Association between preoperative proton pump inhibitor use and postoperative acute kidney injury in patients undergoing major surgery

Figures & data

Figure 1. Flowchart of study participants.

Table 1. Demographic and clinical characteristics of PPI users and Non-users.

Variable	All (n = 21,533)	PPI users (n = 944)	PPI non-users (n = 20,589)	p-Value
Demographic
Age, years, mean (SD)	57.8 (15.0)	62.5 (13.9)	57.6 (15.0)	<0.001
Male gender, n (%)	11,034 (51.2)	514 (54.4)	10,520 (51.1)	0.046
BMI, kg/m^2, mean (SD)	24.5 (3.8)	23.7 (4.0)	24.5 (3.7)	<0.001
Chronic comorbidities
Hypertension, n (%)	7411 (34.4)	395 (41.8)	7016 (34.1)	<0.001
Diabetes mellitus, n (%)	3427 (15.9)	194 (20.6)	3233 (15.7)	<0.001
Cardiovascular disease, n (%)	2086 (9.7)	201 (21.3)	1885 (9.2)	<0.001
Cerebrovascular disease, n (%)	1230 (5.7)	125 (13.2)	1105 (5.4)	<0.001
Peripheral vascular disease, n (%)	801 (3.7)	52 (5.5)	749 (3.6)	0.003
Chronic obstructive pulmonary disease, n (%)	682 (3.2)	46 (4.9)	636 (3.1)	0.002
Baseline laboratory tests
eGFR, ml/(min·1.73 m^2), mean (SD)	92.4 (18.5)	88.6 (19.6)	92.6 (18.5)	<0.001
eGFR < 60 ml/(min·1.73 m^2), n (%)	1203 (5.7)	86 (9.4)	1117 (5.6)	<0.001
Hemoglobin, g/L, mean (SD)	134.9 (17.9)	125.4 (22.2)	135.4 (17.6)	<0.001
Platelet, 10^9/L, mean (SD)	226.0 (69.4)	225.8 (80.2)	226.0 (68.8)	0.922
Serum albumin, g/L, mean (SD)	42.0 (5.2)	39.3 (5.4)	42.2 (5.1)	<0.001
Risk of gastrointestinal bleeding
Previous history of peptic ulcer or gastrointestinal bleeding, n (%)	195 (0.9)	65 (6.9)	130 (0.6)	<0.001
Coagulopathy*, n (%)	125 (0.6)	15 (1.6)	110 (0.5)	<0.001
Medication use before surgery
Antibiotics, n (%)	14,224 (66.1)	857 (90.8)	13,367 (64.9)	<0.001
RAASi, n (%)	947 (4.4)	125 (13.2)	822 (4.0)	<0.001
NSAIDs, n (%)	1015 (4.7)	346 (36.7)	669 (3.2)	<0.001
Diuretics, n (%)	584 (2.7)	148 (15.7)	436 (2.1)	<0.001
Glucocorticoid, n (%)	935 (4.3)	139 (14.7)	796 (3.9)	<0.001
Contrast, n (%)	357 (1.7)	63 (6.7)	294 (1.4)	<0.001
Anticoagulant, n (%)	3212 (14.9)	437 (46.3)	2775 (13.5)	<0.001
H2-receptor antagonist, n (%)	29 (0.1)	4 (0.4)	25 (0.1)	0.036
Statin, n (%)	931 (4.3)	182 (19.3)	749 (3.6)	<0.001
Surgery profiles
Cardiac, n (%)	288 (1.3)	41 (4.3)	247 (1.2)	<0.001
Aortic, n (%)	54 (0.3)	2 (0.2)	52 (0.3)	0.807
Laparoscopy, n (%)	5899 (27.4)	196 (20.8)	5703 (27.7)	<0.001
Emergent, n (%)	538 (2.5)	99 (10.5)	439 (2.1)	<0.001
ASA score ≥3, n (%)	4271 (19.8)	397 (42.1)	3874 (18.8)	<0.001
Duration of surgery ≥120 min, n (%)	17,284 (80.3)	869 (92.1)	16,415 (79.7)	<0.001
Intraoperative blood transfusion, n (%)	1742 (8.1)	157 (16.6)	1585 (7.7)	<0.001

BMI: body mass index; eGFR: estimated glomerular filtration rate; RAASi: renin-angiotensin-aldosterone system inhibitor; NSAIDs: non-steroidal anti-inflammatory drugs; ASA: American Society of Anesthesiologists.

Missing values (n, %): BMI (559, 2.6%), baseline eGFR (523, 2.4%), baseline albumin (1889, 8.8%), baseline hemoglobin (506, 2.3%), baseline platelet (729, 3.4%). Multiple imputation was used to impute missing values.

* Coagulopathy was defined as any of the following: (1) baseline platelet count <50 × 10⁹/L, (2) baseline prothrombin time/international normalization ratio (PT/INR) >1.5, (3) baseline partial-prothrombin time >2 times the control value.

Table 2. Association of PPI use and clinical outcomes in patients undergoing major surgery.

Outcomes	No. patients with event, n (%)		Odds ratio (95% confidence interval)
	PPI users, n (%)	PPI non-users, n (%)	Crude	Adjusted^a
Primary
Postoperative AKI	72 (7.6)	356 (1.7)	4.69 (3.61, 6.10)	1.47 (1.04, 2.07)
Secondary
In-hospital AKI	85 (9.0)	442 (2.1)	4.51 (3.54, 5.75)	1.41 (1.03, 1.94)
In-hospital mortality	9 (1.0)	19 (0.1)	10.42 (4.70, 23.10)	1.63 (0.55, 4.85)

^a Adjusted for demographic parameters (age, gender, and BMI), chronic comorbidities (hypertension, diabetes mellitus, cardiovascular disease, cerebrovascular disease, peripheral vascular disease, and chronic obstructive pulmonary disease), baseline laboratory tests (eGFR, and hemoglobin), risk of gastrointestinal bleeding (previous history of peptic ulcer or gastrointestinal bleeding, and coagulopathy), medication use before surgery (antibiotics, RAASi, NSAIDs, diuretics, and contrast), and surgery profiles (cardiac surgery, laparoscopy, emergent surgery, ASA score ≥3, duration of surgery ≥120 min, and intraoperative blood infusion).

Figure 2. Subgroup analyses of postoperative AKI stratified by age, eGFR, and medication use before surgery. ^aAdjusted for demographic parameters (age, gender, and BMI), chronic comorbidities (hypertension, diabetes mellitus, cardiovascular disease, cerebrovascular disease, peripheral vascular disease, and chronic obstructive pulmonary disease), baseline laboratory tests (eGFR, and hemoglobin), risk of gastrointestinal bleeding (previous history of peptic ulcer or gastrointestinal bleeding, and coagulopathy), medication use before surgery (antibiotics, RAASi, NSAIDs, diuretics, and contrast), and surgery profiles (cardiac surgery, laparoscopy, emergent surgery, ASA score ≥3, duration of surgery ≥120 min, and intraoperative blood infusion).

Data availability statement

The datasets are not publicly available as they are protected by Peking University First Hospital. Access requests should be directed to the corresponding author.