Relationship Between Remote, Ambulatory Pulmonary Artery Pressures, and All-Cause Mortality in Patients With Chronic Heart Failure

Figures

• 

  Figure 1.. Kaplan-Meier survival: effects of baseline pulmonary artery (PA) pressure on subsequent mortality in all enrolled patients (all ejection fractions). Kaplan-Maier survival curves describing the effects of baseline PA pressure on all-cause mortality over 2 years of follow-up for all enrolled patients, all ejection fractions. Survival was examined using baseline PA diastolic, systolic, and mean pressure divided into patients greater than or equal to the median pressure (red) or less than the median pressure (blue). Differences in each case are statistically significant with log rank scores.

• 

  Figure 2.. Effects of baseline pulmonary artery (PA) pressure on subsequent mortality in all enrolled patients (all ejection fractions). Spline analysis examining the probability of all-cause death at 2 years after CardioMems implant predicted by baseline PA pressure for all enrolled patients, all ejection fractions. Survival was examined using baseline PA diastolic, systolic, and mean pressure. The relationship of baseline PA pressure and mortality was modeled using Cox regression with baseline PA pressure as the single covariate with a natural cubic spline effect and 8 degrees of freedom to allow for nonlinearity. Across physiological ranges of PA pressure, there was nearly a linear relationship between PA pressure and all-cause mortality.

• 

  Figure 3.. Kaplan-Meier survival: effects of baseline pulmonary artery (PA) pressure on subsequent mortality. Kaplan-Maier survival curves describing the effects of baseline PA pressure on all-cause mortality over 2 years of follow-up divided into 2 groups: ejection fraction (EF) ≥50% and EF <50%. Survival was examined in heart failure with a preserved EF (HFpEF; top 3 curves) and heart failure with a reduced EF (HFrEF; bottom 3 curves) using baseline PA diastolic, systolic, and mean pressure divided into patients greater than or equal to the median pressure (red) or less than the median pressure (blue). Differences in each case are statistically significant with log rank scores.

• 

  Figure 4.. Effects of baseline pulmonary artery (PA) pressure on subsequent mortality. Spline analysis examining the probability of all-cause death at 2 years after CardioMems implant predicted by baseline PA pressure divided into 2 groups ejection fraction (EF) ≥50% and EF <50%. Survival was examined in heart failure with a preserved ejection fraction (HFpEF; top 3 curves) and heart failure with a reduced ejection fraction (HFrEF; bottom 3 curves) using baseline PA diastolic, systolic, and mean pressure. The relationship of baseline PA pressure and mortality was modeled using Cox regression with baseline PA pressure as the single covariate with a natural cubic spline effect and 8 degrees of freedom to allow for nonlinearity. Across physiological ranges of PA pressure, there was nearly a linear relationship between PA pressure and all-cause mortality. Because of the small difference in the range of PA systolic pressures between patients with HFpEF and HFrEF, there are slight differences in the vertical axis of the middle 2 panels.

• 

  Figure 5.. Kaplan-Meier survival stratified by baseline and 6-month pulmonary artery (PA) diastolic pressure compared with baseline median. Kaplan-Meier survival curves were created that stratified the change from baseline to 6 months into 4 categories. Patients with baseline pressures below the median and 6-month pressure that remained below the median had the highest survival probability; baseline below the median but 6-month pressure rose above the median had an intermediate survival probability; baseline above the median, 6-month pressure fell below the median had an intermediate survival probability; baseline above the median that remained above the median at 6 months had the lowest survival probability.

• 

  Figure 6.. Effects of change from baseline pulmonary artery (PA) pressure on subsequent mortality. Using a Cox regression analysis the effects of an increase of 1 mm Hg PA pressure from baseline to 6 months after CardioMEMS implant on subsequent mortality risk using PA systolic, diastolic, and mean pressure in patients with heart failure with a reduced ejection fraction (HFrEF) vs heart failure with a preserved ejection fraction (HFpEF) was examined. In the continuous analysis method, estimates from Cox multivariable model were made with the inclusion of covariates for change in pressure at 6 months (continuous or categorical), baseline pressure, study, and relevant baseline characteristics.