Nomogram Predicts In-Hospital Mortality in Patients with Emergency Gastrointestinal Bleeding: A Multicenter Retrospective Study

doi:10.21203/rs.3.rs-4905124/v1

Download PDF

Article

Nomogram Predicts In-Hospital Mortality in Patients with Emergency Gastrointestinal Bleeding: A Multicenter Retrospective Study

https://doi.org/10.21203/rs.3.rs-4905124/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Gastrointestinal bleeding (GIB) is a frequent issue encountered in emergency departments, associated with significant rates of incidence and mortality. This study aims to create and validate a reliable nomogram to predict the risk of in-hospital mortality in patients experiencing emergency GIB. Additionally, it seeks to identify the risk factors that influence mortality and to equip the emergency clinical team with a precise predictive tool.

Methods

This study utilized a retrospective cohort design to analyze data from patients with GIB who presented to the emergency departments and were subsequently admitted at three branches of Wuhan Central Hospital: Nanjing Road, Houhu, and Yangchunhu, from January to December 2023. Patient information was collected through the hospital's information system. The LASSO regression method was employed to identify key variables for prediction, and a nomogram was constructed using multivariate logistic regression. The model's ability to discriminate between outcomes was assessed by calculating the area under the curve (AUC). Furthermore, calibration analysis and decision curve analysis (DCA) were performed to evaluate the model's performance.

Results

A total of 847 patients were included, with 75 (8.85%) dying during hospitalization. In-hospital mortality was more common among elderly patients (median age 73 years vs. 65.5 years for survivors, P < 0.001). Deceased patients had lower systolic and diastolic blood pressures, higher heart rates, and higher shock indices upon emergency admission (P < 0.001). They were more likely to arrive by ambulance (P < 0.001) and classified as ESI Level 1 (P < 0.001). Additionally, they had a higher incidence of malignant tumors (P < 0.001), underwent fewer surgeries (P = 0.003), and received fewer hemostasis procedures (P < 0.001). Their total hospitalization costs were also higher (P < 0.001). Logistic regression analysis identified Ambulance ED, Shock Index > 1, ICU admission, malignancy, and hemostatic procedures as independent risk factors for GIB. ROC curve analysis showed an AUC of 0.862 (95% CI: 0.786–0.939) for the training cohort and 0.846 (95% CI: 0.787–0.904) for the validation cohort.

Conclusion

The developed nomogram model effectively predicts in-hospital mortality risk among emergency GIB patients, demonstrating good classification performance and clinical potential. It is recommended that this model be integrated into clinical information systems to support decision-making and optimize patient management.

Health sciences/Gastroenterology

Health sciences/Medical research

Health sciences/Risk factors

Gastrointestinal Bleeding (GIB)

Emergency Department (ED)

In-hospital mortality

Nomogram

Risk assessment

Gastrointestinal bleeding (GIB) is a common presentation in emergency departments (EDs), with hematemesis, melena, and syncope as primary symptoms[1]. Although the survival rates of GIB patients have improved with advancements in pharmacotherapy and endoscopic diagnosis and treatment, it remains a potentially life-threatening gastrointestinal emergency[2]. Emergency clinical teams must have rapid access to effective tools for assessing patient risk, enabling prompt and efficient treatment, and ultimately improving patient outcomes[3].

Current risk predictions for gastrointestinal bleeding patients primarily focus on either upper or lower gastrointestinal bleeding. Researchers have developed several risk assessment tools for upper gastrointestinal bleeding (UGIB)[4–5], which have been proven to accurately predict those patients who require emergency hospitalization and reduce hospitalization rates for UGIB patients by 15–20%[6–7], indicating their potential value in identifying low-risk patients for outpatient and emergency management. However, the effectiveness of these scoring systems in predicting mortality is relatively limited. Although AIMS65 and PNED perform better in predicting mortality compared to GBS and Rockall scores, their area under receiver operating characteristic curves (AUROCs) generally do not exceed 0.80, indicating their limitations in clinical application [8]. Acute lower gastrointestinal bleeding (LGIB) accounts for 20–30% of all gastrointestinal bleeds[9]. Compared to UGIB, LGIB tends to have a milder course, requiring fewer hemostatic interventions and having lower mortality rates [10]. Although several risk scoring systems have been developed to predict the risk of severe bleeding or mortality following LGIB [11–12], recent studies suggest that these scoring systems have suboptimal discrimination in predicting mortality, with AUROCs ranging from 0.66 (BLEED score) to 0.73 (AIMS65) [13]. However, factors influencing these models, such as the bleeding site, cause, or type, are often difficult to quickly ascertain in emergency settings, requiring further clarification during hospitalization.[3, 14]

In clinical practice, distinguishing between UGIB and LGIB quickly can be challenging. For example, hematochezia may indicate upper gastrointestinal bleeding, while black, tarry stools may suggest bleeding from the right colon. Data indicate that among hospitalized GIB patients, approximately 40% of the bleeding originates from the upper gastrointestinal tract, 25% from the lower gastrointestinal tract, and 35% have an indeterminate bleeding site[9]. Although UGIB and LGIB patients are managed by different clinical teams in separate departments when transferred from the emergency department to inpatient care, they are initially managed by a single emergency team. This can make it difficult for emergency physicians to determine the bleeding site at first presentation quickly Therefore, a scoring system that predicts mortality for both conditions is valuable for emergency management teams, especially since precise diagnosis of the bleeding site (UGIB vs. LGIB) is not always possible upon initial assessment. The UK NCEPOD report criticized the practice of segregating hospital care for UGIB and LGIB patients and specifically recommended the development of a risk score to manage all patients presenting with gastrointestinal bleeding[15]. Additionally, in emergency clinical practice, regardless of the source of bleeding, accurately and early identifying high-risk patients for mortality allows for targeted management strategies, including specialized care and early interventions, potentially improving patient outcomes. Identifying patients at very low risk can also reduce unnecessary intensive treatments and optimize the allocation of medical resources, which is the primary responsibility of the emergency clinical team.

Therefore, developing an accurate nomogram for emergency GIB patients is crucial. We aim to develop and validate an accurate nomogram specifically designed to predict in-hospital mortality for emergency GIB patients. This nomogram will integrate multiple clinical variables readily available in emergency settings, reflecting a comprehensive view of the patient's health status and disease severity. By providing more detailed and accurate mortality risk predictions, our nomogram will enable emergency department clinicians and nursing teams to make informed decisions, optimize treatment strategies, and ultimately improve inpatient survival outcomes for this high-risk population.

Study design

This is a retrospective cohort study involving patients with gastrointestinal bleeding who visited the emergency department and were subsequently admitted between January 1, 2023, and December 31, 2023. The study was approved by the Ethics Committee of Wuhan Central Hospital (Approval No.WHZXKYL2024-117). Informed consent was waived due to the anonymized nature of the data. This study adheres to the principles outlined in the Declaration of Helsinki.

Subjects of the study

Patients meeting the following criteria were included in the study: (1) Diagnosed with gastrointestinal bleeding, dyspepsia, hematemesis, melena, hematochezia, or coffee-ground vomiting in the emergency department; (2) Admitted to the hospital from the emergency department for inpatient treatment; (3) Discharged with a diagnosis of gastrointestinal bleeding. Exclusion criteria: Patients under 18 years of age.

Data collection

Our data collection was from the hospital's information systems of Wuhan Central Hospital: Nanjing Road Branch, Houhu Branch, and Yangchunhu Branch, from January to December 2023. The emergency triage system included triage information such as the patient's name, gender, age, chief complaint, diagnosis, mode of arrival, time of arrival, triage level, vital signs, and admitting department. The Hospital Information System (HIS) contains the patient's basic information, such as length of stay, surgical procedures, hemostatic interventions, and their timing, types of comorbidities, discharge status, and total hospital costs. The patient's visit number was used to link the data across different systems. The visit number represents the index for a single hospital visit during the patient's stay. The final baseline data were matched using the data closest to the visit time.

Statistical methods

Data analysis, model construction, and validation were conducted according to the methods described by Wang S, Tu J (2020)[16]. Descriptive statistics were used to summarize the baseline characteristics of the patients. The Kolmogorov-Smirnov test was used to assess the normality of the data. Normally distributed continuous variables were expressed as mean\(\:\pm\:\)standard deviation (x̄\(\:\pm\:\)s) and compared using the independent samples t-test. Non-normally distributed continuous variables were expressed as median (Q1, Q3) and compared using the Wilcoxon rank-sum test. Categorical variables were expressed as frequencies and percentages, and comparisons were made using the χ² test or Fisher's exact test. The Least Absolute Shrinkage and Selection Operator (LASSO) regression, suitable for high-dimensional data (Huang et al., 2016)[17], was used for predictor selection and regularization. Logistic regression analysis was employed to develop a predictive model for the risk of in-hospital mortality among patients with gastrointestinal bleeding admitted through the emergency department, and a nomogram was constructed based on this model. The model's discriminative performance was evaluated by calculating the area under the curve (AUC), with internal validation performed using 1,000 bootstrap resamples. For further validation, patients were randomly divided into two groups: a training set (70%) and a validation set (30%). The DeLong method was used to compare the statistical differences in AUCs. The clinical utility of the model was assessed using decision curve analysis (DCA) to quantify the net benefit at different probability thresholds (Vickers et al., 2008)[18]. Statistical analyses were conducted using EmpowerStats (www.empowerstats.com) and R version 4.2.2 (www.r-project.org). A P-value < 0.05 was considered statistically significant. by

A total of 847 patients participated in this study. The flow of information collection for the target population is shown in Fig. 1. Among these patients, 75 (8.85%) died during their hospitalization after being admitted to the emergency department (ED). In-hospital deaths were more frequent among older patients, with a median age of 73 years, compared to 65.5 years for survivors (P < 0.001). Patients who died had lower systolic and diastolic blood pressure, higher heart rates, and higher shock indices(Initial vital signs monitoring upon emergency department admission) (P < 0.001). They were also more likely to arrive by ambulance (P < 0.001) and to be classified as Emergency Severity Index (ESI) level 1 (P < 0.001). Additionally, the incidence of malignancy was higher among those who died (P < 0.001), and they underwent fewer surgical procedures (P = 0.003) and fewer hemostatic procedures (P < 0.001). These patients also had higher total hospitalization costs (P < 0.001). These findings suggest that age, vital signs, mode of arrival, triage level, presence of malignancy, and types of clinical interventions are important factors associated with in-hospital mortality in this patient population. The demographic and clinical characteristics of the study participants are shown in Table 1.

Table 1

Baseline demographic and clinical characteristics of the study participants
Variables	Total (n = 847)	Survived (n = 772)	Deceased (n = 75)	p
gender, n (%)	575 (68)	520 (67)	55 (73)	0.353
age, Median (Q1,Q3)	67 (55, 76)	65.5 (54, 76)	73 (65, 80.5)	< 0.001
Mode of Arrival to Emergency Department, n (%)				< 0.001
By Ambulance (120)	353 (42)	295 (38)	58 (77)
Walking	169 (20)	169 (22)	0 (0)
Assisted	138 (16)	138 (18)	0 (0)
wheelchair	68 (8)	67 (9)	1 (1)
Stretcher	119 (14)	103 (13)	16 (21)
Heart, Median (Q1, Q3)	92 (80, 106)	91 (80, 105)	100 (85, 116)	0.004
Systolic pressure, Median (Q1, Q3)	122 (107, 141)	123 (108, 141)	115 (93, 130.5)	< 0.001
diastolic pressure, Median (Q1,Q3)	70 (60, 81)	70 (60, 81)	63 (50, 75.5)	< 0.001
Shock Index, Median (Q1, Q3)	0.76 (0.61, 0.92)	0.74 (0.61, 0.91)	0.87 (0.7, 1.14)	< 0.001
Day, Median (Q1, Q3)	6 (4, 10)	6 (4, 9)	6 (2, 13.5)	0.543
Triage Level (ESI), n (%)				< 0.001
ESI Level 1	17 (2)	4 (1)	13 (17)
ESI Level 2	709 (84)	647 (84)	62 (83)
ESI Level 3	72 (9)	72 (9)	0 (0)
ESI Level 4	49 (6)	49 (6)	0 (0)
Type of Health Insurance for Hospitalization, n (%)				0.01
Employee Health Insurance	759 (90)	694 (90)	65 (87)
Resident Health Insurance	40 (5)	32 (4)	8 (11)
Self-Payment	36 (4)	36 (5)	0 (0)
Other	12 (1)	10 (1)	2 (3)
Season of Admission to Emergency Department, n (%)				0.762
Spring	222 (26)	203 (26)	19 (25)
Summer	205 (24)	190 (25)	15 (20)
Fall	225 (27)	204 (26)	21 (28)
Winter	195 (23)	175 (23)	20 (27)
Time of Admission to Emergency Department, n (%)				0.033
08:01–16:00	382 (45)	344 (45)	38 (51)
16:01–24:00	324 (38)	305 (40)	19 (25)
00:01–08:00	141 (17)	123 (16)	18 (24)
Emergency Admission Departments, n (%)				< 0.001
Gastroenterology	572 (68)	544 (70)	28 (37)
Gastrointestinal Surgery	28 (3)	28 (4)	0 (0)
Oncology	19 (2)	7 (1)	12 (16)
Intensive Care Unit (ICU)	90 (11)	67 (9)	23 (31)
Other	138 (16)	126 (16)	12 (16)
Type of Gastrointestinal Bleeding, n (%)				0.687
Upper Gastrointestinal Bleeding	703 (83)	639 (83)	64 (85)
Lower Gastrointestinal Bleeding	144 (17)	133 (17)	11 (15)
Comorbidities, n (%)
Diabetes	190 (22)	172 (22)	18 (24)	0.845
hypertension	313 (37)	286 (37)	27 (36)	0.957
Cerebral Infarction	160 (19)	143 (19)	17 (23)	0.471
Coronary Artery Disease	204 (24)	185 (24)	19 (25)	0.902
malignancy	153 (18)	114 (15)	39 (52)	< 0.001
In-Hospital Interventions, n (%)
Surgical Procedure	626 (74)	582 (75)	44 (59)	0.003
Tanesthesia, n (%)				< 0.001
General Anesthesia	399 (47)	380 (49)	19 (25)
Regional Anesthesia	137 (16)	129 (17)	8 (11)
None	311 (37)	263 (34)	48 (64)
Hemostatic Procedure Performed, n (%)	536 (63)	509 (66)	27 (36)	< 0.001
Hemostatic Procedure within 24 Hours of Emergency Admission, n (%)	316 (37)	296 (38)	20 (27)	0.061
Total Hospitalization Cost, Median (Q1, Q3)	13310.54 (8594.52, 25036.14)	12963.06 (8485.01, 23297.33)	30223.91 (9383.18, 54326.17)	< 0.001

Out of the 21 variables collected from patients, five variables were selected based on non-zero coefficients from the LASSO regression analysis (Fig. 2). These variables include Ambulance ED, Shock Index > 1, Admission to ICU, Malignancy, and Hemostatic Procedure. To develop a predictive model for in-hospital mortality among patients with gastrointestinal bleeding, multivariable logistic regression analysis was performed on these five variables using LASSO regression techniques. We constructed a mortality risk prediction model based on the aforementioned five predictors (Table 2).To assess the model's generalizability, the dataset was split into a training set (70%) and a validation set (30%). In the training set, the model's AUC was 0.8619, with a confidence interval (CI) of [0.7858, 0.9381]; in the validation set, the AUC was 0.8455, with a 95% CI of [0.7871, 0.9039]. These results indicate that the model maintained high and consistent performance across different datasets, demonstrating good statistical stability (Fig. 3). To visually represent the model's predictive capability, a nomogram was constructed based on the training data (Fig. 4). To further assess model calibration, calibration curves were plotted for both the training and validation sets (Fig. 5). In the training set, the calibration curve's slope was 1.02 with an intercept of -0.03, indicating a high agreement between predicted and observed probabilities. In the validation set, the calibration curve's slope was 1.05 with an intercept of -0.05, also showing good calibration. Finally, Decision Curve Analysis (DCA) was used to evaluate the clinical utility of the model (Fig. 6). In the training set, DCA results showed that when the threshold probability was between 0.1 and 0.9, our model significantly improved net benefit compared to the strategies of treating all or none. Specifically, at a threshold probability of 0.1, the model's net benefit was 0.741 higher than the treat-all strategy and 74.1% higher than the treat-none strategy. In the validation set, DCA similarly indicated that our model outperformed both treat-all and treat-none strategies when the threshold probability was between 0.1 and 0.9. At a threshold probability of 0.1, the model's net benefit was 0.814 higher than the treat-all strategy and 81.4% higher than the treat-none strategy. These data suggest that our model not only performs well statistically but also has significant clinical utility in decision-making.

Table 2

Logistic regression model and the Odds ratio of predictors
Variable	β	SE	Wald	OR	95%CI	P
Mode of Arrival to Emergency Department, By Ambulance	1.28	0.42	3.01	3.59	[1.56, 8.71]	0.003
Shock Index>1	0.95	0.39	2.43	2.58	[1.20, 5.56]	0.015
Emergency Admission Departments, ICU	0.98	0.44	2.25	2.67	[1.13, 6.28]	0.025
Malignancy, yes	1.70	0.37	4.57	5.46	[2.64, 11.29]	< 0.001
Hemostatic Surgery, yes	-0.99	0.38	-2.59	0.37	[0.19, 0.78]	0.010
Intercept	4.86	2.40	2.02	-	-	0.043
Notes: Logistic regression model: 1.28×（Mode of Arrival to Emergency Department，By Ambulance ）+0.95×（Shock Index＞1）+0.98×（Emergency Admission Departments，ICU）+1.70×（malignancy，yes）—0.99×（Hemostatic Surgery，yes）+4.86 CI, confidence interval; OR, odds ratio.

In this study, a nomogram was developed to predict in-hospital mortality risk for patients with gastrointestinal bleeding in the emergency department. This nomogram includes five variables: Ambulance ED, Shock Index > 1, Admitted to ICU, Malignancy, and Hemostatic Procedure. The nomogram demonstrated good discriminative ability, calibration, and clinical utility. To our knowledge, most current predictions of mortality risk for patients with gastrointestinal bleeding focus on hospitalized UGIB or LGIB patients. For emergency physicians, the mortality risk for patients with gastrointestinal bleeding admitted through the emergency department, especially when the bleeding location is unclear, is not accurately predicted. In this study, we developed a nomogram predicting all-cause mortality during hospitalization for such patients, based on five variables collected from patients admitted through the emergency department. The variables included in the nomogram were selected through LASSO regression analysis, which is considered superior to univariate analysis for predictor selection [19, 20].

Additionally, we evaluated the clinical significance of these predictive factors. The factors "arrival by ambulance to the emergency department" and "shock index > 1" reflect the urgency and severity of the patient's condition and are associated with poorer outcomes in gastrointestinal bleeding patients [21–22]. The shock index is calculated by dividing the heart rate by the systolic blood pressure, with a normal range considered to be 0.5–0.7. A shock index >1 has a significant potential in predicting short-term adverse outcomes in upper gastrointestinal bleeding patients[23–25]. Research by Dogru U, Yuksel et al. indicates that the shock index can serve as an important quantitative indicator for assessing mortality risk in gastrointestinal bleeding patients [26]. Emergency medical staff should prioritize patients arriving by ambulance and those with a shock index > 1, as these patients may require more urgent care. It is recommended to establish a fast-track system for such patients, optimize emergency procedures, and reduce the time spent in the emergency room to ensure timely and prioritized care for high-risk patients.

Admission to the ICU typically indicates that a patient's condition is very severe, requiring close monitoring and advanced life support. Thus, "Admitted to ICU" as an independent risk factor in predictive models reflects the severity of the patient's condition. Although the ICU provides more medical resources, the mortality rate remains high due to the severity of these patients' conditions. This underscores the importance of early identification and intervention, aiming to take effective measures before the condition worsens. Furthermore, ICU treatment is not solely focused on gastrointestinal bleeding; patients often have other systemic complications whose presence and treatment complexity can also affect prognosis [28]. ICU patients often require invasive procedures such as mechanical ventilation and central venous catheterization, which can increase the risk of infections and other complications, further impacting prognosis [29]. While "Admitted to ICU" is a statistically significant independent risk factor, it does not necessarily mean that ICU treatment itself leads to higher mortality. Rather, it likely reflects the severity of the patient's condition[30]. Therefore, for emergency clinical teams, early identification of high-risk patients who may need ICU admission and proactive intervention before their condition deteriorates may help reduce in-hospital mortality.

Patients with co-existing malignancy generally have poorer overall health and more complex, variable conditions. Therefore, "Malignancy" as an independent risk factor in predictive models reflects the complexity of the patient's condition and poor prognosis. Despite comprehensive treatment and care provided by medical teams, the mortality rate remains high due to the invasive nature of tumors and their systemic impact [31]. This highlights the necessity of fully considering the effects of systemic diseases and the importance of early comprehensive treatment when managing such patients. Additionally, patients with malignancy not only need to address gastrointestinal bleeding but often also suffer from other systemic complications such as anemia, malnutrition, and compromised immune function [32]. The presence of these complications and the complexity of their management can significantly impact the patient's prognosis. As these patients may require multiple treatments, such as chemotherapy and radiation therapy, the side effects of these treatments can further increase the complexity and risk of treatment [33]. Although "Malignancy" is a statistically significant independent risk factor, it does not necessarily mean that the malignancy itself directly leads to higher mortality. It is more likely a reflection of the deterioration of the patient's overall health status. Therefore, for patients with both malignancy and gastrointestinal bleeding, early identification of their complex conditions and the implementation of comprehensive treatment measures may help improve prognosis. In clinical practice, individualized treatment strategies are particularly important for these high-risk patients. Multidisciplinary collaboration, including close coordination between oncology, gastroenterology, and surgery departments, can ensure that patients receive comprehensive and effective treatment, thereby potentially reducing in-hospital mortality.

Regarding "Hemostatic Procedure," the execution of hemostasis surgery in patients with gastrointestinal bleeding is closely related to adverse survival outcomes during hospitalization after being admitted through the emergency department. The success of hemostasis procedures directly impacts bleeding control and the patient's chance of survival [34]. Our study results showed that performing hemostasis surgery on gastrointestinal bleeding patients transferred from the emergency department to inpatient care increased the survival rate by 62.76% compared to those who did not undergo the procedure. This also reflects the severity and complexity of bleeding in non-surgical patients, who may not tolerate surgery due to significant blood loss, difficult-to-control bleeding sites, or other serious comorbidities. Additionally, non-surgical patients may have advanced tumors or other factors affecting surgical decisions, such as overall health status or tumor biology. Although studies have shown that hemostasis within 24 hours significantly reduces mortality [35–36], timely and effective hemostasis interventions during hospitalization remain crucial for saving lives in certain high-risk gastrointestinal bleeding patients who cannot be treated within 24 hours [37–38]. For such patients, it is recommended to stabilize their condition as quickly as possible through multidisciplinary collaboration. By utilizing a combination of pharmacological, endoscopic, and interventional treatments to achieve early hemostasis, the risk of bleeding-related complications and mortality can be minimized.

These five predictive indicators are easily obtainable in clinical settings. The nomogram demonstrates good discriminative ability and calibration, and DCA evaluations show its clinical utility. This freely accessible nomogram for predicting mortality risk could provide the emergency clinical team with a clear and convenient preliminary assessment for managing the in-hospital mortality risk of gastrointestinal bleeding patients admitted through the emergency department. The nomogram developed in this study shows good discriminative ability. Based on our findings, we recommend integrating this model into existing clinical information systems, such as Electronic Health Records (EHR) systems, as a clinical decision support tool. Ensuring that emergency department doctors and nurses can easily access the model's predictive results will help them identify high-risk patients and make more accurate decisions at all stages of emergency care.

However, this study has some limitations. First, it is a retrospective Study, so its clinical utility requires external validation. Second, we did not integrate the model with existing clinical guidelines and practices. Further research is needed to validate this model. Nevertheless, we established a well-performing model based on the current emergency gastrointestinal bleeding patient data. Finally, our study results pertain to in-hospital mortality risk following admission from the emergency department. Further research is needed to determine if this model can predict in-hospital mortality for patients admitted directly from outpatient settings.

This study developed and validated a nomogram model to predict in-hospital mortality risk for patients with gastrointestinal bleeding (GIB) in the emergency department. The model demonstrated good classification performance in the validation set, with an AUC of 0.847, indicating significant clinical application potential. By integrating key clinical variables such as age and shock index, the model provides the emergency clinical team with an effective tool for quickly assessing patient risk and optimizing treatment decisions, especially for GIB patients with unclear bleeding sites. Despite the limitations of being a single-center study, the model shows potential value in improving the prognosis of emergency GIB patients. Future research should focus on external validation of the model and its integration into clinical practice to further confirm its broad applicability.

Author Contribution

Author Contributions StatementY.L., M.W., and L.O. contributed equally to this work.Y.L. conceptualized and designed the study, performed data collection, and conducted statistical analysis. M.W. assisted in data collection, conducted the literature review, and contributed to the interpretation of the results. L.O. was involved in designing the nomogram, drafting the manuscript, and revising it for important intellectual content. D.L.supervised the study, provided critical insights into the methodology, and ensured adherence to ethical standards. W.J. reviewed and edited the manuscript, contributing to its final version.All authors participated in revising the manuscript and approved the final version for submission.

Data Availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Nable, J. V. & Graham, A. C. Gastrointestinal Bleeding. Emerg Med Clin North Am. ;34(2):309 – 25. doi: (2016). 10.1016/j.emc.2015.12.001. Epub 2016 Mar 16. PMID: 27133246.
Balaban, D. V. et al. Predictors for in-hospital mortality and need for clinical intervention in upper GI bleeding: a 5-year observational study. Chirurgia (Bucur). PMID: 24524470. (2014) Jan-Feb;109(1):48–54.
Barkun, A. N. et al. Management of Nonvariceal Upper Gastrointestinal Bleeding: Guideline Recommendations From the International Consensus Group. Ann. Intern. Med. 171 (11), 805–822. 10.7326/M19-1795 (2019). Epub 2019 Oct 22. PMID: 31634917; PMCID: PMC7233308.
Kaya, E., Karaca, M. A., Aldemir, D. & Ozmen, M. M. Predictors of poor outcome in gastrointestinal bleeding in emergency department. World J. Gastroenterol. 22 (16), 4219–4225. 10.3748/wjg.v22.i16.4219 (2016). PMID: 27122672; PMCID: PMC4837439.
Yuan, L. & Yao, W. Development and Validation of a Risk Prediction Model for In-Hospital Mortality in Patients With Acute Upper Gastrointestinal Bleeding. Clin. Appl. Thromb. Hemost. 2023 Jan-Dec ;29:10760296231207806. 10.1177/10760296231207806. PMID: 37828791; PMCID: PMC10576926.
Stanley, A. J. et al. Outpatient management of patients with low-risk upper-gastrointestinal haemorrhage: multicentre validation and prospective evaluation. Lancet. 373 (9657), 42–47. 10.1016/S0140-6736(08)61769-9 (2009). Epub 2008 Dec 16. PMID: 19091393.
Laursen, S. B. et al. Performance of new thresholds of the Glasgow Blatchford score in managing patients with upper gastrointestinal bleeding. Clin. Gastroenterol. Hepatol. 13 (1), 115–21e2 (2015). Epub 2014 Jul 21. PMID: 25058843.
Laursen, S. B. et al. ABC score: a new risk score that accurately predicts mortality in acute upper and lower gastrointestinal bleeding: an international multicentre study. Gut. 70 (4), 707–716. 10.1136/gutjnl-2019-320002 (2021). Epub 2020 Jul 28. PMID: 32723845.
Colorectal Group of the Gastrointestinal Endoscopy Branch of the Chinese Medical Association, Colorectal Group of the Gastroenterologists Branch of the Chinese Physicians Association, National Clinical Medical Research Center for Digestive Diseases. Guidelines for the Diagnosis and Treatment of Lower Gastrointestinal Bleeding (2020) [J]. Chin. J. Gastrointest. Endoscopy. 37 (10), 685–695. 10.3760/cma.j.cn321463-20200618-00544 (2020).
Oakland, K. et al. Acute lower GI bleeding in the UK: patient characteristics, interventions and outcomes in the first nationwide audit. Gut. 67 (4), 654–662. 10.1136/gutjnl-2016-313428 (2018). Epub 2017 Feb 1. PMID: 28148540.
Tominaga, N. et al. A novel prediction tool for mortality in patients with acute lower gastrointestinal bleeding requiring emergency hospitalization: a large multicenter study. Sci. Rep. 14 (1), 5367. 10.1038/s41598-024-55889-7 (2024). PMID: 38438534; PMCID: PMC10912311.
Zhao, Y., Chee, M. Y. M., Sultana, R. & Tan, W. J. Safe discharge for patients admitted for lower gastrointestinal bleeding (LGITB): derivation and validation of a novel scoring system. BMC Gastroenterol. 23 (1), 349. 10.1186/s12876-023-02950-w (2023). PMID: 37814216; PMCID: PMC10561471.
Oakland, K. et al. Derivation and validation of a novel risk score for safe discharge after acute lower gastrointestinal bleeding: a modelling study. Lancet Gastroenterol. Hepatol. 2 (9), 635–643. 10.1016/S2468-1253(17)30150-4 (2017). Epub 2017 Jun 23. PMID: 28651935.
Chen, L., Zheng, H. & Wang, S. Prediction model of emergency mortality risk in patients with acute upper gastrointestinal bleeding: a retrospective study. PeerJ. 9, e11656. 10.7717/peerj.11656 (2021). PMID: 34221734; PMCID: PMC8236237.
Time to Get Control. A review of the care received by patients who had a severe gastrointestinal haemorrhage (NCEPOD, 2015).
Wang, S. & Tu, J. Nomogram to predict multidrug-resistant tuberculosis. Ann. Clin. Microbiol. Antimicrob. 19 (1), 27. 10.1186/s12941-020-00369-9 (2020). PMID: 32505203; PMCID: PMC7276074.
Huang, Y. Q. et al. Development and validation of a radiomics nomogram for preoperative prediction of lymph node metastasis in colorectal cancer. J. Clin. Oncol. 34 (18), 2157–2164. 10.1200/JCO.2015.65.9128 (2016).
Vickers, A. J., Cronin, A. M., Elkin, E. B. & Gonen, M. Extensions to decision curve analysis, a novel method for evaluating diagnostic tests, prediction models and molecular markers. BMC Med. Inf. Decis. Mak. 8 (1), 53. 10.1186/1472-6947-8-53 (2008).
Tibshirani, R. The lasso method for variable selection in the Cox model. Stat Med. ;16(4):385 – 95. doi: (1997). 10.1002/(sici)1097-0258(19970228)16:4<385::aid-sim380>3.0.co;2-3. PMID: 9044528.
Balachandran, V. P., Gonen, M., Smith, J. J. & DeMatteo, R. P. Nomograms in oncology: more than meets the eye. Lancet Oncol. ;16(4):e173-80. doi: 10.1016/S1470-2045(14)71116-7. PMID: 25846097; PMCID: PMC4465353. (2015).
Fil'kov, A. P. & Kozlov, M. I. Opyt lecheniia bol'nykh s gastroduodenal'nymi krovotecheniiami razlichnoĭ étiologii, po materialam sluzhby sanitarnoĭ aviatsii [Experience in the treatment of patients with gastroduodenal hemorrhage of different etiology: based on materials of an air ambulance service]. Vestn Khir Im I I Grek. 158 (3), 63–66 (1999). Russian. PMID: 10481889.
Cai, J. X., Saltzman, J. R., Initial Assessment, R., Stratification & Early Management of Acute Nonvariceal Upper Gastrointestinal Hemorrhage., and Gastrointest. Endosc Clin. N Am. ;28(3):261–275. doi: 10.1016/j.giec.2018.02.001. (2018). Epub 2018 Apr 17. PMID: 29933774.
Rady, M. Y., Smithline, H. A., Blake, H., Nowak, R. & Rivers, E. A comparison of the shock index and conventional vital signs to identify acute, critical illness in the emergency department. Ann. Emerg. Med. 24 (4), 685–690. https://doi.org/10.1016/s0196-0644(94)70279-9 (1994). PMID: 8092595.
Erratum, Rady, M. Y., Nightingale, P., Little, R. A. & Edwards, J. D. 24(6):1208. 7.,in: Ann Emerg Med. Shock index: a reevaluation in acute circulatory failure. Resuscitation. 1992;23(3):227 – 34. PMID: 1321482; (1994). https://doi.org/10.1016/0300-9572(92)90006-x. 8.
Allgöwer, M. & Burri, C. Schockindex [Shock index]. Dtsch Med Wochenschr. ;92(43):1947-50. PMID: 5299769; (1967). https://doi.org/10.1055/s-0028-1106070
Dogru, U. et al. 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. 140 (4), 531–539 (2022 Jul-Aug). PMID: 35544884; PMCID: PMC9491474.
Stanley, A. J. et al. Comparison of risk scoring systems for patients presenting with upper gastrointestinal bleeding: international multicentre prospective study. BMJ. 356, i6432 (2017).
ZULLO, A. et al. Clinical outcomes in cirrhotics with variceal or nonvariceal gastrointestinal bleeding: A prospective, multicenter cohort study. J. Gastroenterol. Hepatol., 2021[2021-10-07]. https://doi.org/10.1111/jgh.15601
LEONTIADIS G I, MOLLOY-BLAND M, MOAYYEDI, P. & HOWDEN C, W. Effect of comorbidity on mortality in patients with peptic ulcer bleeding: systematic review and meta-analysis[J]. Am. J. Gastroenterol. 108 (3), 331–345 (2013).
Kupfer, Y., Cappell, M. S. & Tessler, S. Acute gastrointestinal bleeding in the intensive care unit. The intensivist's perspective. Gastroenterol Clin North Am. ;29(2):275–307, v. doi: (2000). 10.1016/s0889-8553(05)70117-5. PMID: 10836184.
Sun, D. et al. Cancer burden in China: trends, risk factors and prevention. Cancer Biol. Med. 17 (4), 879–895. 10.20892/j.issn.2095-3941.2020.0387 (2020). Epub 2020 Dec 15. PMID: 33299641; PMCID: PMC7721090.
Wu, W. D. et al. Rare cause of upper gastrointestinal bleeding owing to hepatic cancer invasion: a case report. World J. Gastroenterol. 20 (35), 12704–12708. 10.3748/wjg.v20.i35.12704 (2014). PMID: 25253981; PMCID: PMC4168114.
Peterson, D. E. & Cariello, A. Mucosal damage: a major risk factor for severe complications after cytotoxic therapy. Semin Oncol. ;31(3 Suppl 8):35–44. doi: (2004). 10.1053/j.seminoncol.2004.04.006. PMID: 15181607.
Gralnek, I. M. et al. Endoscopic diagnosis and management of nonvariceal upper gastrointestinal hemorrhage (NVUGIH): European Society of Gastrointestinal Endoscopy (ESGE) Guideline - Update 2021. Endoscopy. 53 (3), 300–332. 10.1055/a-1369-5274 (2021). Epub 2021 Feb 10. PMID: 33567467.
Wilkins, T., Wheeler, B. & Carpenter, M. Upper Gastrointestinal Bleeding in Adults: Evaluation and Management. Am Fam Physician. ;101(5):294–300. Erratum in: Am Fam Physician. 2021;103(2):70. PMID: 32109037. (2020).
Lanas, A. et al. Non-variceal upper gastrointestinal bleeding. Nat Rev Dis Primers. ;4:18020. doi: (2018). 10.1038/nrdp.2018.20. PMID: 29671413.
Hawks, M. K. & Svarverud, J. E. Acute Lower Gastrointestinal Bleeding: Evaluation and Management. Am. Fam Physician. 101 (4), 206–212 (2020). PMID: 32053333.
Sengupta, N. et al. Management of Patients With Acute Lower Gastrointestinal Bleeding: An Updated ACG Guideline. Am. J. Gastroenterol. 118 (2), 208–231. 10.14309/ajg.0000000000002130 (2023). Epub 2022 Sep 21. PMID: 36735555.

No competing interests reported.

Basicdata.xlsx

Download PDF

Editor invited by journal
29 Aug, 2024
Submission checks completed at journal
27 Aug, 2024
First submitted to journal
13 Aug, 2024

You are reading this latest preprint version

Nomogram Predicts In-Hospital Mortality in Patients with Emergency Gastrointestinal Bleeding: A Multicenter Retrospective Study

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Research Methods

Study design

Subjects of the study

Data collection

Statistical methods

Results

Discussion

Conclusion

Declarations

Author Contribution

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1