Protective effects of long term antiplatelet and anticoagulant therapy in hospitalized patients with inflammatory bowel disease

Abstract

BACKGROUND

Inflammatory bowel disease (IBD), with its rising prevalence rates is associated with an increased risk of cardiovascular and thromboembolic events. Antiplatelets and/or anticoagulants agents are often prescribed but the literature on the impact of long-term anticoagulation and/or antiplatelet use among patients hospitalized with IBD is scarce. The aim of this study is to assess the outcomes of patients hospitalized with IBD on antiplatelet and/or anticoagulant agents.

AIM

To investigate the effects of long-term use of antiplatelets/anticoagulants on clinical outcomes in patients hospitalized with IBD.

METHODS

We conducted a retrospective cohort study using the Nationwide Inpatient Sample database, including all adult IBD patients hospitalized in the United States from 2016 to 2019. Patient cohorts were stratified based on antiplatelet/anticoagulant therapy status. Multivariate regression analysis was done to assess outcomes, adjusting for potential confounders. The primary outcome was mortality, whereas length of stay (LOS), total parenteral nutrition, acute kidney injury, sepsis, shock, gastrointestinal bleeding, need for colonoscopy/sigmoidoscopy, abdominal surgery and total hospitalization charges were secondary outcomes.

RESULTS

Among 374744 hospitalized IBD patients, antiplatelet or anticoagulant therapy alone was associated with significantly lower in-hospital mortality and reduced healthcare utilization, including shorter LOS and decreased hospitalization costs. Combined therapy was associated with a protective effect on mortality, but did not reach statistical significance. Notably, therapy did not exacerbate disease severity or complications, although higher odds of gastrointestinal bleeding were observed.

CONCLUSION

Our study highlights the potential benefits of long-term anticoagulation/antiplatelet therapy in hospitalized IBD patients, with improved mortality outcomes and healthcare utilization. While concerns regarding gastrointestinal bleeding exist, the overall safety profile suggests a role for these agents in mitigating thromboembolic risks without exacerbating disease severity. Further research is needed to look at optimal treatment strategies and addressing limitations to guide clinical decision-making in this population.

Key Words: Inflammatory bowel disease; Anticoagulation; Antiplatelet therapy; Healthcare utilization; Mortality

Core Tip: In this retrospective study, hospitalized inflammatory bowel disease (IBD) patients on long-term antiplatelet or anticoagulant therapy demonstrated benefits in terms of inpatient mortality, length of stay, and total hospitalization charges compared to those not on such therapy. While the use of anticoagulation alone or in combination with antiplatelet agents was linked to higher odds of lower gastrointestinal bleeding, there was no significant difference in the development of acute kidney injury, sepsis, or the use of total parenteral nutrition. Further research is needed to optimize treatment strategies and inform clinical decision-making in the IBD population.

INTRODUCTION

Inflammatory bowel disease (IBD) is a chronic inflammatory disorder of the gastrointestinal tract (GIT), comprised of ulcerative colitis (UC) and Crohn’s disease (CD). CD can affect the entire GIT with transmural inflammation whereas UC is limited to mucosal layers of the rectum and colon[1]. In a systematic analysis for global burden of disease study 2019, age standardized prevalence rate of IBD was reported 59.25 per 100000 population[2].

In a systematic review of studies reporting cerebrovascular accidents (CVA) and ischemic heart disease (IHD) incidents among patients with IBD, findings indicate a higher risk of CVA and IHD in this population[3]. Yarur et al[4], conducted a prospective longitudinal study revealing an increased incidence of coronary artery disease(CAD) events in IBD patients, despite having a lower prevalence of traditional risk factors for CAD. Similarly, a large Danish nationwide study conducted to evaluate the risk of coronary heart disease in patients with IBD revealed higher CAD risk in IBD patients compared to non IBD population, after adjusting for all the potential confounders[5]. The incidence of IBD is rising with an increase of 47.45% in global prevalent cases of IBD reported between 1990 and 2019. This represented a rise in global prevalent cases of IBD from 3.32 million to 4.9 million[2]. This rising surge means that patients with IBD are also at increased risk of developing early atherosclerosis, cardiovascular diseases such as myocardial infarction, and arrhythmias like atrial fibrillation (AF)[6]. Long term antiplatelet therapy, with its proven morbidity and mortality benefits in IHD patients, is frequently incorporated into the medication regimens of patients with IBD[7].

Findings of a large database study by Nelson et al[8] revealed that patients with IBD were found to have higher odds of associated thromboembolic disorders compared to those without IBD[8]. Long term systemic anticoagulation is therefore oftentimes a part of the medication regimen in patients with IBD with history of venous thromboembolism (in addition to those with AF, mechanical heart valves, etc.). Limited data available on use of aspirin on patents hospitalized with IBD shows aspirin use was associated with lower risk of death, sepsis and shock in an analysis of Nationwide Inpatient Sample (NIS) database[9]. Similar studies on the outcomes of hospitalization among patients with IBD who are on antiplatelet agents vs anticoagulation vs combined antiplatelets and anticoagulation are lacking.

There is paucity of literature on the impact of long-term anticoagulation and/or antiplatelet use among patients hospitalized with IBD[9]. Therefore, we conducted this study to assess outcomes among patients on long term anticoagulation and/or antiplatelets agents hospitalized with IBD.

MATERIALS AND METHODS

Study design and database description

We conducted a retrospective cohort study with adult patients hospitalized with IBD in the United States from the year 2016 to 2019. Patient data was obtained from NIS database which is a part of databases developed for Healthcare-Cost and Utilization Project (HCUP). The NIS is largest publicly available all-payer inpatient healthcare database which estimates United States regional and national inpatient utilization, access, cost, quality and outcomes from all nonfederal acute care hospitals nationwide. NIS approximates a 20-percent stratified sample of all discharges from United States community hospitals, excluding rehabilitation and long-term acute care hospitals. The NIS contains anonymized clinical and resource-use information that is included in a typical discharge with safeguards to protect the privacy of patients, physicians and hospitals. Unweighted, NIS contains data from around 7 million hospital stays and around 35 million weighted estimates of hospitalizations nationwide. Clinical and non-clinical information obtained from NIS include International Classification of Diseases, ninth revision, clinical modification (ICD-10-CM)/procedure coding system diagnosis, procedure, and external causes of morbidity codes from 2016-2019, anonymized patient demographic characteristics (like age, sex, race, median household income for Zip code), hospital characteristics (like ownership), expected payment source, total charges, length of stay, severity and comorbidity measures, etc.[10,11].

Ethical statement

The study was conducted in compliance with the set of ethical principles of Declaration of Helsinki. NIS database contained anonymized clinical and resource-use information with safeguards to protect the privacy of patients, physicians and hospitals. As patient data was de-identified, institutional review board approval and written consent was not required before proceeding with the study.

Study population

Patients aged 18 years or more years, admitted from 2016 to 2019 with principal admission diagnosis of IBD were identified using ICD-10-CM coding system from NIS database. Patients aged < 18 years were excluded from the study. ICD-10 code K50 and K51 was used to query the NIS database 2016-2019 to identify patients admitted with principal diagnosis of CD and UC respectively. Patients admitted with principal admission diagnosis of IBD (CD/UC) were then stratified into four groups: Patients not on any antiplatelet or anticoagulants, patients on long term antiplatelet agent only, patient on long term anticoagulant agents only and patients on both long term antiplatelet and anticoagulant agents. Patients taking long term anticoagulants/antiplatelets on daily outpatient basis were identified using appropriate ICD-10 codes. ICD-10 code Z7902 and Z7982 were used to identify patients on long term antiplatelet agents whereas ICD-10 code Z7901 was used to identify patients on long term anticoagulant agents[9,12,13]. Combination of ICD-10 codes for patients on long term antiplatelet agents and anticoagulant agents were used to query the patient subgroup on both antiplatelet and anticoagulant agents. A total of 374744 hospitalizations with principal admission diagnosis of IBD were included in the study. Flow diagram depicting the distribution of study participants is presented in Figure 1 below.

Figure 1 Flow diagram depicting the distribution of study participants. IBD: Inflammatory bowel disease.

Study variables and outcomes

The study variables encompass patient demographic details, hospital characteristics, and the Charlson Comorbidity Index (CCI). Demographic information collected from the database included age, sex, race, insurance provider, and the median income of the patient’s zip code. Hospital characteristics considered in the study are the hospital's region, teaching status, bed size, and location. The patient's comorbidity status was categorized using the Deyo adaptation of the CCI, which is specifically designed for research that utilizes ICD diagnosis and procedure codes. The Deyo adaptation of the CCI comprises 19 categories of clinical comorbid conditions, each assigned a predetermined score. The CCI is used to predict a patient's 10-year mortality risk based on their CCI score[14].

The primary outcome of the study was mortality where secondary outcomes included length of stay, total parenteral nutrition (TPN), acute kidney injury (AKI), sepsis, shock, lower GI bleeding, non-variceal upper GI bleeding, colonoscopy/sigmoidoscopy (with biopsy), abdominal surgery, total hospitalization charges and discharge disposition. Mortality, length of stay and total hospitalization charge were directly coded within the NIS for each hospitalization. Rest of the secondary outcomes were coded using appropriate ICD-10 codes (available in Supplementary Table 1). Potential confounders addressed in the study included age, sex, race, median income in patient’s zip code, CCI, hospital bed size, hospital region, hospital teaching status, hospital location, pulmonary embolism, deep vein thrombosis and AF.

Statistical analysis

We used Stata^® Version 18 software (StataCorp, Texas, United States) for data analysis and all analyses were conducted using the weighted samples for national estimates in adjunct with HCUP regulations for using the NIS database. HCUP rules and regulations were thoroughly followed in utilizing the NIS database. Hospitalization characteristics, including the mean age and the distribution of variables such as race, gender, hospital location, teaching status, and bed capacity, were gathered from the database. Continuous variables were presented as mean ± SD and analyzed using Student’s t-test, while categorical variables were reported as frequencies and percentages and evaluated using the Pearson chi-square method. To identify potential confounders, we conducted a comprehensive review of existing literature on IBD, antiplatelets, and anticoagulation, sought expert opinions from specialists in these fields, and carried out a univariate screening. All the variables with P value less than 0.2 in the univariate screening were included in multivariate regression model to adjust for confounders for calculation of primary and secondary outcomes. During multivariate logistic and linear regressions, outcomes were adjusted for age, race, Charlson comorbidity groups, hospital location, hospital region, hospital bed size, hospital teaching status, insurance status, quartile of household income, as well as additional comorbidities not included for in the Charlson index including AF, DVT, and Pulmonary Embolism. To compare the outcomes across different groups, Multivariate Linear regression analysis was used for continuous variables and Multivariate Logistic regression analysis was used for binary/dichotomous variables. All P-values <0.05 were considered statistically significant. The statistical methods of this study were reviewed by Suruchi Jai-Kumar-Ahuja from Department of Genomics and Data sciences, Sarepta Therapeutics, Cambridge, MA, United States.

RESULTS

Demographic characteristics

A total of 374744 patients hospitalized with principal diagnosis of IBD were included in the study which were further stratified into four groups based upon use of antiplatelet/anticoagulant agents. Figure 1 depicts the distribution of study population into those four groups. Although numerically small difference, patients not receiving drug therapy were more likely to be females (53.9%) compared to the other three groups. Patients on combination antiplatelet and anticoagulant agents were more likely to have Medicare (72.47%) and higher CCI (3 or more, 37.81%) than rest of the three groups. Numerically small but statistically significant differences were also noted among patient in four groups in terms of median income, hospital region, hospital teaching status, hospital bed size, race and hospital location. Table 1 demonstrates the patient and hospital characteristics.

Table 1 Patient and hospital characteristics (n = 374744), n (%).

Variable	No drug therapy	Antiplatelet therapy	Anticoagulant therapy	Antiplatelet and anticoagulant therapy	P value
Female	53.9	52.06	52.73	43.28	< 0.001
Male	46.1	47.94	47.27	56.72	< 0.001
Mean age in years	44.02	65.55	58.8	67.97	< 0.001
Insurance provider					< 0.001
Medicare	23.61	63.1	53.73	72.47
Medicaid	20.27	7.6	13.19	8.67
Private	50.01	27.9	30.94	17.09
Uninsured	6.11	1.41	02.14	0.77
Charlson comorbidity index					< 0.001
0	69.24	29.97	41.47	17.16
1	18.68	26.87	22.51	24.13
2	6.36	18.62	15.82	20.9
3 or more	5.72	24.54	20.19	37.81
Median income in patients zip code					< 0.0178
$1-51999	25.8	25.15	23.8	22.65
$52000-65999	25.1	27.3	25.85	30.79
$66000-87999	25.73	25.18	26.59	23.92
$88000+	23.38	22.37	23.76	22.65
Hospital region					< 0.001
Northwest	21.42	19.89	22.13	19.65
Midwest	23.93	27.93	28.43	29.35
South	38.18	36.47	32.84	37.56
West	16.46	15.7	16.6	13.43
Hospital teaching status					0.0001
Non-teaching	26.15	29.81	26.27	25.37
Teaching	73.85	70.19	73.73	74.63
Hospital bed size					0.0224
Small	19.27	21.22	19.3	22.39
Medium	27.89	27.24	28.67	31.59
Large	52.84	51.54	52.03	46.02
Race					< 0.001
White	72.46	82.39	79.82	84.62
Black	14.39	9.66	11.89	10.77
Hispanic	08.42	4.33	5.23	2.82
Asian or Pacific Islander	1.5	1.12	0.96	1.03
Native American	0.39	0.52	0.2	0.26
Other	2.85	1.99	1.91	0.51
Hospital location					< 0.001
Rural	6.7	8.91	6.73	9.7
Urban	93.3	91.09	93.27	90.3

Primary outcome

The incidence of in-hospital mortality was highest among patients using combined antiplatelet and anticoagulant agents (0.75%) followed by those using anticoagulant (0.62%) and antiplatelet agents alone (0.42%). After adjusting for patient and hospital level confounders on multivariate analysis, use of combined antiplatelet and anticoagulant agents was not found to be associated as an independent predictor of in-hospital mortality, however there was a trend toward a protective benefit [adjusted odds ratio (aOR): 0.23 with P value = 0.052]. However, use of either antiplatelet or anticoagulant agents was found to be associated with lower in-hospital mortality by 60% [aOR: 0.40 (0.23-0.70) with P value = 0.001] and 59% [aOR: 0.41 (0.22-0.75) with P value = 0.004] respectively compared to those on no drug therapy on multivariate analysis.

Secondary outcomes

Hospital length of stay and total hospitalization charge were used as surrogate measures of healthcare utilization. The mean length of stay (5.87 days) and total hospitalization charges ($54912) were reported higher among the patients taking anticoagulant therapy compared to all other groups. The mean length of stay was lowest among control group of patients not receiving antiplatelet or antiplatelet agents (4.91 days) and mean total hospitalization charges were lowest among patients receiving antiplatelet agents ($45978). The incidence of AKI (21.14%), shock (4.48%), lower GI bleeding (6.72%), colonoscopy (33.33%) was higher among the group using both antiplatelet and anticoagulant agents. TPN use (3.33%) and abdominal surgery (9.13%) was found to highest among group using anticoagulant agents along. On multivariate analysis, use of either antiplatelet or anticoagulant agent alone or in combination was not found to be independent predictor for development of AKI, sepsis, shock, or use of TPN except use of anticoagulant alone which was associated with reduced odds of developing shock by 36 % [aOR: 0.64 (0.43-0.94) with P value = 0.025]. Use of anticoagulant agent alone or combined use of antiplatelet and anticoagulant agents was associated with higher odds of lower GI bleeding [aOR: 1.34 (1.09-1.65) with P value = 0.004 and aOR: 1.56 (1.03-1.65) with P value = 0.035 respectively] compared to those not on any drug therapy on multivariate analysis. However, odds of colonoscopy were higher among groups using antiplatelet agent alone [aOR: 1.16 (1.07-1.26) with P value = < 0.001] and those using combined antiplatelet and anticoagulant agents [aOR: 1.31 (1.04-1.66) with P value = 0.019] compared to those not on any drug therapy. The cohorts of patients on antiplatelet, anticoagulant and combined antiplatelet and anticoagulant were associated with higher likelihood of home discharge with or without home health care [adjusted OR of 1.23 (1.10-1.37) with P value < 0.001, 1.32 (1.14-1.53) with P value < 0.001 and 1.47 (1.07-2.01) with P value = 0.015 respectively] compared to the cohort of patients not receiving either of these agents. Association of discharge to short term hospital was significantly lower in the patients taking long term antiplatelet compared to patients without it with adjusted OR of 0.56 (0.39-0.81) with P value = 0.002. Similarly, for the cohort of patients on anticoagulant group there was a significantly lower association of discharge to skilled nursing facility and other intermediate care facilities with adjusted OR of 0.8 (0.66-0.95) with P value = 0.016 compared to patients not on long term anticoagulants. Table 2 depicts outcomes of hospitalization stratified into four groups based on use of antiplatelet/anticoagulant agents. Table 3 depicts the results of multivariate regression analysis after adjusting for confounders.

Table 2 Outcomes of hospitalization stratified into four groups based on use of antiplatelet/anticoagulant agents (n = 374744), n (%).

Outcomes	No drug therapy	Antiplatelet therapy	Anticoagulant therapy	Antiplatelet and anticoagulant therapy	P value
Mortality	0.27	0.42	0.62	0.75	0.0016
LOS (days)	4.91	5.01	5.87	5.61	< 0.001
TPN	2.23	1.59	3.33	2.74	0.0001
AKI	7.33	16.13	15.2	21.14	< 0.001
Sepsis	2.0	1.27	1.47	1.24	0.0025
Shock	0.7	1.22	1.55	4.48	< 0.001
Colonoscopy	23.37	27.8	25.96	33.33	< 0.001
Lower GI bleed	3.04	3.95	5.34	6.72	< 0.001
Non variceal Upper GI bleed	0.81	0.95	1.04	1.49	0.1905
Abdominal surgery	10.05	7.53	9.13	5.47	< 0.001
Total charge ($)	46800	45978	54912	50569	< 0.001
Home or home health care	92.84	88.17	86.69	82.84	< 0.001
Short term hospital	1.55	1.03	1.70	1.53	< 0.001
SNF or other facility	2.98	9.52	9.48	13.93	< 0.001

LOS: Length of stay; TPN: Total parenteral nutrition; AKI: Acute kidney injury; GI: Gastrointestinal; SNF: Skilled nursing facility.

Table 3 Results of multivariate regression analysis after adjusting for confounders.

Outcome	Unadjusted OR/coefficient (95%CI)	P value	Adjusted OR/coefficient (95%CI)	P value
Mortality (OR)
Antiplatelet	1.55 (0.93-2.60)	0.09	0.40 (0.23-0.70)	0.001
Anticoagulant	2.27 (1.36-3.80)	0.002	0.41 (0.22-0.75)	0.004
Antiplatelet and anticoagulant	2.74 (0.87-8.67)	0.085	0.23 (0.05-1.01)	0.052
LOS days (coefficient)
Antiplatelet	0.096 (-0.06-0.25)	0.249	-0.46 (-0.63 to -0.28)	< 0.001
Anticoagulant	0.96 (0.71-1.20)	0.000	-0.502 (-0.78 to -0.22)	< 0.001
Antiplatelet and anticoagulant	0.698 (0.06-1.32)	0.030	-0.95 (-1.62 to -0.27)	< 0.001
AKI (OR)
Antiplatelet	2.43 (2.21-2.66)	< 0.001	0.95 (0.85-1.05)	0.354
Anticoagulant	2.26 (2.02-2.53)	< 0.001	0.99 (0.86-1.14)	0.942
Antiplatelet and anticoagulant	3.38 (2.66-4.31)	< 0.001	0.92 (0.69-1.22)	0.586
Sepsis (OR)
Antiplatelet	0.63 (0.47-0.84)	0.002	0.78 (0.57-1.06)	0.117
Anticoagulant	0.73 (0.52-1.00)	0.057	0.77 (0.54-1.10)	0.154
Antiplatelet and anticoagulant	0.62 (0.25-1.48)	0.279	0.78 (0.32-1.94)	0.586
Shock (OR)
Antiplatelet	1.76 (1.29-2.39)	< 0.001	0.75 (0.53-1.05)	0.103
Anticoagulant	2.24 (1.62-3.09)	< 0.001	0.64 (0.43-0.94)	0.025
Antiplatelet and anticoagulant	6.68 (4.14-10.76)	< 0.001	1.52 (0.88-2.63)	0.129
TPN (OR)
Antiplatelet	0.7 (0.54-0.92)	0.01	0.76 (0.57-1.01)	0.064
Anticoagulant	1.5 (1.21-1.87)	< 0.001	1.07 (0.83-1.40)	0.569
Antiplatelet and anticoagulant	1.23 (0.64-2.33)	0.522	1.05 (0.53-2.077)	0.877
Lower GI bleed (OR)
Antiplatelet	1.31 (1.10-1.55)	0.002	1.05 (0.87-1.26)	0.578
Anticoagulant	1.79 (1.50-2.14)	< 0.001	1.34 (1.09-1.65)	0.004
Antiplatelet and anticoagulant	2.29 (1.58-3.40)	< 0.001	1.56 (1.03-2.37)	0.035
Non-variceal upper GI bleed (OR)
Antiplatelet	1.18 (0.84-1.66)	0.325	1.13 (0.78-1.65)	0.505
Anticoagulant	1.29 (0.87-1.91)	0.189	1.03 (0.64-1.64)	0.891
Antiplatelet and anticoagulant	1.86 (0.82-4.18)	0.132	1.47 (0.62-3.46)	0.376
Colonoscopy (OR)
Antiplatelet	1.26 (1.17-1.36)	< 0.001	1.16 (1.07-1.26)	< 0.001
Anticoagulant	1.14 (1.04-1.26)	0.003	0.99 (0.90-1.26)	0.95
Antiplatelet and anticoagulant	1.63 (1.32-2.03)	< 0.001	1.31 (1.04-1.66)	0.019
Abdominal surgery (OR)
Antiplatelet	0.73 (0.64-0.82)	< 0.001	0.836 (0.72-0.96)	0.014
Anticoagulant	0.89 (0.78-1.02)	0.123	0.879 (0.74-1.03)	0.118
Both antiplatelet and anticoagulant	0.51 (0.33-0.79)	0.03	0.623 (0.39-0.97)	0.037
Total charge ($) (coefficient)
Antiplatelet	-822.2 (-2777-1132)	0.41	-4889 (-7126 to -2652)	< 0.001
Anticoagulant	8112 (5015-11208)	< 0.001	-7094 (-10979 to -3209)	< 0.001
Antiplatelet and anticoagulant	3768 (-3105-10643)	0.283	-11546 (-19475 to -3617)	< 0.001
Discharge to home or home health care (OR)
Antiplatelet	0.57 (0.51-0.63)	< 0.0001	1.23 (1.10-1.37)	< 0.001
Anticoagulant	0.50 (0.44-0.56)	< 0.0001	1.32 (1.14-1.15)	< 0.001
Antiplatelet and anticoagulant	0.37 (0.28-0.48)	< 0.0001	1.47 (1.07-2.01)	0.015
Discharge to short term hospital (OR)
Antiplatelet	0.66 (0.47-0.92)	0.015	0.56 (0.39-0.81)	0.02
Anticoagulant	1.10 (0.81-1.49)	0.535	0.84 (0.59-1.12)	0.362
Antiplatelet and anticoagulant	1.12 (0.53-2.38)	0.756	0.76 (0.33-1.74)	0.521
Discharge to SNF or other facility (OR)
Antiplatelet	3.42 (0.05-3.08)	< 0.001	0.94 (0.83-1.07)	0.403
Anticoagulant	3.40 (2.96-3.92)	< 0.001	0.80 (0.66-0.95)	0.016
Antiplatelet and anticoagulant	5.26 (3.93-7.05)	< 0.001	0.78 (0.55-1.10)	0.169

OR: Odds ratio; LOS: Length of stay; TPN: Total parenteral nutrition; AKI: Acute kidney injury; GI: Gastrointestinal; SNF: Skilled nursing facility.

DISCUSSION

This retrospective cohort study of patients with IBD who were admitted to the hospital found to have significantly lower inpatient mortality associated with the use of either antiplatelet or anticoagulant agents alone on daily basis. Patients receiving daily combined antiplatelet and anticoagulant therapy on an outpatient basis had a trend towards benefit with regard to inpatient mortality, although this did not reach statistical significance. In addition, use of antiplatelet or anticoagulant agents alone or in combination was associated with significantly decreased burden of healthcare utilization as measured by length of stay and total hospitalization charge. Furthermore, use of antiplatelet or anticoagulant agents alone or in combination was not associated with more severe disease course measured in terms of development of AKI, sepsis, shock or use of TPN (rather use of anticoagulant alone was associated with lower odds of developing shock). Conversely, use of long-term anticoagulant agent alone or in combination with antiplatelet agents was associated with significant higher odds of GI bleeding. Also, use of antiplatelet or anticoagulant alone or in combination was found to be associated with higher likelihood of discharge to home or receive home health care compared to patients not on any drug therapy. Among all IBD patients, nearly one-fourth of the subgroups received care in non-teaching hospitals, while the rest were managed in teaching hospitals. Regarding insurance providers, most patients not receiving long-term therapy had private insurance, whereas those on long-term antiplatelet, anticoagulant, or both therapies were predominantly covered by Medicare.

Venous thromboembolism (VTE) is a common extraintestinal complication in hospitalized IBD patients particularly during active disease flares, with the risk significantly higher compared to the general population, especially in hospital settings[15,16]. Moderate to severe IBD is therefore considered a risk factor for the development of VTE and routine anticoagulant thromboprophylaxis is recommended during IBD hospitalization[15]. Additionally, risk of CVA and IHD is also higher among patients with IBD[3]. In a study reporting utilization of anticoagulant thromboprophylaxis among patients admitted with IBD (from Southern Health Network Database, Australia, 1998-2009), it was found that half of the hospitalized patients suffering venous thromboembolic episodes did not receive thromboprophylaxis. Subsequently, a significantly higher mortality and prolonged length of stay was reported among those patients with a significant proportion requiring ICU admission[17]. Although this study is done among patients with IBD who were on long term therapeutic anticoagulation, findings from this study underscore the importance of anticoagulation/antiplatelet therapy in IBD for reducing mortality and healthcare burden.

Several factors have been implicated behind the increased risk for VTE. Increased plasma levels of factor V, VII, VIII and fibrinogen as acute phase reactants in response to underlying chronic inflammation create a prothrombotic environment. Low levels of tissue plasminogen activator and increased levels of plasminogen activator inhibitor and thrombin activated fibrinolysis inhibitor leads to hypo-fibrinolysis. In addition, decrease in anticoagulant factors like protein C, protein S, antithrombin III and tissue factor pathway inhibitor is also reported in IBD. Hyperhomocysteinemia likely resulting from low levels of folate, B6 and B12 in IBD also creates a thrombogenic environment[18,19]. Damaged intestinal mucosa providing bloodstream access to endotoxin originating from intestinal flora (lipopolysaccharide), elevated inflammatory cytokines (TNFa, IL-1b and IL-6), reactive thrombocytosis, enhanced platelet activation, increased adhesivity to vascular endothelium and tendency of platelets to aggregate are some other factors leading to increased predisposition to thrombus formation[20]. Lastly, factors like dehydration, malnutrition, placement of central venous catheters, severe disease activity, hospitalization, surgery, corticosteroid use, advanced age, obesity, pregnancy, etc. increase the risk of VTE in IBD patients[19,21,22].

Studies have been conducted in an attempt to evaluate the risk of bleeding among patients with IBD on anticoagulation. Scharrer et al[23] in a multicentric cohort study evaluated the risk of major bleeding events among patients with IBD on therapeutic anticoagulation via telephone interview and chart review. Major bleeding was defined as fatal bleeding or symptomatic bleeding in critical area/organ or leading to more than equal to 2 gm/dL drop in hemoglobin or leading to transfusion of more than equal to 2 units of whole blood or red cells. The risk for major but non-fatal bleeding was found increased in patients with IBD on anticoagulation. However, majority of the patients had active disease with mucosal lesion on endoscopy at the time of major bleeding[23]. This is in contrast to a study by Lobo et al[24] where no difference in major bleeding were reported among patients with IBD who developed VTE during acute flare, during remission and patients without IBD. Therapeutic anticoagulation in IBD was therefore, considered as effective and safe in patients without IBD[24]. In our study, with a larger cohort of IBD patients, we observed higher odds of lower GI bleeding among patients on anticoagulant agent alone or on combined antiplatelet and anticoagulant therapy. However, use of antiplatelet or anticoagulant agents alone or in combination was not associated with more severe disease course despite higher odds of lower GI bleeding. There was no significant difference in association with regards to non-variceal upper GI bleeding. Given the diagnostic and therapeutic benefits of colonoscopy, it remains the preferred intervention for lower gastrointestinal bleeding in hospitalized patients[25]. Our study also found higher odds of colonoscopy among patients on antiplatelet therapy alone and those on combined antiplatelet and anticoagulant therapy, which further supports the likelihood of colonoscopic interventions in this subgroup.

Among hospitalized IBD population, patients with UC are more prone to develop life-threatening complications like colonic perforation, gastrointestinal hemorrhage, and toxic megacolon which requires immediate abdominal surgery[26]. We reported lower odds of requirement of emergent abdominal surgery among patients on antiplatelet agents alone and those on combined antiplatelet and anticoagulant agents. Lower odds of requirement of abdominal surgery translates to lower chances of post-surgical complications and subsequent morbidity among patients hospitalized with IBD. Our study demonstrated higher odds of routine discharge to home with or without home health care in all the three groups including antiplatelet, anticoagulant and combined antiplatelet and anticoagulant group compared to patients not receiving either of these drugs. Also, patients taking antiplatelet agents were associated with lower odds of transfer to short term hospital compared to patients not on it. There was a significantly lower association of transfer to skilled nursing facility or any other intermediate care facility among patients taking long term anticoagulant agents compared to those not taking it. These statistically significant discharge disposition results favor the association of beneficiary effects of these long-term agents on overall safety, reduced morbidity and healthcare burden among the patients admitted with IBD.

The study has some inherent limitations. We used the NIS database to obtain information on study population. NIS being an administrative database, has a potential for misclassification of disease using ICD-10 codes or missing codes. However, previous studies have used similar ICD-10 codes to obtain information regarding long term use of antiplatelet/anticoagulant agents used[9,12,13]. The ICD-10 codes were used to query for long term antiplatelets/anticoagulant agents use. However, information regarding indication for antiplatelets agents use i.e. primary vs secondary prevention, single vs dual antiplatelets agents use, compliance could not be obtained. Similarly, information regarding type of the anticoagulant or antiplatelets used, dosage of anticoagulant agents used, specific regimens of anticoagulants (low molecular weight heparin vs direct oral anticoagulants) their indications, duration of use, and compliance could not be obtained which can be potential limitations for this study. The severity of IBD at hospitalization, clinical details about outcomes like GI bleed (occult bleeding vs major bleeding vs fatal bleeding, etc.) could not be obtained due to the administrative nature of the NIS database which can be another limitation. Lastly, information about use of proton pump inhibitors, non-steroidal anti-inflammatory agents, objective laboratory parameters at hospitalization could not be obtained from the database which may influence the outcomes.

CONCLUSION

Despite the limitations of the study, this is one of the first population-based studies done using the largest publicly available national database to evaluate the outcomes of patients with IBD on long term anticoagulation/antiplatelet therapy. The findings of the study demonstrated the benefits in terms of mortality, length of stay and total charges of hospitalization. Although anticoagulation use alone or combined anticoagulation and antiplatelet agent use was associated with higher odds of lower GI bleeding, it was not associated with increased severity of disease course like development of AKI, sepsis, shock or use of TPN. The overall safety profile suggests a role for these agents in mitigating thromboembolic risks without exacerbating disease severity. Further research is needed to look at optimal treatment strategies and addressing limitations to guide clinical decision-making in this population.