Improvement In Discriminatory Ability Research Articles

Benchmarking multi-ancestry prostate cancer polygenic risk scores in a real-world cohort.

Prostate cancer is a heritable disease with ancestry-biased incidence and mortality. Polygenic risk scores (PRSs) offer promising advancements in predicting disease risk, including prostate cancer. While their accuracy continues to improve, research aimed at enhancing their effectiveness within African and Asian populations remains key for equitable use. Recent algorithmic developments for PRS derivation have resulted in improved pan-ancestral risk prediction for several diseases. In this study, we benchmark the predictive power of six widely used PRS derivation algorithms, including four of which adjust for ancestry, against prostate cancer cases and controls from the UK Biobank and All of Us cohorts. We find modest improvement in discriminatory ability when compared with a simple method that prioritizes variants, clumping, and published polygenic risk scores. Our findings underscore the importance of improving upon risk prediction algorithms and the sampling of diverse cohorts.

The role of NMR-based circulating metabolic biomarkers in development and risk prediction of new onset type 2 diabetes

Associations of circulating metabolic biomarkers with type 2 diabetes (T2D) and their added value for risk prediction are uncertain among Chinese adults. A case-cohort study included 882 T2D cases diagnosed during 8-years’ follow-up and a subcohort of 789 participants. NMR-metabolomic profiling quantified 225 plasma biomarkers in stored samples taken at recruitment into the study. Cox regression yielded adjusted hazard ratios (HRs) for T2D associated with individual biomarkers, with a set of biomarkers incorporated into an established T2D risk prediction model to assess improvement in discriminatory ability. Mean baseline BMI (SD) was higher in T2D cases than in the subcohort (25.7 [3.6] vs. 23.9 [3.6] kg/m2). Overall, 163 biomarkers were significantly and independently associated with T2D at false discovery rate (FDR) controlled p < 0.05, and 138 at FDR-controlled p < 0.01. Branched chain amino acids (BCAA), apolipoprotein B/apolipoprotein A1, triglycerides in VLDL and medium and small HDL particles, and VLDL particle size were strongly positively associated with T2D (HRs 1.74–2.36 per 1 SD, p < 0.001). HDL particle size, cholesterol concentration in larger HDL particles and docosahexaenoic acid levels were strongly inversely associated with T2D (HRs 0.43–0.48, p < 0.001). With additional adjustment for plasma glucose, most associations (n = 147 and n = 129 at p < 0.05 and p < 0.01, respectively) remained significant. HRs appeared more extreme among more centrally adipose participants for apolipoprotein B/apolipoprotein A1, BCAA, HDL particle size and docosahexaenoic acid (p for heterogeneity ≤ 0.05). Addition of 31 selected biomarkers to an established T2D risk prediction model modestly, but significantly, improved risk discrimination (c-statistic 0.86 to 0.91, p < 0.001). In relatively lean Chinese adults, diverse metabolic biomarkers are associated with future risk of T2D and can help improve established risk prediction models.

The use of ECG by triage nurses may improve the stratification of patients with cardiological symptoms admitted to the emergency department: a prospective observational study

Abstract Background Triage is the crucial point for overall emergency department (ED) management. Its purpose is to quickly and safely identify among the many patients arriving those who need rapid medical evaluation [1]. Present triage systems are based on symptomatic assessment and clinical impression to assess patient severity on admission to the ED. Some clinical tools that are widely validated in practice, such as ECG, could be implemented in triage to improve the stratification activity of the system [2]. Purpose To evaluate whether the use of ECG can improve nursing triage prediction for patients admitted to the ED for symptoms related to suspected cardiac pathology. Methods A prospective observational single-centered study was conducted. All patients who visited the ED for chest pain, syncope, dyspnoea or palpitation and underwent ECG at the time of triage were considered. The study was conducted by a pool of nurses who had previously received special training in ECG reading. Background characteristics, comorbidities and clinical conditions were recorded for each enrolled patient. In addition, the priority level assigned according to the application of the 5-level Manchester Triage System (1 to 5) and the level of priority after ECG performance (1 to 5) were recorded. The study outcome was composite: the presence of a major cardiac event within 6 weeks after triage, a cardiogenic syncope, the presence of a pulmonary embolism or aortic dissection, or an arrhythmia requiring specific emergency therapy [3]. The priority level assigned in triage and the priority level given after performing the ECG were compared by analysis of the receiver operating characteristic. Results A total of 1,211 patients were included in the study. Of these, 62.4% (756/1,211) presented chest pain, 13.8% (167/1,211) syncope, 18.2% (221/1,211) palpitation and 5.5% (67/1,211) dyspnoea. Overall, 10.5% (127/1,211) had a positive outcome. The concordance between triage nurses and cardiologists in detecting pathological and non-pathological ECGs presented a Cohen's kappa of 0.821 (p&amp;lt;0.001). ECG performed by nurses increased the patient's priority in 7.5% (91/1,211) of cases and reduced it in 39.6% (480/1,211) of cases. The discriminatory ability of the triage system before ECG was performed showed an area under the receiver operating characteristic curve of 0.708 (95% CI 0.659–0.757), whereas the discriminatory ability of triage after ECG was 0.845 (95% CI 0.808–0.883). Conclusions The use of ECG in triage can be a simple and safe tool that improves the assessment of the patient's priority when presenting to the ED for a cardiac symptom. The presence of a good diagnostic concordance and a good improvement in discriminatory ability may support the implementation of ECG in nursing triage. Funding Acknowledgement Type of funding sources: None.

Diffuse coronary artery disease among other atherosclerotic plaque characteristics by coronary computed tomography angiography for predicting coronary vessel-specific ischemia by fractional flow reserve

Background and aimsCoronary computed tomography angiography (CCTA) permits effective identification of diffuse CAD and atherosclerotic plaque characteristics (APCs). We sought to examine the usefulness of diffuse CAD beyond luminal narrowing and APCs by CCTA to detect vessel-specific ischemia. Methods407 vessels (n = 252 patients) from the DeFACTO diagnostic accuracy study were retrospectively analyzed for percent plaque diffuseness (PD). Percent plaque diffuseness (PD) was obtained on per-vessel level by summation of all contiguous lesion lengths and divided by total vessel length, and was logarithmically transformed (log percent PD). Additional CCTA measures of stenosis severity including minimal lumen diameter (MLD), and APCs, such as positive remodeling (PR) and low attenuation plaque (LAP), were also included. Vessel-specific ischemia was defined as fractional flow reserve (FFR) ≤0.80. Multivariable regression, discrimination by area under the receiver operating characteristic curve (AUC), and category-free net reclassification improvement (cNRI) were assessed. ResultsBackward stepwise logistic regression revealed that for every unit increase in log percent PD, there was a 58% (95% CI: 1.01–2.48, p = 0.048) rise in the odds of having an abnormal FFR, independent of stenosis severity and APCs. The AUC indicated no further improvement in discriminatory ability after adding log percent PD to the final parsimonious model of MLD, PR, and LAP (AUC difference: 0.003, 95% CI: −0.003–0.010, p = 0.33). Conversely, adding log percent PD to the base model of MLD, PR, and LAP improved cNRI by 0.21 (95% CI: 0.01–0.41, p < 0.001). ConclusionsAccounting for diffuse CAD may help improve the accuracy of CCTA for detecting vessel-specific ischemia.

A Bayesian Approach to Model-Development: Automatic Learning for Tuning Predictive Performance

The value of manually constructed and tuned Bayesian networks has been demonstrated empirically, however this informal process is limited in terms of what can be reasonably achieved. This paper presents the application of a formal machine learning process, EM learning, to a manually constructed CPN for the assessment of the severity of sepsis. Through learning, the model is tuned to predict 30-day mortality, and displays a significant improvement in discriminatory ability assessed by area under the ROC curve (previous model AUC = 0.647, new model AUC = 0.739, p<0.001).

Abstract B91: Genetic variations in the JAK-STAT pathway are associated with postoperative pulmonary complications after lobectomy in lung cancer patients

Abstract Background: Lung cancer patients who received lobectomy are at high risk of developing postoperative pulmonary complications. Clinical factors have been investigated to predict high risk patients; however, additional predictors are needed to accurately and reliably evaluate risk. Evidence has shown a relationship between the JAK-STAT pathway with pulmonary inflammation and lung injuries. Methods: We conducted a two stage discovery-validation study to evaluate the associations between genetic variations in JAK-STAT signaling pathway and the risk of developing postoperative pulmonary complications following lobectomy. A total of 223 SNPs from 27 genes were genotyped and analyzed in 264 early stage non-small cell lung cancer (NSCLC) patients who received lobectomy. Twelve significant (P&amp;lt;0.05) SNPs were then validated in additional 264 NSCLC patients. Results: Four SNPs showed a consistent effect in both discovery and validation populations and had a pooled p value &amp;lt; 0.05. Among them, PIAS3:rs10910829 was significantly associated with increased risk of developing complications in both the discovery (OR=1.78, 95%CI=1.16-2.74, P=0.008) and validation (OR=1.52, 95%CI=1.03-2.25, P=0.036) populations. PIAS2:rs9304337 was associated with decreased risk in the discovery set (OR= 0.45, 95%CI=0.21-0.95, P=0.037) and reached borderline significance (OR=0.53, 95%CI=0.26-1.05, P=0.071) in the validation set. A dose-response effect of increasing risk of developing complications with increasing number of unfavorable genotypes each patient carried was identified in the discovery population. Those with the highest number of unfavorable genotypes had an 8.82-fold increase (95% CI= 2.58-30.18) in risk compared to those with the lowest number. This effect was replicated in the validation population (P&amp;lt;0.001). A significant improvement in discriminatory ability was observed with the addition of the four genetic loci identified in our study to the epidemiological/clinical-based receiver operating characteristic curve in predicting the risk of developing complications. The area under the curve (AUC) was increased 8% from 0.68 for the initial model to 0.76 with the genetic variables. Bootstrap resampling internally validated the significance of this AUC difference (ΔAUC=0.024-0.129) and this improvement was replicated in the validation dataset. Conclusions: We identified genetic variations in the JAK-STAT pathway associated with risk of developing postoperative pulmonary complications in NSCLC patients. These biomarkers could be translated into the clinic to assist in identifying surgical patients who are at higher risk of developing pulmonary complications. Citation Format: Xia Pu, Michelle A.T. Hildebrandt, Jae Y. Kim, Yuanqing Ye, Ara A. Vaporciyan, Arlene M. Correa, Stephen G. Swisher, Jack A. Roth, Xifeng Wu. Genetic variations in the JAK-STAT pathway are associated with postoperative pulmonary complications after lobectomy in lung cancer patients. [abstract]. In: Proceedings of the Eleventh Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2012 Oct 16-19; Anaheim, CA. Philadelphia (PA): AACR; Cancer Prev Res 2012;5(11 Suppl):Abstract nr B91.

Development and application of a bioindicator for benthic habitat enhancement in the North Carolina Piedmont

This paper describes the development, application, and evaluation of a method for assessing the effectiveness of stream restoration activities in enhancing four lotic habitats based on the presence of habitat specialists. Three genera were identified as specialists for indicators of the enhancement of woody debris, coarse bed substrate, fine roots, and leaf pack habitats. These indicator genera were determined for each habitat type through indicator species analysis, extensive literature review, and consultation with local experts and a statewide distribution database. Water quality influences were isolated by excluding taxa with low tolerance to degraded water quality conditions. The difference in the presence of indicator genera between pairs of upstream-restored reaches was used to evaluate the success of the restoration activities in re-establishing benthic habitats. Application of this methodology to 27 paired reaches in the North Carolina Piedmont indicated that no change in specialists was the most frequent result of restoration, particularly for the woody debris habitats, when each habitat was examined individually. By combining the habitats into a composite score, a distinction by land use emerged, with habitats in urban areas indicating the greatest enhancement, while presence of the indicator genera at the agricultural and rural sites showed no clear trend of improvement or degradation in response to the restoration activities. When this composite IG metric was compared to the EPT taxa richness metric and RBP scores, the dependency of the EPT taxa richness metric on upstream conditions and the improvement in discriminatory ability over the RBP score suggest that this indicator genera (IG) metric provides a distinct signal for representing the biological perspective on the enhancement of benthic habitats by stream restoration activities. While further development of the methodology is desirable, this framework introduces a valuable alternative for evaluating benthic habitat enhancement in various hydrogeographic and land use conditions, and is constructive for guiding restoration designs to maximize biotic integrity.

Minimal approximate hitting sets and rule templates

A set S that has a non-empty intersection with every set in a collection of sets C is called a hitting set of C . If no elements can be removed from S without violating the hitting set property, then we say that S is minimal. Several interesting problems can in part be formulated as that of having to find one or more minimal hitting sets. Many of these problems require proper solutions, but sometimes approximate solutions suffice. We define an r-approximate hitting set as a set that intersects at least a fraction r of the sets in C . This notion is extended to the case, where C is a weighted multiset, and properties of r are explored with respect to simplification of C by absorption of supersets. Also, approximations of reducts from rough set theory are defined by means of minimal r-approximate hitting sets, and some links to the notion of dynamic reducts are established. The most common use of reducts are as templates for the generation of minimal classification rules from empirical data. A genetic algorithm is devised to compute r-approximate hitting sets, and applied to induce classifiers from selected real-world databases. These classifiers are then compared to those generated from proper reducts and dynamic reducts, respectively. The improvement in discriminatory ability yielded by the r-approximate reduct based classifiers was statistically significant ( p<0.05). Furthermore, they were smaller, i.e., had fewer rules.