Deprescription of Aspirin in Patients With GI Bleeding

In This Article

Abstract and Introduction

Abstract

Background: Guidelines recommend against aspirin for primary prevention of cardiovascular events in individuals with a history of gastrointestinal bleeding (GIB). It is unknown how often patients on primary prevention aspirin hospitalised with GIB have aspirin discontinued at discharge.

Aims: To determine the rate of aspirin deprescription and explore long-term outcomes in patients taking aspirin for primary prevention of cardiovascular events.

Methods: We evaluated all patients hospitalised at Yale-New Haven Hospital between January 2014 and October 2021 with GIB who were on aspirin for primary prevention. Our primary endpoint was the frequency of aspirin deprescription at discharge. Our secondary endpoints were post-discharge hospitalisations for major adverse cardiovascular events (MACE) or GIB. Time-to-event analysis was performed using Kaplan–Meier curves and the log-rank test.

Results: We identified 320 patients with GIB on aspirin for primary prevention: median age was 72 (interquartile range [IQR] 61–81) years and 297 (92.8%) were on aspirin 81 mg daily. Only 25 (9.0%) patients surviving their hospitalisation were deprescribed aspirin at discharge. Among 260 patients with follow-up (median 1103 days; IQR 367–1670), MACE developed post-discharge in 2/25 (8.0%) with aspirin deprescription versus 37/235 (15.7%) with aspirin continuation (log-rank p = 0.28). 0/25 patients with aspirin deprescription had subsequent hospitalisation for GIB versus 17/235 (7.2%) who continued aspirin (log-rank p = 0.13).

Conclusions: Aspirin for primary cardiovascular prevention was rarely deprescribed at discharge in patients hospitalised with GIB. Processes designed to ensure appropriate deprescription of aspirin are crucial to improve adherence to guidelines, thereby improving the risk–benefit ratio in patients at high risk of subsequent GIB hospitalisations with minimal increased risk of MACE.

Introduction

Low-dose aspirin (81–325 mg/day) is commonly used for the prevention of atherosclerotic cardiovascular disease but is associated with an estimated 56% increased risk of major gastrointestinal bleeding (GIB) compared with placebo or no aspirin.^[1] Since 2009, the United States Preventative Service Task Force has explicitly discouraged the use of primary prevention aspirin in patients at increased risk of GIB, which includes patients with a prior history of GIB or peptic ulcer disease.^[2,3] In addition, the most recent guidelines from the United States Preventative Service Task Force and the American College of Cardiology/American Heart Association and the United States Preventative Service Task Force have recommended further limiting the use of primary prevention aspirin on the basis of recent large, randomised controlled trials that did not show significant benefits of low-dose aspirin for primary prevention of cardiovascular events^[4–7] but demonstrated increases in GIB. As a result, the American College of Cardiology/American Heart Association and the United States Preventative Service Task Force now does not make a recommendation for the routine use of aspirin for primary cardiovascular prevention.^[8,9] Rather they indicate aspirin may be considered or decisions individualised in those with increased cardiovascular risk but recommend against initiating low-dose aspirin in individuals older than 60–70 years and those with increased GIB risk.

Though the absolute risk of bleeding associated with primary prevention aspirin is low,^[10,11] with an estimated 29 million patients on primary prevention aspirin in the United States alone,^[12] the burden of GIB hospitalisations in patients on primary prevention aspirin is likely significant. Based on the current guideline recommendations outlined above, patients with GIB who are taking aspirin for primary prevention should have their aspirin deprescribed. Deprescription refers to the systematic and supervised process of dose reduction or drug discontinuation with the goal of health improvement and/or reducing the risk of side effects.^[13] As deprescription patterns of primary prevention aspirin among patients hospitalised with GIB remain unknown, in our study, we sought to determine rates of aspirin deprescription among hospitalised patients in a large, tertiary, academic centre and explore long-term outcomes including a second hospitalisation for major adverse cardiovascular events (MACE) and GIB.

Table 1. Baseline characteristics of hospitalised patients with gastrointestinal bleeding on primary prophylaxis aspirin
Demographics
Age	72 (61–81)
Sex, male	149 (46.6%)
Race
White/Caucasian	238 (74.4%)
Black/African American	54 (16.9%)
Asian or Pacific Islander	8 (2.5%)
American Indian or Alaska Native	2 (0.6%)
Other	15 (4.7%)
Unknown	3 (0.9%)
Ethnicity
Hispanic or Latino	18 (5.6%)
Not Hispanic or Latino	300 (93.8%)
Unknown	2 (0.6%)
Daily aspirin dose
81 mg	297 (92.8%)
162 mg	1 (0.3%)
325 mg	22 (6.9%)
Medical comorbidities
Hypertension	278 (86.9%)
Chronic obstructive pulmonary disease	103 (32.2%)
Diabetes mellitus	137 (42.8%)
Chronic kidney disease	83 (25.9%)
Peptic ulcer disease	44 (13.8%)
Helicobacter pylori infection	5 (1.6%)
Cirrhosis	21 (6.6%)
COVID-19	25 (7.8%)
Medications
NSAIDs	16 (5.0%)
SSRIs	45 (14.1%)
Antiplatelets	1 (0.3%)
Anticoagulants	14 (4.4%)

Note: Median and interquartile range were used for continuous variables. N (%) was used for categorical variables.
Abbreviations: NSAIDs, non-selective anti-inflammatory drugs; SSRIs, selective serotonin reuptake inhibitors.

Table 2. Hospitalisation and endoscopic characteristics for patients with gastrointestinal bleeding on primary prophylaxis aspirin
Hospitalisation characteristics
Admitting service
Medicine	289 (90.3%)
Surgery	29 (9.1%)
Other^a	2 (0.6%)
ICU admission	95 (29.7%)
GIB indication	45 (14.1%)
Gastroenterology consultation	251 (78.4%)
Cardiology consultation	38 (11.9%)
Red blood cell transfusion	142 (44.4%)
Endoscopic evaluation	172 (53.8%)
Salvage embolisation	16 (5.0%)
In-hospital mortality	41 (12.8%)
Mortality due to GIB	9 (2.8%)
Hospital length of stay, days	5 (3–9)
ICU length of stay, days^b	3 (1–8)
Endoscopic characteristics
Bleeding cause
Ulcer	61 (35.5%)
Angioectasia	4 (2.3%)
Varices	4 (2.3%)
Dieulafoy	2 (1.2%)
Malignancy	6 (3.5%)
Diverticular	31 (18.0%)
Hemorrhoidal	10 (5.8%)
Other^c	32 (18.6%)
Unknown	22 (12.8%)
Location
Oesophagus	20 (11.6%)
Stomach	30 (17.4%)
Small bowel	35 (20.3%)
Colon	64 (37.2%)
Unknown	23 (7.2%)
Endoscopic procedures performed
EGD	136 (61.0%)
Colonoscopy	81 (36.3%)
Enteroscopy	4 (1.8%)
ERCP	2 (0.9%)
Endoscopic haemostasis performed	60 (26.9%)

Note: Median and interquartile range were used for continuous variables. N (%) was used for categorical variables. One hundred seventy-two unique patients underwent endoscopic evaluation and underwent a total of 223 procedures.
Abbreviations: EGD, esophagogastroduodenoscopy; ERCP, endoscopic retrograde cholangiopancreatography; ICU, intensive care unit; GIB, gastrointestinal bleeding.
^aAdmitted to Neurology (n = 1) and OB-GYN (n = 1).
^bFor 95 patients requiring ICU admission.
^cOther identified causes of bleeding: esophagitis (n = 9), erosive gastropathy (n = 5), ischemic colitis (n = 5), post-sphincterotomy bleed (n = 2), post-polypectomy bleed (n = 2), portal hypertensive gastropathy (n = 2), ulcerative colitis (n = 2), radiation proctitis (n = 1), ulcerated polyp (n = 1), Mallory-Weiss tear (n = 1), orogastric tube trauma (n = 1), diverticulitis (n = 1).

Table 3. Baseline characteristics of patients deprescribed or continued aspirin with follow-up after the index hospitalisation
	Aspirin deprescribed (n = 25)	Aspirin continued (n = 235)	p-value
Age	68 (60–77)	72 (60–80)	0.44
Sex, male	11 (40%)	100 (43%)	1.00
Comorbidities
Hypertension	21 (84%)	203 (86%)	0.76
Chronic obstructive pulmonary disease	4 (16%)	83 (35%)	0.07
Diabetes mellitus	6 (24%)	109 (46%)	0.04
Chronic kidney disease	7 (28%)	62 (26%)	0.82
Peptic ulcer disease	5 (20%)	32 (14%)	0.37
Helicobacter pylori infection	0 (0%)	5 (2%)	1.00
Cirrhosis	1 (4%)	10 (4%)	1.00
COVID-19	0 (0%)	17 (7%)	0.39
Medications
NSAIDs	2 (8%)	8 (3%)	0.25
Antiplatelets	0 (0%)	1 (0.4%)	1.00
Anticoagulants	3 (12%)	7 (3%)	0.06

Note: Median and interquartile range were used for continuous variables. N (%) was used for categorical variables. p-value represents a comparison between aspirin deprescribed and aspirin continued groups.
Abbreviation: NSAID, non-selective anti-inflammatory drug.

My Alerts

Add to Email Alerts

Your Name is required.

Subject is required.

Please enter a Recipient Address and/or check the Send me a copy checkbox.

Your email has been sent.

is an Invalid Email Address.

Deprescription of Aspirin for Primary Prevention Is Uncommon at Discharge in Hospitalised Patients With Gastrointestinal Bleeding

Abstract and Introduction

Abstract

Introduction

Tables

Tables

Tables

References

Authors and Disclosures

Authors and Disclosures

Email This

Feedback