[Reasons of changes in drug therapy after hospital discharge - a qualitative study of perception by general practitioners and patients].

Background: After discharge from hospital there is often change of medication regimen. Usually, the main results of the inpatient stay and the subsequent treatment recommendations are summarised in a "discharge letter". Based on this, the general practitioner decides on how to proceed taking the individual aspects of his/her patient into consideration. The aim of the study is to trace changes of medication and suggested therapy in the discharge letter, from the GP through to the patient and the reasons/influencing factors for any changes in medication undertaken or retained.

Methods: A prospective qualitative study with successively selected patients, who were put on a new long-term medication, at discharge after a stay in a hospital internal medicine unit was undertaken. Semi-structured interviews were conducted with the patients 4-6 weeks after hospital discharge. Subsequently, interviews were conducted with the patient's GP on details of current medication. The interviews were recorded electronically, based on the consensus method and evaluated with respect to changes in medication and influencing factors. In order to detect discrepancies in drug therapy, discharge letters were included in the analysis.

Results: A total of 34 patients and their GPs were interviewed. Few changes of medication changes were registered; however, these were more frequent in the weeks after hospital discharge. Drug therapy recommendations were modified by GPs for different medical or non-medical reasons. Non-medical reasons identified included economic, health policy constraints, personal conviction or non-adhrence of the patient. Reasons for a change in medication by the patient included, questioning of the need for taking the drug, incompatibility, fears and a lack of knowledge about the medication.

Conclusion: The data demonstrate that the transition from inpatient to outpatient care is a sensitive interface. The data do not allow quantitative estimation of the magnitude of this phenomenon. In this study, the reasons for the modification of the drug demonstrated that these findings could be the basis for further studies or the development of interventions for preventing unwanted medication changes.