Diagnostic Accuracy Measures Research Articles

A brief effective screening protocol for identifying cataract patients with binocular vision anomalies

BackgroundTo compare the effectiveness of a brief binocular vision screening protocol to a comprehensive examination for detecting binocular vision anomalies before and after cataract surgery.MethodsA comprehensive binocular vision test battery as a gold standard were administered on recruited patients before the first surgery and at the third visit after surgery on the second eye. A receiver operating characteristic (ROC) curve was plotted to illustrate the diagnostic ability of each test. In addition, a univariate logistic regression analysis was performed to further determine the contribution of each preoperative test to the prediction of pre- and post-surgical binocular vision anomalies.ResultsSignificant differences were shown for the difference in phoria from distance to near measured by the cover test (Area Under Curve [AUC] = 0.96, P < 0.01), step vergence testing of positive fusional vergence at distance (AUC = 0.71, P < 0.01) and near (AUC = 0.77, P < 0.01). The other tests did not show statistically significant differences. The ROC curve generated by combining the difference in distance and near phoria with positive fusional vergence at both distance and near demonstrated a more robust measure of diagnostic accuracy. (AUC = 0.98, P < 0.01).ConclusionDistance and near phoria difference measured by cover test has similar effectiveness as a comprehensive binocular vision testing protocol for the diagnosis of binocular vision anomalies. Distance and near positive fusional vergence measured by step vergence testing also have significant predictive value. A combination of the two tests is an outstanding screening protocol for binocular vision anomalies before cataract surgery.Trial registrationThe study was registered at ClinicalTrials.gov (NCT03592615, Date of registration: July 19, 2018).

Multistep Algorithm to Predict RVOT PVC Site of Origin for Successful Ablation Using Available Criteria: A Two-Center Cross-Validation Study.

Predicting premature ventricular contraction (PVC) origin pre-ablation is a fundamental step, as right ventricular outflow tract (RVOT) PVC often leads to higher success rates. To compare nine published ECG criteria to differentiate between RVOT and non-RVOT origins of PVCs and develop a stepwise algorithm using those criteria to better determine PVC origin to predict ablation success. Two centers were involved in this study, the derivation group and the validation group. The derivation group included 65 patients with PVC left bundle branch block (LBBB) pattern morphology (predominantly negative in lead V1) and inferior axis (predominantly positive in leads II and III), who underwent ablation at Cipto Mangunkusumo Hospital (RSCM) (2017-2022). The validation group included 291 patients who underwent ablation at the Taipei Veteran General Hospital (2020-2023). We calculated and compared six diagnostic accuracy measures from nine previously published ECG morphology criteria to develop an algorithm to enhance the accuracy of predicting RVOT PVC origin for successful ablation. Our multistep algorithm using Criteria 5, 8, and 1 enhanced diagnostic performance compared to using each criterion alone. The accuracy, sensitivity, and specificity in the derivation group were 86.2%, 93.6%, and 66.7%, respectively; those in the validation group were 85.9%, 90.8%, and 64.7%, respectively. The ROC curve AUCs were 0.802 and 0.775, respectively. In cases of inferior axis and LBBB pattern PVCs, a multistep algorithm using multiple criteria increases the accuracy of predicting RVOT PVC origin instead of using a single criterion.

Salivary Proteomics for Detecting Novel Biomarkers of Periodontitis: A Systematic Review.

Salivary content is regarded as a powerful diagnostic window for oral and systemic diseases and the proteomic profile could be useful to distinguish between different periodontal conditions. The aim of the present systematic review was to assess distinctive salivary proteins identified through untargeted proteomics in periodontitis patients compared to periodontally healthy and gingivitis subjects, as well as to provide a qualitative methodological assessment of the current literature. Relevant studies identified from Medline via PubMed, Scopus, Embase, and Cochrane Library databases were retrieved to answer the following PECO question: "In systemically healthy individuals, are there any differences in salivary protein expression profiles assessed in proteomics studies between patients with periodontitis and periodontally healthy or gingivitis subjects?" Moreover, diagnostic utility of the identified markers was sought via a targeted literature search and further quantitative assessment. A modified version of the QUADAOMICS tool was used for the quality assessment of the included studies. After screening 461 relevant articles, a total of 13 studies were selected. The number of identified discriminant salivary proteins ranged from 2 to 4161. However, it was possible to identify proteins that were consistently over- or under-expressed in periodontitis patients in at least 3 studies. Among these, complement C3, profilin-1, SA100A8, and fibrinogen were consistently reported as increased in periodontitis, while cystatin-SN and leukocyte elastase inhibitor were more elevated in periodontally healthy controls. Only 4 studies reported diagnostic accuracy measures, with SA100A8 showing an area under the curve of 0.71 (95% CI: 0.66-0.75) in meta-analysis. Untargeted proteomics techniques identified some key biological molecules which were consistently reported to be over- or under-expressed in periodontitis. These findings could be useful to support novel candidate biomarkers for periodontitis. The high level of heterogeneity in methods and reporting urge to develop standardized protocols to be implemented in this research field (PROSPERO CRD42022299826).

T2Bacteria panel used simultaneously with blood cultures helps to determine the etiology ofbloodstream infections.

Bloodstream infections (BSI) result in significant morbidity and mortality rates, and delayed administration of appropriate antimicrobial treatment is a major predictor of poor outcomes. T2 magnetic resonance (T2MR®) (T2 Biosystems®, Lexington, MA, USA) is an innovative technology that can rapidly identify pathogens from a sample of whole blood in a remarkably short time frame of 3-5 h. We are evaluating if the T2Bacteria Panel (T2BP) contributes to the etiological diagnosis of bloodstream infections when combined with standard blood cultures (BC). The study was performed between December 2018 and March 2019, and a total of 28 patients with suspected BSI were included. The most notable finding of our study was that the addition of T2BP to BC in a diagnostic workflow led to a statistically significant higher rate of T2BP-targeted bacteria identification in patients with suspected BSI (46.4% versus 7.1%, P = 0.001) when compared to BC alone. Considering the measures of diagnostic accuracy, T2BP showed 100.00% sensitivity, 88.24% specificity, 100% negative predictive value (NPV), and 84.62% positive predictive value (PPV). Our findings give valuable insights for microbiologists and clinicians into this molecular method and its advantages in routine diagnostics of BSI.

Empirical Testing of a Middle-Range Theory for Ineffective Breathing Pattern in Children With Congenital Heart Disease.

To test a middle-range theory (MRT) for the nursing diagnosis of ineffective breathing pattern in children with congenital heart disease (CHD) based on analysis of two general propositions. This cross-sectional study is guided by STROBE. The propositions represent hypotheses about the relationships between the concepts of this MRT to be tested empirically, and thus, log-linear models were used to verify the structure of the proposition related to the stimuli. Diagnostic accuracy measures, univariate logistic regressions and the Mann-Whitney test were used to analyse the structure of the propositions related to behaviours. The analysis of the propositions related to the stimuli (eight concepts, four of which were classified as focal stimuli and four as contextual stimuli) suggested a reclassification of the stimulus "deformities in the thoracic wall" which became too focal. In the analysis of the propositions related to behaviours (17 concepts, five of which were classified as acute confirmatory, nine as acute clinical deterioration and three classified as chronic), guided changes in the operationalisation of concepts were suggested after comparing clinical findings; thus, acute confirmatory behaviours now have 10 concepts, while acute clinical deterioration behaviours and chronic behaviours continued with nine and three concepts, respectively, but with reclassifications between them. Changes in the operationalisation of the classification of the elements of the two propositions occurred after comparing the clinical findings with the theoretical model. By establishing precise causal relationships and describing how IBP manifests itself over time in children with CHD, empirical testing of this MRT helps nurses understand clinical reasoning based on temporal logic and spectral interaction between diagnostic components, which in turn will improve the use and accuracy of nursing diagnoses. Children and adolescents with CHD were recruited for this study sharing their clinical history and physical lung examination.

Diagnostic Test Accuracy of Fat Pad Sampling and Bone Marrow Biopsies in the Diagnosis of AL Amyloidosis: A Systematic Review and Meta-Analysis

Introduction: Tissue biopsies are needed to establish a diagnosis of systemic AL amyloidosis. Biopsy can be obtained from an involved organ with suspected amyloid deposition (target biopsy) or from sites in which amyloid commonly deposits (surrogate biopsy). The choice of biopsy site depends on institutional and physician preference. Two commonly used sites for surrogate biopsies are abdominal fat pad and bone marrow. The present systematic review and meta-analysis aims to evaluate the diagnostic test accuracy of fat pad and bone marrow biopsies as surrogate biopsy sites in the diagnosis of AL amyloidosis. Methods: As part of the ASH AL Amyloidosis Diagnostic Guideline, we performed systematic reviews for seven prioritized PICO questions related to the screening, diagnosis and organ involvement evaluation of AL amyloidosis. We conducted electronic searches on PubMed, Embase, and the Cochrane Central Register of Controlled Trials from inception until January 2024. Two reviewers independently and in duplicate performed title and abstract screening and full text article screening on LASER AI, with conflicts resolved by a third reviewer. Studies were eligible if they included patients that had a fat pad sample or bone marrow biopsy and a target organ biopsy for suspicion of AL amyloidosis allowing for the calculation of diagnostic test accuracy measures. Diagnostic test accuracy measures of interest were sensitivity, specificity, positive and negative predictive value. Statistical analysis was performed on OpenMeta[Analyst]. Grading of Recommendations Assessment, Development and Evaluation (GRADE) was used to assess the certainty of evidence. Results: After deduplication, we retrieved 29,237 studies.. 20 studies with 2955 patients reporting on bone marrow biopsy, 25 with 4702 patients on fat pad sampling were included. The pooled sensitivity for bone marrow biopsy was 55.1% (95% CI 45.8-64.0, I2 = 94%) (Certainty of evidence: Low ⨁⨁◯◯), For fat pad sampling sensitivity was 76.6% (95% CI 72.1-80.8, I2 = 84.2%) (Certainty of evidence: Moderate ⨁⨁⨁◯). Among the included studies, 1 evaluated the sensitivity of fat pad sampling and bone marrow biopsy combined and found it to be 89%. No included studies evaluated specificity for bone marrow biopsy or fat pad sample. Conclusion: Combination of fat pad and bone marrow biopsy together significantly improve sensitivity for amyloid detection compared to either fat pad sampling or bone marrow biopsy alone.

Predicting radiation pneumonitis in lung cancer using machine learning and multimodal features: a systematic review and meta-analysis of diagnostic accuracy.

To evaluate the diagnostic accuracy of machine learning models incorporating multimodal features for predicting radiation pneumonitis in lung cancer through a systematic review and meta-analysis. Relevant studies were identified through a systematic search of PubMed, Web of Science, Embase, and the Cochrane Library from October 2003 to December 2023. Additional studies were located by reviewing bibliographies and relevant websites. Two independent researchers screened titles, abstracts, and full-text articles according to predefined inclusion and exclusion criteria. Data extraction was performed using standardized forms, and study quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 tool. The primary outcomes, including combined sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and area under the curve (AUC), were calculated using STATA MP-64 software(Stata Corporation LLC, College Station, USA) with a random-effects model. Meta-analysis was conducted to synthesize diagnostic accuracy measures, and analyses of heterogeneity and publication bias were performed. A total of 1,406 patients with primary lung cancer were included in this systematic review, drawing data from 9 studies. The pooled analysis revealed a sensitivity of 0.74 [0.58-0.85] and a specificity of 0.91 [0.87-0.95] for machine learning models in diagnosing radiation pneumonitis. The positive likelihood ratio (PLR) was 8.69 [5.21-14.50], the negative likelihood ratio (NLR) was 0.28 [0.16-0.49], and the diagnostic odds ratio (DOR) was 30.73 [11.96-78.97]. The area under the curve (AUC) was 0.93 [0.90-0.95], indicating excellent diagnostic performance. Meta-regression analysis identified that the number of machine learning models, year of publication, and study design contributed to heterogeneity among studies. No evidence of publication bias was found. Overall, machine learning models incorporating multimodal characteristics demonstrated 75% accuracy in predicting moderate to severe radiation pneumonitis. In conclusion, by integrating the current machine learning (ML) algorithm's ability in big data mining, a predictive model can be constructed by combining multi-modal features such as genetics, imaging, and cell factors. By selecting multiple machine learning algorithm frameworks and competing for the best combination model based on research goals, the reliability and accuracy of the radiation pneumonitis prediction model can be greatly improved. PROSPERO (CRD42024497599).

Different view of the diagnostics test accuracy measures and optimal cut-off point selection procedure under tree or umbrella ordering

ABSTRACT In the realm of medical diagnostic testing, diagnostic tests can assume either binary forms, distinguishing between diseased and non-diseased states, or ordinal forms, categorizing states from non-diseased to various stages (1 to K). Another significant classification scheme for multi-class scenarios is tree or umbrella ordering, which entails several unordered sub-classes (subtypes) within a biomarker. Within tree or umbrella ordering, the classifier assesses whether the marker measurement for one class surpasses or falls below those for the other classes. Although Receiver Operating Characteristic (ROC) curves and summary measures have been adapted to accommodate tree and umbrella ordering, these approaches often yield cut-off points that generate highly sensitive tests for certain disease subtypes while compromising specificity for others. This may not be ideal for all diseases. Hence, in this investigation, we explore diverse measures of diagnostic test accuracy and optimal cut-off point selection procedures under tree or umbrella ordering to foster more specific tests. We present numerical examples and simulation studies and demonstrate the approach using real data on lung cancer.

Assessing diagnostic accuracy: 18F-FDG PET-CT scans in low-grade infection detection among post-traumatic long bone non-unions; a literature review and clinical data

IntroductionThe diagnosis of low-grade infection in post-traumatic long bone non-unions poses challenges due to the absence of clinical signs. This study aimed to review the available literature on the diagnostic accuracy of imaging techniques for low-grade infections and assess the diagnostic accuracy of 18F-FDG PET-CT scans for low-grade infection in post-traumatic long bone non-unions. MethodsA mini-review was conducted using Pubmed in March 2024. A retrospective study was conducted including adult patients with a long bone non-union, suspected of infection. All patients underwent 18F-FDG PET-CT scans as the index test before surgical intervention, with peri‑operative cultures obtained during surgery serving as the reference standard. Quantitative analyses were performed on the standardized uptake value (SUV) measurements obtained from the 18F-FDG PET-CT scans. Diagnostic accuracy measures including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. Receiver operating characteristic (ROC) curve analysis was employed to evaluate the discriminatory ability of SUV measurements. ResultsLiterature suggests that 18F-FDG PET-CT is the most accurate imaging technique to detect low-grade infections. The study included a total of 51 18F-FDG PET-CT scans and cultures from 50 patients with long bone non-unions. The diagnostic accuracy was found to be 0.67 (95 % CI 0.44–0.87). The PPV and NPV were calculated as 0.79 (95 % CI 0.53–1.00) and 0.52 (95 % CI 0.30–0.73), respectively. Quantitative analyses of SUV measurements demonstrated a low level of accuracy, with all area under the curve (AUC) values < 0.75 and ROC curves showing a trajectory fairly parallel to the diagonal line. Conclusion and discussionThe findings of this study indicate that in post-traumatic long bone non-unions, where a low-grade fracture-related infection (FRI) is suspected, the 18F-FDG PET-CT has a performance that is advantageous over other imaging techniques. A careful interpretation of the scan results is warranted, possibly including the quantitative analysis on tracer uptake as an adjunct. Nevertheless, the diagnostic accuracy in this condition is not as good as in early-onset FRI cases, and this should be taken into account when treating these challenging cases.

Prioritising primary care patients with unexpected weight loss for cancer investigation: diagnostic accuracy study (update)

ObjectiveTo quantify the predictive value of unexpected weight loss for cancer according to patient’s age, sex, smoking status, and concurrent clinical features (symptoms, signs, and abnormal blood test results).DesignDiagnostic accuracy...

Development of a transformer-based deep learning algorithm for diabetic peripheral neuropathy classification using corneal confocal microscopy images.

Diabetic peripheral neuropathy (DPN) is common and can go unnoticed until it is firmly developed. This study aims to establish a transformer-based deep learning algorithm (DLA) to classify corneal confocal microscopy (CCM) images, identifying DPN in diabetic patients. Our classification model differs from traditional convolutional neural networks (CNNs) using a Swin transformer network with a hierarchical architecture backbone. Participants included those with (DPN+, n = 57) or without (DPN-, n = 37) DPN as determined by the updated Toronto consensus criteria. The CCM image dataset (consisting of 570 DPN+ and 370 DPN- images, with five images selected from each participant's left and right eyes) was randomly divided into training, validation, and test subsets at a 7:1:2 ratio, considering individual participants. The effectiveness of the algorithm was assessed using diagnostic accuracy measures, such as sensitivity, specificity, and accuracy, in conjunction with Grad-CAM visualization techniques to interpret the model's decisions. In the DPN + group (n = 12), the transformer model successfully predicted all participants, while in the DPN- group (n = 7), one participant was misclassified as DPN+, with an area under the curve (AUC) of 0.9405 (95% CI 0.8166, 1.0000). Among the DPN + images (n = 120), 117 were correctly classified, and among the DPN- images (n = 70), 49 were correctly classified, with an AUC of 0.8996 (95% CI 0.8502, 0.9491). For single-image predictions, the transformer model achieved a superior AUC relative to the ResNet50 model (0.8761, 95% CI 0.8155, 0.9366), the Inception_v3 model (0.8802, 95% CI 0.8231, 0.9374), and the DenseNet121 model (0.8965, 95% CI 0.8438, 0.9491). Transformer-based networks outperform CNN-based networks in rapid binary DPN classification. Transformer-based DLAs have clinical DPN screening potential.

Sodium fluorescein and 5-aminolevulinic acid fluorescence- guided biopsy in brain lesions: a systematic review and meta-analysis.

Stereotactic brain biopsies are highly efficient for diagnosing intracerebral pathologies, particularly when surgical resection is infeasible. Fluorescence-based agents such as 5-aminolevulinic acid (5-ALA) and fluorescein sodium (NaFl) can enhance diagnostic accuracy and safety, improving the visualization of lesional tissues. This meta-analysis aimed to evaluate their effect on diagnostic yield and complication rates of brain biopsies. This study adhered to Cochrane and PRISMA guidelines. We assessed studies for diagnostic yield and complication rates. Data was analyzed using a random-effects model in RStudio. Diagnostic accuracy measures such as sensitivity and predictive values were calculated based on fluorescence visibility in biopsy samples. Thirty-two non-randomized studies were included, comprising 947 patients, with a mean age ranging from 37 to 77years, and a mean sample number ranging from 1 to 15 specimens. Diagnostic yields were high: 93% for NaFl and 96% for 5-ALA. Major complications occurred in 3% of procedures with both agents, while minor complications were reported in 7% and 5% with NaFl and 5-ALA respectively. The Negative-predictive-value (NPV) of 5-ALA and NaFl were 8-11% and 60-80% respectively. NaFl demonstrates higher sensitivity and specificity at 84% and 100% compared to 5-ALA's 66%. and 85% respectively. 5-ALA and NaFl provide high diagnostic yields with acceptable safety profiles in stereotactic biopsies. NaFl showed higher sensitivity and specificity. NaFl outperforms 5ALA in terms of NPV making it more efficient for small lesions near eloquent regions or major blood vessels. The significance of these findings can be further ascertained through randomized trials.

Early prediction of gestational diabetes mellitus using first trimester maternal serum pregnancy-associated plasma protein-a: A cross-sectional study.

The oral glucose tolerance test with 75 g glucose is commonly regarded as the gold standard (GS) for the detection of gestational diabetes mellitus (GDM). However, one limitation of this test is its administration in the late second trimester of pregnancy in some countries (e.g., Iran). This study aimed to evaluate the accuracy of pregnancy-associated plasma protein-A (PAPP-A) for predicting GDM in the early first trimester of pregnancy using a novel statistical modeling technique. The study population consisted of 344 pregnant women who participated in the first trimester screening program for GDM. Maternal serum PAPP-A levels were measured between 11 and 13 gestational weeks for all participants. A Bayesian latent profile model (LPM) under the skew-t (ST) distribution was employed to estimate the diagnostic accuracy measures of PAPP-A in the absence of GS test outcomes. The mean (standard deviation) age of the participants was 28.87 ± 5.20 years. The median (interquartile range (IQR)) PAPP-A MoM was 0.91 (0.69-1.34). Utilizing the LPM under the ST distribution while adjusting for covariates, the sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) of PAPP-A were 92% (95% credible interval [CrI]: 0.89, 0.98), 81% (95% CrI: 0.76, 0.91), and 0.91 (95% CrI: 0.83, 0.97), respectively. Notably, the pregnant women with GDM had significantly lower PAPP-A values ( -0.52, 95% CrI [-0.61, -0.46]). Generally, our findings confirmed that PAPP-A could serve as a potential screening tool for the identification of GDM in the early stages of pregnancy.

Assessing Biofilm at the Bedside: Exploring Reliable Accessible Biofilm Detection Methods.

Biofilm is linked through a variety of mechanisms to the pathogenesis of chronic wounds. However, accurate biofilm detection is challenging, demanding highly specialized and technically complex methods rendering it unapplicable for most clinical settings. This study evaluated promising methods of bedside biofilm localization, fluorescence imaging of wound bacterial loads, and biofilm blotting by comparing their performance against validation scanning electron microscopy (SEM). In this clinical trial, 40 chronic hard-to-heal wounds underwent the following assessments: (1) clinical signs of biofilm (CSB), (2) biofilm blotting, (3) fluorescence imaging for localizing bacterial loads, wound scraping taken for (4) SEM to confirm matrix encased bacteria (biofilm), and (5) PCR (Polymerase Chain Reaction) and NGS (Next Generation Sequencing) to determine absolute bacterial load and species present. We used a combination of SEM and PCR microbiology to calculate the diagnostic accuracy measures of the CSB, biofilm blotting assay, and fluorescence imaging. Study data demonstrate that 62.5% of wounds were identified as biofilm-positive based on SEM and microbiological assessment. By employing this method to determine the gold truth, and thus calculate accuracy measures for all methods, fluorescence imaging demonstrated superior sensitivity (84%) and accuracy (63%) compared to CSB (sensitivity 44% and accuracy 43%) and biofilm blotting (sensitivity 24% and accuracy 40%). Biofilm blotting exhibited the highest specificity (64%), albeit with lower sensitivity and accuracy. Using SEM alone as the validation method slightly altered the results, but all trends held constant. This trial provides the first comparative assessment of bedside methods for wound biofilm detection. We report the diagnostic accuracy measures of these more feasibly implementable methods versus laboratory-based SEM. Fluorescence imaging showed the greatest number of true positives (highest sensitivity), which is clinically relevant and provides assurance that no pathogenic bacteria will be missed. It effectively alerted regions of biofilm at the point-of-care with greater accuracy than standard clinical assessment (CSB) or biofilm blotting paper, providing actionable information that will likely translate into enhanced therapeutic approaches and better patient outcomes.

Real-world performance of a single-use, analyser-free, molecular point-of-care test for COVID-19 used in the emergency department: Results of a prospective trial (ED-POC)

A novel single-use, analyser-free, molecular point-of-care test for SARS-CoV-2 (Veros COVID-19 test, Sherlock Biosciences) could reduce time to results and improve patient care and flow in the emergency department (ED), but its performance in this setting is unknown. Adults aged ≥18 years presenting to Southampton General Hospital (UK) with suspected COVID-19 were tested with the Veros COVID-19 test in addition to standard of care near-patient PCR. Measures of diagnostic accuracy were calculated for the Veros COVID-19 test stratified by Ct value. Discrepant results underwent viral culture. Between Jan 16 and May 2, 2023, 400 patients were enrolled with a median (IQR) age of 60 (34-77) and 141 (35·3%) were SARS-CoV-2 positive by PCR. The Veros test gave valid results on the first test in 384 (96·0%), and sensitivity and specificity were 127/141 (90·1%, 95%CI 83·9-94·5) and 258/259 (99·6%, 95%CI 97·9-100) overall. For those with high or moderate viral load (Ct ≤30), sensitivity was 125/129 (96·9%, 95%CI 92·3-99·2). One (7·1%) of 14 PCR positive/Veros test negative samples was culture positive. Median (IQR) time from sample collection to result was 19 (18-20) mins with the Veros test versus 73 (59-92) mins with PCR (p<0·0001). The Veros COVID-19 test generated results in near real-time, around 1h sooner than rapid, near-patient, analyser-based PCR, and accuracy was excellent for samples with moderate and high viral loads. The Veros test represents a step-change in molecular diagnostics for infection and could significantly reduce time to results and improve patient management in EDs and other settings.

Accuracy of machine learning in predicting outcomes post-percutaneous coronary intervention: a systematic review.

Recent studies have shown potential in introducing machine learning (ML) algorithms to predict outcomes post-percutaneous coronary intervention (PCI). We aimed to critically appraise current ML models' effectiveness as clinical tools to predict outcomes post-PCI. Searches of four databases were conducted for articles published from the database inception date to 29 May 2021. Studies using ML to predict outcomes post-PCI were included. For individual post-PCI outcomes, measures of diagnostic accuracy were extracted. An adapted checklist comprising existing frameworks for new risk markers, diagnostic accuracy, prognostic tools and ML was used to critically appraise the included studies along the stages of the translational pathway: development, validation, and impact. Quality of training data and methods of dealing with missing data were evaluated. Twelve cohorts from 11 studies were included with a total of 4,943,425 patients. ML models performed with high diagnostic accuracy. However, there are concerns over the development of the ML models. Methods of dealing with missing data were problematic. Four studies did not discuss how missing data were handled. One study removed patients if any of the predictor variable data points were missing. Moreover, at the validation stage, only three studies externally validated the models presented. There could be concerns over the applicability of these models. None of the studies discussed the cost-effectiveness of implementing the models. ML models show promise as a useful clinical adjunct to traditional risk stratification scores in predicting outcomes post-PCI. However, significant challenges need to be addressed before ML can be integrated into clinical practice.

Diagnosing Sarcopenia with AI-Aided Ultrasound (DINOSAUR)-A Pilot Study.

Background: Sarcopenia has been recognized as a determining factor in surgical outcomes and is associated with an increased risk of postoperative complications and readmission. Diagnosis is currently based on clinical guidelines, which includes assessment of skeletal muscle mass but not quality. Ultrasound has been proposed as a useful point-of-care diagnostic tool to assess muscle quality, but no validated cut-offs for sarcopenia have been reported. Using novel automated artificial intelligence (AI) software to interpret ultrasound images may assist in mitigating the operator-dependent nature of the modality. Our study aims to evaluate the fidelity of AI-aided ultrasound as a reliable and reproducible modality to assess muscle quality and diagnose sarcopenia in surgical patients. Methods: Thirty-six adult participants from an outpatient clinic were recruited for this prospective cohort study. Sarcopenia was diagnosed according to Asian Working Group for Sarcopenia (AWGS) 2019 guidelines. Ultrasonography of the rectus femoris muscle was performed, and images were analyzed by an AI software (MuscleSound® (Version 5.69.0)) to derive muscle parameters including intramuscular adipose tissue (IMAT) as a proxy of muscle quality. A receiver operative characteristic (ROC) curve was used to assess the predictive capability of IMAT and its derivatives, with area under the curve (AUC) as a measure of overall diagnostic accuracy. To evaluate consistency between ultrasound users of different experience, intra- and inter-rater reliability of muscle ultrasound parameters was analyzed in a separate cohort using intraclass correlation coefficients (ICC) and Bland-Altman plots. Results: The median age was 69.5 years (range: 26-87), and the prevalence of sarcopenia in the cohort was 30.6%. The ROC curve plotted with IMAT index (IMAT% divided by muscle area) yielded an AUC of 0.727 (95% CI: 0.551-0.904). An optimal cut-off point of 4.827%/cm2 for IMAT index was determined with a Youden's Index of 0.498. We also demonstrated that IMAT index has excellent intra-rater reliability (ICC = 0.938, CI: 0.905-0.961) and good inter-rater reliability (ICC = 0.776, CI: 0.627-0.866). In Bland-Altman plots, the limits of agreement were from -1.489 to 1.566 and -2.107 to 4.562, respectively. Discussion: IMAT index obtained via ultrasound has the potential to act as a point-of-care evaluation for sarcopenia screening and diagnosis, with good intra- and inter-rater reliability. The proposed IMAT index cut-off maximizes sensitivity for case finding, supporting its use as an easily implementable point-of-care test in the community for sarcopenia screening. Further research incorporating other ultrasound parameters of muscle quality may provide the basis for a more robust diagnostic tool to help predict surgical risk and outcomes.

Development and validation of a clinical prediction rule for walking independence in hospitalized older adults with a vertebral compression fracture.

No reports on factors or Clinical prediction rules (CPRs) associated with walking independence among patients with vertebral compression fractures (VCFs) are available. Evidence regarding epidemiological walking independence rates is also sparse. Here, we sought to (i) obtain epidemiological data on the probability of inpatients with VCFs achieving walking independence, and (ii) develop and validate a CPR to determine walking independence in hospitalized patients with VCFs. We conducted a retrospective cross-sectional observational study of patients aged ≥60years who were hospitalized for VCF at four hospitals in Japan in 2019-2022. The outcome was walking independence at discharge. We performed a binomial logistic regression analysis to assess predictors of walking independence. Five independent variables were entered: age, American Society of Anesthesiologists physical status, cognitive function, Berg Balance Scale (BBS), and 10-m walking test. Among the independent variables that were significant, we converted the continuous variables to binary data by calculating cut-off values and then created the CPR. The area under the curve (AUC) was calculated as the measure of the CPR's diagnostic accuracy, and internal validation was conducted by bootstrapping. Of the 240 patients, 188 (78.3%) achieved walking independence. Cognitive function and the BBS score (with a cut-off of 45 points) were identified as significant predictors. We created a CPR using these two items (0-2 points). The CPR's AUC was 0.92 (0.874-0.967), and internal validation by bootstrapping yielded a mean AUC of 0.919 with a slope of 0.965. The walking independence rate of patients with a VCF during hospitalization was 78.3%, with cognitive function and BBS being predictors. The developed CPR performed well enough to retrospectively predict walking independence in VCF patients. The BBS cut-off value and the CPR may serve as useful indicators for clinicians to predict VCF patients' walking independence.

Assessment of the Diagnostic Accuracy of CT as Compared to MRI in Detecting Metastases in Patients With Colorectal Cancer.

This study aimed to compare the diagnostic accuracy of computed tomography (CT) and magnetic resonance imaging (MRI) in detecting metastases of colorectal cancer (CRC) in a hospital in Najran, Saudi Arabia. A total of 51 patients with CRC were included in the study. The radiological findings of metastatic lesions and the diagnostic accuracy measures of CT compared to MRI were analyzed. The results showed that CT had a false negative rate of 7.8%, a false positive rate of 7.8%, a true negative rate of 27.5%, and a true positive rate of 56.9% in detecting metastases. Diagnostic accuracy measures varied based on the number of metastatic lesions, with higher sensitivity observed for cases with fewer lesions. Gender, timing of imaging in relation to surgical intervention, and administration of nonsurgical therapy showed significant associations with diagnosis mismatch between CT and MRI. The site of metastases and the site of the primary tumor in the colon also demonstrated significant associations with diagnosis mismatch. The size of the largest metastasis detected by MRI was significantly associated with diagnosis mismatch. The overall diagnostic accuracy of CT in detecting any metastases, compared to MRI as the reference standard, was estimated to have a sensitivity of 87.8%, a specificity of 77.8%, a positive predictive value of 87.8%, and a negative predictive value of 77.8%. This study provides valuable insights into the comparative diagnostic performance of CT and MRI in detecting metastases of CRC, highlighting the importance of considering patient characteristics, disease outcome, and tumor characteristics in the interpretation of imaging results.

Diagnostic Accuracy of Clinical Sign Algorithms to Identify Sepsis in Young Infants Aged 0 to 59 Days: A Systematic Review and Meta-analysis.

Accurate identification of possible sepsis in young infants is needed to effectively manage and reduce sepsis-related morbidity and mortality. Synthesize evidence on the diagnostic accuracy of clinical sign algorithms to identify young infants (aged 0-59 days) with suspected sepsis. MEDLINE, Embase, CINAHL, Global Index Medicus, and Cochrane CENTRAL Registry of Trials. Studies reporting diagnostic accuracy measures of algorithms including infant clinical signs to identify young infants with suspected sepsis. We used Cochrane methods for study screening, data extraction, risk of bias assessment, and determining certainty of evidence using Grading of Recommendations Assessment Development and Evaluation. We included 19 studies (12 Integrated Management of Childhood Illness [IMCI] and 7 non-IMCI studies). The current World Health Organization (WHO) 7-sign IMCI algorithm had a sensitivity of 79% (95% CI 77%-82%) and specificity of 77% (95% CI 76%-78%) for identifying sick infants aged 0-59 days requiring hospitalization/antibiotics (1 study, N = 8889). Any IMCI algorithm had a pooled sensitivity of 84% (95% CI 75%-90%) and specificity of 80% (95% CI 64%-90%) for identifying suspected sepsis (11 studies, N = 15523). When restricting the reference standard to laboratory-supported sepsis, any IMCI algorithm had a pooled sensitivity of 86% (95% CI 82%-90%) and lower specificity of 61% (95% CI 49%-72%) (6 studies, N = 14278). Heterogeneity of algorithms and reference standards limited the evidence. IMCI algorithms had acceptable sensitivity for identifying young infants with suspected sepsis. Specificity was lower using a reference standard of laboratory-supported sepsis diagnosis.