Patiromer for Patients With Heart Failure at Risk of Hyperkalemia: CARE-HK in HF Registry

Figures

• 

  Fig 1.. Flowchart for matching patiromer-treated patients and those not treated with a potassium binder. HK, hyperkalemia; MRA, mineralocorticoid receptor antagonist; NTPB, not treated with potassium binder; PS, propensity score; PT, patiromer treated; RASi, renin–angiotensin system inhibitor.

• 

  Fig 2.. Piecewise linear mixed model of serum potassium values before and after patiromer initiation of all patiromer-treated patients (n = 234). Linear mixed model with sK+ values as the dependent variable, time (continuous, in months, calculated as the difference between sK+ measurement date and patiromer initiation date divided by 30.4), time spline (equal to 0 before patiromer initiation, and equal to time after patiromer initiation) as fixed effects and patient as random effects. Patiromer initiation date is used as a knot. The blue line indicates the regression line. The red dashed line indicates the sK+ concentration threshold for hyperkalemia at 5 mEq/L. The circles in the figures illustrate all sK+ readings; all results for a patient are included, so there are multiple sK+ readings displayed. sK+ values measured before 3 months or after 6 months from patiromer initiation were excluded from the analysis. Values below 2 mEq/L and above 9 mEq/L were considered to be clinically invalid and are excluded from the analysis. Duplicated values of sK+ reported on the same day for the same patient were excluded from analysis, only distinct values were considered in the models. sK+, serum…

• 

  Fig 3.. Piecewise linear mixed model of sK+ before and after patiromer initiation, with subanalyses by sK+ category of all patiromer-treated patients (n = 196). Linear mixed model with sK+ values as the dependent variable, time (continuous, in months, calculated as the difference between sK+ measurement date and patiromer initiation date divided by 30.4), time spline (equal to 0 before patiromer initiation, and equal to time after patiromer initiation) as fixed effects and patient as random effects. sK+ subgroup variable (categorical) and its interaction with time and time spline were added as fixed effects. Patiromer initiation date is used as a knot. The blue line indicates the regression line. The red dashed line indicates the sK+ concentration threshold for hyperkalemia at 5 mEq/L. The circles in the figures illustrate all sK+ readings; all results for a patient are included, so there are multiple sK+ readings displayed. sK+ values measured before 3 months or after 6 months from patiromer initiation were excluded from the analysis. Values below 2 mEq/L and above 9 mEq/L were considered to be clinically invalid and are excluded from the analysis. Duplicated values of sK+ reported on…

• 

  Fig 4.. RASi/MRA optimization in each 6-month period in the propensity score-matched patiromer-treated vs not treated with potassium binder patients after a hyperkalemia episode. *Percentages denote missing data at each time interval relative to the baseline-matched cohort (PT n = 211; NTPB n = 361 at 0–6 months). Optimal: ≥50% of the guideline-recommended dose for ≥1 ACEi/ARB/ARNi, and ≥50% of the guideline-recommended dose for MRA. Suboptimal: <50% of the guideline-recommended dose for all ACEi/ARB/ARNi, or <50% of the guideline-recommended dose for MRA, including untreated patients. For propensity score-matched PT patients, RASi/MRA optimization is observed at 6-month intervals after the last HK episode before patiromer initiation. For propensity score-matched NTPB patients, RASi/MRA optimization is observed at 6-month intervals after the first hyperkalemia episode of the analyzed period. P value calculated with conditional logistic regression to compare proportions in matched patients. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; ARNi, angiotensin receptor-neprilysin inhibitor; MRA, mineralocorticoid receptor antagonist; NTPB, not treated with…