Impact of the type of centre on management of AF patients: surprising evidence for differences in antithrombotic therapy decisions.

Atrial fibrillation (AF) patients may receive treatment from specialists or from general medicine physicians representing different levels of care within a structured health care system. This "choice" is influenced by patient flow within a health care system, patient preference, and individual access to health care resources. We analysed how the postgraduate training and work environment of treating physicians affects management decisions in AF patients. Patient characteristics and treatment decisions were analysed at the time of enrolment into the registry of the German Atrial Fibrillation NETwork (AFNET). A total of 9,577 patients were enrolled from 2004 to 2006 in 191 German centres that belonged to the following four levels of care: 13 tertiary care centres (TCC) enrolled 3,795 patients (39.6%), 58 district hospitals (DH) enrolled 2,339 patients (24.4%), 62 office-based cardiologists (OC) enrolled 2,640 patients (27.6%), and 58 general practitioners or internists (GP) enrolled 803 patients (8.4%). Patients with new-onset AF were often treated in DH. TCC treated younger patients who more often presented with paroxysmal AF. Older patients and patients in permanent AF more often received outpatient care. Consistent with recommendations, younger patients and patients with non-permanent AF received rhythm control therapy more often. In addition, the type of centre affected the decision for rhythm control. Stroke risk was similar between centre types (mean CHADS2 scores 1.6 -1.9). TCC (68.8%) and OC (73.6%) administered adequate antithrombotic therapy more often than DH (55.1%) or GP (52.0%, p<0.001 between groups). Upon multivariate analysis, enrolment by TCC or OC was associated with a 1.60 (1.20-2.12, p=0.001) fold chance for adequate antithrombotic treatment. This difference between centre types was consistent irrespective of the type of stroke risk estimation (ESC 2001 guidelines, CHADS2 score), and also consistent when the recently suggested CHA2DS2-VASc score was used to estimate stroke risk. In conclusion, management decisions in AF are influenced by the education and clinical background of treating physicians in Germany. Inpatients receive more rhythm control therapy. Adequate antithrombotic therapy is more often administered in specialist (cardiologist) centres.