Skip to main content
. 2019 Nov 15;18:156. doi: 10.1186/s12933-019-0960-8

Table 2.

Predictors of time to first hHF (SMQ definition)

Baseline predictor Hazard ratio [95% CI] Relative importance P-value
Prior heart failure (Y vs. N) 4.89 [3.90; 6.14] 54.6 < 0.0001
Hepatic impairment (Y vs. N) 3.08 [2.15; 4.41] 11.0 < 0.0001
eGFR (log regression) 0.44 [0.34; 0.58] 10.0 < 0.0001
Atrial fibrillation (Y vs. N) 1.95 [1.50; 2.55] 7.2 < 0.0001
Total insulin dose (U/kg) at week 1 1.53 [1.27; 1.84] 5.9 < 0.0001
Prior myocardial infarction (Y vs. N) 1.54 [1.23; 1.91] 4.3 0.0001
Macular edema (Y vs. N) 3.77 [1.40; 10.2] 2.0 0.0087
A1C (squared regression) 1.00 [1.00; 1.01] 1.8 0.0137
Proteinuria (microalbuminuria and gross proteinuria) 1.36 [1.06; 1.73] 1.8 0.0140
Systolic blood pressure at baseline 1.01 [1.00; 1.01] 1.5 0.0251

Variables identified by stepwise selection − FAS. Relative importance is calculated as 100 × Chi square/Total Chi square, where the Chi squares are from a model simultaneously considering all effects mentioned in the table

A1C glycosylated hemoglobin, CI confidence interval, eGFR estimated glomerular filtration rate, FAS full analysis set, hHF hospitalization for heart failure, N no, SMQ standardized Medical Dictionary for Regulatory Activities Query, U units, Y yes