Proton-pump inhibitors are associated with increased risk of prosthetic joint infection in patients with total hip arthroplasty: a case-cohort study

Figures & data

Table 1. Patient demographics. Values are count (%) unless otherwise specified

	Cases	Controls
Variable	n = 75	n = 302
Age, mean years (SD)	69 (10)	68 (11)
Female sex	42 (56)	188 (62)
BMI, mean (SD)	30 (5.4)	27 (4.2)
Obesity (BMI > 30)	28 (38)	57 (19)
Proton pump inhibitor	32 (43)	75 (25)
Acetylsalicylic acid	17 (23)	46 (15)
Vitamin K antagonist	16 (21)	14 (4.6)
Immunosuppressive drugs	7 (9.3)	12 (4.0)
Polypharmacy ^a	37 (49)	109 (36)

Figure 1. Graph showing risk of prosthetic joint infection (PJI) according to PPI use (weighted 1-minus-survival Kaplan–Meier plot).

Table 2. Univariable weighted Cox proportion hazard regression model

Risk factor	HR (95% CI)
Age, years	1.0 (1.0–1.0)
Sex (male)	0.8 (0.5–1.1)
Obesity (BMI > 30)	2.5 (1.6–4.0)
Proton pump inhibitor	2.4 (1.4–4.0)
Acetylsalicylic acid	1.6 (0.9–2.8)
Vitamin K antagonist	5.4 (3.1–9.2)
Immunosuppressive drugs	2.4 (1.1–5.3)
Polypharmacy ^a	1.7 (1.1–2.7)

^a Defined as the daily use of ≥ 5 different medications.

HR: hazard ratio. CI: confidence interval.

Table 3. Multivariable weighted Cox proportion hazard regression model for PPI

Factor	HR (95% CI)
Proton pump inhibitor use
Crude ^a	2.4 (1.4–4.0)
Proton pump inhibitor use adjusted for
Model 1: age	2.3 (1.4–4.0)
Model 2: sex	2.4 (1.3–4.3)
Model 3: BMI	1.9 (1.1–3.3)
Model 4: acetylsalicylic acid use	2.3 (1.2–4.3)
Model 5: vitamin K antagonist use	2.2 (1.0–5.0)
Model 6: immunosuppressive drug use	2.4 (1.1–5.4)
Model 7: polypharmacy.^b	2.2 (1.1–4.1)
Model 8: all above	1.9 (0.4–10)

^a Crude = HR from univariable model (Table 2).

^b Defined as the daily use of ≥ 5 different medications.