Why a Bayesian approach to safety analysis in pharmacovigilance is important

Large databases of routinely collected data are a valuable source of information for detecting potential associations between drugs and adverse events (AE). A pharmacovigilance system starts with a scan of these databases for potential signals of drug-AE associations that will subsequently be examined by experts to aid in regulatory decision-making. The signal generation process faces some key challenges: (1) an enormous volume of drug-AE combinations need to be tested (i.e. the problem of multiple testing); (2) the results are not in a format that allows the incorporation of accumulated experience and knowledge for future signal generation; and (3) the signal generation process ignores information captured from other processes in the pharmacovigilance system and does not allow feedback. Bayesian methods have been developed for signal generation in pharmacovigilance, although the full potential of these methods has not been realised. For instance, Bayesian hierarchical models will allow the incorporation of established medical and epidemiological knowledge into the priors for each drug-AE combination. Moreover, the outputs from this analysis can be incorporated into decision-making tools to help in signal validation and posterior actions to be taken by the regulators and companies. We discuss in this paper the apparent advantage of the Bayesian methods used in safety signal generation and the similarities and differences between the two widely used Bayesian methods. We will also propose the use of Bayesian hierarchical models to address the three key challenges and discuss the reasons why Bayesian methodology still have not been fully utilised in pharmacovigilance activities.     

A Characterization of the Normal Family of Distributions Based on the Bias of the Maximum Likelihood Estimator

This paper characterizes the family of Normal distributions within the class of exponential families of distributions, via the structure of the bias of the maximum likelihood estimator Θ _n of the canonical parameter Θ . More specifically, when E _θ ( Θ n ) – Θ = (1/ n ) Q ( Θ ) + o (1/ n ), the equality Q ( Θ ) = 0 proves to be a property of the Normal distribution only. The same conclusion is obtained for the one-dimensional case bt assuming that Q ( Θ ) is a polynomial of Θ .     

An investigation of youths' attributions for unemployment and their political attitudes

This study investigated some hypotheses concerning the determinants of attributions for unemployment and the psychological consequences of unemployment in a randomly selected sample of 204 male school leavers. There was weak evidence for the hypothesis that the unemployed make more external attributions for unemployment, but none showing them to make less internal attributions. There was weak evidence that political party preference was related to internal and external attributions, and stronger evidence that a general tendency to see life as determined by chance was linked to external attributions. The unemployed were found to be more likely to support violent political action, to support the Labour Party, and to believe chance controlled their lives. Finally, external attributions were seen as more important than internal ones by all respondents. The results suggest that other factors, e.g. the mass media, may be influential in determining the nature of attributions and this is interpreted in the context of Moscovici's social representation theory.     

Analytic approaches to problems of a middle-sized defence department

The application of operational research to the problems of Canadian defence was very similar to that in Great Britain and the United States during World War II and through the 1950s. But because of the rapidly increasing costs of major weapon systems, since the 1960s systems analysis has been very thoroughly applied to the study of the large procurement programs, including far-reaching examination of their economic impact on the nation. Other problems of special interest to Canada which received analytical attention included peacekeeping and military involvement in activities such as maritime surveillance and search and rescue. Analytical studies have been extended into examination of logistics and manpower, and into economic, sociological, and strategic problems of concern to the Canadian defence department, including arms control. This type of research represents a considerable excursion from the original operational research, but is needed in the Canadian Department of National Defence.     

Cultural Analysis of Religious Change and Movements*

In this article I review a cultural perspective on religion and suggest that cultural analysis resolves current debates over the nature of religion as either collectivist or individualist. I use one type of cultural analysis, institutionalism, to present an interpretive overview of religious change and movements in historical contexts of global instrumental rationality, in particular the expansion of state authority. The usefulness of this approach is revealed in interpreting Protestantism in the United States, Islamic fundamentalism, and movements and trends in global Roman Catholicism. While not reflecting precisely the views of the authors of this collection, this article introduces the studies of the recent restructuring of religion in the United States (Robert Wuthnow), Islamic fundamentalisms in Iran and Syria (Mansoor Moaddel). and global Roman Catholicism (Jose Casanova).     

Talcott Parsons'Marshall Lectures: Contemporay, But Flawed

The occasion of the publication of Parsons'Marshall Lectures is used to demonstrate Parsons'prescience and the contemporary relevance of his ideas. It is argued that these lectures are similar in intent to the recent work of Amitai Etzioni and his objective of creating the field of socio-economics. The two works are compared and their relative strengths and weaknesses discussed. Parsons is shown to be well ahead of his time, not only in his effort to integrate social and economic concerns, but in the broader integrative intent of his overall theory. While Parsons was not able to begin to engineer a rapprochement with economics, such an intellectual integration may be more possible today because economics appears to be in greater crisis and sociology seems to have more to offer to disaffected economists.     

TOWARD A REVISED FRAMEWORK FOR EXAMINING BELIEFS ABOUT THE CAUSES OF POVERTY

This study uses survey data from Baltimore, Maryland to assess predictions from the dominant ideology thesis and the public arenas theory concerning causal beliefs about three specific types of poverty—welfare dependency, homelessness and impoverished migrant laborers. Findings indicate that the incidence and determinants of causal beliefs are more complex than has been reached in prior "generic' poverty research and the findings provide greater support for the public arenas theory. Three findings are of particular note: (1) sample respondents have distinctive causal beliefs for different types of poverty: for welfare dependency, individualistic beliefs are dominant; for homelessness, structural causes are emphasized; a causal "middle ground' is most popular for impoverished migrant laborers (2) status characteristics do not operate as determinants of causal beliefs in a straightfor-ward fashion for any of the three poverty types and, (3) variables that measure types of exposure to, and perceived racial composition of, the poor are also significant determinants of causal beliefs. The effects of variables that measure perceived racial composition are particularly strong and the pattern of "color coding' suggests that racial prejudice shapes causal beliefs about the plight of the poor. The implications of the findings for the mobility opportunities of the three impoverished groups are discussed.     

Longitudinal non-euclidean networks: Applying Galileo

This article discusses the theoretical utility of using a non-Euclidean spatial manifold when describing social networks. It proposes that a variant of metric MDS, the Galileo System, can be particularly useful in analyzing social networks and their changes over time, partially because it does not impose Euclidean assumptions on the data. Two sets of longitudinal network data are examined with Galileo. One is the American air traffic network from 1968–1981. The other is ten groups engaged in a computer conference over a 24 month period. In both cases, the results indicate that a Riemannian spatial manifold is required to describe the network structure. Consistent theoretically valid results based upon the non-Euclidean components of spatial manifold are obtained. Further, they could be readily explained by exogenous factors. The implications of these results for network analysis are discussed.