The study tests whether implementing professional evidence-based guidelines and best practices for diabetes care -through participation of multidisciplinary teams in a Breakthrough collaborative- results in an improvement of diabetes care, its health outcomes and economic costs. Determinants of success are studied. Data on diabetes will also be used to better understand Breakthrough as an implementation or improvement method.

Diabetes mellitus as a chronic disease is a major and growing health care problem. Studies on the prevention of severe complications provide evidence for the necessity of tight control. Different interventions and models to achieve strict control and reduce diabetes related risks of complications are available. These are, however, not implemented in daily practice. Our study focuses on this implementation problem: it tests whether implementing professional evidence-based guidelines and best practices -through participation of multidisciplinary teams in the Breakthrough collaborative- results in an improvement of diabetes care, its health outcomes and economic costs.

Data on diabetes will also be used to explore and better understand the Breakthrough model as an implementation method. Only uncontrolled observational studies have, so far, described the outcomes of Breakthrough collaboratives. They also describe significant differences between teams in specific improvements made in patient care and organisational performance, resulting in different implementation and medical costs. There is hardly any information regarding these costs and the cost-effectiveness of collaboratives, and little knowledge about how they could be made more effective. Insight is also needed into the factors that influence the success of individual teams. There are no data regarding the sustainability of improvements.

Inclusion Criteria:

• each patient, male or female, visiting the hospital outpatient clinic or general practice with type II diabetes mellitus during a specified period. Patients should be able to read and write the Dutch language.

Exclusion Criteria:

• pregnancy
• patients with a life expectancy < 1 year

6 intervention and 9 control sites representing multiporfessional provider teams from 13 outpatient clinics (47 internists) and 37 general practices (42 general practitioners)

NB To assess the effects of the collaborative, patient outcomes, professional performance, and structural aspects of care management are measured. Patients visiting the clinics were invited to participate by filling out questionnaires and providing written consent to have their medical records examined. Below information on patient inclusion is provided.