Assessing inter-rater reliability when the raters are fixed: two concepts and two estimates.

Intraclass correlation (ICC) is an established tool to assess inter-rater reliability. In a seminal paper published in 1979, Shrout and Fleiss considered three statistical models for inter-rater reliability data with a balanced design. In their first two models, an infinite population of raters was considered, whereas in their third model, the raters in the sample were considered to be the whole population of raters. In the present paper, we show that the two distinct estimates of ICC developed for the first two models can both be applied to the third model and we discuss their different interpretations in this context.

Decision curve analysis revisited: overall net benefit, relationships to ROC curve analysis, and application to case-control studies.

ABSTRACT: BACKGROUND: Decision curve analysis has been introduced as a method to evaluate prediction models in terms of their clinical consequences if used for a binary classification of subjects into a group who should and into a group who should not be treated. The key concept for this type of evaluation is the "net benefit", a concept borrowed from utility theory. METHODS: We recall the foundations of decision curve analysis and discuss some new aspects. First, we stress the formal distinction between the net benefit for the treated and for the untreated and define the concept of the "overall net benefit". Next, we revisit the important distinction between the concept of accuracy, as typically assessed using the Youden index and a receiver operating characteristic (ROC) analysis, and the concept of utility of a prediction model, as assessed using decision curve analysis. Finally, we provide an explicit implementation of decision curve analysis to be applied in the context of case-control studies. RESULTS: We show that the overall net benefit, which combines the net benefit for the treated and the untreated, is a natural alternative to the benefit achieved by a model, being invariant with respect to the coding of the outcome, and conveying a more comprehensive picture of the situation. Further, within the framework of decision curve analysis, we illustrate the important difference between the accuracy and the utility of a model, demonstrating how poor an accurate model may be in terms of its net benefit. Eventually, we expose that the application of decision curve analysis to case-control studies, where an accurate estimate of the true prevalence of a disease cannot be obtained from the data, is achieved with a few modifications to the original calculation procedure. CONCLUSIONS: We present several interrelated extensions to decision curve analysis that will both facilitate its interpretation and broaden its potential area of application.

Essays in equilibrium asset pricing

Introduction This dissertation consists of three essays in equilibrium asset pricing. The first chapter studies the asset pricing implications of a general equilibrium model in which real investment is reversible at a cost. Firms face higher costs in contracting than in expanding their capital stock and decide to invest when their productive capital is scarce relative to the overall capital of the economy. Positive shocks to the capital of the firm increase the size of the firm and reduce the value of growth options. As a result, the firm is burdened with more unproductive capital and its value lowers with respect to the accumulated capital. The optimal consumption policy alters the optimal allocation of resources and affects firm's value, generating mean-reverting dynamics for the M/B ratios. The model (1) captures convergence of price-to-book ratios -negative for growth stocks and positive for value stocks - (firm migration), (2) generates deviations from the classic CAPM in line with the cross-sectional variation in expected stock returns and (3) generates a non-monotone relationship between Tobin's q and conditional volatility consistent with the empirical evidence. The second chapter proposes a standard portfolio-choice problem with transaction costs and mean reversion in expected returns. In the presence of transactions costs, no matter how small, arbitrage activity does not necessarily render equal all riskless rates of return. When two such rates follow stochastic processes, it is not optimal immediately to arbitrage out any discrepancy that arises between them. The reason is that immediate arbitrage would induce a definite expenditure of transactions costs whereas, without arbitrage intervention, there exists some, perhaps sufficient, probability that these two interest rates will come back together without any costs having been incurred. Hence, one can surmise that at equilibrium the financial market will permit the coexistence of two riskless rates that are not equal to each other. For analogous reasons, randomly fluctuating expected rates of return on risky assets will be allowed to differ even after correction for risk, leading to important violations of the Capital Asset Pricing Model. The combination of randomness in expected rates of return and proportional transactions costs is a serious blow to existing frictionless pricing models. Finally, in the last chapter I propose a two-countries two-goods general equilibrium economy with uncertainty about the fundamentals' growth rates to study the joint behavior of equity volatilities and correlation at the business cycle frequency. I assume that dividend growth rates jump from one state to other, while countries' switches are possibly correlated. The model is solved in closed-form and the analytical expressions for stock prices are reported. When calibrated to the empirical data of United States and United Kingdom, the results show that, given the existing degree of synchronization across these business cycles, the model captures quite well the historical patterns of stock return volatilities. Moreover, I can explain the time behavior of the correlation, but exclusively under the assumption of a global business cycle.

An Analysis of the Benefit of Using HEV Genotype 3 Antigens in Detecting Anti-HEV IgG in a European Population

Introduction Quatre génotypes pathogènes de l'hépatite E (HEV) sont actuellement connus. Ils présentent des caractéristiques épidémiologiques différentes. Les génotypes 1 et 2 infectent uniquement l'homme et sont à l'origine d'épidémies dans des pays en voie de développement. Les génotypes 3 et 4 se présentent sous forme de zoonose, endémiques chez des cochons et autres mammifères dans des pays industrialisés. Ces derniers génotypes sont à l'origine de cas sporadiques d'hépatite E autochtones. La majorité des tests de sérologie actuellement commercialisés se basent sur des virus de génotype 1 et 2. Le bénéfice de l'utilisation d'un test sérologique basé sur le génotype 3 dans des pays industrialisés n'a pas été étudié jusqu'à présent. Dans cette étude, les performances de tests sérologiques basés sur des antigènes de plusieurs génotypes de l'HEV ont été comparées. Méthode Les tests ont été appliqués à deux populations distinctes: une population de 20 patients, chez qui une infection aiguë d'hépatite E, génotype 3, a été documentée par PCR sanguine, et une population de 550 donneurs de sang de la région de Lausanne. Le dépistage des IgGs anti-HEV a été effectué dans le sérum des deux populations par trois «Enzyme Immuno Assays» (EIA) à savoir MP Diagnostics, Dia.Pro et Fortress. Les échantillons positifs avec au moins un des EIA ont été testés par un «Immunodot Assay», le recomLine HEV IgG/IgM. Tous les EIA sont basés sur des antigènes des génotypes 1 et 2, alors que l'immunodot se base sur des antigènes des génotypes 1 et 3. Résultats Tous les échantillons des cas d'hépatite E documentés et 124 sur 550 échantillons des donneurs de sang étaient positifs avec au moins un des tests sérologique. Parmi les cas confirmés par PCR, 45 %, 65 %, 95 % et 55 % étaient respectivement positifs avec le test de MP Diagnostics, Dia.Pro, Fortress et recomLine. Parmi les échantillons positifs des donneurs de sang avec au moins un des tests, 120/124 (97 %) étaient positifs avec le test Fortress, 19/124 (15 %) étaient positifs avec tous les EIA et 51/124 (41 %) étaient positifs avec le recomLine. Parmi les cas d'hépatite E confirmés, 11/20 (55 %) étaient positifs avec le recomLine et parmi ceux-ci, une réactivité plus forte pour le génotype 3 était observée dans 1/11 (9 %) et une réactivité identique dans 5/11 (45.5 %) cas. Conclusions Même si le recomLine contient des protéines dérivées de l'HEV génotype 3, sa sensibilité est inférieure à l'EIA de Fortress dans les cas d'hépatite E aiguë de génotype 3. De plus, chez environ 45 % des patients, le recomLine ne parvient pas à identifier une infection comme étant causé par un virus du génotype 3. Dans la population de donneurs de sang, nous avons observe de grandes variations dans les séroprévalences mesurées, allant de 4.2 % à 21.8 % selon les tests sérologiques employés.

Populations with elevated mutation load do not benefit from the operation of sexual selection.

Theory predicts that if most mutations are deleterious to both overall fitness and condition-dependent traits affecting mating success, sexual selection will purge mutation load and increase nonsexual fitness. We explored this possibility with populations of mutagenized Drosophila melanogaster exhibiting elevated levels of deleterious variation and evolving in the presence or absence of male-male competition and female choice. After 60 generations of experimental evolution, monogamous populations exhibited higher total reproductive output than polygamous populations. Parental environment also affected fitness measures - flies that evolved in the presence of sexual conflict showed reduced nonsexual fitness when their parents experienced a polygamous environment, indicating trans-generational effects of male harassment and highlighting the importance of a common garden design. This cost of parental promiscuity was nearly absent in monogamous lines, providing evidence for the evolution of reduced sexual antagonism. There was no overall difference in egg-to-adult viability between selection regimes. If mutation load was reduced by the action of sexual selection in this experiment, the resultant gain in fitness was not sufficient to overcome the costs of sexual antagonism.