Association of Direct Oral Anticoagulant-Proton Pump Inhibitor Co-Therapy with Adverse Outcomes: A Population-Based Cohort Study

Abstract

Background:

Direct-acting oral anticoagulants (DOACs) are first-line therapy for atrial fibrillation and for venous thromboembolism prevention and treatment. In addition, proton pump inhibitors (PPIs) are recommended by the ACC Expert Consensus for patients receiving DOAC.

Objective:

To determine the risk of clinically important thromboembolic and bleeding events and death in patients concomitantly prescribed DOACs and PPIs compared with DOAC alone.

Methods:

We undertook a population-based cohort study of Ontario residents aged 66 years or older who were newly dispensed a DOAC from 2009 to 2020. The primary outcome was a composite of clinically relevant bleeding, thrombotic events, and all-cause death. We then used a time-dependent Cox regression model to estimate the adjusted hazard of outcomes during time periods where patients were exposed vs not exposed to PPI.

Results:

We included 283 771 new DOAC users (mean age 78.3 years, 49.4% female). The age/sex-standardized outcome event rate of the composite outcome was higher in the DOAC-PPI cohort [20.2 (95% CI 20.0-20.5) per 100 person-years] than in the DOAC-alone cohort [15.2 (95% CI 15.1-15.4) per 100 person-years]. In time-dependent Cox regression analyses examining time periods during DOAC-PPI co-therapy vs DOAC alone, risks were elevated for all outcomes during periods of PPI co-therapy, suggesting residual confounding. In a secondary analysis of patients receiving DOAC-PPI co-therapy (n = 115 493), periods of DOAC-PPI co-therapy were associated with a greater hazard for death (HR 2.24, 95% CI 2.14-2.35), but a lower risk for both thrombosis (HR 0.93, 95% CI 0.89-0.96) and clinically relevant bleeding (HR 0.79, 95% CI 0.76-0.81).

Conclusion and Relevance:

Our cohort study of older adults suggests that DOAC-PPI co-therapy is associated with decreased upper GI bleeding but increased mortality. Since the implications are major and may be due to residual confounding, we recommend that the findings be verified in randomized trials before clinical application.

Keywords

direct oral anticoagulants proton pump inhibitors bleeding disorders and coagulopathies cardiology gastroenterology vascular medicine population-based study

Introduction

Direct acting oral anticoagulants (DOACs) are first-line therapy for atrial fibrillation and for venous thromboembolism prevention and treatment.^1
-4 One of the proposed advantages of DOACs compared with warfarin has been fewer clinically significant drug interactions based on limited evidence.^2
-4 Proton pump inhibitors (PPIs) are one of the drug families commonly co-administered with anticoagulants, as is recommended by the ACC Expert Consensus for patients receiving DOAC co-prescribed with antiplatelets or agents with antiplatelet properties.^5,6 PPIs are effective for managing acid-related gastrointestinal (GI) disorders including peptic ulcer disease and related GI bleeding, and gastroesophageal reflux disease, since they have a favorable safety profile, evidence has been extrapolated to support their off-label use for prevention of DOAC-related GI bleeding.⁷,^8
-12 It was reported that the utilization of DOACs in the United States has increased significantly for annual prescription numbers from 1079 in 2011 to 1.5 million in 2019,¹³ and the PPI absolute number of units purchased from September 2018 to August 2019 reached 99 021 030.¹⁴ Similarly, in 2018, the direct factor Xa inhibitors, and PPIs were each among the 10 drug classes topping public drug program spending on seniors in Canada: $316.2 million and $180.8 million, respectively.¹⁵

According to a recent systematic review, PPIs have a protective effect against warfarin-related GI bleeding (OR = 0.69; 95% CI 0.64-0.73), with no significant effect on thromboembolic events or mortality.¹⁶ Population-based studies by Ray et al^11,17 have reported that PPI co-therapy was associated with reduced risk of OAC (including DOAC)-related upper GI bleeding. However, evidence for the co-therapy of PPIs and DOACs that includes the outcomes of thromboembolic events and mortality is lacking. Accordingly, controversy remains regarding the necessity of combining PPIs with DOACs, even if it is recommended by the ACC Expert Consensus for patients receiving ≥2 antithrombotic agents.^5,6 For instance, 2 studies have reported no evidence of a protective effect of PPIs against dabigatran-related GI bleeding.^18,19 The only relevant randomized trial showed that pantoprazole given concomitantly with low-dose rivaroxaban did not reduce upper GI bleeding.²⁰ Co-prescription of PPIs may reduce the efficacy of dabigatran due to an increase in gastric pH that may reduce drug absorption.²¹ In the RE-LY trial, concomitant use of PPIs reduced dabigatran serum levels by 15%, but there was no significant impact on dabigatran efficacy in terms of plasma concentrations.²² A pilot RCT reported that a 2-week period of PPI withdrawal led to a significant increase in dabigatran trough and peak plasma levels in patients with AF.²³

It is important for clinicians to know whether there are clinically relevant interactions between PPIs and DOACs, particularly beyond the data on protection from upper GI bleeding.^11,24,25 For example, could the trade-off from upper GI bleeding rates be increased thromboembolic events, or deaths from other causes. The objective of this study was to determine the risk of clinically important thromboembolic and bleeding events and death in patients concomitantly prescribed DOACs and PPIs compared with those prescribed DOACs alone.

Methods

Study Design and Data Sources

Our study was a population-based cohort study of administrative healthcare data from Ontario, Canada’s most populous province. The detailed study protocol is published.²⁶ The databases we used are listed in Table A1. This report follows the REporting of studies Conducted using Observational Routinely collected health Data (RECORD-PE) guidelines ( Appendix A ).²⁷ The work was conducted in compliance with Institutional Review Board/Human Subjects Research Committee requirements.

The study was performed at ICES (www.ices.on.ca), where Ontario’s administrative healthcare databases are held securely and linked at the person level using a coded version of the provincial health insurance number. We used prescription drug claims in the Ontario Drug Benefit Database, which contains comprehensive records of prescriptions dispensed to all Ontarians aged 65 years or older; the Canadian Institute for Health Information (CIHI) Discharge Abstract Database, which captures diagnostic and procedural information about hospital admissions; the Ontario Health Insurance Plan (OHIP) Registered Persons Database, which contains demographic and vital status data; and the OHIP Claims History Database to identify physicians’ services. Researchers routinely use these databases to study the clinical consequences of drug-drug interactions.^28,29 International Classification of Diseases, 9th and 10th Revision, codes were used to capture the clinical diagnoses associated with healthcare encounters (Table A1).

Study Population

We studied Ontario residents aged 66 years or older who were newly dispensed a DOAC (dabigatran, rivaroxaban, apixaban, or edoxaban) from 1 June 2009 to 31 March 2020. As prescription drug insurance is available for all adults from their 65^th birthday, we included individuals aged 66 years or older to ensure complete medication records with a 1-year look back. We excluded patients with: (1) a missing or invalid Ontario Health Insurance Plan (OHIP) number; (2) missing age or sex; (3) age < 66 years at entry (the date of the index dispensation for DOAC); (4) a history of fewer than 365 days of OHIP coverage at entry; (5) prescription claims for multiple different DOACs at entry; (6) a prescription claim for a DOAC in the 365 d preceding the index date; (7) a prescription claim for warfarin or heparin in the 100 days preceding or including the index date; and (8) patients who received their initial DOAC prescription within 7 days of total joint arthroplasty, on the grounds that such courses was likely to be short perioperative therapy. A study flow diagram is provided in Figure 1.

Figure 1.

Study flow diagram.

Study Outcomes

The primary outcome was a composite of all clinically relevant bleeding (defined as bleeding bringing the patient to an Emergency Department [ED]), thrombotic events, and all-cause death. The diagnosis and procedure codes used to define the outcomes can be found in Table A1. Thrombotic events were defined as a hospitalization or ED visit for any thromboembolic event, including myocardial infarction (MI), TIA, systemic embolism, peripheral embolism, ischemic stroke, deep vein thrombosis (DVT), pulmonary embolism (PE), coronary artery bypass graft surgery (CABG), and percutaneous coronary interventions, as captured in the hospital discharge abstracts (CIHI-DAD) or ED records (NACRS). GI bleeding included either upper or lower GI sources. Secondary outcomes included all clinically relevant bleeding, including GI, intracranial, and other bleeding that required hospitalization or an emergency room visit.

Study Drug Exposures

Exposures to PPIs and DOACs were treated as time-varying variables, with each day assessed for exposure. The drug exposure period was defined according to the date the prescription was filled and the prescription duration (dispensed days of supply). We identified a period of continuous DOAC use for each patient, beginning with the first pharmacy claim for a DOAC following the patient’s 66th birthday (index date). Our definition of continuous use was a subsequent prescription within 1.5 times the days supplied of the previous DOAC prescription, using a minimum grace period of 30 days. The risk of DOAC-related bleeding, thromboembolic events, or death was captured only while patients were taking the index DOAC. Thus, all study analyses were restricted to periods of anticoagulant treatment during follow-up.

Co-Variables

The potential confounders included patient demographics, inferred indications for DOAC therapy, comorbidities, components of HAS-BLED³⁰ and CHA₂DS₂-VASc Scores,³¹ and the Deyo-Charlson Comorbidity Index.³² Other variables included previous use of warfarin, PPI use at baseline and potentially interacting drugs during the follow-up period. The potentially interacting drugs were treated as time-varying exposures, with continuous use defined as subsequent prescription within 1.5 times the days’ supply of the previous prescription, plus a grace period of 30 days. Potentially interacting drugs included amiodarone, antiplatelet agents, nonsteroidal anti-inflammatory drugs, statins, antimicrobials, and selective serotonin receptor inhibitors. Detailed information on covariates can be found in the protocol.²⁶ The details of the co-variables can be found in Appendix A .

Statistical Analysis

Patients were followed from DOAC initiation until: (1) discontinuation of the index DOAC, defined as 1.5 times the days’ supply plus a grace period of 30 days with no following prescription; (2) switching to another DOAC, warfarin or heparin; (3) loss of OHIP coverage; (4) the date of a study endpoint (bleeding, thrombosis or death); or (5) the last day of the study period (March 31, 2020), whichever occurred first.

The categorical variables are presented as counts and percentages and the continuous variables as means and standard deviations (SD). We compared baseline characteristics of the exposure groups—those exposed to DOAC-PPI co-therapy at any time during study follow-up versus those exposed to DOAC alone—using standardized differences, with values >0.1 considered important. We compared outcomes during time periods during which patients were exposed vs not exposed to PPI as a time-varying variable.

A time-dependent Cox regression model was used to estimate the adjusted hazard of outcomes expressed using hazard ratios (HRs) and 95% confidence intervals (CIs). A secondary analysis was performed among the subset of DOAC patients who were exposed to PPIs at some point during the study period, comparing risks during the days when patients were exposed versus unexposed to PPIs. All statistical analyses were performed at ICES McMaster using SAS version 9.4 (Statistics Analysis System, Inc., Cary, North Carolina, USA).

Use of the data in this project was authorized under section 45 of Ontario’s Personal Health Information Protection Act, which does not require review by a Research Ethics Board.

Results

Participants

We identified 469 049 patients with a DOAC prescription during the 11-year study period. After applying the exclusion criteria, 283 771 new DOAC users remained—details appear in the Study Flow Diagram (Figure 1). A total of 140 280 (49.4%) patients were female and the mean (SD) age of patients at entry was 78.3 (7.7) years. Most lived in urban centers (246 118/283 771, 86.7%). The numbers receiving Apixaban, Dabigatran, Edoxaban, and Rivaroxaban were 115 701 (40.8%), 46 387 (16.3%), 3030 (1.1%), and 118 653 (41.8%), respectively. More than half of the study patients (184 165/283 771, 64.9%) had AF. Patients were followed for an average (SD) 514.0 (609.4) days. Patients’ sociodemographic and clinical characteristic details are presented in Table 1 and Table A2.

Table 1.

Baseline Characteristics of Patients Taking DOAC-PPI Co-Therapy Versus DOAC Alone.

Variable	DOAC therapy(n = 168 278)	DOAC-PPI Co-therapy(n = 115 493)	Standardized Difference^a
Days of follow-up, Mean ± SD	436.89 ± 568.78	626.43 ± 647.93	0.31
Person-years of follow-up, Mean ± SD	1.20 ± 1.56	1.72 ± 1.78	0.41
Age Group (years)
66-74	64 286 (38.2%)	37 263 (32.3%)	0.12
75-84	66 797 (39.7%)	49 062 (42.5%)	0.06
85+	37 195 (22.1%)	29 168 (25.3%)	0.07
Year of cohort entry, n (%)
2009-2011	3206 (1.9%)	683 (0.6%)	0.12
2012-2013	32 499 (19.3%)	24 301 (21.0%)	0.05
2014-2015	35 395 (21.0%)	27 489 (23.8%)	0.07
2016-2017	41 076 (24.4%)	29 934 (25.9%)	0.03
2018-2019	47 863 (28.4%)	30 176 (26.1%)	0.05
2020-	8239 (4.9%)	2910 (2.5%)	0.13
Sex, n (%)
Female	79 674 (47.3%)	60 606 (52.5%)	0.1
Male	88 604 (52.7%)	54 887 (47.5%)	0.1
Type of DOAC
Apixaban	65 301 (38.8%)	50 400 (43.6%)	0.1
Dabigatran	26 748 (15.9%)	19 639 (17.0%)	0.03
Edoxaban	2116 (1.3%)	914 (0.8%)	0.05
Rivaroxaban	74 113 (44.0%)	44 540 (38.6%)	0.11
Census Neighborhood Income Quintile
Not available	553 (0.3%)	370 (0.3%)	0
1 (Low)	32 456 (19.3%)	25 243 (21.9%)	0.06
2	34 564 (20.5%)	24 783 (21.5%)	0.02
3	33 178 (19.7%)	22 873 (19.8%)	0
4	31 828 (18.9%)	21 119 (18.3%)	0.02
5 (High)	35 699 (21.2%)	21 105 (18.3%)	0.07
Rural residence
Not available	379 (0.2%)	265 (0.2%)	0
No	146 113 (86.8%)	100 005 (86.6%)	0.01
Yes	21 786 (12.9%)	15 223 (13.2%)	0.01
Indications
1. AF in 5 years preceding index	103 987 (61.8%)	80 178 (69.4%)	0.16
2. VTE in 5 years preceding index	13 142 (7.8%)	9976 (8.6%)	0.03
3. Valve replacement/repair in 5 years preceding index	1740 (1.0%)	2095 (1.8%)	0.07
4. At least 2 or more of AF, VTE, valve replacement, hip/knee replacement	5594 (3.3%)	5966 (5.2%)	0.09
Comorbidities
1. GI bleeding in 5 years	4711 (2.8%)	8261 (7.2%)	0.2
2. Chronic kidney disease in 3 years	9613 (5.7%)	9438 (8.2%)	0.1
3. Moderate to severe eGFR in 3 years	6405 (3.8%)	8316 (7.2%)	0.15
4. End stage renal disease in 180 days	300 (0.2%)	430 (0.4%)	0.04
5. Liver disease in 3 years	2108 (1.3%)	2524 (2.2%)	0.07
6. Alcoholism in 3 years	1817 (1.1%)	1754 (1.5%)	0.04
7. Dementia	16 823 (10.0%)	13 860 (12.0%)	0.06
8. Diabetes	51 403 (30.5%)	43 481 (37.6%)	0.15
9. Hypertension	134 944 (80.2%)	101 058 (87.5%)	0.2
10. CHF	37 286 (22.2%)	36 770 (31.8%)	0.22
11. Cancer diagnosis	10 242 (6.1%)	8982 (7.8%)	0.07
Charlson Index
Mean ± SD	1.14 ± 1.62	1.51 ± 1.81	0.22
Median (IQR)	0 (0-2)	1 (0-2)	0.24
Charlson score
0	54 220 (32.2%)	34 784 (30.1%)	0.05
1	20 814 (12.4%)	18 055 (15.6%)	0.09
2	14 940 (8.9%)	14 180 (12.3%)	0.11
3	16 949 (10.1%)	20 222 (17.5%)	0.22
Missing	61 355 (36.5%)	28 252 (24.5%)	0.26
CHAD-VASC score
Mean ± SD	3.67 ± 1.34	4.11 ± 1.37	0.33
Median (IQR)	4 (3-4)	4 (3-5)	0.33
HAS-BLED score
Mean ± SD	2.36 ± 0.83	2.65 ± 0.88	0.34
Median (IQR)	2 (2-3)	3 (2-3)	0.34
Medications
1. Warfarin in 100 days preceding index	17 694 (10.5%)	13 971 (12.1%)	0.05
2. PPI within 100 days preceding index	2155 (1.3%)	3123 (2.7%)	0.1

Abbreviation: AF, atrial fibrillation; CHF, congestive heart failure; DAD, Discharge Abstract Database; DOAC, direct oral anticoagulant; eGFR, estimated glomerular filtration rate; IQR, interquartile range; NSAID, Non-steroidal anti-inflammatory drugs; OCR, Ontario Cancer Registry; OHIP, Ontario Health Insurance Plan; PPI, proton pump inhibitors; SD, standard deviation; SSRI, selective serotonin reuptake inhibitors; VTE, venous thromboembolism.

Values ≥ 0.1 are considered important.

We identified 115 493 and 168 278 patients in the DOAC-PPI (defined as patients taking a PPI at any point during their available follow-up) and DOAC-only cohorts, respectively. The DOAC-PPI cohort of 115 493 patients contains 198 648 person-years of observation of patients taking DOAC and PPI together. The 279 670 person-years of observation for no PPI co-therapy included a total of 332 342 patients, 167 672 patients in the DOAC-only cohort, as well as 74 497 patients in the DOAC-PPI cohort who had periods where they had DOAC without PPI. A comparison of baseline characteristics between the DOAC-PPI treated and DOAC only treated patients in Table 1 shows expected differences for multiple potential confounders, including age, history of AF, history of bleeding, Charlson Comorbidity Index score, CHA₂DS₂-VASC score, and HAS-BLED score, with the PPI-treated patients being older, with greater comorbidity burden and having characteristics that put them at relatively greater risk for subsequent bleeding, thrombosis, and death.

Age/Sex-Standardized Outcome Event Rates

A total of 60 639 composite outcome events occurred during follow-up, with 23 903 and 36 736 events during the DOAC-PPI observation times and DOAC-only observation times, respectively. The age/sex-standardized composite outcome event rate of the composite outcome was higher during the DOAC-PPI observation times [20.2 (95% CI 20.0-20.5) per 100 person-years] than that during the DOAC-only observation times [15.2 (95% CI 15.1-15.4) per 100 person-years]. Results for the other outcomes are presented in Table 2. The age/sex-standardized outcome event rates were consistently greater during DOAC-PPI co-therapy.

Table 2.

Age/Sex-Standardized Outcome Event Rates During DOAC-PPI Co-Therapy Versus DOAC Alone.

	DOAC therapy			DOAC-PPI Co-therapy
	Number of events, n	Person-years	Age/Sex Standardized rate/100 person-years (95% CI)	Number of events, n	Person-years	Age/Sex-Standardized rate/100 person-years (95% CI)
Total (composite outcome)	36 736	245 455.51	15.21 (15.06-15.37)	23 903	115 885.67	20.21 (19.95-20.47)
Thrombotic events	11 706	259 016.92	4.55 (4.47-4.63)	7243	125 501.38	5.75 (5.62-5.89)
All-cause death	15 492	265 651.01	6.01 (5.91-6.10)	12 339	131 370.62	9.06 (8.90-9.22)
Clinically relevant bleeding	15 764	254 383.80	6.27 (6.17-6.37)	9553	122 405.36	7.71 (7.55-7.86)
GI bleeding	6145	264 170.97	2.37 (2.31-2.43)	4260	122 405.36	3.25 (3.15-3.35)
Upper GI bleeding	4922	265 360.73	1.89 (1.84-1.94)	3480	129 047.80	2.64 (2.55-2.73)

Abbreviation: DOAC, direct oral anticoagulant; CI, confidence interval; GI, gastrointestinal; PPI, proton pump inhibitor.

Time-Dependent Cox Regression Analyses

In the fully-adjusted time-dependent Cox regression analyses comparing patient time periods exposed to PPIs with periods exposed to DOAC alone, risks were elevated for all outcomes during periods of PPI co-therapy, with the adjusted hazard ratios (HR,95% CI) ranging from 1.04 (95% CI 1.02-1.07) for clinically relevant bleeding to 1.20 (95% CI 1.16-1.22) for all-cause death (Table 3).

Table 3.

Adjusted Hazard Ratios (95% CI) for Study Outcomes for DOAC-PPI Co-Therapy.

	Composite outcome	All-cause death	Thromboembolism,	Clinically relevant bleeding	GI bleeding	Upper GI bleeding
Primary analysis	1.096 (1.077-1.115)	1.192 (1.162-1.222)	1.112 (1.077-1.147)	1.043 (1.015-1.072)	1.109 (1.063-1.157)	1.120 (1.068-1.174)
Secondary analysis ^a	1.070 (1.043-1.097)	2.239 (2.137-2.346)	0.924 (0.886-0.964)	0.785 (0.758-0.813)	0.564 (0.537-0.592)	0.524 (0.497-0.553)

Abbreviation: CI, confidence interval; GI, gastrointestinal; HR, hazard ratio; PPI, proton pump inhibitors.

Among patients who had periods with and without PPIs during their DOAC therapy.

In the secondary analysis of the subset of 115 493 patients who received at least 1 day of DOAC-PPI co-therapy, compared with periods when patients were not exposed to PPI, periods of DOAC-PPI co-therapy were associated with a greater hazard for death (HR 2.24, 95% CI 2.14-2.35), but a significantly lower risk for both thrombosis (HR 0.93, 95% CI 0.89-0.96) and clinically relevant bleeding (HR 0.79, 95% CI 0.76-0.81), risk reductions of 7 and 21%, respectively (Table 3).

Discussion

In this population-based cohort study, DOAC-PPI co-therapy was associated with increased risk for all study outcomes, independently and collectively. This suggests important residual confounding. In our secondary analysis of patients who were exposed to DOAC-PPI co-therapy at some time during the study period, periods of PPI co-therapy were associated with fewer clinically relevant bleeds and thromboembolic events, but with double the risk of death. The relatively greater risk of death during periods of DOAC-PPI co-therapy suggests that PPI treatment was a surrogate for periods of higher risk likely due to worse illness beyond bleeding and thromboembolic events. More detailed data including laboratory results, markers of frailty, residence in long term care, use of chemotherapy, socioeconomic factors, and so on would help clarify prognosis but might require electronic medical record data.

The main purpose for using PPIs in patients with DOACs is to decrease the risk of potential OAC-related GI bleeding while having a neutral effect on thromboembolic events. There is no evidence or pharmacologic rationale including drug interaction with DOACs for PPIs causing an increase in thromboembolic events or death while reducing GI bleeding. The pharmacokinetic interactions proposed mainly affect dabigatran which is no longer commonly used, and the interaction is mainly theoretical and would decrease the DOAC effect.³³ Multiple guidelines based on trials and long term safety evidence, recommend PPIs to decrease GI bleeding and gastroesophageal reflux disease (GERD) events such that the drugs are amongst the most widely used families in the world.^34,35

Other studies have addressed DOAC—PPI interactions and their impact on bleeding. However, none has targeted the comparison of all of the key clinical events related to oral anticoagulant use—thromboembolic, hemorrhagic, and death.^{9,11,20,22,24,25,33} A Korean population-based study published in 2022 found that PPI co-therapy with rivaroxaban for patients with AF who had previously had a GI bleed, was associated with a significantly lower risk of major GI bleeding, but in dabigatran, apixaban, and edoxaban users, PPI co-therapy did not show a significant reduction in the risk of major GI bleeding.²⁴ Ray et al¹¹ examined the incidence of upper GI bleeding hospitalizations for individual anticoagulants without and with PPI co-therapy and found the risk of upper GI tract bleeding hospitalizations to be lower with PPI co-therapy (IRR, 0.66 [95% CI 0.62-0.69]). Recently published systematic reviews which summarize the evidence on association of anti-secretory drugs with upper GI bleeding in patients using OACs suggest that patients taking OAC and PPI co-therapy achieved a lower total, major, or upper GI bleeding compared with OAC alone.^36,37 These findings are similar to our secondary analysis of DOAC patients who served as their own controls by having periods with and without DOAC-PPI co-therapy. However, it is unclear what is contributing to the greater risk of death during periods of DOAC-PPI co-therapy. Death rates for patients with disorders requiring OAC therapy can be substantial—from 7% over 2 years in patients with AF to 8% to 10% in-hospital mortality in patients with cancer.^38,39 A recent systematic review suggests that the most frequent causes of death in patients taking DOACs are cancer (42%), venous thromboembolism (20%), infections (13%), bleeding (6%), heart disease (4%), and stroke (2%).⁴⁰ These are common causes of death, given the diseases that patients taking DOACs tend to have. We are not aware of data that DOACs themselves increase mortality rates from cancer, infections or vascular events.

Strengths of our study include its large, ethnically diverse population-based cohort in Ontario (approximately 250 ethnic origins were reported by Ontarians in the 2016 Census)⁴¹ and rigorous methodology. Our study also has important limitations. First, as a retrospective cohort study of administrative data, it suffers from the intrinsic problem of confounding, including unmeasured confounding. Second, our study concentrated on seniors since they are the most relevant population for this question and, until recently, the only segment of the Ontario population with universal drug coverage. Consequently, we are uncertain about the generalizability of the results to younger people. Third, some PPIs and other relevant drugs (eg, Aspirin and NSAIDs) are available over the counter, leading to possible information bias, although coverage of PPI by the provincial drug plan makes it less likely PPI would be purchased out of pocket. Finally, more sophisticated methodologies beyond the scope of this article (such as propensity score-matching) might have helped reduced confounding.

In terms of the clinical and policy implications of our findings, they suggest clinicians and policymakers need not make important changes in practice. However, future studies, preferably randomized trials, on this topic are a high priority given the huge population at risk of potentially severe clinical outcomes.

Conclusion and Relevance

In this population-based cohort study of older adults, analyses including self-controlled time dependent analyses showed that periods of DOAC-PPI co-therapy were associated with reduced GI bleeding and a small decrease in thromboembolic events, but increased mortality compared with periods of DOAC alone. Since the implications are major and may be due to residual confounding, we recommend that the findings be verified in randomized trials before clinical application.

Footnotes

Appendix A

Acknowledgements

This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and the Ministry of Long-Term Care. This study used data adapted from the Statistics Canada Postal Code Conversion File^OM, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under license from ^©Canada Post Corporation and Statistics Canada. Parts of this material are based on data and/or information compiled and provided by the Ontario Ministry of Health and the Canadian Institute for Health Information. We thank IQVIA Solutions Canada Inc. for the use of their Drug Information File. The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.

Author Contributions

Conceptualization: MW, AH, and MP; Data curation: FN; Formal analysis: FN; Funding acquisition: MW and AH; Investigation: All authors; Methodology: MW, AH, FN, and MP; Project administration: AH and MP; Resources: MP and FN; Software: MP and FN; Supervision: AH and MP; Validation: DS, AH, and MP; Writing—original draft: MW; Writing—review and editing: All authors.

Data Availability

The data set from this study is held securely in coded form at ICES. While legal data-sharing agreements between ICES and data providers (eg, health care organizations and government) prohibit ICES from making the data set publicly available, access may be granted to those who meet pre-specified criteria for confidential access, available at (email: das@ices.on.ca). The full data set creation plan and underlying analytic code are available from the authors upon request, with the understanding that the computer programs may rely upon coding templates or macros that are unique to ICES and are therefore either inaccessible or may require modification.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funded by the Canadian Institutes of Health Research (CIHR) Award Number 365834 to Dr Anne Holbrook, a studentship award to Mei Wang from the Research Institute of St. Joe’s Hamilton and a CanVECTOR Research Start-Up Award.

Ethical Considerations

Use of the data in this project was authorized under section 45 of Ontario’s Personal Health Information Protection Act, which does not require review by a Research Ethics Board.

ORCID iDs

Mei Wang

Anne M. Holbrook

References

Andrade

Aguilar

Atzema

, et al The 2020 Canadian cardiovascular society/Canadian heart rhythm society comprehensive guidelines for the management of atrial fibrillation. Can J Cardiol. 2020;36:1847-1948. doi:10.1016/j.cjca.2020.09.001

Joglar

Chung

Armbruster

, et al 2023 ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation: a report of the American College of Cardiology/American Heart Association Joint committee on clinical practice guidelines. Circulation. 2024;149:e1-e156. doi:10.1161/cir.0000000000001193

Hindricks

Potpara

Dagres

, et al 2020 ESC guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): the task force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42:373-498. doi:10.1093/eurheartj/ehaa612

Steffel

Collins

Antz

, et al 2021 European heart rhythm association practical guide on the use of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation. Europace. 2021;23:1612-1676. doi:10.1093/europace/euab065

Kumbhani Dharam

Cannon Christopher

Beavers

Craig J

, et al 2020 ACC expert consensus decision pathway for anticoagulant and antiplatelet therapy in patients with atrial fibrillation or venous thromboembolism undergoing percutaneous coronary intervention or with atherosclerotic cardiovascular disease. J Am Coll Cardiol. 2021;77:629-658. doi:10.1016/j.jacc.2020.09.011

Chen

Stecker

E A

Warden

Direct oral anticoagulant use: a practical guide to common clinical challenges. J Am Heart Assoc. 2020;9:e017559. doi:10.1161/jaha.120.017559

Savarino

Marabotto

Zentilin

, et al Proton pump inhibitors: use and misuse in the clinical setting. Expert Rev Clin Pharmacol. 2018;11:1123-1134. doi:10.1080/17512433.2018.1531703

Burnett

Mahan

Vazquez

Oertel

Garcia

Ansell

Guidance for the practical management of the direct oral anticoagulants (DOACs) in VTE treatment. J Thromb Thrombolysis. 2016;41(1):206-232. doi:10.1007/s11239-015-1310-7

Chan

Lau

Leung

, et al Prevention of dabigatran-related gastrointestinal bleeding with gastroprotective agents: a population-based study. Gastroenterology. 2015;149(3):586-595. doi:10.1053/j.gastro.2015.05.002

10.

O’Dea

Whetteckey

Ting

A prospective, randomized, open-label study to evaluate two management strategies for gastrointestinal symptoms in patients newly on treatment with dabigatran. Cardiol Ther. 2016;5(2):187-201.

11.

Ray

Chung

Murray

, et al Association of oral anticoagulants and proton pump inhibitor cotherapy with hospitalization for upper gastrointestinal tract bleeding. JAMA. 2018;320:2221-2230. doi:10.1001/jama.2018.17242

12.

Tang

Xiao

Logistic regression analysis of risk factors for upper gastrointestinal bleeding induced by PCI in combination with double antiplatelet therapy for STEMI patients. Acta Gastroenterol Belg. 2020;83(2):245-248.

13.

Costa

LSD

Alsultan

Hincapie

Guo

. Trends in utilization, reimbursement, and price for DOACs and warfarin in the US Medicaid population from2000 to 2020. J Thromb Thrombolysis. 2023;55:339-345. doi:10.1007/s11239-022-02727-0

14.

Gunning

Chu

Nakhla

Kim

Suda

Tadrous

Major shifts in acid suppression drug utilization after the 2019 ranitidine recalls in Canada and United States. Dig Dis Sci. 2023;68:3259-3267. doi:10.1007/s10620-023-07958-6

15.

Lee

Jani

Cheng

Hunt

Prescribed drug spending in Canada, 2019: a focus on public drug programs. Healthc Q. 2020;23:10-12.

16.

Wang

Zeraatkar

Obeda

, et al Drug-drug interactions with warfarin: a systematic review and meta-analysis. Br J Clin Pharmacol. 2021;87:4051-4100. doi:10.1111/bcp.14833

17.

Ray

Chung

Murray

, et al Association of proton pump inhibitors with reduced risk of warfarin-related serious upper gastrointestinal bleeding. Gastroenterology. 2016;151:1105-1112.e10. doi:10.1053/j.gastro.2016.08.054

18.

Bang

Joo

Kim

, et al The role of acid suppressants in the prevention of anticoagulant-related gastrointestinal bleeding: a systematic review and meta-analysis. Gut Liver. 2020;14:57-66. doi:10.5009/gnl19009

19.

Nantsupawat

Soontrapa

Nantsupawat

, et al Risk factors and prevention of dabigatran-related gastrointestinal bleeding in patients with atrial fibrillation. J Arrhythm. 2018;34:30-35. doi:10.1002/joa3.12015

20.

Moayyedi

Eikelboom

Bosch

, et al Pantoprazole to prevent gastroduodenal events in patients receiving rivaroxaban and/or aspirin in a randomized, double-blind, placebo-controlled trial. Gastroenterology. 2019;157:403-412.e5. doi:10.1053/j.gastro.2019.04.041

21.

Ogawa

Echizen

Drug-drug interaction profiles of proton pump inhibitors. Clin Pharmacokinet. 2010;49:509-533. doi:10.2165/11531320-000000000-00000

22.

Liesenfeld

Lehr

Dansirikul

, et al Population pharmacokinetic analysis of the oral thrombin inhibitor dabigatran etexilate in patients with non-valvular atrial fibrillation from the RE-LY trial. J Thromb Haemost. 2011;9:2168-2175. doi:10.1111/j.1538-7836.2011.04498.x

23.

Schnierer

Samoš

Bolek

, et al The effect of proton pump inhibitor withdrawal on dabigatran etexilate plasma levels in patients with atrial fibrillation: a washout study. J Cardiovasc Pharmacol. 2020;75(4):333-335. doi:10.1097/FJC.0000000000000791

24.

Lee

Kwon

Choi

, et al Proton pump inhibitor co-therapy in patients with atrial fibrillation treated with oral anticoagulants and a prior history of upper gastrointestinal tract bleeding. Cardiovasc Drugs Ther. 2022;36(4):679-689. doi:10.1007/s10557-021-07170-6

25.

Lee

Kim

, et al Risk of upper gastrointestinal bleeding in patients on oral anticoagulant and proton pump inhibitor co-therapy. PLoS ONE. 2021;16(6):e0253310. doi:10.1371/journal.pone.0253310

26.

Wang

Paterson

Thabane

, et al Association of direct oral anticoagulant-proton pump inhibitor cotherapy with adverse outcomes: protocol for a population-based cohort study. BMJ Open. 2022;12:e057991. doi:10.1136/bmjopen-2021-057991

27.

Langan

Schmidt

Wing

, et al The reporting of studies conducted using observational routinely collected health data statement for pharmacoepidemiology (RECORD-PE). BMJ. 2018;363:k3532. doi:10.1136/bmj.k3532

28.

Wright

Gomes

Mamdani

Horn

Juurlink

DN.

The risk of hypotension following co-prescription of macrolide antibiotics and calcium-channel blockers. CMAJ. 2011;183:303-307. doi:10.1503/cmaj.100702

29.

Juurlink

Gomes

, et al A population-based study of the drug interaction between proton pump inhibitors and clopidogrel. CMAJ. 2009;180:713-718. doi:10.1503/cmaj.082001

30.

Methavigul

Revised HAS-BLED score for bleeding prediction in atrial fibrillation patients with oral anticoagulants. J Arrhythm. 2022;38(3):380-385. doi:10.1002/joa3.12709

31.

Chen

, et al CHA2DS2-VASc score as an independent outcome predictor in patients hospitalized with acute ischemic stroke. PLoS ONE. 2022;17(7):e0270823. doi:10.1371/journal.pone.0270823

32.

Deyo

Cherkin

Ciol

MA.

Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613-619. doi:10.1016/0895-4356(92)90133-8

33.

Ferri

Colombo

Tenconi

Baldessin

Corsini

Drug-drug interactions of direct oral anticoagulants (DOACS): from pharmacological to clinical practice. Pharmaceutics. 2022;14(6):1120. doi:10.3390/pharmaceutics14061120

34.

Alshamsi

Belley-Cote

Cook

, et al Efficacy and safety of proton pump inhibitors for stress ulcer prophylaxis in critically ill patients: a systematic review and meta-analysis of randomized trials. Crit Care. 2016;20:120. doi:10.1186/s13054-016-1305-6

35.

Szabó

Mátics

Hegyi

, et al PPIs prevent aspirin-induced gastrointestinal bleeding better than H2RAs. A systematic review and meta-analysis. J Gastrointestin Liver Dis. 2017;26(4):395-402. doi:10.15403/jgld.2014.1121.264.hra

36.

Ahn

Lee

Choi

, et al Protective effect of proton-pump inhibitor against gastrointestinal bleeding in patients receiving oral anticoagulants: a systematic review and meta-analysis. Br J Clin Pharmacol. 2022;88(11):4676-4687. doi:10.1111/bcp.15478

37.

Kurlander

Barnes

Fisher

, et al Association of antisecretory drugs with upper gastrointestinal bleeding in patients using oral anticoagulants: a systematic review and meta-analysis. Am J Med. 2022;135(10):1231-1243. doi:10.1016/j.amjmed.2022.05.031

38.

Yamashita

Suzuki

Inoue

, et al Two-year outcomes of more than 30 000 elderly patients with atrial fibrillation: results from the all nippon AF in the elderly (ANAFIE) registry. Eur Heart J Qual Care Clin Outcomes. 2022;8:202-213. doi:10.1093/ehjqcco/qcab025

39.

Lyman

Culakova

Poniewierski

Kuderer

NM.

Morbidity, mortality and costs associated with venous thromboembolism in hospitalized patients with cancer. Thromb Res. 2018;164(suppl 1):S112-S118. doi:10.1016/j.thromres.2018.01.028

40.

Gómez-Outes

Terleira-Fernández

Lecumberri

Suárez-Gea

Calvo-Rojas

Vargas-Castrillón

Causes of death in patients with venous thromboembolism anticoagulated with direct oral anticoagulants: a systematic review and meta-analysis. Semin Thromb Hemost. 2018;44(4):377-387. doi:10.1055/s-0038-1642644

41.

Ontario population 2025. World Population Review. Retrieved March 21, 2025. Accessed April 18, 2025. https://worldpopulationreview.com/canadian-provinces/ontario