a b s t r a c t

Introduction: Acute gastrointestinal (GI) bleeding is one of the most important and the common causes of pa- tients visiting the emergency department. Dehydrated state leads to increased blood urea nitrogen (BUN) and decreased Albumin level. Many scoring systems had been developed to predict outcomes for patients with GI bleeding. Among the many scoring systems, the AIMS65 score was a simple and accurate risk assessment scoring tool. Therefore, in this study, we evaluated the prognostic performance of the blood urea nitrogen to serum albu- min ratio (B/A ratio) compared to the AIMS65 score tool in elderly patients with GI bleeding.

Material and methods: This was a retrospective cohort study of patients with GI bleeding in our institution from February 2018 through May 2020. Baseline characteristic data were obtained. The data were compared with the prevalence of ICU admission and in-hospital mortality. The B/A ratio and the AIMS65 score as predictors of ICU admission and in-hospital mortality was evaluated using the area under the receiver operating characteristic (AUROC) curve.

Results: In the 596 patients included in the study, of whom 188 (31.5%) were admitted to the ICU and 36 (6.0%) died during hospitalization. Multivariate logistic regression analysis revealed that the B/A ratio was significant predictors of ICU admission and in-hospital mortality. In addition, the B/A ratio was significant higher in ICU ad- mission patients and non-survivors. The AUROC value of the B/A ratio was 0.682 and that of the AIMS65 score was 0.699 for predicting ICU admission. For predicting in-hospital mortality, the AUROC value was 0.770 and 0.763, respectively.

Conclusion: The B/A ratio is as simple and useful tool for predicting outcomes for old aged GI bleeding patients as the AIMS65 score.

(C) 2020

1. Introduction

Acute gastrointestinal (GI) bleeding is one of the most significant emergency conditions worldwide [1-4]. It was the fifth most common GI disease (of 18) and a major cause of ED visits [5]. In the United States, there were more than 800,000 ED visits every year, half of which required hospitalization [5]. Improvements in endoscopic thera- pies and pharmacological treatments, such as acid-suppressing agents (proton pump inhibitors), have reduced mortality associated with GI bleeding [6-9]. Despite diagnostic and therapeutic developments, GI bleeding still has a 5-10% mortality rate [10] and is a health care burden for patients, physicians, and health care systems. [11]. Nevertheless, risk stratification is essential to improve patient prognosis.

* Corresponding author.

E-mail address: [email protected] (S.H. Lee).

The world population, including Korea, is aging at an unprece- dented rate. The proportion of the population that is over 65 years of age has increased from 7.0% in 1999 to 13.8% in 2017 and is ex- pected to increase to 20.3% by the year 2025 [12]. Some reports sug- gest that being over the age of 65 was an important variable affecting poor outcomes from GI bleeding [13,14]. Therefore, we determined the factors that can predict the prognosis for Patients over the age of 65 who experience GI bleeding.

There are many scoring systems to predict outcomes for patients with GI bleeding, including mortality, length of hospital stay, and the need for endoscopy or other interventions. The Glasgow-Blatchford score (GBS) [15] and the Rockall Score [16] are the most widely used scoring tools. However, these scoring systems are complex and dif- ficult to use, and the RS requires endoscopic findings. Recently, the sim- ple and accurate AIMS65 score system was introduced [13,17,18]. Patients over the age of 65 are often unaware of their underlying disease and illness (melena, presence of syncope, etc.) due to intoxication, no

https://doi.org/10.1016/j.ajem.2020.12.022

0735-6757/(C) 2020

health caretaker, neurodegenerative disease, or a confused mentality due to aggravated GI bleeding.

Hypovolemia and dehydration were common in patients with se- vere GI bleeding. In the dehydrated state, increased urea reabsorption by the kidney leads to an increased blood urea nitrogen (BUN) level [19]. Although some studies suggest that the prognosis of pneumonia patients can be predicted with serum albumin level and the BUN: serum albumin ratio (B/A ratio) [19-21], no studies have used the B/A ratio in patients with GI bleeding.

Therefore, we evaluated the Predictive performance of the B/A ratio in Geriatric patients with GI bleeding.

1. Material and methods
   1. Study design

This was a single-center, retrospective study that used the Electronic medical records of patients who presented in the ED with GI bleeding symptoms, such as hematemesis, melena, and hematochezia. The study was approved by the institutional review board of our hospi- tal, and the requirement for written informed consent was waived.

• 1. Study setting and population

This study included patients who visited a tertiary-care university hos- pital that had 65,000 annual emergency visits from February of 2018 through May of 2020. The inclusion criteria were: 1) those aged 65 years or older who visited the ED and 2) patients complaining of any GI bleeding symptoms (melena, hematemesis, and/or hematochezia). The exclusion criteria were: 1) patients who visited the ED for non-medical purposes,

The primary outcome was in-hospital mortality, defined as any death occurring during admission. The secondary outcome was how the ICU admission compared to the GW admission or discharge.

2.4. Statistical analysis

Continuous variables were presented as the mean with a standard de- viation (SD), while categorical variables were presented as count (percent- age). Baseline demographic and clinical characteristics were summarized using the independent t-test for continuous variables and Pearson’s chi- square test for categorical variables. The performance of the AIMS65 score and B/A ratio for predicting outcome was evaluated with the area under the receiver operating characteristic (AUROC) curve. An AUROC curve between 0.8 and 0.9 is considered good, a curve between 0.7 and

0.8 is considered adequate, and a curve between 0.6 and 0.7 is considered poor [22]. For each scoring system, a cut-off value was calculated that max- imized the sum of sensitivity and specificity [23]. All statistical analyses were performed using SPSS 26.0 (SPSS Inc., Chicago, IL). The AUROC curve analysis was performed with the DeLong method by using MedCalc Statistical Software version 19 (MedCalc Software bvba, Ostend, Belgium). The significance level was set at p-value <0.05.

1. Results
   1. Characteristics of the study population

A total of 596 patients enrolled in our ED during the study period. The demographic characteristics, vital signs, initial laboratory findings,

Table 1

Demographic, laboratory and clinical variables of patients (Total n = 596)

2) patients who have other spontaneous infectious disease or trauma dur-

ing ED that are confounding risk factors for a poor prognosis, and 3) pa-

tients with incomplete EMRs data.

2.3. Data collection and outcome measurement

Patients with GI bleeding were confirmed in the ED by board- certified emergency physicians based on the presence of melena, hematemesis, and/or hematochezia. Before admission or discharge from the hospital, the final diagnosis was reviewed by board-certified emergency physicians or board-certified gastrointestinal physicians if the endoscopy was performed. Data were collected from the EMRs of the eligible patients. Data included patient demographics (sex and age), initial vital signs in the ED (systolic blood pressure (SBP), diastolic blood pressure , pulse rate (PR), respiratory rate (RR), body tem- perature, and mental status), laboratory findings (white blood cell (WBC), hemoglobin, hematocrit, platelet, C-reactive protein (CRP), BUN, creatinine, albumin, total bilirubin, aspartate aminotransferase , alanine aminotransferase (ALT), amylase, glucose, sodium, potas- sium, troponin-T, prothrombin time (international normalized ratio [INR]), and partial thromboplastin time ), and clinical details, including the ED treatment results (discharge, general ward [GW] ad- mission, and intensive care unit [ICU] admission), length of hospital stay, and mortality.

The AIMS65 score was composed of age, albumin level, SBP, PT (INR), and mental status. This scoring system assigned one point for each of the following variables: age older than 65 years, albumin level < 30 g/L, SBP < 90 mmHg, PT (INR) > 1.5, and altered mental status.

We defined the ICU positive group (ICU (+) group) as patients ad- mitted to the ICU and the ICU negative group (ICU (-) group) as those patients discharged or admitted to the GWs. We also defined the mortality positive group (mortality (+) group) as patients who died in the hospital and the mortality negative group (mortality (-) group) as the discharged patients.

Variable Value

Age (years) 78.0 +- 8.0

Sex

Male 327 (54.9)

Female 269 (45.1)

Systolic Blood Pressure (mmHg) 119.8 +- 28.2

Diastolic Blood Pressure (mmHg) 62.9 +- 16.6

Pulse rate (beats/min) 89.9 +- 20.4

Respiratory rate (breath/min) 19.7 +- 2.4

Body temperature (?C) 36.2 +- 3.8

Mental status

Alert 561 (94.1)

Verbal response 13 (2.2)

Painful response 19 (3.2)

Unresponsive 3 (0.5)

Laboratory test

White blood cell (10⁹/L) 10.4 +- 14.8

Hemoglobin (g/dL) 9.3 +- 3.0

Hematocrit (%) 26.9 +- 8.3

Platelet(10⁹/L) 217.9 +- 111.5

C-reactive protein (mg/dL) 2.5 +- 4.5

Blood urea nitrogen (mg/dL) 36.8 +- 26.6

Creatinine (mg/dL) 1.6 +- 1.6

Albumin (g/dL) 3.1 +- 0.7

Total bilirubin (mg/dL) 0.7 +- 1.5

Aspartate aminotransferase (IU/L) 36.6 +- 61.1

Alanine aminotransferase (IU/L) 22.8 +- 48.0

Amylase (IU/L) 73.6 +- 61.4

Glucose (mg/dL) 173.5 +- 99.1

Sodium (mmol/L) 136.5 +- 4.9

Potassium (mmol/L) 4.3 +- 0.8

Troponin-T (ng/mL) 0.06 +- 0.25

PT (INR) 1.4 +- 1.4

PTT (sec) 27.2 +- 10.9

Hospital day (days) 7.8 +- 12.6

ICU admission 188 (31.5)

In-hospital mortality 36 (6.0)

BUN/Albumin ration 13.0 +- 11.4

The values are given as mean +- standard deviation. The values are given as number (%).

and clinical outcomes are in Table 1. The mean age +- SD of the 596 pa- tients was 78.0 +- 8.0 years, and 327 patients (54.9%) were men. There were 188 patients admitted to the ICU (31.5%), and 36 patients died in the hospital (6.0%).

• 1. Comparing clinical factors for in-hospital mortality

The mortality (+) group had significantly lower blood pressure (SBP and DBP) and body temperature compared to the mortality (-) group. The mortality (+) group had a significantly greater percentage of pa- tients with an altered mental status. The mortality (+) group had signif- icantly higher CRP, creatinine, total bilirubin, AST, ALT, amylase, potassium, PT (INR), and PTT levels. This group had significantly lower hemoglobin and platelet values. Additionally, the B/A ratio was higher in the mortality (+) group than the mortality (-) group. A multivariate logistic regression analysis demonstrated that hemoglobin, hematocrit, ALT, and the B/A ratio were significant independent variables for predicting death. (Table 2).

• 1. Comparing clinical factors for ICU admission

The ICU (+) group had significantly lower blood pressure (SBP and DBP) and body temperature compared to the ICU (-) group. The ICU (+) group had a statistically significant higher percentage of patients with altered mental status. Significant differences between the ICU (-) group and ICU (+) group were observed for the hemoglobin (9.7 +- 3.0 versus 8.4 +- 2.7), hematocrit (28.1 +- 8.4 versus 24.3 +-

7.5), amylase (69.5 +- 45.7 versus 82.4 +- 85.5), glucose (163.9 +- 92.0

versus 194.4 +- 110.3), potassium (4.2 +- 0.6 versus 4.5 +- 1.0), and PT (INR) (1.3 +- 1.2 versus 1.6 +- 1.7) values. The B/A ratio was significantly higher in the ICU (+) group than in the ICU (-) group (11.0 +- 9.5

versus 17.4 +- 13.7, P < 0.001). Multivariate logistic regression analysis demonstrated that SBP, hematocrit, and B/A ratio were significant inde- pendent variables for predicting ICU admission. (Table 3).

• 1. Predictive performance of the B/A ratio compared to the AIMS65 score

The AUROC curve for predicting ICU admission is in Fig. 1. The AUROC for the AIMS65 score predicting ICU admission was 0.699 (95% confidence interval [CI], 0.660 to 0.736), and for B/A ratio, it was 0.682 (95% CI, 0.643 to 0.719). There was no statistically significant difference between the two predicting systems. The cutoff value for separating risk groups for ICU admission was above 2 points on the AIMS65 scale (sen- sitivity: 68.6%, specificity: 66.4%) and the B/A ratio was 9.70 (sensitivity: 68.1%, specificity: 61.9%).

The AUROC curve for predicting in-hospital mortality is shown in Fig. 2. The AUROC for the AIMS65 score was 0.763 (95% CI: 0.727 to 0.797), and the B/A ratio was 0.770 (95% CI 0.734 to 0.803). The B/A ratio was superior to the AIMS65 score, without a statistically significant difference. The cutoff value of the AIMS65 score was 2 points (sensitiv- ity: 52.8%, specificity: 88.0%) and that of the B/A ratio was 15.1 (sensitiv- ity: 72.2%, specificity: 76.2%), which demonstrated higher sensitivity and specificity. (See Fig. 3.)

1. Discussion

It is essential to create an appropriate treatment plan for patients with GI bleeding to avoid devastating results. According to our findings, the B/A ratio was an independent predictor of the poor prognosis of ge- riatric patients with GI bleeding. Many scoring systems predict poor prognosis, however, most are not useful in ED because they require EGD findings, underlying disease, and physical examination, which

Table 2

Logistic regression analysis of ICU admission predictors

	Univariate analysisa					Multivariate analysis+
Variable	ICU (-) group n = 408	ICU (+) group n = 188		P-value		OR	B		P-value
Age (years)	77.9 +- 8.2	78.1 +- 7.5		0.819
Sex; Male	211 (51.7)	116 (61.7)		0.023
Systolic Blood Pressure (mmHg)	124.7 +- 24.8	109.2 +- 32.0		<0.001		0.99 (0.98, 1.00)	-0.012		0.005
Diastolic Blood Pressure (mmHg)	65.5 +- 15.3	57.4 +- 17.9		<0.001		1.00 (0.98, 1.02)	0.001		0.887
Pulse rate (beats/min)	88.2 +- 19.0	87.3 +- 23.2		0.632
Respiratory rate (breath/min)	19.7 +- 1.7	19.7 +- 3.4		0.977
Body temperature (?C)	36.6 +- 2.0	35.5 +- 6.1		0.025		0.99 (0.91, 1.06)	-0.015		0.702
Altered mental status	11 (2.7)	24 (12.8)		<0.001		1.38 (0.54, 3.49)	0.320		0.500
Laboratory test White blood cell (109/L)	10.0 +- 17.4	11.2 +- 6.2		0.413
Hemoglobin (g/dL)	9.7 +- 3.0	8.4 +- 2.7		<0.001		1.28 (0.69, 2.38)	0.249		0.428
Hematocrit (%)	28.1 +- 8.4	24.3 +- 7.5		<0.001		0.97 (0.94, 0.99)	-0.035		0.018
Platelet(109/L)	222.6 +- 119.1	207.9 +- 92.5		0.136
C-reactive protein (mg/dL)	2.4 +- 4.5	2.7 +- 4.5		0.456
Creatinine (mg/dL)	1.5 +- 1.5	1.8 +- 1.7		0.054
Total bilirubin (mg/dL)	0.7 +- 1.3	0.8 +- 1.9		0.254
Aspartate aminotransferase (IU/L)	33.1 +- 43.8	44.1 +- 87.2		0.062
Alanine aminotransferase (IU/L)	21.5 +- 41.5	25.7 +- 60.0		0.332
Amylase (IU/L)	69.5 +- 45.7	82.4 +- 85.5		0.030		1.00 (1.00, 1.00)	0.002		0.234
Glucose (mg/dL)	163.9 +- 92.0	194.4 +- 110.3		0.001		1.00 (1.00, 1.00)	0.002		0.156
Sodium (mmol/L)	136.6 +- 4.6	136.3 +- 5.7		0.441
Potassium (mmol/L)	4.2 +- 0.6	4.5 +- 1.0		<0.001		1.16 (0.83, 1.60)	0.144		0.386
Troponin-T (ng/mL)	0.04 +- 0.10	0.08 +- 0.42		0.276
PT (INR)	1.3 +- 1.2	1.6 +- 1.7		0.027		1.15 (0.95, 1.38)	0.136		0.149
PTT (sec)	26.7 +- 10.9	28.3 +- 10.9		0.105
Hospital day (days)	6.6 +- 10.8	10.3 +- 15.5		0.002		1.02 (1.00, 1.04)	0.022		0.023
In-Hospital Mortality	19 (4.7)	17 (9.0)		0.040		0.51 (0.18, 1.42)	-0.676		0.197
BUN/Albumin ratio	11.0 +- 9.5	17.4 +- 13.7		<0.001		1.03 (1.00, 1.05)	0.030		0.007

+ Data in parentheses are 95% confidence intervals, conducted on variables with a p value of <0.05 on univariate analysis. OR odds ratio, B regression coefficient.

Boldface type indicates statistical significance (p < 0.05).

^a Data are mean (standard deviation) or number (%).

Table 3

Logistic regression analysis of Mortality predictors

	Univariate analysisa						Multivariate analysis+
Variable	Mortality (-) group		Mortality (+) group
	n = 560		n = 36		P-value		OR	B		P-value
Age (years)	77.9 +- 7.8		78.5 +- 9.4		0.661
Sex; Male	302 (53.9)		25 (69.4)		0.074
Systolic Blood Pressure (mmHg)	120.6 +- 27.6		108.0 +- 34.5		0.010		0.99 (0.97, 1.03)	-0.002		0.903
Diastolic Blood Pressure (mmHg)	63.4 +- 16.0		55.3 +- 22.7		0.005		0.99 (0.95, 1.03)	-0.014		0.485
Pulse rate (beats/min)	87.7 +- 20.0		91.2 +- 26.4		0.317
Respiratory rate (breath/min)	19.7 +- 2.2		20.0 +- 4.4		0.537
Body temperature (?C)	36.3 +- 3.2		34.4 +- 8.8		0.013		1.37 (0.58, 3.24)	0.313		0.477
Altered mental status	27 (4.8)		8 (22.2)		<0.001		0.10 (0.01, 2.09)	-2.271		0.139
Laboratory test White blood cell (109/L)	10.3 +- 15.1		12.5 +- 6.9		0.449
Hemoglobin (g/dL)	9.4 +- 2.9		7.8 +- 2.8		0.003		10.23 (1.23, 84.96)	2.325		0.031
Hematocrit (%)	27.2 +- 8.3		22.9 +- 7.8		0.003		0.39 (0.18, 0.85)	-0.951		0.017
Platelet(109/L)	222.4 +- 111.2		148.8 +- 93.9		<0.001		1.00 (1.00, 1.01)	-0.001		0.819
C-reactive protein (mg/dL)	2.3 +- 4.5		4.1 +- 4.7		0.030		0.96 (0.85, 1.09)	-0.036		0.570
Creatinine (mg/dL)	1.5 +- 1.5		2.8 +- 2.6		<0.001		0.95 (0.69, 1.32)	-0.051		0.761
Total bilirubin (mg/dL)	0.7 +- 1.1		1.8 +- 4.1		0.016		1.15 (0.99, 1.33)	0.138		0.065
Aspartate aminotransferase (IU/L)	32.8 +- 46.5		95.6 +- 158.2		<0.001		1.00 (0.98, 1.01)	-0.004		0.642
Alanine aminotransferase (IU/L)	20.3 +- 34.0		63.4 +- 139.0		0.002		1.01 (1.00, 1.03)	0.014		0.008
Amylase (IU/L)	72.0 +- 54.7		100.6 +- 129.4		0.022		1.00 (0.99, 1.01)	0.001		0.740
Glucose (mg/dL)	172.9 +- 99.6		182.5 +- 92.0		0.572
Sodium (mmol/L)	136.5 +- 4.8		136.4 +- 6.9		0.909
Potassium (mmol/L)	4.3 +- 0.8		4.8 +- 0.9		<0.001		1.01 (0.43, 2.37)	0.011		0.981
Troponin-T (ng/mL)	0.05 +- 0.26		0.07 +- 0.07		0.761
PT (INR)	1.3 +- 1.2		2.4 +- 2.8		<0.001		0.42 (0.17, 1.04)	-0.878		0.061
PTT (sec)	26.3 +- 8.1		41.3 +- 27.8		<0.001		1.05 (0.96, 1.14)	0.044		0.299
Hospital day (days)	7.7 +- 12.5		9.8 +- 14.58		0.346
ICU admission	171 (30.5)		17 (47.2)		0.040		1.12 (0.32, 3.87)	0.108		0.864
BUN/Albumin ratio	12.1 +- 9.9		26.9 +- 20.6		<0.001		1.06 (1.01, 1.11)	0.054		0.030

+ Data in parentheses are 95% confidence intervals, conducted on variables with a p value of <0.05 on univariate analysis. OR odds ratio, B regression coefficient.

Boldface type indicates statistical significance (p < 0.05).

^a Data are mean (standard deviation) or number (%).

may vary among physicians [24,25]. There must be an objective predic- tive value for geriatric patients with GI bleeding. This is the first study to evaluate the association between the B/A ratio and the poor prognosis of geriatric patients with GI bleeding. The results demonstrate that the B/A ratio independently predicted the in-hospital mortality of this patient cohort. A positive association was observed between a high B/A ratio and poor prognosis, as evidenced by in-hospital mortality and ICU ad- mission rates. This is consistent with another report demonstrating that a high B/A ratio measured in the ED was independently associated

Fig. 1. Flow chart of patients enrolled in the study. ED: Emergency Department; GI bleeding: Gastrointestinal bleeding; EMR: Electronic Medical Records.

with in-hospital mortality, multi-organ failure, and sepsis [26-28]. In ad- dition, we presented the cut-off value of poor prognosis and demon- strated a comparative value superior to the AIMS 65 scoring system.

Fig. 2. ICU admission receiver operating characteristic (ROC) curve for AIMS65 and B/A ratio.

Fig. 3. In-hospital mortality receiver operating characteristic (ROC) curve for AIMS65 and B/A ratio.

Several studies have reported on a prognosis factor in patients with GI bleeding, including serum BUN, creatine, and albumin levels [29,30]. Albumin synthesis is inhibited in chronic disease [31]. Serum albumin levels can indirectly reflect Nutritional status, such as dehydration in pa- tients with mucosal conditions [32]. Previous research indicates that azotemia is associated with acute GI bleeding [30,33]. Although the in- crease in serum BUN level may be caused by dehydration, azotemia may occur in patients with GI bleeding due to the absorption of blood products in the GI tract [34]. BUN, one of the most widely used scoring systems for GI bleeding, is a variable in GBS, [15] and albumin is in- cluded in the AIMS65 score [13].

Compared to other scoring systems, such as GBS or RS, the AIMS65 score is simple and easy to apply in the ED. The AIMS65 score is com- posed of the serum albumin level, prothrombin time (international nor- malized ratio [INR]), systolic blood pressure, mental status, and age (is 65 years and older). The AIMS65 score is dependent on age, which tends to result in higher scores in elderly patients. Over the last few de- cades, the extensive use of Non-steroidal anti-inflammatory drugs (NSAIDS) and anti-thrombotic agents has resulted in the GI bleeding pa- tients being older [35-37]. Therefore, in this study, we evaluated the prognostic performance of the B/A ratio compared to the AIMS65 score tool only in those patients with GI bleeding who are aged 65 or older. Also, a disadvantage of the AIMS65 scoring system is that mental status is judged subjectively by the physician [38,39]. On the other hand, the B/A ratio is based on objective laboratory results, such as BUN and albumin.

Among the various multivariate logistic analysis variables, including demographic data, vital signs, and laboratory tests, the B/A ratio was sig- nificant in both ICU admission and in-hospital mortality (Tables 2, 3). The B/A ratio had a superior AUROC value, with a 72.2% sensitivity and a 76.2% specificity for in-hospital mortality, a 68.1% sensitivity and a 62.0% specificity for ICU admission to the that of the AIMS65 score. There were no statistically significant differences between the two sys- tems. The B/A ratio analyzed using the ROC curve showed better perfor- mance for in-hospital mortality than ICU admission (0.770 vs. 0.682). According to these results, only the simple equation between BUN and albumin can predict poor prognosis of geriatric patients with GI bleed- ing, and it is simple, economically efficient, and widely available. Fur- thermore, the B/A ratio is superior to the AIMS65 score, and the equation is based on objective laboratory findings.

• 1. Limitations

First, this was a retrospective, single-center study, and laboratory re- sults and Treatment outcomes were based on existing clinical records. There may be selection bias due to the limited sample size in a single in- stitute. Caution should be used in generalizing our results because they cannot represent all patients with GI bleeding. Further multi-centered, prospective design studies are required. Second, older patients tend to have co-morbidities, which may affect prognosis. Third, the relatively small sample size of in-hospital mortality cases compared with survi- vors may be insufficient to identify all factors. Next, our large-scale ter- tiary academic hospital received patients from another nearby hospital, but we did not confirm if albumin was administered before transfer. Last, the BUN baseline level was not considered in patients with chronic kidney disease, even though it was likely high.

1. Conclusion

In conclusion, the B/A ratio was associated with ICU admission and mortality in elderly patients with GI bleeding. The B/A ratio was not only a useful tool for predicting outcomes, but was easier to apply in the ED than the AIMS65 score. We recommend the B/A ratio for predicting GI bleeding severity in elderly patients. A B/A ratio over

15.1 requires aggressive, prompt, and timely treatment to prevent dev- astating outcomes.

Grant or other financial support

none.

Previous presentation

none.

Credit Author Statement

Manuscript title: Predictive performance of BUN/albumin ratio in elderly patients with gastrointestinal bleeding.

All persons who meet authorship criteria are listed as authors, and all authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript. Fur- thermore, each author certifies that this material or similar material has not been and will not be submitted to or published in any other publica- tion before its appearance in the American Journal of Emergency Medicine.

Declaration of Competing Interest

None declared.

References

1. Lau JY, Barkun A, Fan DM, Kuipers EJ, Yang YS, Chan FK. Challenges in the manage- ment of acute Peptic ulcer bleeding. Lancet. 2013;381(9882):2033-43. https://doi. org/10.1016/s0140-6736(13)60596-6.
2. Sostres C, Lanas A. Epidemiology and demographics of upper gastrointestinal bleed- ing: prevalence, incidence, and mortality. Gastrointest Endosc Clin N Am. 2011;21 (4):567-81. https://doi.org/10.1016/j.giec.2011.07.004.
3. Longstreth GF. Epidemiology of hospitalization for acute upper gastrointestinal hemorrhage: a population-based study. Am J Gastroenterol. 1995;90(2):206-10.
4. Stanley AJ, Ashley D, Dalton HR, Mowat C, Gaya DR, Thompson E, et al. outpatient management of patients with low-risk upper-gastrointestinal haemorrhage: multicentre validation and prospective evaluation. Lancet. 2009;373(9657):42-7. https://doi.org/10.1016/s0140-6736(08)61769-9.
5. Peery AF, Crockett SD, Murphy CC, Lund JL, Dellon ES, Williams JL, et al. Burden and Cost of Gastrointestinal, Liver, and Pancreatic Diseases in the United States: Update 2018. Gastroenterology. 2019;156(1). https://doi.org/10.1053/j.gastro.2018.08.063 254-72.e11.
6. Baradarian R, Ramdhaney S, Chapalamadugu R, Skoczylas L, Wang K, Rivilis S, et al. Early intensive resuscitation of patients with upper gastrointestinal bleeding de- creases mortality. Am J Gastroenterol. 2004;99(4):619-22. https://doi.org/10.1111/ j.1572-0241.2004.04073.x.
7. Lee JG, Turnipseed S, Romano PS, Vigil H, Azari R, Melnikoff N, et al. Endoscopy- based triAge SIgnificantly reduces hospitalization rates and costs of treating upper GI bleeding: a randomized controlled trial. Gastrointest Endosc. 1999;50(6): 755-61. https://doi.org/10.1016/s0016-5107(99)70154-9.
8. Leontiadis GI, Sharma VK, Howden CW. Proton pump inhibitor treatment for acute peptic ulcer bleeding. Cochrane Database Syst Rev. 2006(1). https://doi.org/10. 1002/14651858.CD002094.pub3 Cd002094.
9. Barkun AN, Bardou M, Kuipers EJ, Sung J, Hunt RH, Martel M, et al. International con- sensus recommendations on the management of patients with nonvariceal upper gastrointestinal bleeding. Ann Intern Med. 2010;152(2):101-13. https://doi.org/10. 7326/0003-4819-152-2-201001190-00009.
10. Laine L, Yang H, Chang SC, Datto C. Trends for incidence of hospitalization and death due to GI complications in the United States from 2001 to 2009. Am J Gastroenterol. 2012;107(8):1190-5 quiz 6 https://doi.org/10.1038/ajg.2012.168.
11. Targownik LE, Nabalamba A. Trends in management and outcomes of acute nonvariceal upper gastrointestinal bleeding: 1993-2003. Clin Gastroenterol Hepatol. 2006;4(12):1459-66. https://doi.org/10.1016/j.cgh.2006.08.018.
12. Korea Statistics. Population projections for Korea. Statistics Korea. 2019http://kosis. kr/statHtml/statHtml.do?orgId=101&tblId=DT_1BPA003&vw_cd=MT_ ZTITLE&list_id=A41_10&seqNo=&lang_mode=ko&language=kor&obj_var_id= &itm_id=&conn_path=MT_ZTITLE.
13. Saltzman JR, Tabak YP, Hyett BH, Sun X, Travis AC, Johannes RS. A simple risk score accurately predicts in-hospital mortality, length of stay, and cost in acute upper GI bleeding. Gastrointest Endosc. 2011;74(6):1215-24. https://doi.org/10.1016/j.gie. 2011.06.024.
14. Lee YJ, Min BR, Kim ES, Park KS, Cho KB, Jang BK, et al. Predictive factors of mortality within 30 days in patients with nonvariceal upper gastrointestinal bleeding. Korean J Intern Med. 2016;31(1):54-64. https://doi.org/10.3904/kjim.2016.31.1.54.
15. Blatchford O, Murray WR, Blatchford M. A risk score to predict need for treatment for upper-gastrointestinal haemorrhage. Lancet. 2000;356(9238):1318-21. https:// doi.org/10.1016/s0140-6736(00)02816-6.
16. Rockall TA, Logan RF, Devlin HB, Northfield TC. Risk assessment after acute upper gastrointestinal haemorrhage. Gut. 1996;38(3):316-21. https://doi.org/10.1136/ gut.38.3.316.
17. Kim MS, Choi J, Shin WC. AIMS65 scoring system is comparable to Glasgow- Blatchford score or Rockall score for prediction of clinical outcomes for non- variceal upper gastrointestinal bleeding. BMC Gastroenterol. 2019;19(1):136. https://doi.org/10.1186/s12876-019-1051-8.
18. Robertson M, Majumdar A, Boyapati R, Chung W, Worland T, Terbah R, et al. Risk stratification in acute upper GI bleeding: comparison of the AIMS65 score with the Glasgow-Blatchford and Rockall scoring systems. Gastrointest Endosc. 2016;83(6): 1151-60. https://doi.org/10.1016/j.gie.2015.10.021.
19. Ryu S, Oh SK, Cho SU, You Y, Park JS, Min JH, et al. Utility of the blood urea nitrogen to serum albumin ratio as a prognostic factor of mortality in Aspiration pneumonia patients. Am J Emerg Med. 2020. https://doi.org/10.1016/j.ajem.2020.02.045.
20. Ugajin M, Yamaki K, Iwamura N, Yagi T, Asano T. Blood urea nitrogen to serum albu- min ratio independently predicts mortality and severity of community-acquired pneumonia. Int J Gen Med. 2012;5:583-9. https://doi.org/10.2147/ijgm.S33628.
21. Viasus D, Garcia-Vidal C, Simonetti A, Manresa F, Dorca J, Gudiol F, et al. Prognostic value of serum albumin levels in hospitalized adults with community-acquired pneumonia. J Infect. 2013;66(5):415-23. https://doi.org/10.1016/j.jinf.2012.12.007.
22. Hajian-Tilaki K. Receiver operating characteristic curve analysis for medical diagnostic test evaluation. Caspian J Intern Med. 2013;4(2):627-35.
23. Perkins NJ, Schisterman EF. The inconsistency of “optimal” cutpoints obtained using two criteria based on the receiver operating characteristic curve. Am J Epidemiol. 2006;163(7):670-5. https://doi.org/10.1093/aje/kwj063.
24. Verghese A, Charlton B, Kassirer JP, Ramsey M, Ioannidis JP. Inadequacies of Physical Examination as a Cause of Medical errors and Adverse Events: A Collection of Vi- gnettes. Am J Med. 2015;128(12). https://doi.org/10.1016/j.amjmed.2015.06.004 1322-4.e3.
25. Reilly BM. Physical examination in the care of medical inpatients: an observational study. Lancet. 2003;362(9390):1100-5. https://doi.org/10.1016/s0140-6736(03) 14464-9.
26. Wang S, Li T, Li Y, Zhang J, Dai X. Predictive value of four different scoring systems for septic patient’s outcome: a retrospective analysis with 311 patients. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2017;29(2):133-8. https://doi.org/10.3760/cma.j.issn.2095- 4352.2017.02.008.
27. Tokgoz Akyil F, Yalcinsoy M, Hazar A, Cilli A, Celenk B, Kilic O, et al. Prognosis of hos- pitalized patients with community-acquired pneumonia. Pulmonology. 2018. https://doi.org/10.1016/j.rppnen.2017.07.010.
28. Feng DY, Zhou YQ, Zou XL, Zhou M, Yang HL, Chen XX, et al. Elevated blood urea nitrogen-to-serum albumin ratio as a factor that negatively affects the mortality of patients with Hospital-acquired pneumonia. Can J Infect Dis Med Microbiol. 2019; 2019:1547405. https://doi.org/10.1155/2019/1547405.
29. Cheng HC, Yang EH, Wu CT, Wang WL, Chen PJ, Lin MY, et al. Hypoalbuminemia is a predictor of mortality and rebleeding in peptic ulcer bleeding under proton pump inhibitor use. J Formos Med Assoc. 2018;117(4):316-25. https://doi.org/10.1016/j. jfma.2017.07.006.
30. Wu KH, Shih HA, Hung MS, Hsiao CT, Chen YC. The association between blood urea nitrogen to creatinine ratio and mortality in patients with upper gastrointestinal bleeding. Arab J Gastroenterol. 2018;19(4):143-7. https://doi.org/10.1016/j.ajg. 2018.11.002.
31. Chojkier M. Inhibition of albumin synthesis in chronic diseases: molecular mecha- nisms. J Clin Gastroenterol. 2005;39(4 Suppl 2):S143-6. https://doi.org/10.1097/ 01.mcg.0000155514.17715.39.
32. Nicholson JP, Wolmarans MR, Park GR. The role of albumin in critical illness. Br J Anaesth. 2000;85(4):599-610. https://doi.org/10.1093/bja/85.4.599.
33. Stellato T, Rhodes RS, McDougal WS. Azotemia in upper gastrointestinal hemor- rhage. A review Am J Gastroenterol. 1980;73(6):486-9.
34. Sanaka M, Kuyama Y, Nishinakagawa S, Yamanaka M. Age-specific diagnostic values of the blood urea nitrogen/creatinine ratio in distinguishing upper from lower gas- trointestinal bleeding. J Clin Gastroenterol. 1998;26(2):151-2. https://doi.org/10. 1097/00004836-199803000-00014.
35. Kyaw MH, Chan FK. Pharmacologic options in the management of upper gastrointes- tinal bleeding: focus on the elderly. Drugs Aging. 2014;31(5):349-61. https://doi. org/10.1007/s40266-014-0173-5.
36. Thomopoulos KC, Vagenas KA, Vagianos CE, Margaritis VG, Blikas AP, Katsakoulis EC, et al. Changes in aetiology and clinical outcome of Acute upper gastrointestinal bleeding during the last 15 years. Eur J Gastroenterol Hepatol. 2004;16(2):177-82. https://doi.org/10.1097/00042737-200402000-00009.
37. Theocharis GJ, Thomopoulos KC, Sakellaropoulos G, Katsakoulis E, Nikolopoulou V. Changing trends in the epidemiology and clinical outcome of acute upper gastroin- testinal bleeding in a defined Geographical area in Greece. J Clin Gastroenterol. 2008; 42(2):128-33. https://doi.org/10.1097/01.mcg.0000248004.73075.ad.
38. Pope JV, Edlow JA. Avoiding misdiagnosis in patients with neurological emergencies. Emerg Med Int. 2012;2012:949275. https://doi.org/10.1155/2012/949275.
39. Moeller JJ, Kurniawan J, Gubitz GJ, Ross JA, Bhan V. Diagnostic accuracy of neurolog- ical problems in the emergency department. Can J Neurol Sci. 2008;35(3):335-41. https://doi.org/10.1017/s0317167100008921.