Assessing health status is fundamental when weighing treatment options for atrial fibrillation (AF) patients. Most health-related quality-of-life (HRQoL) data stem from self-ratings, whereas treatment decisions are based upon physicians' estimations.

The degree of congruence between patients' and physicians' assessments of the patients' subjective health status was used as an indicator of good communication and shared understanding. A total of 334 patients with paroxysmal AF without significant concomitant heart diseases and their physicians were asked in a prospective blinded study to rate the patients' HRQoL. The Short Form-12 was used for self-ratings; the Short Form-8, for physician ratings. Using baseline data, intraclass correlations and Bland-Altman graphs were used to assess concordance; cross-sectional multivariate regression analyses assessed patient characteristics associated with discordance.

On average, physicians rated their patients' HRQoL higher than patients did (∆mental component score [MCS] = −3.23, P < .0001, and ∆physical component score [PCS] = −2.21, P = .0001). Intraclass correlations and Bland-Altman graphs showed unsatisfactory concordance. Physical inactivity (∆ = 4.84) had the greatest bivariate effect on PCS discordance, and major depressive disorder (∆ = 7.01), on MCS discordance. In the regression analyses, depression was significantly associated with discord in the MCS (β = −0.94, P < .001) and the PCS (β = −0.37, P < .002). Sleeping disorder was associated with discord in the MCS (β = −4.13, P < .002), and physical inactivity, with discord in the PCS (β = −1.47, P = .006).

In patients with AF, even in the absence of significant concomitant cardiac diseases, depression, followed by sleeping disorder and physical inactivity, was significantly associated with discordance. These findings should be considered by physicians when choosing treatment strategies.