Survival characteristics
The all-cause mortality in patients in mixed CMP (21.1%) was similar to
that observed in ICM (20.1%; p=0.8) and higher than in NICM (15.6%;
p=0.2). Time-adjusted survival estimated using Kaplan-Meir curves
revealed hazards of 1.57 (95% CI: 0.91, 2.68; p=0.1) for mixed CMP
compared to NICM. (Figure 2 ) The mean age at death in patients
with mixed CMP (79±8years) was significantly higher than in ICM
(73±12years; p=0.01) and NICM (66±14years; p<0.001). Analysis
of the cause of death revealed higher proportion of non-cardiac deaths
in patients with mixed CMP (52.9%), compared to ICM (26.7%; p=0.04)
and NICM (18.2%; p=0.02). The distribution of heart failure related
deaths and sudden cardiac deaths were similar between all the 3 groups.
(Table 2 )
Cox-regression analysis (Table 3) revealed the following
significant predictors of mortality- age (HR: 1.04; 95% C.I:
1.02-1.06), LVEF (HR: 0.96; 95% C.I: 0.93-0.99), CKD (HR: 2.9; 95%
C.I: 1.9-4.5), NYHA class (HR: 1.7; 95% C.I:1.1-2.4) and CAD (HR: 1.9;
95% C.I: 1.1-3.2). This model accounted for various confounding
variables including age, gender, clinical variables, presence or absence
of moderate-severe CAD and documented nonischemic triggers. Compared to
the survivors in the whole cohort, the non-survivors had significantly
(p<0.05) higher mean age (69.1±11.8y vs 62.7±13y), lower LVEF
(29.7±6.6% vs 36.2±11.3%), higher NYHA class III (51.5% vs 19.8%),
lower GFR (65.8±27.8 vs 85.5±25.2), and significantly higher incidences
of comorbidities- hypertension (64.4% vs 52.9%), chronic kidney
disease (46.5% vs 10%), malignancy (23.8% vs 11%). The distribution
of ICM, NICM and mixed CMP was similar. (Supplemental Table 3 )
The proportion of patients receiving therapies was significantly higher
in the non-survivors compared to the survivors (50% vs 32.3%, p=
0.001). Among the patients receiving device therapies, significantly
higher proportion of patients received shocks in non-survivors compared
to survivors (79.6% vs 63.7%, p=0.04).