Background

Atrial fibrillation (AF) and acute heart failure (HF) often co-incide and begets each other. It is unclear which factors determine the duration of hospital stay for cardiac recompensation in these high risk patients. One therapeutic strategy is to restore sinus rhythm (SR). However, little is known about the proportion of AF patients discharged in sinus rhythm following cardiac recompensation without or with attempted cardioversion.

Methods

We retrospectively analysed patient data from 500 consecutive patients with the main diagnoses decompensated HF and AF admitted to our cardiology ward via the emergency room between 2016 and 2019. Clinical endpoints were the duration of hospital stay (days, d) and the rate of cardioversion into sinus rhythm. Data were analysed by Student’s t-test, chi-square or Kruskal-Wallis with pairwise post-hoc testing as appropriate. Following correction for data completeness N= 474 patients were included in the final analysis.

Results

Patients’ mean age was 72±11 years , 66% male. HF type was HFrEF in 218 (46%), HFmrEF in 91 (19%) or HFpEF in 165 (35%) patients, respectively. Average time in hospital was 9.2±7.6 days and was not significantly different in the elderly (≥ 75 yrs). Tachycardia (≥ 100/min, N=133, 28%), hypotension (syst. blood pressure ≤ 90 mmHg, N=22, 4.7%) or NTproBNP level on admission had no discriminative value for the duration of hospital stay. Patients with advanced renal impairment (eGFR < 30 ml/min/kg) had a significantly longer hospital stay (9.65d vs. 9.16d with eGFR 30-59 and 9.14d with eGFR ≥ 60).

Cardioversion was successful in 76% of attempted cases (pharmacological 39%, electrical 83%, catheter ablation 91%). Interestingly, 133 patients (43%) converted spontaneously to SR during in-hospital recompensation. Patients converting spontaneously into SR were significantly (P<0.01) younger (71.1 vs. 75.1 yrs), male (46% of males vs. 36% of females) and had a lower NTproBNP upon admission (2,443 vs. 3,343 pg/ml). At discharge 222 patients (48.1%) were  in SR. An attempt to cardiovert (N=130, 27%) had no effect on the duration of hospital stay (9.0 vs. 9.3 d, n.s.). Duration of hospital stay was not significantly different in patients discharged in SR (9.6 d) vs. patients in AF (9.0 d).

Conclusion

In summary, in an unselected patient cohort one third of patients presenting with acute HF and AF had HFpEF. Rhythm controll with electrical cardioversion or catheter ablation were more effective than medical therapy. However, a considerable proportion of patients with acute HF and AF converted spontaneously into SR during cardiac recompensation. Attempted cardioversion at any time during the hospitalization did not shorten the duration of recompensation. Based on our results rate control is a viable option in hemodynamically stable patients with acute HF with AF. The impact of rhythm control at an early timepoint on the duration of acute recompensation and subsequent re-hospitalization needs to be evaluated in prospective studies.