The statistician's role in the prevention of missing data

Considerable statistical research has been performed in recent years to develop sophisticated statistical methods for handling missing data and dropouts in the analysis of clinical trial data. However, if statisticians and other study team members proactively set out at the trial initiation stage to assess the impact of missing data and investigate ways to reduce dropouts, there is considerable potential to improve the clarity and quality of trial results and also increase efficiency. This paper presents a Human Immunodeficiency Virus (HIV) case study where statisticians led a project to reduce dropouts. The first step was to perform a pooled analysis of past HIV trials investigating which patient subgroups are more likely to drop out. The second step was to educate internal and external trial staff at all levels about the patient types more likely to dropout, and the impact this has on data quality and sample sizes required. The final step was to work collaboratively with clinical trial teams to create proactive plans regarding focused retention efforts, identifying ways to increase retention particularly in patients most at risk. It is acknowledged that identifying the specific impact of new patient retention efforts/tools is difficult because patient retention can be influenced by overall study design, investigational product tolerability profile, current standard of care and treatment access for the disease under study, which may vary over time. However, the implementation of new retention strategies and efforts within clinical trial teams attests to the influence of the analyses described in this case study. Copyright © 2012 John Wiley & Sons, Ltd.     

The quality and comparability of child care data in U.S. surveys

This paper examines the quality and comparability of child care data obtained from eight waves of data from four nationally representative data sources: the National Longitudinal Survey of Youth (1986 and 1988), the Survey of Income and Program Participation (1986, 1987, 1988, and 1990), the National Child Care Survey, and the National Survey of Families and Households. We examine whether different study designs and survey techniques for asking questions about child care produce similar results on both the levels and determinants of child care. We identified four main sources of difference in the data sets that could impact the quality and comparability of child care research: when the interview is conducted; screening questions used to determine who is asked about child care; the population of parents and children represented in the survey; and the way child care questions are asked. Our findings indicate that summer interviews and screening on mother's work status produce the largest differences in the levels and effects of child care across these studies. Even after removing the effects of summer interviews and screening questions, however, substantial differences exist across the studies.     

Settlement has many faces: physicians, attorneys and medical malpractice

We conduct an analysis of the jurisdictional dispute over the management of medical malpractice lawsuits, focusing on the process through which liability is defined. We utilize a North Carolina sample of physicians who have been sued, their defense counsel, and counsel for the plaintiff in the case. A comparison of the perspectives of these three parties reveals that over half of the physicians who settle perceive themselves as not liable. Defense counsel are more adept at predicting both negotiated resolutions and whether or not money will be paid than either plaintiffs' counsel or physicians. Almost two-thirds of physicians who thought they were not liable expressed a desire for vindication. Almost half the time when the physicians denied liability money was nonetheless paid to resolve the claim. Physician responses to the outcome of their cases focus on the need for reform, especially in terms of a call for peer or expert review. We identify and discuss culture conflict between law and medicine. For lawyers "settlement" is not a negative thing, but for physicians it implies fault. We challenge existing literature which analyzes the settlement of medical malpractice claims solely in terms of rational economic models, and we argue that social psychological variables are equally important.     

The significance of definitions of success in program evaluation

In program evaluation, determining whether a program is reaching its goals is key to evaluating program success. Having an understanding of the stated goals of each organization within a system rather than simply knowing the formal system goals can aid in program development and inform policy making. Evaluability assessments were used initially as a means of learning how Philadelphia delinquency prevention programs define success. Subsequent surveys of programs were conducted to ascertain whether any changes in definitions of success had occurred. Examination of programs' definitions of success over time has facilitated program development and policy making.     

THE CONSEQUENCES OF COSTLY DEFAULT

This paper investigates the consequences of assuming that default on loans or corporate debt is costly; that is, the act of default imposes a deadweight cost on the economy. The analysis deals with two simple capital market models. Conditions for capital market equilibrium nominal interest rates and probability of default to exist are given and the comparative statics of these equilibrium variables with respect to changes in expectations, productivity of investment, cost of default and riskless interest rates are examined. In many cases these comparative static results are unambiguous in sign.     

Synthetic-Fuel Plants: Potential Tumor Risks to Public Health

This article quantifies potential public health risks from tumor-producing pollutants emitted from two synthetic-fuel plants (direct liquefaction--Exxon Donor Solvent: and indirect liquefaction--Lurgi Fischer-Tropsch) located at a representative site in the eastern United States. In these analyses gaseous and aqueous waste streams were characterized; exposures via inhalation, terrestrial and aquatic food chains, and drinking water supplies were modeled. Analysis suggested that emissions of "polycyclic aromatic hydrocarbons," "aromatic amines," "neutral N, O, S heterocyclics," "nitriles," and "other trace elements" pose the largest quantifiable risks to public health. Data and analysis for these pollutant categories should be refined to more accurately match compound-specific estimated exposure levels with tumorigenic potency estimates. Before these results are used for regulatory purposes, more detailed analysis for selected pollutant classes are needed, and more sophisticated aquatic exposure models must be developed. Also, differences in geographic scales among the environmental transport models used need to be rectified.     

Portfolio analysis and the product life cycle

This paper explains how the product life cycle and Boston Consulting Group's portfolio matrix can be combined to provide a more comprehensive framework for strategic analysis. The integrated model is more powerful than either concept taken separately because it provides an exhaustive system for classifying a diverse assortment of business units or product/market categories.