Multivariate Analysis
A multivariate analysis was performed (Table 2 ) using moderate to
severe mitral valve disease, use of class III antiarrhythmics, use of
loop diuretics, persistent AF, LAP, and 30-day AF recurrence. Patients
with moderate to severe mitral valve disease (odds ratio (OR) 8.7
[95% confidence interval 3.3 – 22.6]; p = 0.03) and taking a loop
diuretic (OR 4.8 [2.6 – 9.1]; p = 0.01) were more likely to have a
HF event. Those taking class III antiarrhythmics were less likely to
have a HF event (OR 0.24 [0.1 – 0.5]; p = 0.04) as seen in theFigure 2 . A second multivariate analysis incorporating variable
associated with HF events in the retrospective study (procedure time,
fluid administration and preexisting HF diagnosis) revealed no
persistent relationship with HF events. Finally, a third multivariate
analysis was performed that included demographic information putatively
associated with HF events. Of these variables (Table 3 ), only the
presence of mitral valve disease was significant (OR 6.6 [1.3 –
34]; p = 0.03).