Estimates of Chronic Kidney Diseases Associated with Proton-Pump Inhibitors Using a Retrospective Hospital-Based Cohort in Thailand

Figures & data

Figure 1 Flowchart of Participant’s Selection.

Figure 2 On-treatment scheme with grace period = 90 days.

Table 1 Baseline Demographic and Health Characteristics of Overall Cohort Between Proton-Pump Inhibitors (PPIs) and Histamine-2-Receptor Antagonist (H2RAs) Users, at Index Time

		Before Propensity Score Matching (N=63,595)			After 1:1 Propensity Score Matching (N=8174)
		H2RAs	PPIs	Absolute Standardized Difference^Ŧ	H2RAs	PPIs	Absolute Standardized Difference^Ŧ
Number of Participants (%)		4118 (6.48)	59,477 (93.52)		4087 (50%)	4087 (50%)
Age group
	<=60 (%)	2976 (72.27)	40,208 (67.60)	0.102	2957 (72.4)	2962 (72.5)	0.003
	>60 (%)	1142 (27.73)	19,269 (32.40)	0.102	1130 (27.6)	1125 (27.5)	0.003
Sex
	Female (%)	2720 (66.05)	37,458 (62.98)	0.064	2706 (66.2)	2819 (69.0)	0.059
	Male (%)	1398 (33.95)	22,019 (37.02)	0.064	1381 (33.8)	1268 (31.0)	0.059
Entry Years (Fiscal years**)
	2011 (%)	3486 (84.71)	48,932 (82.29)	0.066	3468 (84.9)	3487 (85.3)	0.013
	2012 (%)	629 (15.29)	10,533 (17.71)	0.066	619 (15.1)	600 (14.7)	0.013
Index Years (Calendar years)
	2011 (%)	690 (16.76)	18,887 (31.75)	0.355	687 (16.8)	710 (17.4)	0.015
	2012 (%)	861 (20.91)	15,988 (26.88)	0.140	852 (20.8)	830 (20.3)	0.013
	2013 (%)	1117 (27.12)	7911 (13.30)	0.349	1103 (27.0)	1108 (27.1)	0.003
	2014 (%)	491 (11.92)	5540 (9.31)	0.085	489 (12.0)	499 (12.2)	0.008
	2015 (%)	424 (10.30)	4390 (7.38)	0.103	422 (10.3)	449 (11.0)	0.021
	2016 (%)	303 (7.36)	3753 (6.31)	0.042	303 (7.4)	269 (6.6)	0.033
	2017 (%)	232 (5.63)	3008 (5.06)	0.026	231 (5.7)	222 (5.4)	0.010
Charlson Co-morbidityIndex (SD)		1.52 (0.82)	1.30 (0.67)	0.296^Ŧ	1.52 (0.82)	1.48 (0.77)	0.054
Diagnoses	Malignant neoplasms of digestive organs (C15-C26) (%)	20 (0.49)	173 (0.29)	0.031	20 (0.5)	25 (0.6)	0.017
	Infectious gastroenteritis and colitis, unspecified (A09) (%)	15 (0.36)	369 (0.62)	0.037	15 (0.4)	30 (0.7)	0.050
	Hypertension (I10-I16)	79 (1.92)	2325 (3.91)	0.119^Ŧ	71 (1.7)	114 (2.8)	0.071
	Cardiovascular diseases (I20-I25, I60-I69) (%)	15 (0.36)	60 (0.10)	0.055	15 (0.4)	2 (0.0)	0.070
	Diabetes Mellitus (E08-E11, E13) (%)	12 (0.29)	318 (0.53)	0.038	12 (0.3)	30 (0.7)	0.062
	Metabolic Disorders (E70-E88)	5 (0.12)	52 (0.09)	0.011	5 (0.1)	7 (0.2)	0.013
	Osteoporosis (M81) (%)	0 (0.00)	12 (0.02)	0.020	0 (0.0)	1 (0.0)	0.022
Co-Prescription	Angiotensin-Converting Enzyme Inhibitors (ACEIs) (%)	195 (4.73)	3242 (5.45)	0.033	186 (4.6)	261 (6.4)	0.081
	Angiotensin Type II Receptor Antagonist (AIIA) (%)	74 (1.80)	1315 (2.21)	0.030	72 (1.8)	90 (2.2)	0.032
	Clopidogrel (%)	166 (4.03)	918 (1.54)	0.152^Ŧ	152 (3.7)	157 (3.8)	0.006
	Diuretics (%)	257 (6.24)	4418 (7.43)	0.047	253 (6.2)	323 (7.9)	0.067
	Non-steroidal Anti-inflammatory drugs (NSAIDs) (%)	974 (23.65)	35,196 (59.17)	0.773^Ŧ	965 (23.6)	950 (23.2)	0.009
	Steroids (%)	1059 (25.72)	5953 (10.01)	0.419^Ŧ	1046 (25.6)	1077 (26.4)	0.017
	Insulin (%)	33 (0.80)	451 (0.76)	0.005	33 (0.8)	50 (1.2)	0.041
	Beta-2 adrenergic Agonist (%)	204 (4.95)	2055 (3.45)	0.075	203 (5.0)	311 (7.6)	0.109
	Digoxin (%)	20 (0.48)	466 (0.78)	0.038	19 (0.5)	25 (0.6)	0.020
	Aminoglycosides (%)	27 (0.66)	363 (0.61)	0.006	27 (0.7)	28 (0.7)	0.003
	Polyene Antifungals (%)	5 (0.12)	24 (0.04)	0.029	5 (0.1)	6 (0.1)	0.007
eGFR* [mL/min/1.73m²]	≥ 90	2606 (63.28)	34,230 (57.55)	0.117^Ŧ	2589 (63.3)	2565 (62.8)	0.012
	< 90	1512 (36.70)	25,247 (42.40)	0.117^Ŧ	1498 (36.7)	1522 (37.2)	0.012
Number of hospital visits (SD)		6.70 (8.20)	7.93 (10.36)	0.132^Ŧ	6.71 (8.20)	7.62 (9.45)	0.103
Serum Magnesium [mg/dl] (SD)	N (%) = 1170 (1.84)	1.71 (0.40)	1.70 (0.49)	0.010	1.71 (0.41)	1.70 (0.40)	0.012
Number of serum creatinine tests [median (IQR)]	N (%) = 22,015 (34.62)	2 (1–4)	2 (1–4)	0.037	2 (1–4)	2 (1–4)	0.075
BMI [kg/m2] (SD)***	N (%) = 9209 (14.48)	26.68 (6.58)	24.90 (4.89)	0.307^Ŧ	26.70 (6.59)	24.54 (5.17)	0.365^Ŧ

Notes: ^ŦStandardized difference= difference in means or proportions divided by standard error; imbalanced defined as absolute value greater than 0.10; *CKD_Epi formulation; **Thai fiscal year 2011 is during October 1, 2010 and September 30, 2011 and Thai fiscal year 2011 is during October 1, 2011 and September 30, 2012; ***BMI was obtained only 14.48 per-cent of all participants that it was not used in adjusted models.

Abbreviations: H2RAs, Histamine Receptor 2 Antagonists; PPIs, Proton pump inhibitors; SD, Standard Deviation; eGFR, Estimated Glomerular Filtration Rate; IQR, Inter Quartile Range; BMI, Body Mass Index.

Table 2 Number of Events, Follow-Up and Incidence Rate for the Study Mediators and Outcomes Between PPI and H2RA Users (Grace Period = 90 Days)

Measures	Before Propensity Score Matching (N=63,595)		After 1:1 Propensity Score Matching (N=8174)
	H2RAs	PPIs	H2RAs	PPIs
Total number of participants (%)	4118 (64.75)	59,477 (93.52)	4087 (50)	4087 (50)
Total Follow-up (person-years)	1802	22,594	1784.16	1742.58
Median Follow-up years (IQR)	0.25 (0.24)	0.25 (0.01)	0.25 (0.24)	0.25 (0)
Number of participants with events
Acute Kidney Injury [AKI]	0	1	0	0
Hypomagnesemia [Hypomag]	16	310	16	43
Chronic Kidney Diseases [CKD]	20	725	19	73
Crude incidence rate per 100 person-years (95% CI)
Acute Kidney Injury [AKI]	0.000 (0.000–0.209)	0.000 (0.000–0.002)	-	-
Hypomagnesemia [Hypomag]	0.888 (0.518–1.473)	1.372 (1.224–1.530)	0.897 (0.512–1.456)	2.467 (1.785–3.323)
Chronic Kidney Diseases [CKD]	1.110 (0.678–1.714)	3.209 (2.979–3.451)	1.065 (0.641–1.663)	4.188 (3.283–5.266)

Table 3 Multivariable Cox Proportional Hazards: Overall Survival Following CKD for PPIs and H2RAs Users by Null, and Stratified Cox-PH, with Grace Periods = 90 Days

Model	Crude HR [95% CI]	Adjusted HR [95% CI]
Null	4.242 [2.718–6.619]	-
Stratified Cox-PH adjusted for baseline covariates*	-	3.753 [2.385–5.905]
1:1 Propensity score-matched [PSM] Cox-PH**	-	5.164 [3.110–8.574]

Notes: *Adjusted for age category, sex, baseline eGFR(imputed), Steroids, Clopidogrel, and stratified, Charlson Comorbidity index (CCI), Hypertension, Steroid, and stratified by NSAID, entry year, and hospital visits assumingly different baseline hazard function between NSAID, entry year and hospital visit intervals at baseline. **PSM with weight score adjusting by age, NSAIDs, steroids, clopidogrel, CCI, Hypertension, Diseases of esophagus, stomach and duodenum (K20-K31), eGFR(imputed), Index years, and number of hospital visits.

Table 4 Subgroup Analysis Depicting the HR with 95% CI for CKD Events (Before and After Propensity Score Matching (PSM))

Subgroups	Before PSM		After PSM
	HR	95% CI	HR	95% CI
N	4118 (6.48)	59,477 (93.52)	4087 (50.00)	4087 (50.00)
Age at baseline
Non-elderly (<=60)	3.550	2.051–6.146	4.766	2.591–8.767
Elderly (>60)	4.569	2.000–10.439	5.753	2.284–14.489
Concomitant use of NSAID
No	3.040	1.685–5.484	3.805	1.963–7.377
Yes	4.544	2.208–9.353	7.670	3.421–17.196
Medication Possession Ratio
<=50%	3.880	2.336–6.445	5.211	3.003–9.042
>50%	4.401	1.184–16.356	4.842	1.310–17.901
Generic PPIs
Omeprazole vs H2RAs	3.783	2.404–5.952	5.206	3.135–8.644
Non-omeprazole vs H2RAs	3.612	1.263–10.326	NA*	NA*

Notes: *Not applicable due to small number of events in this stratification.