PT  - JOURNAL ARTICLE
AU  - Sílvia C Mendonça
AU  - Catherine L Saunders
AU  - Jenny Lund
AU  - Jonathan Mant
AU  - Duncan Edwards
TI  - Temporal trends in incidence of atrial fibrillation in primary care records: a population-based cohort study
AID  - 10.1136/bmjopen-2020-042518
DP  - 2020 Dec 01
TA  - BMJ Open
PG  - e042518
VI  - 10
IP  - 12
4099  - http://bmjopen.bmj.com/content/10/12/e042518.short
4100  - http://bmjopen.bmj.com/content/10/12/e042518.full
SO  - BMJ Open2020 Dec 01; 10
AB  - Objectives Atrial fibrillation (AF) is a heart condition associated with a fivefold increased risk of stroke. The condition can be detected in primary care and treatment can greatly reduce the risk of stroke. In recent years, a number of policy initiatives have tried to improve diagnosis and treatment of AF, including local National Health Service schemes and the Quality and Outcomes Framework. We aimed to examine trends in the incidence of recorded AF in primary care records from English practices between 2004 and 2018.Design Longitudinal cohort study.Setting English primary care electronic health records linked to Index of Multiple Deprivation data.Participants Cohort of 3.5 million patients over 40 years old registered in general practices in England, contributing 22 million person-years of observation between 2004 and 2018.Primary and secondary outcome measures Incident AF was identified through newly recorded AF codes in the patients’ records. Yearly incidence rates were stratified by gender, age group and a measure of deprivation.Results Incidence rates were stable before 2010 and then rose and peaked in 2015 at 5.07 (95% CI 4.94 to 5.20) cases per 1000 person-years. Incidence was higher in males (4.95 (95% CI 4.91 to 4.99) cases per 1000 person-years vs 4.12 (95% CI 4.08 to 4.16) in females) and rises markedly with age (0.58 (95% CI 0.56 to 0.59) cases per 1000 person-years in 40–54 years old vs 21.7 (95% CI 21.4 to 22.0) cases in over 85s). The increase in incidence over time was observed mainly in people over the age of 75, particularly men. There was no evidence that temporal trends in incidence were associated with deprivation.Conclusions Changes in clinical practice and policy initiatives since 2004 have been associated with increased rates of diagnosis of AF up until 2015, but rates declined from 2015 to 2018.The study uses data from the Clinical Practice Research Datalink (CPRD). CPRD does not allow the sharing of patient-level data. The data specification for the CPRD data set is available at: https://www.cprd.com/sites/default/files/CPRD_GOLD_Full_Data_Specification_v2.0_0.pdf. Additional information on the IMD patient-level linkage data is available at: https://cprd.com/sites/default/files/Documentation_SmallAreaData_Patient_set18_v2.7.pdf. The code list used in this study is part of a set of lists made available at www.phpc.cam.ac.uk/pcu/cprd_cam/codelists/.