Table 2. Association of Sodium Polystyrene Sulfonate Use vs Nonuse With Hospitalization or Emergency Department Visit for an Adverse Gastrointestinal Event Within 30 Days.

Variable	No. of Events (% Total)	Time to Event, Median (IQR), d	Incidence Rate (95% CI) per 1000 Person-Years	HR (95% CI)
Total Cohort (27 704 With Use and 1 826 162 With Nonuse)
Use	63 (0.2)	12 (6.2)	28.33 (21.33-35.32)	5.61 (4.34-7.26)
Nonuse	756 (0.04)	16 (9.2)	5.05 (4.69-5.41)
Matched Cohort (20 020 With Use and 20 020 With Nonuse)a
Use	37 (0.2)	12 (8.2)	22.97 (15.57-30.37)	1.94 (1.10-3.41)
Nonuse	18 (0.1)	14 (3.2)	11.01 (5.92-16.09)
Matched With Laboratory Values (7557 With Use and 18 492 With Nonuse)b
Use	15 (0.2)	12 (6.2)	24.54 (12.11-36.95)	2.91 (1.38-6.12)
Nonuse	13 (0.1)	8 (4.2)	8.59 (3.92-13.26)

Abbreviations: HR, hazard ratio; IQR, interquartile range.

SI conversion factor: To convert potassium level to millimoles per liter, multiply by 1.0.

^aMatched (greedy algorithm, without replacement) 1:1 to individuals without use on the logit of the high-dimensional propensity score (±0.2 of the standard deviation) and the following: age, sex, diabetes, congestive heart failure, prior acute kidney injury, chronic dialysis, history of hyperkalemia, previous nephrologist visit, medication use (β-blocker or renin-angiotensin-aldosterone system blockade), and index date (within 1 year). Additional adjustment was for place of residence.

^bMatched (greedy algorithm, without replacement) 1:4 to individuals without use on the logit of the high-dimensional propensity score (±0.2 of the standard deviation) and the following: baseline estimated glomerular filtration rate (<15, 15-29, 30-44, 45-59, 60-89, or ≥90 mL/min/1.73 m²), baseline serum potassium level (<5 or ≥5 mEq/L), age, sex, diabetes, congestive heart failure, prior acute kidney injury, chronic dialysis, medication use (renin-angiotensin-aldosterone system blockade or low-molecular weight heparin), and index date (within 1 year). Additional adjustment was for place of residence.