Table 1.
Baseline patient characteristics, before and after propensity score matching
| Before matching | After matching | ||||
|---|---|---|---|---|---|
| No CKD, randoma | P-value | CKD, matched | P-value | No CKD, matched | |
| (N = 2399) | (N = 2399) | (N = 2399) | |||
| Mean (±SD) | |||||
| Age (years) | 59.9 (±11.0) | <0.0001 | 65.4 (±8.8) | 0.776 | 65.4 (±8.9) |
| Ejection fraction (%) | 31.5 (±12.0) | 0.140 | 32.0 (±12.7) | 0.789 | 31.9 (±12.5) |
| Systolic blood pressure (mmHg) | 126 (±19) | <0.0001 | 128 (±21) | 0.905 | 128 (±20) |
| Body mass index (kg/m2) | 27.6 (±5.5) | 0.011 | 27.2 (±5.2) | 0.801 | 27.2 (±5.2) |
| Serum potassium (mEq/l) | 4.32 (±0.42) | 0.031 | 4.35 (±0.45) | 0.871 | 4.35 (±0.40) |
| Median duration of HF (months) | 29 (±36) | 0.164 | 31 (±37) | 0.971 | 31 (±39) |
| N (%) | |||||
| Female sex | 456 (19%) | <0.0001 | 564 (24%) | 0.760 | 574 (24%) |
| Non-white race | 473 (20%) | <0.0001 | 280 (12%) | 0.451 | 297 (12%) |
| Ischaemic aetiology for HF | 1580 (66%) | <0.0001 | 1699 (71%) | 0.726 | 1710 (71%) |
| NYHA functional classes III–IV | 648 (27%) | <0.0001 | 770 (32%) | 0.598 | 753 (31%) |
| Prior myocardial infarction | 1469 (61%) | 0.012 | 1553 (65%) | 0.525 | 1574 (66%) |
| Hypertension | 1045 (44%) | 0.014 | 1130 (47%) | 0.795 | 1140 (48%) |
| Diabetes mellitus | 626 (26%) | 0.022 | 697 (29%) | 0.849 | 703 (29%) |
| Digoxin (pre-trial use) | 1026 (43%) | 0.861 | 1032 (43%) | 0.397 | 1002 (42%) |
| Digoxin (trial use) | 1183 (49.3%) | 0.544 | 1204 (50.2%) | 0.624 | 1186 (49.4%) |
| ACE inhibitors | 2258 (94%) | 0.369 | 2243 (94%) | 0.685 | 2235 (93%) |
| Non-potassium-sparing diuretics | 1748 (73%) | <0.0001 | 1902 (79%) | 0.915 | 1906 (79%) |
| Potassium-sparing diuretics | 152 (6%) | 0.070 | 184 (8%) | 0.788 | 190 (8%) |
| Potassium supplement | 615 (26%) | 0.018 | 688 (29%) | 0.873 | 683 (29%) |
| Jugular venous distension | 281 (12%) | 0.051 | 326 (14%) | 0.496 | 310 (13%) |
| Pulmonary congestion by x-ray | 307 (12.8%) | 0.085 | 348 (14.5%) | >0.999 | 348 (14.5%) |
aTo avoid inflated significance due to the larger sample size of the pre-match cohort, we identified a random sample of 2399 patients with no CKD from the pre-match cohort of 4261 patients without CKD and merged them with 2399 matched CKD patients to assemble a pre-match cohort of 4798 patients, which is the same as that of the matched cohort.
ACE = angiotensin-converting enzyme; CKD = chronic kidney disease; HF = heart failure; NYHA = New York Heart Association.