Introduction

Acute upper gastrointestinal bleeding (AUGIB) is a common, life-threatening clinical emergency with considerable morbidity and mortality. It is postulated that AUGIB can lead to hypotension and hypoperfusion of the vital organs. In response to these changes, the body activates its compensatory mechanism deriving in higher levels of oxidative stress, inflammation, hypercoagulable state and vascular endothelial injury. All these put these patients in a higher risk of myocardial injury (MI), which may lead to increased mortality in AUGIB patients (1,2). However, older studies show similar mortality rates for those with AUGIB who develop MI (3,4).
Moreover, general abnormalities caused by AUGIB might well have delayed consequences with a significant impact on mortality (5).

Aims & Methods

Aim: To analyze the risk factors for in-hospital (ACVC) and delayed (DCVC) cardiovascular complications after AUGIB.
Methods: Retrospective analysis of a prospective registry of UGIB patients admitted to ‘Virgen de las Nieves University Hospital’ (2013-2020). Main outcomes were ACVC (including stroke, lung embolism, myocardial injury and heart failure) and DCVC-6months (including coronary syndrome, thromboembolic event, heart failure and arrhythmia). Statistical analysis was performed by SPSS.

Results

We included 1345 AUGIB patients, mean age 64.44 +/- 16.19, 32.3% females, 23 with ACVC (1.7%), 8 of these patients died. 79 develop DCVC (5.9%), 18 of these patients died.
Table show differences between patients with and without ACVC and DCVC.

ACVC	With	Without	HR (95%CI)	p
Age COPD Coronary disease Stroke High blood pressure Antiplatelets Urea at admission (mg/dL) Active bleeding Need for intervention Reebleeding Lenght hospital stay Inpatient In-hospital mortality	76.39 30.4% 34.8% 17.4% 69.6% 39.1% 147.17 56.5% 91.3% 26.1% 16.96 34.8% 36.4%	64.23 10.1% 9.6% 5.8% 43.7% 20.3% 86.51 27.5% 73.7% 9.9% 8.97 15.2% 10%	1.45 (1.10 - 2.33) 3.87 (1.56 - 9.59) 5.02 (2.10 - 12.07) 3.40 (1.13 - 10.25) 2.95 (1.20 - 7.21) 2.52 (1.08 - 5.88) 4.85 (1.75 - 11.23) 3.43 (1.49-7.90) 4.62 (1.70 - 12.52) 3.23 (1.25 - 8.33) 5.23 (2.34 - 14.53) 2.97 (1.24 - 7.10) 5.15 (2.12 - 12.52)	0.043 0.007 0.001 0.045 0.018 0.037 0.010 0.004 0.047 0.023 0.015 0.018 0.001
DCVC	With	Without	HR (95%CI)	p
Age COPD Cronic Kidney Disease Heart Failure Coronary disease Atrial fibrilation Valvular disease Stroke High blood pressure Anticoaugulants Urea at admission (mg/dL) Delayed bleeding Delayed mortality	74.46 19% 29.1% 39.2% 22.8% 49.4% 37% 15.2% 55.7% 49.4% 113.94 40.5% 23.1%	63.67 9.8% 13.7% 8.2% 9.4% 15.8% 7.3% 5.1% 43.5% 18.9% 83.36 12.9% 6.5%	1.56 (1.15 - 2.46) 2.16 (1.19 - 3.91) 2.59 (1.55 – 4.33) 7.25 (4.41 - 11. 93) 2.85 (1.63 - 5.00) 5.18 (3.24 - 8.28) 7.45 (3.20 - 17.31) 3.35 (1.72 - 6.53) 1.63 (1.03 - 2.58) 4.18 (2.62 - 6.65) 3.47 (1.83 - 6.25) 4.58 (2.83 - 7.42) 4.34 (2.44 - 7.72)	0.035 0.020 0.001 <0.001 0.001 0.001 < 0.001 0.001 0.046 <0.001 < 0.001 < 0.001 < 0.001

Independent risk factor for ACVC were age (HR 1.08, 95%CI 1.03-1.15, p=0.003), previous coronary disease (HR 2.98, 95%CI 1.05-9.48, p=0.045), urea (HR 1.007, 95%CI 1.002-1.013, p=0.011) and rebleeding (HR 3.85, CI95% 1.17-2.71, p= 0.027).
Independent risk factor for DCVC was previous heart failure (HR 5.92, 95%CI 2.17-16.17, p=0.001).

Conclusion

Patients with AUGIB who develop in-hospital cardiovascular events are older, have more comorbidities, more severe bleeding and worse outcomes. The independent risk factors for AUGIB induced cardiovascular event identified are age, previous coronary disease, urea and rebleeding. In the same way, patients with delayed cardiovascular event are older, have more comorbidities and worse delayed outcomes. The independent risk factor for AUGIB induced delayed cardiovascular event identified is previous heart failure.

References

1. Hao J, Dang P, Quan X, Chen Z, Zhang G, Liu H, Shi T, Yan Y. Risk factors, prediction model, and prognosis analysis of myocardial injury after acute upper gastrointestinal bleeding. Front Cardiovasc Med. 2022 Dec 8;9:1041062
2. Jiang GJ, Gao RK, Wang M, Xie TX, Zhan LY, Wei J, Sun SN, Ji PY, Tan DY, Lyu JJ. A Nomogram Model for Predicting Type-2 Myocardial Infarction Induced by Acute Upper Gastrointestinal Bleeding. Curr Med Sci. 2022 Apr;42(2):317-326.
3. Emenike E, Srivastava S, Amoateng-Adjepong Y, al-Kharrat T, Zarich S, Manthous CA. Myocardial infarction complicating gastrointestinal hemorrhage. Mayo Clin Proc. 1999 Mar;74(3):235-41.
4. Bhatti N, Amoateng-Adjepong Y, Qamar A, Manthous CA. Myocardial infarction in critically ill patients presenting with gastrointestinal hemorrhage: retrospective analysis of risks and outcomes. Chest. 1998 Oct;114(4):1137-42.
5. Jiménez-Rosales R, Valverde-López F, Vadillo-Calles F, Martínez-Cara JG, López de Hierro M, Redondo-Cerezo E. Inhospital and delayed mortality after upper gastrointestinal bleeding: an analysis of risk factors in a prospective series. Scand J Gastroenterol. 2018 Jun;53(6):714-720.

Introduction

Antithrombotic drugs (AT) pose an increased risk of upper gastrointestinal bleeding (UGIB), and their discontinuation carries the risk of thromboembolic events, presenting a challenge for clinicians (1,2). Clinical practice guidelines advocate for a restrictive approach, recommending non-discontinuation and/or early reintroduction based on the drug and its indication (3).
Despite available literature, controversy persists regarding the outcomes of patients with UGIB on antithrombotic therapy, both in the short and long term, as well as the potential impact of drug discontinuation (4,5).

Aims & Methods

Aim:
To determine the impact of ATs on in-hospital and delayed outcomes. Also, to determine whether suspension and resumption influenced delayed outcomes.
Methods:
Prospective registry analysis of UGIB patients between 2013-2021. Anticoagulants and antiplatelets were classified as AT drugs. Outcomes examined included in-hospital mortality and delayed-6 months cardiovascular, bleeding and mortality events.

Results

1345 patients were included, with 21.7% on anticoagulants and 19.1% on antiplatelets.
Patients on ATs were older (73 vs. 59 years; p<0.0001) and had a higher rate of comorbidities, has longer delay of endoscopic performance (11±11h vs. 9.6±8h; p=0.027), with gastric erosions being more usual (14.2% vs.9.1%; p=0.006). Table shows main outcomes.

	Non-AT (n=839)	AT (n=506)	P	Aspirin (n=195)	Thienopyridine (n=29)	Double antiaggregating (n=24)	VKA (n=179)	NACO (n=49)
Endoscopy therapy	48.3%	38.5%	0.001	75 (38.5%)	12 (41.4%)	7 (29.2%)	62 (34.6%)	20 (40.8%)
Rescue therapy	6.1%	3.8%	0.083	8 (4.1%)	17 (58.6%)	0	4 (2.2%)	1 (2%)
Rebleeding	10.5%	9.7%	0.702	20 (10.3%)	1 (3.4%)	18 (75%)	12 (6.7%)	4 (8.2%)
Transfusion	56.3%	66.2%	0.000	127 (65.1%)	17 (58.6%)	0	115 (64.2%)	41 (83.7%)
In-hospital mortality	12%	8%	0.028	14 (7.2%)	0	1 (4.2%)	13 (7.3%)	1 (2%)
Delayed cardiovascular event	3.4%	13.7%	0.000	18 (9.2%)	5 (17.2%)	1 (4.2%)	31 (17.3%)	12 (24.5%)
Delayed hemorrhagic event	12.9%	22.9%	0.000	43 (22.1%)	5 (17.2%)	6 (25%)	41 (22.9%)	12 (24.5%)
Delayed mortality	6%	9.4%	0.034	17 (8.7%)	2 (6.9%)	2 (8.3%)	19 (10.6%)	4 (8.2%)

AT suspension <72h was associated with increased in-hospital mortality (14.9% vs. 2.3%; p=0.001), especially <24h (20.9% vs. 2.6%; p<0.0001).
Delayed mortality was higher in patients on ATs who suspended them more than 7 days (17% vs.8.7%; p= 0.033), with no differences when last <72h. Suspension/reintroduction have no effect on other delayed outcomes.
Multivariate analysis identified ASA≥3 (HR 12.20;95%CI 3.69-56-92; p<0.001), disseminated malignancy (HR 4.73; 95%CI 1.92-11.42; p=0.001) and rebleeding (HR 19.25; 95%CI 9.07-43.04) as independent risk factors for in-hospital mortality, with AT therapy (HR 0.27; 95%CI 0.13-0.55; p<0.001), NSAIDS (HR 0.26; 95%CI 0.07-0.75; p=0.023), endoscopic therapy (HR 0.49; 95%CI 0.24-0.96; p=0.040) and hemoglobin levels being protectives factors HR 0.80; 95%CI 0.69.0.93; p=0.004).
Logistic regression analysis for delayed mortality showed that age (HR 1.04; 95%CI 1.02-1.07; p=0.001), disseminated malignancy (HR 3.49; 95%CI 1.81-7.86; p=0.004), esophageal varices HR 3.34; 95%CI 1.62-6-64; p=0.001), ischemic heart disease (HR 2.44; 95%CI 1.10-5.19; p=0.023) and a being an inpatient (HR 1.92; 95%CI 1.00-3.98; p=0.042) were independent risk factors for the outcome.

Conclusion

Despite increased comorbidities and older age, patients with AUGIB on ATs had lower in-hospital mortality, with no differences in other relevant outcomes. In fact, our results suggest that AT treatment provides a survival benefit, possibly related to a higher risk of bleeding in milder lesions.
Regarding delayed mortality, it was higher in patients on AT, determined by the comorbidity leading to their prescription.
Interestingly, early resumption of AT was associated with higher in-hospital mortality, while reintroduction after 7 days increased delayed one.

References

1. Lanas A, Carrera-Lasfuentes P, Arguedas Y, et al. Risk of upper and lower gastrointestinal bleeding in patients taking nonsteroidal anti-inflammatory drugs, antiplatelet agents, or anticoagulants. Clin Gastroenterol Hepatol 2015;13:906-12 e2.
2. Chan FK, Leung Ki EL, Wong GL, et al. Risks of Bleeding Recurrence and Cardiovascular Events With Continued Aspirin Use After Lower Gastrointestinal Hemorrhage. Gastroenterology 2016;151:271-7.
3. Gralnek IM, Stanley AJ, Morris AJ, et al. Endoscopic diagnosis and management of nonvariceal upper gastrointestinal hemorrhage (NVUGIH): European Society of Gastrointestinal Endoscopy (ESGE) Guideline - Update 2021. Endoscopy 2021;53:300-332.
4. Dunne PDJ, Laursen SB, Laine L, et al. Previous Use of Antithrombotic Agents Reduces Mortality and Length of Hospital Stay in Patients With High-risk Upper Gastrointestinal Bleeding. Clin Gastroenterol Hepatol 2019;17:440-447 e2.
5. Marmo R, Occhipinti V, Zullo A, et al. Improved Survival for Patients With Acute Upper Gastrointestinal Bleeding While on Antithrombotic Therapy: A Multicenter Prospective Cohort Study. J Clin Gastroenterol 2023;57:278-284.

Introduction

Several scoring systems have been established to identify upper gastrointestinal bleeding (UGIB) patients at high risk of adverse outcomes. Recently, focus have been posed in the Shock Index since the NCEPOD reported this use to assess bleeding severity and found an association with mortality. Moreover, they report that there was inconsistent use of recognized scoring systems in their examined cases of upper GI bleeding and postulated their use (1).
After this, Stanley and colleagues examine the performance of SI in comparisons with preexisting scores predicting UGIB outcomes in a cohort of 3012 patients and reported that SI was inferior to existing pre‐endoscopy scores (2). Recently, other study positioned SI as the best ones to predict 30, 180 and 360-day mortality in UGIB patients over 65 years, generating the controversy (3).
We aim to compare the performance of SI versus previous scores and versus traditional vital constants to predict outcomes in UGIB.

Aims & Methods

We aim to compare the performance of SI versus previous scores and versus traditional vital constants to predict outcomes in UGIB.
Methods: Retrospective analysis of a prospective registry of UGIB patients admitted to ‘Virgen de las Nieves University Hospital’ between 2013-2020.
Main outcomes were intervention (a composite endpoint including need of transfusion, endoscopy therapy, interventional radiology or surgery) and in-hospital mortality.
Scores calculated were SI, MAP(ASH), ABC, Glasgow-Blatchford score (GBS), AIMS65 and Rockall score. We also included pulse and blood pressure (BP) in the comparison.
MedCalc was used to calculate ROC Curve of scores in determining outcomes.

Results

AUROCs are shown in the table. Comparisons of ROC curves for relevant outcomes show all the pre-existing scores superior to SI with statical difference. In relation to vital signs, only for intervention SI was superior to pulse with statical difference. For mortality, no differences were found between SI and vital signs.

AUROC (95%CI)	Need of intervention	In-hospital mortality
Pulse	0.538 (0.590 - 0.643)	0.618 (0.591 - 0.645)
Systolic Blood Pressure	0.645 (0.511 - 0.565)	0.587 (0.559 - 0.614)
Shock Index	0.626 (0.596 - 0.655)	0.612 (0.582 - 0.642)
MAP(ASH)	0.789 (0.764 - 0.813)	0.758 (0.731 - 0.784)
ABC	0.723 (0.695 - 0.749)	0.804 (0.779 - 0.828)
Glasgow-Blatchford score	0.780 (0.754 - 0.805)	0.708 (0.680 - 0.736)
AIMS65	0.682 (0.654 - 0.710)	0.738 (0.710 - 0.765)
Rockall score	0.760 (0.734 - 0.785)	0.743 (0.715 - 0.769)

Conclusion

In view of our results, we have to conclude, as Stanley et al. (2), that Shock Index is inferior to pre-existing scores in predicting outcomes in UGIB. In comparison to vital signs Shock Index not offer any advantages with the exception to pulse for intervention. Therefore, we maintain our previous recommendation (4) to use MAP(ASH), or failing that ABC, for predicting intervention and mortality in UGIB since it is easier to calculate with an excellent performance.

References

1. National Confidential Enquiry into Patient Outcome and Death (NCEPOD). Time to Get Control? A review of the care received by patients who had a severe gastrointestinal haemorrhage. 2015. http://www.ncepod.org.uk/2015gih.html. Accessed May 10, 2017.
2. Saffouri E, Blackwell C, Laursen SB, Laine L, Dalton HR, Ngu J, Shultz M, Norton R, Stanley AJ. The Shock Index is not accurate at predicting outcomes in patients with upper gastrointestinal bleeding. Aliment Pharmacol Ther. 2020 Jan;51(2):253-260. doi: 10.1111/apt.15541. Epub 2019 Oct 23. PMID: 31642558.
3. Dogru U, Yuksel M, Ay MO, Kaya H, Ozdemır A, Isler Y, Bulut M. The effect of the shock index and scoring systems for predicting mortality among geriatric patients with upper gastrointestinal bleeding: a prospective cohort study. Sao Paulo Med J. 2022 Jul-Aug;140(4):531-539. doi: 10.1590/1516-3180.2021.0735.13102021. PMID: 35544884; PMCID: PMC9491474.
4. Jimenez-Rosales R, Lopez-Tobaruela JM, Lopez-Vico M, Ortega-Suazo EJ, Martinez-Cara JG, Redondo-Cerezo E. Performance of the New ABC and MAP(ASH) Scores in the Prediction of Relevant Outcomes in Upper Gastrointestinal Bleeding. Journal of Clinical Medicine. 2023; 12(3):1085.

Introduction

Massive gastrointestinal bleeding (MGB) is a life-threatening condition that requires prompt recognition and resuscitation to increase the chance of good outcomes (1).

Aims & Methods

Aims: Analyze risk factors, comorbidities and outcomes of MGB.
Methods: Retrospective analysis of a prospective registry of upper gastrointestinal bleed (UGIB) patients admitted to ‘Virgen de las Nieves University Hospital’ between 2013-2020.
MGB was defined as UGIB with transfusion of >2 pRBCs in the emergency room or during the first 24h of hospital stay and signs of shock (2). Main outcomes were 30-day mortality, rebleeding, bleeding persistence or severe complication throughout admission and complications within the first 6 months after discharge (including cardiovascular, hemorrhagic or deaths).
Statistical analysis was performed by SPSS.

Results

We included 1213 UGIB patients, 171 with MGB.
Multivariate logistic regression analysis showed that, being admitted for another condition (HR 1.72,95%CI 1.04-2.58,p=0.035), having a systemic disease (HR 2.30,95%CI 1.22-5.32,p=0.045), a disseminated malignancy (HR 2.07,95%CI 1.02-4.21,p=0.045) or suffering a severe comorbidity (HR 2.21,95%CI 1.25-3.91,p=0.009) were independent risk factor for MGB.
Among 171 MGB patients, 21% died, with 10 deaths directly from hypovolemic shock secondary to UGIB, and 11% due to complications.
Table shows significance differences between deaths and survivors.

Relevant Comorbidities ASA 3-4 (%)	Deaths 96.7%	Survivors 78.5%	p 0.0018
Active bleeding (%)	76.3	43.3	0.007
Interventional radiology (%)	13.3	3.7	0.034
SBP (mean±SD mmHg)	79±8	84±9	0.025
Albumin (mean±SD mg/dl)	2.4±0.5	2.8±0.6	0.011
Calcium (mean±SD mg/dl)	7.7±0.6	8.2±0.8	0.032
Platelets (mean±SD /mm3)	158*103±101*103	206*103±106*103	0.027
Severe complications during hospitalization	76.7	22.1	<0.0001
Bleeding persistence (%)	63.3	17.6	<0.0001
Rebleeding (%)	40	17	0.005

Multivariate analysis showed that active bleeding in endoscopy (HR 4.80,95%CI 1.06-21.62,p=0.041) and severe complications (HR 14.03,95%CI 3.30-59.60,p<0.0001) were independent predictors for mortality, being calcium levels a protective factor (HR 0.41,95%CI 0.16-0.94,p=0.045).
Among the survivors after the index episode,14% died after 6 months, being main causes of death recurrent bleeding(5.2%), neoplasms(4.4%), cirrhosis complications(2%), cardiovascular disease(1.5%) and sepsis(1.5%).
Delayed bleeding was caused by recurrent UGIB in 86.6%, lower GIB in 10% and from a non-GI origin in 3.4%. Delayed cardiovascular events were heart failure in 46%, thromboembolic events in 23%, myocardial ischemia in 15.4% and atrial fibrillation in 15.6%.

Conclusion

Our study is a prospectively collected single center registry of patients with massive UGIB, in which we observed that being an inpatient and harboring severe comorbidities, particularly systemic diseases and malignancies, were risk factors for this condition. Mortality among those patients was related to CKD, cirrhosis, severe comorbidities and excessive alcohol consumption. We also found an increased delayed mortality, probably related to a worsening health condition in which GI bleeding heralded poor outcomes, some of them potentially preventable with close follow-up.

References

1. Nagata N, Sakurai T, Shimbo T, et al. Acute Severe Gastrointestinal Tract Bleeding Is Associated With an Increased Risk of Thromboembolism and Death. Clin Gastroenterol Hepatol. 2017;15:1882-1889 e1881.
2. D'Amore K, Swaminathan A. Massive Gastrointestinal Hemorrhage. Emerg Med Clin North Am. 2020;38:871-889.

Introduction

Non-variceal upper gastrointestinal bleeding (NVUGI) represents a significant gastrointestinal disorder with elevated mortality rates (1). Non-steroidal anti-inflammatory drugs (NSAIDs) have been firmly established as a risk factor for NVUGI, yet their precise impact on clinical outcomes remains insufficiently defined (2,3). Conversely, proton pump inhibitors (PPIs) have shown efficacy in reducing the incidence of NVUGI associated with NSAIDs. However, the extent of their influence on the clinical outcomes of NVUGI remains inadequately substantiated. Therefore, it is necessary to study the impact of these drugs on the clinical outcomes of patients with NVUGI.

Aims & Methods

The primary objective of our study is to compare the outcomes of NVUGI patients using NSAIDs with or without PPIs against those who did not use either medication. We conducted a retrospective analysis at a Spanish tertiary hospital, encompassing NVUGI cases from 2013 to 2022. NVUGI was defined as hematemesis or melena accompanied by a two-point decrease in hemoglobin or hemodynamic instability. Key clinical outcomes included the necessity for intervention (endoscopic, radiological, or surgical) and in-hospital mortality. Additionally, we compared baseline characteristics and pertinent clinical results such as rebleeding and delayed cardiovascular events. Statistical analysis was performed using SPSS.

Results

​A total of 1200 cases were included, with 6.6% undergoing concomitant treatment with PPIs and NSAIDs, and 15.7% utilizing NSAIDs alone. Patients who took NSAIDs alone were younger (57 years vs. 66 in both other groups) and had a lower rate of comorbidities. Table 1 presents risk scores and outcome comparisons among the three groups. Statistical differences were observed, revealing lower acute mortality (9.6% vs. 7 vs. 2.8%, p=0.011) and reduced need for intervention (74% vs. 69.6 vs. 57.8%, p<0.001) in the group using NSAIDs alone. In multivariate analysis, MAP(ASH) > 2 (OR: 8, CI: 3-21, p<0.001) and ASA > 2 (OR: 4.4, CI: 1.7-11.4, p:0.002) emerged as independent risk factors for acute mortality, with an elevated MAP(ASH) score (OR: 8.8, CI: 5.9-13, p<0.001) being the sole independent factor explaining the need for intervention.

	Naive	NSAIDs and PPIs	NSAIDs without PPIs	P
ASA 3-4	596 (64,3%)	50 (63,3%)	63 (33,7%)	<0,001
MAPASH>2	448 (59,3%)	39 (65%)	52 (34,4)	<0,001
Need for intervention	691 (74%)	55 (69,6%)	108 (57,8%)	<0,001
Rebleeding	96 (10,3%)	9 (11,4%)	6 (3,2%)	0,008
Acute mortality	86 (9,6%)	5 (7%)	5 (2,8%)	0,011
Mortality at 6 months	77 (8,8%)	7 (9,3%)	7 (3,8%)	0,075
Delayed Cardiovascular event	77 (8,7%)	3 (3,9%)	4 (2,2%)	0,004
Delayed rebleeding	145 (16,4%)	15 (19,5%)	6 (3,3%)	<0,001

Conclusion

Patients abstaining from NSAIDs or those concurrently undergoing PPI therapy exhibited a higher incidence of intervention or acute mortality compared to those solely using NSAIDs. Through multivariate analysis, our study indicates that bleeding severity and baseline comorbidities, expressed by elevated MAP(ASH) scores and ASA classification, are pivotal factors influencing the clinical outcomes of NVUGI. In summary, the use of NSAIDs or PPIs to safeguard against gastrointestinal lesions induced by these drugs did not demonstrate a discernible impact on the short-term outcomes of patients with non-variceal upper gastrointestinal bleeding.

References

1. Lakatos L, Gonczi L, Lontai L, Izbeki F, Patai A, Racz I, et al. Incidence, predictive factors, clinical characteristics and outcome of non-variceal upper gastrointestinal bleeding – a prospective population-based study from hungary. Journal of Gastrointestinal and Liver Diseases. 2021 Sep 1;30(3):327–33.
2. Durak MB, Şimşek C, Yüksel İ. Clinical Outcomes of Older Patients with Non-Variceal Upper Gastrointestinal Bleeding Taking Anti-Thrombotic or Non-Steroidal Anti-Inflammatory Agents. Turkish Journal of Gastroenterology. 2023 Sep 1;34(9):918–24.
3. Åhsberg K, Höglund P, Kim WH, Von Holstein CS. Impact of aspirin, NSAIDs, warfarin, corticosteroids and SSRIs on the site and outcome of non-variceal upper and lower gastrointestinal bleeding. Scand J Gastroenterol. 2010 Dec;45(12):1404–15.

Introduction

Upper gastrointestinal bleeding(UGIB) is a common condition that entails morbidity and mortality. Therefore, an early prediction of negative outcomes would be clinically relevant selecting patients who might benefit from a closer follow up. In this way, several pre-endoscopic scoring systems have been developed for risk assessment in the emergency room(1).
One of the limitations of scoring systems is that sensitivity and specificity alone (usually combined in the receiver operator characteristic curve) may help comparing the accuracy of different systems but cannot be used to accurately estimate the probability for an individual patient. A more appropriate approach is to establish the post‐test probability considering the expected pre‐test probability of poor outcomes in the population(2).

Aims & Methods

Given the absence of previous similar studies in UGIB, we aim to investigate the role of the most common pre-endoscopy scoring systems in modifying the pre-test probability of adverse outcomes.
Retrospective analysis of a prospective registry of UGIB patients admitted to 'Virgen de las Nieves University Hospital' (2013-2024). The outcomes included were intervention (composite endpoint including need of transfusion, endoscopy therapy, interventional radiology or surgery) and in-hospital mortality.
Pre-endoscopy scores analyzed were MAPASH,ABC and GBS as they have demonstrated their utility in predicting adverse events. The thresholds used for the definition of low-risk patients(test negative) were ABC≤3,MAPASH≤2 and GBS≤1(3).
First, the prevalence of the main outcomes was calculated to obtain the pre‐test probability. Next, characteristics of different scoring systems were used to obtain the likelihood ratio(LR) and to estimate the post‐test probability using an online calculator (http://araw.mede.uic.edu/cgi-bin/testcalc.pl).

Results

1649 UGIB patients were included.
Prevalence of intervention(pre-test probability) was 75%.
The sensitivity and specificity of MAPASH were 71.4% and 78.4%,respectively. +LR and -LR 3.31 and 0.36,respectively. Post‐test probability was 91% when positive and 52% when negative.
The sensitivity and specificity of ABC were 68.1% and 66.8%,respectively. +LR and -LR 2.05 and 0.48,respectively. Post‐test probability was 86% when positive and 59% when negative.
The sensitivity and specificity of GBS were 98.8% and 8.7%,respectively. +LR and -LR 1.08 and 0.14,respectively. Post‐test probability was 76% when positive and 29% when negative.
Prevalence of in-hospital mortality(pre-test probability) was 11%.
The sensitivity and specificity of MAPASH were 90.1% and 44.8%, respectively. +LR and -LR 1.63 and 0.22,respectively. Post‐test probability was 17% when positive and 3% when negative.
The sensitivity and specificity of ABC were 95.9% and 44.8%,respectively. +LR and -LR 1.74 and 0.09,respectively. Post‐test probability was 18% when positive and 1% when negative.
The sensitivity and specificity of GBS were 99.4% and 3.6%,respectively. +LR and -LR 1.03 and 0.17, respectively. Post‐test probability was 11% when positive and 2% when negative.

Conclusion

While MAPASH and ABC effectively prioritize high-risk patients for intervention, their ability to rule out intervention in low-risk patients is limited. Additionally, they are more useful excluding mortality among low-risk patients than accurately predicting fatal outcomes. GBS,in contrast, are useful identifying low risk patients for intervention and mortality.
Given these findings, clinicians should interpret these scores with caution and incorporate additional clinical parameters into decision-making.

References

1. Boustany A, Alali AA, Almadi M, Martel M, Barkun AN. Pre-Endoscopic Scores Predicting Low-Risk Patients with Upper Gastrointestinal Bleeding: A Systematic Review and Meta-Analysis. J Clin Med. 9 de agosto de 2023;12(16).
2. Capurso G, Ponz de Leon Pisani R, Lauri G, Archibugi L, Hegyi P, Papachristou GI, et al. Clinical usefulness of scoring systems to predict severe acute pancreatitis: A systematic review and meta‐analysis with pre and post‐test probability assessment. United European Gastroenterol J. 27 de septiembre de 2023;11(9):825-36.
3. Jimenez-Rosales R, Lopez-Tobaruela JM, Lopez-Vico M, Ortega-Suazo EJ, Martinez-Cara JG, Redondo-Cerezo E. Performance of the New ABC and MAP(ASH) Scores in the Prediction of Relevant Outcomes in Upper Gastrointestinal Bleeding. J Clin Med. 30 de enero de 203;12(3):1085.