Serum Potassium And Cause-Specific Mortality In A Large Contemporary Peritoneal Dialysis Cohort

Unlike hemodialysis (HD), peritoneal dialysis (PD) is a continuous therapy and does not induce myocardial stunning. However, since the death risk in HD and PD patients is similar we tested the hypothesis that low serum potassium levels contribute more to the high cardiac risk seen with PD resulting in cardiac risk equivalent to that seen with HD in patients treated in DaVita facilities between 7/1/2001 through 06/30/2006 in the United States (PD, 10,468; HD, 111,651). PD patients were significantly more likely to have serum potassium <4meq/L (adjusted odds ratio, 3.30 (3.05, 3.56)) than HD patients and there was a U-shaped relationship between time-averaged serum potassium and all-cause and cardiovascular mortality (adjusted hazards ratio for all-cause mortality, potassium <3.5meq/L, 1.51 (1.29, 1.76); ≥5.5meq/L, 1.52 (1.32, 1.75)). However, a higher risk for infection-related mortality was seen only with low serum potassium. The population-attributable risks for all-cause mortality for serum potassium <4.0 and ≥5.5meq/L were 3.6% and 1.9% respectively in PD-treated patients; the corresponding risks in HD patients were 0.8% and 1.5% respectively. Hence, we can conclude that abnormalities in serum potassium – particularly low levels contribute disproportionately to the high death risk both cardiac and non-cardiac in PD patients compared to those treated with HD. Whether aggressive correction of abnormalities in serum potassium levels will reduce death risk is presently not known but seems prudent to pursue