To the Editor,

We have read with interest the article by Barrios et al.1 on management of atrial fibrillation (AF) in primary care (PC) clinics in Spain. The authors demonstrate the numerous comorbidities present in these patients as well as the room for improvement in anticoagulation therapy. We cannot but note that this study is of interest because of the large number of subjects evaluated to obtain the prevalence of AF among patients attended in PC. The result obtained was 6.1% for all patients who attended the clinic. Given the large sample size, their study population was probably a fairly representative sample for determining the real percentage for a population attended in PC.

What caught our attention was that Barrios et al.1 report the predictors of progression to permanent AF, but their prognostic impact could not be evaluated given the cross-sectional nature of the study. Our group has, however, shown these predictors to be important in terms of morbidity and mortality (death or hospitalization) in the AFBAR study.2 We found that a change in the type of AF during follow-up almost tripled the possibility of such an event. However, in our series, given the small number of patients with such a change, a multivariate analysis was not undertaken to determine the predictors of change in type of AF.

Likewise, we noted 2 aspects in the methods and results which we believe require further elaboration. First, given the subsequent comparisons established between the different types of patients, it is unclear what factors are associated with a switch to permanent AF: do the authors refer to those who had recent-onset AF, those with paroxysmal AF, those with persistent AF, or those who entered permanent AF? The second aspect is the period when the change takes place. Is this from the moment in which the diagnosis was made until inclusion (which, we would imagine, could be years) or from the 6 months prior to inclusion? In the AFBAR study, with a follow-up of 7 months, a change in type of AF occurred in 7% of the patients, whereas in the Val-FAAP study, it was reported in 31.6%. This leads us to think that the time considered is from when the first diagnosis of AF was made.

In addition, the article mentions the prevalence of AF in a previous study by our group, conducted in 2000.3 That study included 6325 consecutive patients who attended a PC clinic for any reason. Of these, 3.86% were diagnosed with AF. However, we would like to take this opportunity to update this information with data obtained from when the AFBAR2 registry was compiled in 2008. Thirty-five PC physicians in northeast Galicia, catering to a population of 44 973 patients aged over 18 years, participated in the registry. The number of patients identified with diagnosis of AF was 1045, that is, a prevalence of 2.32% for the overall population aged over 18 years. This figure has not been published previously and, although it might not be fully representative of the Spanish population, it seems more reliable as it refers to the overall population and not only those who attend the clinic for some complaint.

We believe that studies such as this one1 or that of Riesgo et al.4 are a good opportunity to collect data pertaining to such a common condition in PC clinics and one that is associated with a high morbidity and mortality. The more we know the better.

.

Corresponding author: rafavidal@hotmail.com