Imperial College School of Medicine MRC PhD Studentship in Medical Statistics We are seeking an enthusiastic and highly motivated individual to join a prestigious and stimulating research environment centred on the Department of Medical Statistics & Evaluation. The department's focus includes leading-edge work in medical statistics, statistical computing, clinical trials and meta-analysis. Candidates are invited to propose projects in these areas, in addition to those listed below. Studentships will commence in October, 1998. MRC RESEARCH/COLLABORATIVE STUDENTSHIP PROJECTS 1. Methods for investigating the relationship between underlying risk and treatment benefit in meta-analysis. Supervisor: Professor Simon Thompson. PROJECT DESCRIPTION: The usual assumption in meta-analysis of clinical trials, that the extent of treatment benefit does not depend on the underlying risk of the patients in the different trials, is likely to be false in many situations. Indeed the nature of the relationship between underlying risk and treatment benefit will delineate which patients will benefit most, and which least, from treatment and will be of crucial importance in health economic assessments. Therefore robust methods to investigate such relationships are needed. We have recently shown that naive analyses, which for example use the observed risk of events in the control group of each trial as a measure of underlying risk, are seriously flawed for reasons related to regression to the mean. We have also developed a method, for treatment effects expressed as odds ratios, which avoids these biases and can be implemented in the Bayesian software BUGS. However the method is limited in a number of ways, and its extension is the purpose of this proposed research.The method needs to be developed to consider treatment effects on scales other than odds ratios, including relative risk, absolute risk difference, and number needed to treat (NNT). Applications to types of outcomes other than binary is also necessary, including means of continuous data (for example for analysing economic costs) and censored survival data. These extensions are not straightforward because of technical statistical issues, but are crucial for the methods to become fully useful in both epidemiological and clinical settings. These developments will require the use of various software, including BUGS and the multilevel package MLn, but one aim will be to produce accessible software that will enable applied researchers to undertake these analyses. Some very recent publications have also proposed other solutions than our own. Comparison between these methods will be undertaken, both empirically using data from existing meta-analyses and in terms of the different assumptions of the methods. Empirical research based on the Cochrane database of systematic reviews will be carried out to ascertain the extent to which meta-analyses show currently unrecognised relationships between underlying risk and treatment effects, and the implications that these have on the conclusions that should be drawn. In the cardiovascular field, available data sets include meta-analyses of antiplatelet therapy, fibrinolytic therapy, magnesium treatment, and of cholesterol reduction. Many other data sets are available through the Cochrane Collaboration. 2. Transformation in the analysis of hierarchical medical data, with focus on fetal monitoring. Supervisor: Professor Patrick Royston. PROJECT DESCRIPTION: Datasets with a hierarchical or multilevel structure are increasingly important in medicine. Examples include growth curves, cluster-randomised trials, multiperiod clinical studies and observational studies with repeated measures. The multilevel random-effects model is the analytical tool of choice. When one or more predictors are continuous, appropriate regression models are needed. Polynomials are almost invariably chosen, but they are often inadequate. The project will explore the use of fractional polynomials in multilevel modelling. These involve transformations of the predictors and offer greater flexibility and parsimony than ordinary polynomials. Issues such as how to detect and deal with heterogeneous curve shapes will be explored. Secondly, for continuous outcome variables, the multilevel model assumes a Gaussian distribution for relevant parameters. Transformation of the response variable may be needed to satisfy this condition. However transformation affects all aspects of the model, including the shapes of the response curves and their heterogeneity and the distribution of quantities at all levels of the hierarchy. The project will investigate the effects of response transformation on different parts of the model. The aim will be to develop techniques which will help the analyst decide whether and how to transform the response and understand the effects thereof. A particular application is the analysis of longitudinal fetal size data to produce `conditional reference intervals', which are intended to help the clinician detect fetuses whose growth is faltering. Transformation of predictor and response variables is needed here. The project will also consider how best to present the predictions from such models for ease of understanding and use by the clinician. Several datasets are available to the project. INFORMAL ENQUIRIES may be made to Professor Simon Thompson, Medical Statistics & Evaluation (email: [log in to unmask], telephone: 0181 383 1572), and to Professor Patrick Royston, Medical Statistics & Evaluation ([log in to unmask], telephone: 0181 383 8425). APPLICATIONS: please send a letter and CV, including the names of two referees, to Sandra Griffin, Imperial College School of Medicine, Department of Medical Statistics & Evaluation, The Commonwealth Building, Hammersmith Hospital, Du Cane Road, London W12 0NN. EMAIL: [log in to unmask], FAX: 0181 383 8573. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%