TY - JOUR T1 - Variation in anticoagulation for atrial fibrillation between English clinical commissioning groups: an observational study JF - British Journal of General Practice JO - Br J Gen Pract DO - 10.3399/bjgp18X697913 SP - bjgp18X697913 AU - John Robson AU - Kate Homer AU - Zaheer Ahmed AU - Sotiris Antoniou Y1 - 2018/07/03 UR - http://bjgp.org/content/early/2018/07/03/bjgp18X697913.1.abstract N2 - Background Despite improvement in anticoagulation for atrial fibrillation (AF), substantial variation in anticoagulation persists between clinical commissioning groups (CCGs) and regions in England.Aim To identify reasons for variation between English CCGs in anticoagulation for AF.Design and setting A 4-year observational study from 2012/2013 to 2015/2016, of the national Quality and Outcomes Framework.Method Multiple regression and Pearson’s correlation coefficients were used to analyse anticoagulation for AF in relation to older age, Index of Multiple Deprivation, prescription of non-vitamin K antagonist oral anticoagulants (NOACs), and exception reporting, as well as stroke hospital admission and mortality.Results The proportion of eligible patients in England prescribed anticoagulants for AF without exceptions for clinical complexity or patient dissent increased from 65.1% in 2012/2013 to 77.9% in 2015/2016. In 2015, 290 920 additional eligible people were anticoagulated in association with use of the CHA2DS2VASc rather than CHADS2 score. From 2012 to 2015, exception reporting almost halved from 20% to 10.2%. Variation in CCG anticoagulation was not associated with deprivation or NOAC use. There was a strong negative association between exception reporting representing patient complexity and anticoagulation performance, accounting for 57% of the variation in anticoagulation without exceptions (multiple regression coefficient = −0.81; 95% confidence intervals = −0.92 to −0.71; P<0.001).Conclusion Anticoagulation for AF has improved substantially in England in association with considerable increases in the eligible population as a result of decreased exception reporting and the use of the CHA2DS2VASc score. There is still substantial room for improvement in most CCGs because, even allowing for exceptions, nine out of 10 CCGs failed to achieve 90% anticoagulation. ER -