Increased mortality in patients with secondary diagnosis of atrial fibrillation: Report from Chinese AF registry

Abstract Background The relationship between mortality and the primary diagnosis in AF patients is poorly recognized. The purpose of the study is to compare the differences on mortality in patients with a primary or secondary diagnosis of AF and to identify risk factors amenable to treatment. Methods This was a prospective cohort study using data from the Chinese AF registry. For admitted patients, a follow‐up was completed to obtain the outcomes during 1 year. Results A total of 2015 patients with confirmed AF were included. AF was the primary diagnosis in 40.9% (n = 825) of them. 78.9% (n = 939) of the secondary AF diagnosis patients and 55.5% (n = 458) of the primary AF diagnosis patients were sustained AF. Compared with primary AF diagnosis group, the secondary AF diagnosis group was older with more comorbidities. At 1 year, the unadjusted mortality was much higher in the secondary AF diagnosis groups compared with the primary AF diagnosis groups. In Cox regression analysis with adjustment for confounding factors, patients with secondary AF diagnosis were associated with an increased mortality (relative risk 1.723; 95% CI: 1.283 to 2.315, p < .001). On multivariate analysis, age ≥ 75, LVSD, COPD, and diabetes were independent predictors of mortality in patients with primary AF diagnosis, while for the secondary AF diagnosis group, the risk factors were age ≥ 75, heart failure, and previous history of stroke. Conclusions Patients presenting to ED with secondary diagnosis of AF were suffering from higher mortality risks compared with primary AF diagnosis patients. Physicians should distinguish these two groups in clinical practice.

Data from the Framingham study demonstrated a 1.5-fold to 1.9fold risk of mortality in patients with AF after adjustment for the preexisting cardiovascular conditions (Benjamin et al., 1998).
However, there is relatively little research concerning on the relationship between mortality and the primary diagnosis of these TA B L E 1 Overview of data from patients with sustain or non-sustain AF and an alternative primary ED diagnosis Age (

| MATERIAL S AND ME THODS
The Chinese AF registry was a multicenter, prospective, observational study enrolled patients from 20 participating hospitals be-

| Statistical analysis
Categorical variables were expressed as frequencies and percentage, and the normally distributed continuous variables were presented as mean with standard deviation (SD). Different patient strata were compared by chi-squared tests for categorical variables and by the t test for continuous variables. Cox proportional hazards regression analyses were used to identify whether patients with AF as the secondary diagnosis were associated with increased 1-year mortality and the independent predictors of mortality in each group. The models included age (as a seconddegree polynomial), sex, body mass index (as a second-degree polynomial), type of AF, history of myocardial infarction, coro- At enrollment, blood pressure was higher, but heart rate was much lower in the two secondary AF diagnosis groups compared with the primary AF diagnosis groups. The mean CHADS 2 score was higher in the secondary AF diagnosis group than primary AF diagnosis group. As the increasing score of CHADS 2 marking scheme, the proportion of patients with a secondary diagnosis of AF was increasing ( Figure 1). In patients with secondary diagnosis of AF, the top 7 definite primary ED diagnosis are listed in Table 2, along with the 7 most common presenting chief complaints.
During hospitalization, ACE inhibitors, diuretics, digoxin were all given significantly more often in patients with secondary diagnosis of AF, whereas they less frequently received β-blocker and CCB, especially in group 1. There was no difference between the groups with regard to ARB, lipid-lowering medication, or antithrombotic therapy ( Table 1).
The crude results indicated that all-cause mortality was significantly higher in secondary AF diagnosis group than in the primary AF diagnosis group at 1 year, while the cardiovascular mortality was no significant difference between these groups ( to the cardiovascular mortality, patients with a history of COPD and without a history of HF were associated with decreased risk in secondary AF diagnosis group (Table 3). Heart failure, infection or stroke and pulmonary embolus were the most common causes of death among both of the two groups ( Figure 3).

| D ISCUSS I ON
This analysis from the Chinese AF registry shows that ED patient with a secondary diagnosis of AF had remarkably high mortality rate compared with those with a primary diagnosis of AF. The all-cause mortality risk was increased by 72% in secondary AF diagnosis patients compared with primary AF diagnosis patients after adjustment for the confounders.
To the best of our knowledge, this is the first large outcome study to explore the impact of primary diagnosis on mortality in AF patients. In a pilot study, Atzema et al.(Atzema, Lam, Young, & Kester-Greene, 2013) described the characteristics and outcomes in a small group of AF patients and found that the crude mortality was three times higher in secondary AF diagnosis patients than in those with a primary diagnosis of AF. Nevertheless, due to the limitation of this single-center, retrospective study with a portion of incomplete data, the authors emphasized that the conclusions might be inconclusive and further study was warranted. Here we performed a well-designed, multicenter, prospective work to demonstrate a more convincible result as expected and to further explore the potential risks of mortality in each group.
Patient with a secondary diagnosis of AF was much older and had a worse condition with more concomitant disease compared with primary AF diagnosis patients. This was consistent with the former study (Andersson et al., 2013;Atzema et al., 2013). As we all know, AF is particularly common in elderly people, and any condition that predisposes to left atrial enlargement will associate a rising incidence of AF (Schoonderwoerd, Smit, Pen, & Van Gelder, 2008). Apparently, a number of classical factors, such as heart failure, hypertension, valvular disease, diabetes mellitus, cardiomyopathy, obesity, or thyroid disease, are powerful stimulus for the initiation and development of AF, and this is quite familiar in the clinical practice. In our study, beta-blockers were less often used in patients with a secondary diagnosis of AF, which may relate to the higher age and more comorbidities. Treatment with digoxin was more frequent among secondary AF diagnosis group and that may reflect the lower heart rates on admission. Due to its narrow therapeutic index and a potential to contribute to life-threatening arrhythmia, the use of digoxin for rate control in AF patients remains controversial (Hallberg et al., 2007). Especially, two recent post hoc analysis of the AFFIRM data got opposed conclusions on digoxin use and all-cause mortality (Gheorghiade et al., 2013;Whitbeck et al., 2013). In the present study, digoxin was not associated with mortality neither in primary AF diagnosis group nor in secondary AF diagnosis group in multivariable Cox analysis, and we expect further study to investigate the role of digoxin in the contemporary management of AF patients. There was no significant difference between these two groups on antithrombotic therapy. However, it was worth noting that the oral anticoagulants prescription in our population was much lower than reported from previous study (Nieuwlaat et al., 2005). Under-treatment with anticoagulation agents is a great challenge, especially in secondary AF diagnosis group which was at high risk of thrombosis.
After adjustment for confounders, the all-cause mortality risk for patients with a secondary diagnosis of AF remained significantly higher than those with primary AF diagnosis, indicating that secondary AF diagnosis was an independent risk of mortality. The observed difference between patients with AF as a primary diagnosis and as a secondary diagnosis indicated the great influence of concomitant diseases on mortality risk. In our analysis, the top one reason for admission in secondary AF diagnosis patients and the major cause of death for the total study population was heart failure. Atrial fibrillation and heart failure are two of the most prevalent cardiovascular disease conditions. They often coexist and lead to significant morbidity and mortality. Many patients with advanced heart failure develop AF as the severity of heart failure increases. The SOLVD trial (Dries et al., 1998) , 2006). Moreover, incident heart failure has an adverse impact on prognosis in AF independently of other cardiovascular diagnoses and risk factor. AF, particularly when the heart rate is poorly controlled, can lead to the development of dilated cardiomyopathy and heart failure (Suzuki et al., 2012). Clearly, atrial fibrillation is a complex condition and frequently associated with admissions for hypertension, stroke, heart failure, acute coronary syndrome, or infection.
Physicians could not ignore the interaction about AF and its concomitant disease. Therefore, we emphasize the importance of focusing on patients as an entirety rather than a single disease entity.
We also analyzed risk factors of mortality in the study popu-

| LI M ITATI O N S
We used a large administrative database from the Chinese AF registry for analysis, of which 2.7% were patients with atrial flutter. Typical atrial flutter has a well-defined macro-reentrant circuit in the right atrium as its major mechanism and therefore can be relatively easily cured by ablation. Nevertheless, AF and atrial flutter usually coexist and patients with atrial flutter develop AF even subsequently to successful ablation (Perez et al., 2009). Moreover, response to therapy and management approaches for atrial flutter in improvement of survival and reduction of cardiovascular complication is similar to those of AF. So they can be treated as one entity in trials designed to investigate the outcomes. In addition, the anticoagulation rate in the present study was much lower than reported from previous literature (Casciano, Singer, Kwong, Fox, & Martin, 2012). Due to the nature of an observational study that management decisions were made by individual physicians, the snapshot of anticoagulation could just reflect the current status and we underline that an appropriate management of anticoagulation therapy in AF patients was warranted.

| CON CLUS IONS
Patients with secondary diagnosis of AF were associated with an increased 1-year mortality compared with those with primary AF diagnosis. Physicians should distinguish these two groups and pay attention to their risk factors on treatment.

CO N FLI C T O F I NTE R E S T
None.

AUTH O R CO NTR I B UTI O N S
Dr Shao contributed to statistical analyses, data interpretation, and drafting and revisions of the manuscript. Dr Yang: contributed to study design and hypothesis, data interpretation, and drafting of the manuscript. Dr Zhu, Dr Yu, and Dr Liu: contributed to drafting and revision of the manuscript.

E TH I C A L A PPROVA L
All procedures performed in this study were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki declaration and its later amendments.