Data analysis of medical doctors and clinical pharmacists’ cooperation

Maja Madarasi (2011) Data analysis of medical doctors and clinical pharmacists’ cooperation. EngD thesis.

Preview

PDF Download (1829Kb)

Abstract

The thesis describes the analysis of questionnaires filled out by doctors, nurses and pharmacists at the University Medical Centre Ljubljana about cooperation between them. The data were first processed in the statistical program R with OrdEval method, which is included in the CORElearn package and compared with other statistical techniques. Evaluation of attributes was carried out with ReliefF and MDL algorithms. Predictive models were created in the program Orange with 10-fold cross validation and leave-one-out cross-validation methods. We discovered that answers to certain questions are very important, because they influence the assessment of cooperation. The best classification accuracy was obtained with the classification tree, so we used it to predict the success of cooperation in clinical departments, who have not cooperated with the clinical pharmacists yet.

Actions (login required)