A randomized phase I Bayesian dose escalation design for the combination of anti‐cancer drugs

Nowadays, treatment regimens for cancer often involve a combination of drugs. The determination of the doses of each of the combined drugs in phase I dose escalation studies poses methodological challenges. The most common phase I design, the classic ‘3+3' design, has been criticized for poorly estimating the maximum tolerated dose (MTD) and for treating too many subjects at doses below the MTD. In addition, the classic ‘3+3' is not able to address the challenges posed by combinations of drugs. Here, we assume that a control drug (commonly used and well‐studied) is administered at a fixed dose in combination with a new agent (the experimental drug) of which the appropriate dose has to be determined. We propose a randomized design in which subjects are assigned to the control or to the combination of the control and experimental. The MTD is determined using a model‐based Bayesian technique based on the difference of probability of dose limiting toxicities (DLT) between the control and the combination arm. We show, through a simulation study, that this approach provides better and more accurate estimates of the MTD. We argue that this approach may differentiate between an extreme high probability of DLT observed from the control and a high probability of DLT of the combination. We also report on a fictive (simulation) analysis based on published data of a phase I trial of ifosfamide combined with sunitinib.Copyright © 2014 John Wiley & Sons, Ltd.     

On the distribution of some ordered variables

Abstract

Motivated by a recent article published by Adam and Tawn, we characterize the distribution of two random variables X, Y ordered linearly like X < Y. We suppose that the random variables follow a bivariate extreme value distribution.     

INCOME AND WEALTH CONCENTRATION IN SPAIN FROM A HISTORICAL AND FISCAL PERSPECTIVE

This paper presents series on top shares of income and wealth in Spain using personal income and wealth tax return statistics. Top income shares are highest in the 1930s, fall sharply during the first decade of the Franco dictatorship, then remain stable and low till the 1980s, and have increased since the mid 1990s. The top 0.01% income share in Spain estimated from income tax data is comparable to estimates for the United States and France over the period 1933–1971. Those findings, along with a careful analysis of all published tax statistics, suggest that income tax evasion and avoidance among top income earners in Spain was much less prevalent than previously thought. Wealth concentration has been about stable from 1982 to 2005 as surging real estate prices have benefited the middle class and compensated for a slight increase in financial wealth concentration in the 1990s. We use our wealth series and a simple model to analyze the effects of the wealth tax exemption of stocks for owners‐managers introduced in 1994. We show that the reform induced substantial shifting from the taxable to tax exempt status, hence creating efficiency costs. (JEL D31, H31, O15)     

"Killed by its mother": infanticide in Providence County, Rhode Island, 1870 to 1938

This article analyzes infanticide based on the Coroners' Records for Providence County, Rhode Island, from the 1870s to 1938 to determine doctors' and coroners' attitudes toward mothers who killed. The nineteenth century witnessed a medical discourse on the possibility of postpartum insanity as a cause of infanticide. While some women claimed temporary insanity, and some doctors and coroners legitimated this defense, its application to mothers who killed was arbitrary. They determined who deserved this diagnosis based on the woman's character, her forthrightness, and extenuating circumstances. Infanticide divided the profession nationally and at the local level and prevented doctors or coroners from speaking in a united voice on the issue. This article does not attempt to follow cases of infanticide through to jury verdicts. Instead, it provides an opportunity to analyze the circumstances women faced that led them to kill their newborns, and to analyze the responses of doctors and coroners to these mothers who killed. Unlike the findings of other studies, neither physicians nor coroners in Rhode Island were united in a claim of ignorance to save these women from guilty verdicts.     

Too Big to Learn: The Effects of Major Acquisition Failures on Subsequent Acquisition Divestment

We examine whether firms learn from their major acquisition failures. Drawing from a threat‐rigidity theoretical framework, we suggest that firms do not learn from their major acquisition failures. Furthermore, we hypothesize that host‐country experience reinforces the negative effects of major acquisition failures. Our research hypotheses are tested using an event history analysis of 741 acquisitions undertaken by French listed and non‐listed firms in the USA between January 1988 and December 2008. We use failure divestment (divestment resulting from acquisition failure) as a proxy for acquisition performance. Consistent with our theoretical framework, we find that major acquisition failures have a negative impact on future acquisition performance. Furthermore, we find that such negative effects are reinforced by firms’ host‐country experience.     

A Simplified Approach to Risk Assessment Based on System Dynamics: An Industrial Case Study

Seveso plants are complex sociotechnical systems, which makes it appropriate to support any risk assessment with a model of the system. However, more often than not, this step is only partially addressed, simplified, or avoided in safety reports. At the same time, investigations have shown that the complexity of industrial systems is frequently a factor in accidents, due to interactions between their technical, human, and organizational dimensions. In order to handle both this complexity and changes in the system over time, this article proposes an original and simplified qualitative risk evaluation method based on the system dynamics theory developed by Forrester in the early 1960s. The methodology supports the development of a dynamic risk assessment framework dedicated to industrial activities. It consists of 10 complementary steps grouped into two main activities: system dynamics modeling of the sociotechnical system and risk analysis. This system dynamics risk analysis is applied to a case study of a chemical plant and provides a way to assess the technological and organizational components of safety.     

How Reassuring are Risk Comparisons to Pollution Standards and Emission Limits?

Pollutant emissions or ambient levels are often justified by reference to a higher benchmark, such as a public health standard or permit limit. However, does this risk comparison persuade the public audiences to whom it is frequently directed that such pollution levels are “acceptable”? A substantial proportion of people living within one mile of New Jersey's industrial facilities, perhaps as much as half, is indeed reassured by a comparison to such benchmarks. Positive attitudes toward discharge limits were linked to speaking English at home; positive attitudes toward drinking water standards were associated with seeing local benefits of industry as outweighing its risks, not speaking English, and relative youth (49 years old or less). Three scenarios involving drinking water contamination varied pre‐ and post‐treatment levels of the pollutant, though all post‐treatment levels were below the standard. In all cases great concern was expressed, with no significant differences across scenarios; net benefits, being white, and non‐English speaking were linked to lower concern. Relative trust seems a better explanation of different attitudes toward benchmark comparisons than varying levels of knowledge, but both hypotheses have drawbacks that merit further testing. These results imply that people with negative views of industry or government, the ones officials might most wish to reassure, will tend to doubt that regulatory benchmarks offer a valid risk comparison.     

An R Package for Value at Risk and Expected Shortfall

Value at risk and expected shortfall are the two most popular measures of financial risk. But the available R packages for their computation are limited. Here, we introduce an R contributed package written by the authors. It computes the two measures for over 100 parametric distributions, including all commonly known distributions. We expect that the R package could be useful to researchers and to the financial community.     

Nonparametric Estimation of the Probability of Detection of Flaws in an Industrial Component,from Destructive and Nondestructive Testing Data,Using Approximate Bayesian Computation

We consider the problem of estimating the probability of detection (POD) of flaws in an industrial steel component. Modeled as an increasing function of the flaw height, the POD characterizes the detection process; it is also involved in the estimation of the flaw size distribution, a key input parameter of physical models describing the behavior of the steel component when submitted to extreme thermodynamic loads. Such models are used to assess the resistance of highly reliable systems whose failures are seldom observed in practice. We develop a Bayesian method to estimate the flaw size distribution and the POD function, using flaw height measures from periodic in‐service inspections conducted with an ultrasonic detection device, together with measures from destructive lab experiments. Our approach, based on approximate Bayesian computation (ABC) techniques, is applied to a real data set and compared to maximum likelihood estimation (MLE) and a more classical approach based on Markov Chain Monte Carlo (MCMC) techniques. In particular, we show that the parametric model describing the POD as the cumulative distribution function (cdf) of a log‐normal distribution, though often used in this context, can be invalidated by the data at hand. We propose an alternative nonparametric model, which assumes no predefined shape, and extend the ABC framework to this setting. Experimental results demonstrate the ability of this method to provide a flexible estimation of the POD function and describe its uncertainty accurately.     

Écrire la loi. Un travail de bureau pour hauts fonctionnaires du ministère de l’Écologie

This article analyses the process of the drafting of the Bachelot Bill on technological risks, passed on 30 July 2003 but initiated by Lionel Jospin's government immediately following the AZF factory disaster in September 2001. It focuses on the practical work of the civil servants at the Ministry of Ecology responsible for setting the orientations of the reform, transcribing them into legal provisions and ensuring that they would be supported and passed in Parliament. The analysis approaches the task by tracking both the successive versions of the bill and the multiple documents used in the preparation of these versions. This immersion in the administrative production of law provides a better understanding of the relations between the civil service and the political authorities, and shows how the boundary between these two worlds of practice is difficult to draw in the legal drafting process.