A new parametric method based on S-distributions for computing receiver operating characteristic curves for continuous diagnostic tests

The effect of the area proportion of the metastatic lesion within the central metastatic lymph node on response to therapy in papillary thyroid carcinoma

Lymph node (LN) metastasis has been strongly associated with locoregional recurrence and decreased survival time of patients with papillary thyroid carcinoma (PTC). Although the characteristics of the metastatic LNs (mLN) have been determined, including size, number, micro-metastasis and extra-nodal extension (ENE), further analysis is warranted. The present study introduced a new parameter known as the area proportion of the metastatic lesion within the central mLNs (APmCLN). The objective was to evaluate the impact of the APmCLN on response to therapy in patients with PTC. In total, 355 patients with PTC treated with total thyroidectomy and neck dissection, post-operative radioactive iodine and thyroid-stimulating hormone suppression were retrospectively studied. The patients were classified into two groups: Group A (APmCLN ≤75%) and group B (APmCLN >75%). The association of various clinicopathological characteristics between these two groups was investigated. Univariate and multivariate analyses were used to evaluate risk factors associated with a non-Excellent response to therapy and recurrence-free survival (RFS). The analysis showed that APmCLN >75% was significantly associated with extra-thyroidal extension, clinically apparent nodes (cN1), pathological N1b (pN1b), ENE, greater number and larger size of central mLN and larger size of the central LN metastatic lesion. Furthermore, it was reported that chronic lymphocytic thyroiditis, larger central mLN size and APmCLN >75% were independent risk factors for a non-excellent response to therapy. Finally, it was determined that the rate of excellent response to therapy was significantly higher in pathological N1 (pN1) patients with APmCLN ≤75% (108/144, 75.0%) compared with patients with APmCLN >75% (27/47, 57.4%) (P=0.022). However, there was no significant difference (P=0.247) between patients with APmCLN ≤75% and pN0 (132/164, 80.5%). RFS was 89.4% in patients with pN1-APmCLN >75%, whereas those with pN1-APmCLN ≤75% and pN0 did not experience a relapse. Patients with PTC with APmCLN >75% should be regarded as high-risk and may require more aggressive treatment and careful follow-up.

Likelihood Ratios as Value Proposition for Diagnostic Laboratory Tests

The clinical and health economic value of clinical laboratory diagnostics has been debated increasingly in recent years without resulting in practical recommendations for measuring the effectiveness of diagnostic tests. One way to achieve such a goal could be to enrich the mere data of laboratory test results with additional information about their likelihood ratios for diagnosis. The diagnostic significance of test results can be judged subjectively based on the experience of the treating physician or expressed objectively in the form of likelihood ratios. The provision of likelihood ratios by the laboratory would increase the impact of laboratory diagnostics in healthcare and thus have positive economic value. Consequently, likelihood ratios should be taken into account in reimbursement strategies.

An Integrated Bayesian Nonparametric Approach for Stochastic and Variability Orders in ROC Curve Estimation: An Application to Endometriosis Diagnosis

In estimating ROC curves of multiple tests, some a priori constraints may exist, either between the healthy and diseased populations within a test or between tests within a population. In this paper, we proposed an integrated modeling approach for ROC curves that jointly accounts for stochastic and variability orders. The stochastic order constrains the distributional centers of the diseased and healthy populations within a test, while the variability order constrains the distributional spreads of the tests within each of the populations. Under a Bayesian nonparametric framework, we used features of the Dirichlet process mixture to incorporate these order constraints in a natural way. We applied the proposed approach to data from the Physician Reliability Study that investigated the accuracy of diagnosing endometriosis using different clinical information. To address the issue of no gold standard in the real data, we used a sensitivity analysis approach that exploited diagnosis from a panel of experts. To demonstrate the performance of the methodology, we conducted simulation studies with varying sample sizes, distributional assumptions and order constraints. Supplementary materials for this article are available online.

Biochemical Systems Theory: A Review

Biochemical systems theory (BST) is the foundation for a set of analytical andmodeling tools that facilitate the analysis of dynamic biological systems. This paper depicts major developments in BST up to the current state of the art in 2012. It discusses its rationale, describes the typical strategies and methods of designing, diagnosing, analyzing, and utilizing BST models, and reviews areas of application. The paper is intended as a guide for investigators entering the fascinating field of biological systems analysis and as a resource for practitioners and experts.

A more flexible parametric estimation of univariate reference intervals: a new method based on the GS-distribution

BACKGROUND

Reference interval estimation is an important issue in clinical laboratories. Present methods are based either on data transformation or on non-parametric approaches.

METHODS

We present a new technique based in a family of statistical distributions known as GS-distributions that provide a suitable model for continuous unimodal variables. We compare, both by simulation studies an on actual data, the reference intervals estimated by using non-parametric methods and data transformations suggested by the IFCC and those obtained by fitting a GS-distribution. Simulated data are generated from various distributions to evaluate the accuracy of these methods. In each case, confidence intervals for the resulting reference intervals are obtained by bootstrap.

RESULTS

In all the cases, the GS-distribution based method provides comparable or more accurate results than the non-parametric methods. In most cases, the proposed method produces better results than those obtained by transforming the original data.

CONCLUSIONS

Our results suggest that the method for computing reference intervals based on GS-distribution is a valid alternative for the current non-parametric methods.

New scaphocephaly severity indices of sagittal craniosynostosis: a comparative study with cranial index quantifications

OBJECTIVE

To describe a novel set of scaphocephaly severity indices (SSIs) for predicting and quantifying head- and skull-shape deformity in children diagnosed with isolated sagittal synostosis (ISS) and compare their sensitivity and specificity with those of the traditional cranial index (CI).

METHODS

Computed tomography head scans were obtained from 60 patients diagnosed with ISS and 41 age-matched control patients. Volumetric reformations of the skull and overlying skin were used to trace two-dimensional planes defined in terms of skull-base plane and internal or surface landmarks. For each patient, novel SSIs were computed as the ratio of head width and length as measured on each of these planes. A traditional CI was also calculated and a receiver operating characteristic curve analysis was applied to compare the sensitivity and specificity of the proposed indices with those of CI.

RESULTS

Although the CI is a sensitive measure of scaphocephaly, it is not specific and therefore not a suitable predictor of ISS in many practical applications. The SSI-A provides a specificity of 95% at a sensitivity level of 98%, in contrast with the 68% of CI. On average, the sensitivity and specificity of all proposed indices are superior to those of CI.

CONCLUSIONS

Measurements of cranial width and length derived from planes that are defined in terms of internal or surface landmarks and skull-base plane produce SSIs that outperform traditional CI measurements.

Bayesian semi-parametric ROC analysis

This paper describes a semi-parametric Bayesian approach for estimating receiver operating characteristic (ROC) curves based on mixtures of Dirichlet process priors (MDP). We address difficulties in modelling the underlying distribution of screening scores due to non-normality that may lead to incorrect choices of diagnostic cut-offs and unreliable estimates of prevalence of the disease. MDP is a robust tool for modelling non-standard diagnostic distributions associated with imperfect classification of an underlying diseased population, for example, when a diagnostic test is not a gold standard. For posterior computations, we propose an efficient Gibbs sampling framework based on a finite-dimensional approximation to MDP. We show, using both simulated and real data sets, that MDP modelling for ROC curve estimation closely parallels the frequentist kernel density estimation (KDE) approach.

Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews

BACKGROUND AND OBJECTIVES

Studies of diagnostic accuracy most often report pairs of sensitivity and specificity. We demonstrate the advantage of using bivariate meta-regression models to analyze such data.

METHODS

We discuss the methodology of both the summary Receiver Operating Characteristic (sROC) and the bivariate approach by reanalyzing the data of a published meta-analysis.

RESULTS

The sROC approach is the standard method for meta-analyzing diagnostic studies reporting pairs of sensitivity and specificity. This method uses the diagnostic odds ratio as the main outcome measure, which removes the effect of a possible threshold but at the same time loses relevant clinical information about test performance. The bivariate approach preserves the two-dimensional nature of the original data. Pairs of sensitivity and specificity are jointly analyzed, incorporating any correlation that might exist between these two measures using a random effects approach. Explanatory variables can be added to the bivariate model and lead to separate effects on sensitivity and specificity, rather than a net effect on the odds ratio scale as in the sROC approach. The statistical properties of the bivariate model are sound and flexible.

CONCLUSION

The bivariate model can be seen as an improvement and extension of the traditional sROC approach.

PoPS: a computational tool for modeling and predicting protease specificity

Proteases play a fundamental role in the control of intra- and extra-cellular processes by binding and cleaving specific amino acid sequences. Identifying these targets is extremely challenging. Current computational attempts to predict cleavage sites are limited, representing these amino acid sequences as patterns or frequency matrices. Here we present PoPS, a publicly accessible bioinformatics tool (http://pops.csse.monash.edu.au/) that provides a novel method for building computational models of protease specificity, which while still being based on these amino acid sequences, can be built from any experimental data or expert knowledge available to the user. PoPS specificity models can be used to predict and rank likely cleavages within a single substrate, and within entire proteomes. Other factors, such as the secondary or tertiary structure of the substrate, can be used to screen unlikely sites. Furthermore, the tool also provides facilities to infer, compare and test models, and to store them in a publicly accessible database.