Table 1.

Baseline characteristics of study cohort before and after propensity score matching.

Before PS matchingAfter PS matchinga
All (n = 63,605)Statin (n = 15,990)Nonstatin (n = 47,615)ASDbStatin (n = 11,678)Nonstatin (n = 11,192)ASDb
Age at triple therapy (years)55.6 (±14.6)62.6 (±11.1)53.5 (±14.9)0.6661.7 (±11.0)63.6 (±13.8)0.18
Male sex (n, %)29,629 (46.6%)8,041 (50.3%)21,588 (45.3%)0.095,714 (48.9%)5,313 (47.5%)0.01
Duration of follow-up (years)7.6 (5.1–10.3)8.0 (5.5–10.5)7.4 (4.9–10.2)7.9 (5.5–10.3)7.1 (4.7–9.8)
Smoking (n, %)1,647 (2.6%)561 (3.5%)1,086 (2.3%)0.08394 (3.4%)327 (2.9%)0.02
Alcohol (n, %)556 (0.9%)78 (0.5%)478 (1.0%)0.0151 (0.4%)45 (0.4%)0.01
History of GU (n, %)1,463 (2.3%)448 (2.8%)1,015 (2.1%)0.05322 (2.8%)286 (2.6%)0.03
History of DU (n, %)1,913 (3.0%)444 (2.8%)1,469 (3.1%)0.02318 (2.7%)270 (2.4%)0.01
DM (n, %)7,436 (11.7%)4,652 (29.0%)2,784 (5.8%)0.442,827 (24.2%)1,821 (16.3%)0.07
Dyslipidemia (n, %)5,082 (8.0%)3,974 (24.9%)1,108 (2.3%)0.391,897 (16.2%)851 (7.6%)0.08
Hypertension (n, %)13,173 (20.7%)6,776 (42.4%)6,397 (13.4%)0.474,271 (36.6%)3,221 (28.8%)0.06
IHD (n, %)5,756 (9.0%)4,189 (26.2%)1,567 (3.3%)0.372,054 (17.6%)1,092 (9.8%)0.05
AF (n, %)2,439 (3.8%)1,107 (6.9%)1,332 (2.8%)0.16770 (6.6%)653 (5.8%)0.04
CHF (n, %)2,554 (4.0%)1,300 (8.1%)1,254 (2.6%)0.18831 (7.2%)612 (5.5%)0.02
Stroke (n, %)4,005 (6.3%)2,422 (15.1%)1,583 (3.3%)0.281,485 (12.7%)929 (8.3%)0.01
CRF (n, %)1,416 (2.2%)764 (4.8%)652 (1.4%)0.14487 (12.7%)362 (8.3%)0.40
Cirrhosis (n, %)1,049 (1.6%)115 (0.7%)934 (2.0%)0.0275 (0.6%)86 (0.8%)0.03
Aspirin (n, %)11,116 (17.5%)7,684 (48.1%)3,432 (7.2%)0.634,215 (36.1%)2,287 (20.4%)0.01
Metformin (n, %)8,993 (14.1%)6,200 (38.8%)2,793 (5.9%)0.573,772 (32.3%)2,253 (20.1%)0.06
NSAIDs/COX-2 inhibitors (n, %)14,692 (23.1%)4,435 (27.7%)10,257 (21.5%)0.101,418 (12.1%)1,383 (12.4%)0.01
PPIs (n, %)7,715 (12.1%)2,955 (18.5%)4,760 (10.0%)0.181,224 (10.5%)1,020 (9.1%)0.02
  • Note: Age of receiving triple therapy was expressed as mean (years) ± 1 SD. Duration of follow-up was expressed as median (years) with IQR. Categorical variables were expressed as number (%). Drug use was defined as use for more than 180 days, and expressed as number (%).

  • Abbreviations: AF, atrial fibrillation; ASD, absolute standardized difference; CHF, congestive heart failure; CRF, chronic renal failure; DM, diabetes mellitus; DU, duodenal ulcer; GU, gastric ulcer; IHD, ischemic heart disease; PS, propensity score.

  • aPS matching was performed after trimming of the extreme PS strata (5th and 95th percentiles). Nonstatin users were matched to statin users on PS within a caliper width of 0.1. All variables were included in the model for PS estimation.

  • bVariables with an ASD >0.20 are considered to be imbalanced.