Transplant Comorbidity Index (HCT-CI) remain suboptimal for predicting mortality after allogeneic hematopoietic cell transplantation (HCT) in pediatric patients. These risk scores are calculated based on a small number of variables collected before HCT. Their predictive performance is limited because of the inherent procedural uncertainties, the statistical methodologies used, and the limited number and quality of variables included. By combining multi-institutional, comprehensive, and patient-centric data, machine learning (ML)-based techniques provide an opportunity to develop more accurate risk prediction algorithms by incorporating longitudinal data collected before and after HCT. To our knowledge, no risk prediction algorithms incorporate longitudinal data collected after HCT into the prediction of overall survival (OS) for pediatric patients. We hypothesized that by including high-dimensional longitudinal data such as objective clinical and laboratory assessments from electronic health records (EHRs), models built with ML could help identify patients at high risk of mortality. To test this hypothesis, we developed an ML model incorporating such longitudinal data, which were readily available from EHRs, to predict the 1-year OS among pediatric patients undergoing allogeneic HCT, and we compared its prediction performance with that of a model involving only the baseline variables normally used in standard risk prediction models. Patients and Methods: This study included a cohort of 770 pediatric patients who underwent their first allogeneic HCT at St. Jude Children's Research Hospital between 2000 and 2020. Combining the baseline data (e.g., disease, donor, and graft characteristics) and the longitudinal evaluation data (e.g., blood cell counts, serum chemistries, measurable residual disease) collected within 30 days before or after HCT, we built a predictive model that uses penalized logistic regression to identify patients at high risk for mortality within 1 year after HCT. We compared the proposed ML model incorporating longitudinal measurements with a model trained only on baseline data and analyzed the importance of the different input features. Results: The area under the curve (AUC), either the receiver operating characteristic (ROC) curve or the precision-recall (PR) curve, measures the prediction accuracy of the estimated model. The AUC ranges from 0.0 and 1.0, with a higher AUC indicating a more accurate prediction. The proposed ML model has achieved an ROC AUC of 0.760 (95% confidence interval [CI]: 0.739-0.779), significantly outperforming the baseline model, for which the ROC AUC is 0.698 (95% CI: 0.675-0.718). This predictive advantage of the proposed ML model is also reflected in the PR AUC of 0.680 (95% CI: 0.651-0.712), as compared with 0.513 (95% CI: 0.482-0.546). The feature importance analysis suggests that in prediction modeling, the learned model relies heavily on the longitudinal evaluations, including the concentrations of serum magnesium, aspartate aminotransferase (AST), and ferritin, and the platelet count before and after HCT in addition to certain baseline variables such as the underlying diagnosis and graft type. Conclusion: Leveraging readily available longitudinal data from EHRs, we developed an ML model for 1-year OS prediction in pediatric patients undergoing allogeneic HCT. Our model shows that the results of frequently monitored laboratory and clinical tests performed within 30 days before or after HCT can potentially identify patients with a high risk of mortality after HCT. We are currently validating this model by using external datasets. Such ML models could be included in the EHR system as clinical decision support tools (see an R Shiny app created for data analysis and an illustration of the results at https://sjbiostat.shinyapps.io/pedsHCT), enabling better dynamic stratification of patient risk, optimization of clinical care, and improved therapeutic outcomes. The findings of this study reflect the fact that ML approaches are hugely augmentative and vital to efforts to transform large pools of data into valuable information as a result of their ability to extract predictive features efficiently from high-dimensional and heterogeneous data sources. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Hydroxyurea Pharmacokinetics and Fetal Hemoglobin Response in Ugandan Children with Sickle Cell Anemia

Introduction Over 300,000 infants are born with sickle cell anemia (SCA) in Africa every year, and the under-5 mortality of this population is estimated to be 50-90% without treatment. Hydroxyurea is a safe, oral medication that decreases morbidity and mortality in SCA by increasing protective fetal hemoglobin (HbF) production and reducing inflammation. Hydroxyurea dosing to maximum tolerated dose (MTD) offers the optimal treatment with greatest laboratory responses and clinical benefits. Patients have variable hydroxyurea absorption and elimination rates, so the traditional approach to determining MTD is through incremental dose escalation to achieve mild myelosuppression and reduction in absolute neutrophil count (ANC). In the United States, pharmacokinetic (PK)-guided dosing of hydroxyurea for children with SCA has been shown to decrease time to MTD and to augment treatment responses, with most patients achieving over 30% HbF. Maximizing clinical benefits and minimizing logistical burdens of dosing hydroxyurea would be advantageous for areas with high SCA prevalence, but PK parameters of African children with SCA who might have higher rates of malnutrition, infections, and organ damage that affect drug absorption, distribution, and clearance are yet to be determined. Here we describe hydroxyurea PK exposure-response in Ugandan children with SCA from the Optimizing Hydroxyurea Therapy in Children with Sickle Cell Anemia In Malaria Endemic Areas: The NOHARM MTD Study (NOHARM MTD, NCT03128515). Methods NOHARM MTD was a randomized controlled trial that compared dosing strategies between hydroxyurea at fixed, weight-based dosing (~20 mg/kg/day) and dose escalation to MTD in children with SCA living in Uganda. NOHARM MTD participants underwent PK testing after 18 months of treatment. Children were administered their prescribed hydroxyurea dose in the morning at clinic, and blood was obtained by fingerstick using Mitra® microsampler devices after 20, 60, and 180 minutes. The microsampler devices were frozen at -80C in Uganda and then transported to Cincinnati Children's Hospital for analysis of hydroxyurea concentration via liquid chromatography with tandem mass spectrometry. Using our previously published population PK model, individual hydroxyurea exposures expressed as area-under-the-curve (AUC) were generated by Bayesian estimation using the PK modeling software NONMEM (Version 7.5, ICON Development Solutions, Ellicott City, USA). A hydroxyurea exposure-response analysis was completed with HbF% as the primary pharmacodynamic outcome. Results A total of 90 NOHARM MTD participants completed PK testing; samples from 12 were excluded from full analysis due to an elevated baseline concentration suggesting the participant already received the daily hydroxyurea dose before PK testing (9), supraphysiologic concentration (1), or mislabeling (2). PK profiles were analyzed for the remaining 78 participants (mean age 6.5 years, SD 0.9); of these, 41 were on fixed dose hydroxyurea and 37 were at MTD with mean doses of 18.7 mg/kg/day (SD 1.6) and 28.3 mg/kg/day (SD 4.3), respectively (Table). Those at MTD had a higher AUC with a significantly higher %HbF (mean 19% HbF on fixed dose vs 29% at MTD, p=0.0002) and lower ANC (mean 4.7 x109/L at fixed dose vs 3.2 at MTD, p=0.0045). Eighteen participants (23%) had malnutrition with a body mass index Z-score less than -1.0, and 12 (15%) participants had renal hyperfiltration with an estimated glomerular filtration rate over 150 mL/min/1.73m2; neither condition was associated with differences in AUC. Conclusion This study demonstrates the feasibility of collecting hydroxyurea PK data with timed, sparse sampling in Uganda, and the exposure-response analysis confirms that higher hydroxyurea exposure in Ugandan children with SCA is associated with greater induction of protective HbF and reduction in ANC. Malnutrition and glomerular hyperfiltration did not affect hydroxyurea exposure in this cohort. Further studies of PK-guided dosing to optimize response and time to response in this population are needed. Future directions include building capacity for local analysis of hydroxyurea concentrations using HPLC technology, real-time dose calculation via a PK model-informed algorithm, and prospective evaluation of PK-guided dosing. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Evaluation of the quantitative performance of non-enhanced dual-energy CT X-map in detecting acute ischemic brain stroke: A model observer study using computer simulation

PurposeA simulation study was performed to evaluate the quantitative performance of X-map images—derived from non-enhanced (NE) dual-energy computed tomography (DECT)—in detecting acute ischemic stroke (AIS) compared with that of NE-DECT mixed images. MethodsA virtual phantom, 150 mm in diameter, filled with tissues comprising various gray- and white-matter proportions was used to generate pairs of NE-head images at 80 kV and Sn150 kV at three dose levels (20, 40, and 60 mGy). The phantom included an inserted low-contrast object, 15 mm in diameter, with four densities (0%, 5%, 10%, and 15%) mimicking ischemic edema. Mixed and X-map images were generated from these sets of images and compared in terms of detectability of ischemic edema using a channelized Hotelling observer (CHO). The area under the curve (AUC) of the receiver operating characteristic that generated CHO for each condition was used as a figure of merit. ResultsThe AUCs of X-map images were always significantly higher than those of mixed images (P < 0.001). The improvement in AUC for X-map images compared with that for mixed images at edema densities was 9.2%-12.6% at 20 mGy, 10.1%-17.7% at 40 mGy, and 14.0%-19.4% at 60 mGy. At any edema density, X-map images at 20 mGy resulted in higher AUCs than mixed images acquired at any other dose level (P < 0.001), which corresponded to a 66% dose reduction on X-map images. ConclusionsThe simulation study confirmed that NE-DECT X-map images have superior capability of detecting AIS than NE-DECT mixed images.