Design of Experiments (DoE) is an important tool in the “Improve” phase of lean six sigma. Although applications of DoE in healthcare are not common, interesting examples include assessing the effect of social norms on changing behaviour, improving the smooth running of an emergency department and increasing the profitability of pharmacy provision. Factors which may be influential in a designed experiment can be identified in many ways, such as brainstorms, content analysis of focus groups, statistical process control and data mining decision trees. Designed experiments can be used in conjoint analysis questionnaires, for example to tackle the problem of “did-not-attends (DNAs)”. The talk will review some successful case studies and discuss the limitations of using DoE for improvement in healthcare.
Keywords: Designed experiments DoE; Healthcare; Data mining; Lean six sigma
Biography: Dr Coleman is Technical Director of the i-Statistics Research Unit, a self-financing consultancy and training group in Newcastle University. She enjoys the challenges of applied statistics and has worked in many sectors including utilities and healthcare. Her current interests include designing experiments in service sectors, data mining customer data and encouraging statistical skills and appreciation in a wide range of practitioners.