Physician variation inthemanagement ofpatients withatrial fibrillation

Objective-To investigate variations in themanagementofpatients withatrial fibrillation amongconsultant physicians. Design-Questionnaire survey. Subjects-Consultant physiciansin England, Wales, andScotland. Results-214 consultant physicians (88 cardiologists and 126non-cardiologists) weresurveyed betweenMay and July 1994.Mostphysicians (47.7%) reported thattheysaw onetofivepatients with atrial fibrillation weekly. Some 52% of cardiologists and40% ofnon-cardiologistsconsidered thatthemain factor influencing theirdecision ofwhetheror nottoanticoagulate wastheclinical history--that is,heartfailure, valvedisease, orstroke. When encountering apatient admitted acutely withnewonsetatrial fibrillation, significantly more cardiologists (66%v 52%,x2= 6-89, P = 0.03) wouldinmnediately startanticoagulant treatment, most favouring intravenous heparin.Most physicians wouldalsointroduce antiarrhythmic treatment ordigoxin, butmore cardiologists wouldattemptimmediate pharmacological (39%v 18%ofnon-cardiologists, P < 0.001) or laterelectrical (86%v 69%,x2= 11-75 P = 0.003) cardioversion tosinusrhythm,whilenoncardiologists tendedto prefer"rate control" withdigoxin. Althoughmany physicians wouldnot continueantiarrhythmictreatment post-cardioversion, morecardiologists thannon-cardiologists woulddoso(thecommonest choice being classIIIagents) (31%v 17%,P = 0.04). Fewernon-cardiologists wouldcontinue anticoagulant treatment post-cardiover