RSS PRIMARY HEALTH CARE STUDY GROUP meeting Tuesday 27th March 2007 Royal Statistical Society, Errol St, London 2-5pm (directions http://www.rss.org.uk/about/direction.html) All welcome No pre-registration is necessary Title: Developing the methodological mind in primary care research 2pm REBECCA CANNINGS (Cardiff University) Examining the relationship between antibiotic prescribing and complications from respiratory tract infections: moving away from simple correlations Antibiotics are being prescribed less often in primary care, but there is concern amongst GPs that reduced levels are associated with an increase in complications from common infections. This hypothesis will be tested using both ecological and individual level data for a sample of practices. I will also discuss how the model will be extended and improved. 2.25pm SARA MOTTRAM (Keele University) The use of ordinal regression models: an example from a study of mobility limitation Data are often collected on ordinal scales and dichotomised for perceived ease of analysis and interpretation. The potential uses of ordinal regression techniques will be discussed, in particular the partial proportional odds model, which will be compared to a binary regression model using an example from a study of mobility limitation. 2.50pm PAUL HEWSON (University of Plymouth) Spatial multivariate multilevel analysis of selected QoF indicators at the PCT level The 120 indicator "Quality outcomes Framework" reports NHS surgery achievement, measuring compliance with defined practice as well as management issues. Despite many obvious caveats with such observational data, some patterns in disease management throughout the country can be investigated taking account of the multivariate, spatial and multilevel nature of these data. 3.15pm Tea 3.45pm Discussion in groups 4.30pm Rounding up and conclusion 5pm Close