Table 4.

Multivariate analyses for relation between AF (versus sinus rhythm) and in-hospital mortality on imputed data of patients recorded as HF (N = 96,593)

	HR	95% CI		p value
Atrial fibrillation on ECG	1.15	1.09	1.21	< 0.0001
Breathlessness
NYHA I	1.00
NYHA II	1.01	0.87	1.18	0.896
NYHA III	1.10	0.96	1.27	0.18
NYHA IV	1.32	1.14	1.52	< 0.0001
Peripheral oedema
None	1.00
Mild	1.03	0.94	1.13	0.498
Moderate	1.09	1.01	1.18	0.023
Severe	1.33	1.22	1.45	< 0.0001
Palliative care follow-up	1.64	1.43	1.89	< 0.0001
Heart failure liaison service	0.51	0.44	0.59	< 0.0001
GP follow-up	0.23	0.19	0.27	< 0.0001
Care of the elderly follow-up	0.44	0.37	0.52	< 0.0001
Cardiology follow-up	0.33	0.28	0.39	< 0.0001
Previous AMI	1.05	0.99	1.11	0.123
History of diabetes	1.01	0.96	1.06	0.799
History of hypertension	0.97	0.92	1.01	0.156
History of IHD	1.08	1.03	1.13	0.003
History of valvular heart disease	1.16	1.10	1.23	< 0.0001
Age categories (years)
Min-54	1.00
55–64	1.38	1.13	1.70	0.002
65–74	1.88	1.55	2.28	< 0.0001
75–84	2.33	1.90	2.84	< 0.0001
85 +	2.72	2.21	3.36	< 0.0001
Male	1.06	1.02	1.11	0.003
LVH	0.88	0.80	0.97	0.01
ACEi/ARB use	0.77	0.72	0.83	< 0.0001
Beta-blocker	0.80	0.75	0.86	< 0.0001
Thiazide	0.91	0.78	1.06	0.205
Loop diuretic	0.70	0.65	0.76	< 0.0001
Digoxin	0.70	0.63	0.79	< 0.0001
Renal failure	1.15	1.06	1.25	0.003
Length of time in hospital (per 5 days)	1.02	1.02	1.03	< 0.0001

HR hazard ratio, CI confidence interval, ECG electrocardiograph, NYHA New York Heart Association, AMI acute myocardial infarction, GP general practitioner, IHD ischaemic heart disease, LVH left ventricular hypertrophy, ARB angiotensin receptor blockers, ACI angiotensin-converting enzyme inhibitor