Published in:
Open Access 01-12-2012 | Research article
Predictors of dropout in the German disease management program for type 2 diabetes
Published in: BMC Health Services Research | Issue 1/2012
Login to get accessAbstract
Background
To improve and assess the effectiveness of disease management programs (DMPs), it is critical to understand how many people drop out of disease management programs and why.
Methods
We used routine data provided by a statutory health insurance fund from the regions North Rhine, North Wurttemberg and Hesse. As part of the German DMP for type 2 diabetes, the insurance fund received regular documentation of all members participating in the program. We followed 10,989 patients who enrolled in the DMP between July 2004 and December 2005 until the end of 2007 to study how many patients dropped out of the program. Dropout was defined based on the discontinuation of program documentation on a particular patient, excluding situations in which the patient died or left the insurance fund. Predictors of dropout, assessed at the time of program enrolment, were explored using logistic regression analysis.
Results
5.5% of the patients dropped out of the disease management program within the observation period. Predictors of dropout at the time of enrolment were: region; retirement status; the number of secondary diseases; presence of a disabling secondary disease; doctor's recommendations to stop smoking or to seek nutritional counselling; and the completion and outcome of the routine foot and eye exams. Different trends of dropout were observed among retired and employed patients: retired patients of old age, who possibly drop out of the program due to other health care priorities and employed people of younger age who have not yet developed many secondary diseases, but were recommended to change their lifestyle.
Conclusions
Overall, dropout rates for the German disease management programs for type 2 diabetes were low compared to other studies. Factors assessed at the time of program enrolment were predictive of later dropout and should be further studied to provide information for future program improvements.