Missing data: Discussion points from the PSI missing data expert group

The Points to Consider Document on Missing Data was adopted by the Committee of Health and Medicinal Products (CHMP) in December 2001. In September 2007 the CHMP issued a recommendation to review the document, with particular emphasis on summarizing and critically appraising the pattern of drop‐outs, explaining the role and limitations of the ‘last observation carried forward’ method and describing the CHMP's cautionary stance on the use of mixed models. In preparation for the release of the updated guidance document, statisticians in the Pharmaceutical Industry held a one‐day expert group meeting in September 2008. Topics that were debated included minimizing the extent of missing data and understanding the missing data mechanism, defining the principles for handling missing data and understanding the assumptions underlying different analysis methods. A clear message from the meeting was that at present, biostatisticians tend only to react to missing data. Limited pro‐active planning is undertaken when designing clinical trials. Missing data mechanisms for a trial need to be considered during the planning phase and the impact on the objectives assessed. Another area for improvement is in the understanding of the pattern of missing data observed during a trial and thus the missing data mechanism via the plotting of data; for example, use of Kaplan–Meier curves looking at time to withdrawal. Copyright © 2009 John Wiley & Sons, Ltd.     

Gambling decisions: an early intervention program for problem gamblers

An implementation and one-year follow-up of the Gambling Decisions program attempted to answer several important questions. First, is controlled gambling a viable treatment option for some gamblers? Can earlier stage problem gamblers be separated for treatment from those with more severe problems? Finally, would problem gamblers utilize a community health agency for treatment of their excessive gambling? A pretest/posttest design was chosen where the efficacy of the program was assessed using repeated measures ANOVA analysis. Results showed that an average loss of $608 over a 4-week period was reduced to $113 immediately after the 6-week program and to a loss of $73 at 12 months. The average number of hours spent gambling per 4 weeks was significantly reduced from 23.5 at pretest to 6.5 at the 12 month posttest. Significant decreases were also observed in the number of days per week that clients gambled, and clients reported significant reductions in everyday life problems related to gambling after completing the program.     

Career story: Consultant Forensic Statistician

Ian Evett works for the Forensic Science Service as a consultant statistician. He tells Helen Joyce about the contribution a statistician can make to analysing handwriting, glass, DNA and other forensic materials, as well as to understanding the fundamentals of evidence interpretation.     

Creating pathways to participation: a community-based developmental prevention project in Australia

Pathways to Prevention is a developmental prevention project focused on the transition to school in a disadvantaged multicultural urban area in Queensland. The project incorporates two elements: The Preschool Intervention Program (PIP) promotes communication and social skills related to school success; and the Family Independence Program (FIP) (parent training, facilitated playgroups, support groups, etc) promotes family capacity to foster child development. Using a quasi-experimental design (N= 597), improvements in boys' but not girls' behaviours over the preschool year were found. FIP reached more than a quarter of the target population, including many difficult-to-reach families experiencing high stress. Case studies and other qualitative data suggest positive outcomes.     

Trust: Missing Piece(s) in the Safety Puzzle

While trust is increasingly recognized as a factor that impacts on safety behavior, the exact nature of trust and its role in shaping organizational safety is poorly understood. This special issue contains six articles that examine the relationship between trust and safety behavior in a range of high-risk work contexts. The issue begins with two articles that introduce the complex nature of trust and the positive and negative roles that trust can play in shaping an organization's safety culture. This background is then developed by two articles that explore the role of trust and distrust in safety performance, and uncover a range of significant but often counterintuitive relationships between forms of trust and safe behavior. Finally, the issue concludes with two articles that examine the role that leadership may play in developing trust. These articles examine the conditions important for the development of trust in leaders, and the trust-promoting actions that leaders can employ to influence employees' engagement in safety participation.     

Three Big Breakthroughs for Social Research

The past two years have brought three developments almost certain to change radically the way social scientists and policy analysts view the conduct of their enterprise. These have been the release of samples of data collected by the ABS as unit records on magnetic tape, major modifications to the Statistical Package for the Social Sciences and the development of the completely new Australian Standard Classification of Occupations. The impact of each individually will be enormous, but the extraordinarily complementary nature of these almost simultaneous innovations is emphasised to demonstrate that together this impact will be revolutionary.     

Identifying Dangerous Trucking Firms

The paper develops a statistical procedure for predicting the safety performance of motor carriers based on characteristics of firms and results of two government safety enforcement programs. One program is an audit of management safety practices, and the other is a program to inspect drivers and vehicles at the roadside for compliance with safety regulations. The technique can be used to provide safety regulators with an empirical approach to identify the most dangerous firms and provide a priority list of firms against which educational and enforcement actions should be initiated. The government needs to use such an approach rather than directly observing accident rates because the most dangerous firms are generally small and, despite relatively high accident rates, accidents remain rare events. The technique uses negative-binomial regression procedures on a dataset of 20,000 firms. The definition of poor performance in roadside inspection is based on both the rate of inspections per fleet mile and the average number of violations found during an inspection. This choice was made because selection for inspection has both a random and nonrandom component. The results of the study suggest that both of the government's safety programs help identify the most dangerous firms. The 2.5% of firms that do poorly in both programs have an average accident rate twice that of the mean for all other firms.     

Tests for Comparing Mark-Specific Hazards and Cumulative Incidence Functions

It is of interest in some applications to determine whether there is a relationship between a hazard rate function (or a cumulative incidence function) and a mark variable which is only observed at uncensored failure times. We develop nonparametric tests for this problem when the mark variable is continuous. Tests are developed for the null hypothesis that the mark-specific hazard rate is independent of the mark versus ordered and two-sided alternatives expressed in terms of mark-specific hazard functions and mark-specific cumulative incidence functions. The test statistics are based on functionals of a bivariate test process equal to a weighted average of differences between a Nelson-Aalen-type estimator of the mark-specific cumulative hazard function and a nonparametric estimator of this function under the null hypothesis. The weight function in the test process can be chosen so that the test statistics are asymptotically distribution-free. Asymptotically correct critical values are obtained through a simple simulation procedure. The testing procedures are shown to perform well in numerical studies, and are illustrated with an AIDS clinical trial example. Specifically, the tests are used to assess if the instantaneous or absolute risk of treatment failure depends on the amount of accumulation of drug resistance mutations in a subject's HIV virus. This assessment helps guide development of anti-HIV therapies that surmount the problem of drug resistance.     

The Child Client

In court cases involving children the child is often a party to proceedings and is thus entitled to Legal Aid and representation by a solicitor. This representation of the child happens not just in youth courts under the Criminal Justice Act 1991 but also in Family Courts in proceedings under the Children Act 1989. This article looks at how a solicitor ascertains whether a child client is competent to give instructions and offers information on the techniques employed by the solicitor in communicating with the child, especially in proceedings under the Children Act.