Doctors’ Prescribing Patterns in the Midi-Pyrénées rRegion of France: Point-process Aggregation
People can be located according to their residence, their place of work, their doctor’s office, their pharmacy, and so forth. It is sometimes of interest to look for patterns in people’s locations in relation to their behaviours.
In this article, we are particularly interested in the cost of patients’ prescriptions per doctor consultation. In particular, we consider the common problem brought about by aggregation when only the less-precise locational information and the associated variable ’average prescription amount per consultation’ are available.
We build a spatial regression model for the spatially aggregated data depending on covariates. We fit initially a non-spatial version of the model to the doctor-prescribing data. We then consider spatial dependence in the data after the large-scale variation has been accounted for, and propose a final model that explains doctors’ prescribing patterns.
Key wordsEDA ESDA Region of Midi-Pyrenees (France) Spatial analysis of doctors’ prescribing patterns Spatial regression
Unable to display preview. Download preview PDF.
- N.A.C. Cressie. Statistics for Spatial Data (rev. edn.). Wiley, 1993.Google Scholar
- N.A.C. Cressie, O. Perrin and C. Thomas-Agnan. Likelihood-based estimation for gaussian mrfs. Statistical Methodology (forthcoming), 2004.Google Scholar
- D. Stoyan and H. Stoyan. Fractals, Random Shapes and Point Fields. Wiley, 1994.Google Scholar