Skip to main content

Table 2 Multivariate cox proportional hazard models

From: Renal dysfunction, restrictive left ventricular filling pattern and mortality risk in patients admitted with heart failure: a 7-year follow-up study

Variable

Hazard ratio (95% confidence interval)

P value

Model 1: eGFR (Univariate)

−2LogL: 5172

 

eGFReffect of a 10 ml increase

0.83 (0.79-0.87)

<0.001

Model 2: eGFR + traditional confounders

−2LogL: 4614

 

eGFReffect of a 10 ml increase

0.94 (0.89-0.99)

0.024

Model 3: RF (Univariate)

−2LogL: 5223

 

RF

1.38 (1.14-1.66)

<0.001

Model 4: RF + traditional confounders

−2LogL: 4601

 

RF

1.54 (1.26-1.89)

<0.001

Model 5: eGFR + RF + traditional confounders

−2LogL: 4595

 

eGFReffect of a 10 ml increase

0.93 (0.89-0.99)

0.012

RF

1.57 (1.28-1.93)

<0.001

  1. Significant traditional confounders: age, sex, wall motion index, ischemic heart disease, diabetes, chronic obstructive pulmonary disease and NYHA class. Covariates were removed by backward elimination if P > 0.10.
Back to article page

BMC Nephrology

ISSN: 1471-2369

Contact us