https://orcid.org/0009-0006-9646-1224
Figures
Abstract
Objective
Blood urea nitrogen to albumin ratio (BAR) has served as a predictive marker for patients in the Intensive Care Unit (ICU), and has been studied in patients with sepsis, post-cardiac surgery, severe COVID-19, and acute exacerbation of chronic obstructive pulmonary disease (AECOPD). This objective indicator has demonstrated capability in prognostic prediction.However, research on the prognostic value of BAR in acute pancreatitis (AP) patients are scarce,the goal was to explore the relationship between BAR and total mortality in AP admitted to ICU.
Methods
A Retrospective analysis was performed utilizing the Medical Information Market for Intensive Care (MIMIC IV) database. Patients with AP admitted to ICU were included and grouped based on BAR. Univariate and multivariate Cox regression analysis were utilized to explore the relationship between BAR and total mortality. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve was applied to assess the predictive value of BAR. Cumulative hazard risk accumulation curve verified BAR’s predictive capability for short- and long-term mortality. Heterogeneity between different subgroups was excluded by subgroup analysis.
Results
Total 514 AP patients were divided into high-BAR (BAR ≥ 7.62) and low-BAR group (BAR < 7.62). The duration of ICU stay was significantly extended in the high BAR group. In the Cox proportional hazard model, whether adjusting for confounding factors or not, the high BAR was an independent risk factor for total mortality. AUC for BAR was 0.78 (95% C1: 0.72–0.84) at 28 days and 0.70 (95%: Cl: 0.64–0.75) at 360 days.
Citation: Li W, Li N, Zhan Y, Gu J (2025) Blood urea nitrogen to albumin ratio as predictor of mortality among acute pancreatitis patients in ICU: A retrospective cohort study. PLoS One 20(5): e0323321. https://doi.org/10.1371/journal.pone.0323321
Editor: Keiko Hosohata, Osaka University of Pharmaceutical Sciences, JAPAN
Received: September 23, 2024; Accepted: April 5, 2025; Published: May 15, 2025
Copyright: © 2025 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.The data in this article is also available from the MIMIC IV database (v2.2) at https://mimic.mit.edu.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Acute pancreatitis (AP) is a prevalent gastrointestinal disorder observed in intensive care unit (ICU), with over 300,000 annual visits to the emergency department in the United States. Its incidence is estimated to be 110–140 per 100,000 people, and the hospitalization cost per hospitalization is close to $ 7,000. The annual cost of hospitalization has exceeded $2.5 billion [1–3]. AP is marked by acinar cell injury affecting exocrine function. The inappropriate activation of trypsinogen in acinar cells triggers the activation of other digestive enzymes, kallikrein-kinin system and complement cascades. This leads to self-digestion of pancreatic tissue. Activated trypsin and inflammatory factors continue to escalate, leading to the manifestation of edema, hemorrhage, and necrosis in pancreatic tissue. The involvement of additional organs and tissues result in organ dysfunction [4]. Studies indicate that around 80% of patients develop mild to moderate symptoms, while 20% experience severe pancreatitis with a 20% mortality rate. It can be quite challenging to anticipate disease progression at an early stage and the management of complications not only increases patient suffering but also consumes more medical resources [5,6]. Thus, effective assessment of AP severity is essential in early phase and timely intervention measures are of critical importance in reducing mortality in AP.
Blood urea nitrogen (BUN) and serum albumin (ALB) are easily accessible and provide objective results in routine clinical blood tests. Prior research has established the BUN to ALB ratio (BAR) as a significant forecast index in sepsis, post-cardiac surgery, severe COVID-19,and acute exacerbation of chronic obstructive pulmonary disease (AECOPD) [7–11]. The predictive potential of BAR for disease severity can assist in clinical treatment decisions and patient management. No studies have examined the role of BAR in predicting the prognosis of adult AP patients in ICU. In order to make up for this gap, we analysed the admission and follow-up data from MIMIC-IV database to investigate the correlation between BAR and total mortality in AP patients.
2. Methods
2.1. Database
Primary dataset for our study was sourced from MIMIC-IV database (version 2.2). Including medical records of adults with ICU at Beth Israel Deaconess Medical Center in Boston, Massachusetts, from 2008 to 2019 [12]. All patient info were anonymized records and can be accessed publicly after authorization. Database authorization is granted (No: 10323541),upon completion of the online training,thus exempting ethical approval and informed consent requirements.
2.2. Data collection
We used the Structured Query Language (SQL) on the Navicat (16.2.11) software to extract variable information. In order to reduce treatment interference and ensure the homogeneity of the baseline data, BAR was used as the main research study variable to extract the first vital signs, initial laboratory test data within 24 hours of ICU admission, demographic information, clinical scores, treatment plan, complications, and prognosis of patients after ICU admission. Data from a total of 813 AP sufferers were collected from 2012 to 2019. Inclusion criteria: (1) ages between 18–90 years; (2) ICU admission for treatment; (3) survival time exceeding 24 hours since admission to ICU (The first ICU record data was selected for a sufferer with multiple admissions). The following criteria are excluded: (1) Missing BUN and ALB data at ICU admission; (2) Patients with end-stage renal disease, cirrhosis, or malignant tumors. Eventually, 514 patients participated in our research (Fig 1).
2.3. Primary outcomes
The primary endpoint was total mortality within 28, 60, 90, and 360 days following patient transfer to ICU. The secondary outcome was the length of ICU stay, ICU mortality, and total length of hospitalization.
2.4. Management data
To mitigate bias, indicators with a missing value exceeding 20% were excluded (e.g., height, PCO2, PO2, TCO2, FCa2+). For variables with data missing no more than 20% the multiple imputation method supplements it (e.g., weight, NBPm, T, WBC, RBC, Plt, HB, RDW, HCT, K + , pH, Lac, PT, APTT, INR, Tbil, ALT, AST).
2.5. Statistical analysis
Our study utilized t-tests and Mann-Whitney U tests for continuous variables, presenting non-normally distributed baseline characteristic as median (interquartile range, IQR). Chi-square and Fisher’s exact tests were employed for categorical variable comparisons, displayed as numbers (percentages). The Receiver Operating Characteristic (ROC) curve was generated to establish the BAR threshold for patients who died within 28 days of ICU admission. With the cutoff value, patients were divided into high-BAR (BAR ≥ 7.62) and low-BAR groups (BAR < 7.62), analysis of clinical characteristics and endpoint events. To clarify independent predictors of in-hospital mortality at 28, 60, 90, and 360 days, we performed both univariate and multivariate Cox regression analyses. HR with 95% confidence interval (CI) is given. The cumulative hazard risk accumulation curve was utilized to assess survival rate variations among the BAR groups. The area under the curve (AUC) of the ROC curve was utilized to evaluate and compare the predictive efficacy of BAR and other scoring systems (e.g., LAR, CAR, RAR, SOFA, and OASIS) on all-cause death in AP patients. SPSS Statistics software Version 29.0.0 (International Business Machines Corporation, Armonk, NY, USA) was utilized for investigation.
3. Results
3.1. Baseline variables
Table 1 presents baseline characteristics of AP within 28 days following ICU admission, categorized by survivor and expired. Ultimately,514 inpatients were involved in this study, including 218 (42%) female and 296 (58%) male. Median age was 58 years(46.00–70.75). The mortality rate within 28 days in the ICU was 7.0%, with the deceased group being older [73.50 (63.00–81.75) vs 56.00 (45.00–68.00), P < 0.001], lower body temperature [36.60 (36.44–36.94) vs 36.94 (36.56–37.44), P < 0.001], higher SOFA score [9.00 (5.75–12.00) vs 4.00 (2.00–7.00), P < 0.001], and higher OASIS score [41.00 (35.00–46.00) vs 31.00 (26.00–38.00), P < 0.001]. The incidence of AKI and sepsis in the expired group was 86% and 89%, For therapeutic interventions, ventilators were used in 33% of the non-survival group and CRRT in 15%. On laboratory indices, the BAR in the expired group was significantly elevated [14.08 (8.74–20.12) vs. 6.15 (3.58–10.29), P < 0.001]. Comparison of survival and mortality groups revealed RDW, [14.40 (13.50–15.58) VS 15.20 (14.35–16.55), P = 0.003], Cr, [1.00 (0.70–1.60) VS 1.50 (1.08–2.25), P < 0.001], BUN, [18.00 (11.00–29.00) VS 32.50 (19.75–43.75), P < 0.001], sodium, [138.00 (135.00–141.00) VS 140.00 (137.00–145.25), P = 0.007], chloride, [104.00 (100.00–109.00) VS 109.50 (104.00–112.25),P = 0.001], APTT, [30.10 (26.83–34.41) VS 34.30 (28.40–41.18), P = 0.021], lac, [1.80 (1.20–2.63) VS 2.05 (1.58–3.93), P = 0.025], and anion gap, [15.00 (13.00–18.00) VS 18.00 (13.00–21.25), P = 0.029]. The survival values are lower than mortality group (P < 0.05) which assumes statistical significance.
3.2 Relationship between BAR and clinical variables
The clinical characteristics and end-point events were analyzed. The area under the ROC curve (AUC) within 28 days was 0.78 (95% CI: 0.72–0.86), the sensitivity was 0.86 (0.73–1.00), and the specificity was 0.62 (0.44–0.82). The BAR cut-off value that maximizes the risk ratio is 7.62 (Fig 2).
Participants were categorized into high-BAR containing 303 patients (BAR ≥ 7.62) and low-BAR containing 211 patients (BAR < 7.62) based on cut-off value. Presentation in Table 2, the two groups exhibit a statistical significance in age, NBPm, T, WBC, PLT, RDW, Crea, Na, K, Cl, pH, LAC, AG, PT, SOFA, OASIS, and SAPSII (P < 0.05). No notable differences were found in remaining covariates (P > 0.05).
3.3. Univariate and Multivariate Cox regression analysis in different models
Univariate Cox analysis was performed on the risk factors associated with mortality within 28 days of ICU admission, aiming to elucidate the potential relationship between BAR and mortality. The results indicated that BAR was identified as a significant independent risk factor, and BAR ≥ 7.62 group exhibited an elevated mortality rate (P < 0.05) (S1 Table). Prognostic indicators at different time periods (28, 60, 90, and 360 days) were analysed by the Cox proportional hazards model. The results indicated that in crude model 0, elevated BAR (>7.62) showed positive correlation with a higher risk of mortality over different time periods [28 days (HR =
9.5 95% CI: 3.69–24.43, P = 0), 60 days (HR = 5.67, 95% CI: 2.99–10.76, P = 0), 90 days (HR = 5.59, 95% CI: 3.15–9.94, P = 0), 360 days (HR = 3.7, 95% CI: 2.37–5.78, P = 0)]. Elimination of confounding factors across different models, in the multivariable model 1 (age, gender,), patients in BAR ≥ 7.62 group had a significantly higher risk of death [28days (HR = 6.83, 95% CI: 2.62–17.82, P = 0), 60days (HR = 4.9, 95% CI: 2.08–7.68, P = 0), 90days (HR = 3.97, 95% CI: 2.21–7.13, P = 0), 360days (HR = 2.7, 95% CI: 1.7–4.26, P = 0)]. Multivariate models 2–6 included additional possible confounders (temperature, complications, clinical score, auxiliary examination, and treatment intervention). Patients in the BAR ≥ 7.62 group were all at higher risk of death (Table 3).
3.4. ROC and Cumulative hazard risk accumulation curves in different intervals
ROC for total mortality in AP patients were analysed and compared with BAR, RAR, CAR, LAR, SOFA and OASIS scores (Fig 3). The comprehensive analysis indicates that value for BAR surpasses that of RAR, CAR, LAR, SOFA, and OASIS scores. Death within ICU 28 days, BAR (AUC = 0.78 95% CI: 0.72–0.84) was better than LAR (AUC = 0.70 95% CI: 0.63–0.78), CAR (AUC = 0.73 95% CI: 0.65–0.82), RAR(AUC = 0.76 95% CI: 0.66–0.84), SOFA (AUC = 0.74 95% CI: 0.64–0.81), and OASIS(AUC = 0.78 95%CI: 0.68–0.83). Death within ICU 360 days, BAR (AUC = 0. 70 95% CI: 0. 65–0. 76) was also higher than LAR (AUC = 0. 61 95% CI: 0.54–0.66), CAR (AUC = 0. 67 95% CI: 0.62–0.73), RAR (AUC = 0.66 95% CI: 0.61–0.72), SOFA (AUC = 0.67 95% CI: 0.61–0.73), OASIS (AUC = 0.67 95% CI: 0.61–0.73). The AUC of BAR was lower than OASIS score for death within ICU 60 days [BAR (AUC = 0.75 95% CI: 0.70–0.81), OASIS (AUC = 0.76 95% CI: 0.70–0.83), Therefore, BAR has a significant advantage in predicting total mortality after ICU admission (S2 Table).The cumulative risk of deaths occurring in the BAR ≥ 7.62 group was significantly higher than BAR < 7.62 group (P < 0.05) at various time intervals (Fig 4).
≥ 7.62, Cumulative hazard risk accumulation curves are displayed for the following intervals: (A) 28 days, (B) 60 days, (C) 90 days, (D) 360 days.
3.5. Subgroup analyses of BAR at different periods
Subgroup analyses were performed in order to exclude the heterogeneity caused by confounding factors such as age, gender, and comorbidity (Fig 5). In predicting 360-day all-cause mortality post ICU admission,BAR was influenced by diabetes(P for interaction = 0.024).In contrast, no interactions were identified between the BAR and all subgroups across 28, 60, and 90-days intervals.(P for interaction > 0.05) (S3 Table).The results indicate BAR serves as a stable independent predictor.
4. Discussion
In recent years, the predictive effect of serum markers on the prognosis of ICU patients is a major research hotspot, such as red blood cell distribution width to albumin ratio(RAR), lactate/albumin ratio (LAR), creatinine to albumin ratio (CAR), glucose to lymphocyte ratio (GLR) on the prognosis of AP, and the predictive effects of BAR on the prognosis of patients with sepsis, chronic obstructive pulmonary disease (COPD), and post-cardiac surgery [13–16]. However, the role of BAR in the relationship between death and survival of AP patients after ICU admission is still unknown. Our study provides conclusive evidence that BAR emerged as an independent predictor for overall mortality within 28, 60, 90 and 360 days after ICU admission. According to the cumulative hazard risk accumulation curves, patients with AP and BAR ≥ 7.62 had obviously superior mortality than low BAR at 28, 60, 90, and 360 days. Meanwhile, according to the AUC, the predictive effect of BAR on patients’ poor prognosis surpasses that of RAR, LAR, and CAR in existing studies, and also better than traditional SOFA and OASIS scores. By subgroup analysis, we excluded the impact of heterogeneity. BAR demonstrated a superior capacity to predict total death risk with no significant interaction with subgroups (P > 0.05). Previous studies indicate that BAR cutoff values serve as a predictor for various patient populations such as post-cardiac surgery, 6.41, Sepsis, 8.0, COVID-19,4.94, AECOPD (ALB unit g/L),0.249, and patients admitted to the surgical ICU,9.69 [7–11]. The values differ significantly, variations in cutoff values stem from disease mechanisms and inclusion criteria. However, BAR is an independent predictive metric that positively correlates with increased mortality. The validation of our research findings is also substantiated. The BAR cutoff value cannot be a universal indicator similar to assessing anemia based on hemoglobin levels, it can be customized to create individualized standards depending on specific disease conditions.
Studies have shown that ALB has a variety of biological effects, such as maintaining osmotic pressure, participating in drug binding and transport, scavenging free radicals, antioxidant, maintaining microvascular integrity, and thus reducing intravascular fluid leakage [17–20]. Hypoalbuminemia is a confirmed risk factor for poor prognosis in seriously ill patients [21]. A waterfall-like inflammatory response is present in the early stage of AP patients. The inflammatory response accelerates the process of protein hydrolysis, thereby reducing the levels of ALB, the plasma colloidal osmotic pressure correlates positively with ALB level. At the same time, a large amount of fluid leaks out of the blood vessels into the tissue interstitial space, resulting in insufficient effective volume of blood and reduced cardiac output, causing multiple organ hypoperfusion and activation of the renin-angiotensin-aldosterone system (RAAS) and the sympathetic nervous system [22]. Angiotensin and epinephrine stimulate renal vasoconstriction, reducing glomerular filtration rate (GFR) and renal blood flow, thereby enhancing urea reabsorption [23].
BUN is a nitrogen-containing compound with a small molecular weight, generated through the metabolic hydrolysis of proteins and excreted by the kidneys [24], which reflects absorption of protein, nutritional status, endogenous proteolytic metabolism, liquid balance, urea synthesis, and renal function of serious patients [25–27].
ALB and BUN are independent risk factors for many serious diseases. However, ALB levels are affected by age, calorie intake, inflammatory response, and chronic diseases, whereas BUN is influenced by protein intake, blood capacity, renal tubular reabsorption and urea excretion. The predictive value of a single variable for prognosis is significantly limited [28]. Existing studies show that BUN is less effective than BUN to ALB radio for predicting sepsis prognosis [7], while lactate to ALB ratio outperforms ALB in forecasting ICU patient outcomes [29]. Therefore, our study used BUN/ALB combined indicators to enhance prognostic predictive value.
Previous research indicate that concentration of BUN and ALB can be affected by renal function, sepsis, age, etc, which are more prominent in elderly patients [30–34]. We performed univariate COX regression analyses and found that BAR was an independent predictor of total mortality for AP patients. In the multivariate Cox regression models, we adjusted for confounders such as age, gender, temperature, complications, clinical scores, auxiliary examination, treatment intervention, and BAR remained strongly associated with poor prognosis. In subgroup analyses, the BAR ≥ 7.62 group still demonstrated a consistent effect with other subgroups for patients aged > 65 years with comorbid CKD. Current research evaluates the prognostic predictive ability of BAR, SOFA, and OASIS via AUC, revealing no consistent advantage in one direction [35–37]. Similarly, our research observed discontinuities in the one-sided advantages of BAR and OASIS. Perhaps the statistical errors from a retrospective study utilizing the MIMIC database, characterized by single-center and limited sample size, may explain why the AUC of BAR at 60 days does not exhibit the continuous linear relationship seen with OASIS at 28, 90, and 360 days. The statement above does not negate that BAR as an independent predictor. In addition, there was an interaction between AP patients with diabetes and BAR at 360 days after admission to ICU (P for interaction = 0.024), The final outcome may be biased due to sample volume limitations, particularly after grouping.
Based on research utilizing real-world datasets, the BAR demonstrates a significant correlation with mortality among ICU patients diagnosed with AP. However, this study is not without its limitations. Firstly, the current investigation constitutes a single-center retrospective cohort study, future research should use multi-center large sample data to validate the predictive reliability of BAR for AP patients or conduct multi-center prospective studies. Second, this study only analysed the data of the first admission to ICU and did not dynamically assess the impact of BAR values at different time points on clinical outcomes of AP patients. Third, the MIMIC-IV database provides data on inpatients from 2008 to 2019, with a long time span. Considering the updates and changes in medical monitoring technology and disease treatment methods, the organ damage observed in patients with AP varies significantly among individuals. The implementation of individualized treatment plans, coupled with inconsistent baseline characteristics, will inevitably influence the outcomes. Fourth, the database includes both ICD-9 and ICD-10 codes. Unlike ICD-10, ICD-9 lacks detailed etiology diagnoses, restricting subgroup analyses for AP based on etiology.Finally, this study is a retrospective analysis of public databases, and patients with a high number of missing values of some indicators have not been included for analysis, so selection bias and confounding bias is inevitable.
5. Conclusions
BAR is an objective and independent predictor of both short- and long-term total mortality in AP patients. A prompt, efficient, and uncomplicated assessment of the severity and prognosis, which facilitates ICU doctors to develop treatment plans for poor patient outcomes.
Supporting information
S1 Table. Univariate COX analysis for expired within 28 days after AP admission.
https://doi.org/10.1371/journal.pone.0323321.s001
(DOCX)
S2 Table. ROC for total mortality in AP were analysed and compared with BAR, RAR, CAR, LAR, SOFA and OASIS scores in 28, 60, 90, and 360days.
https://doi.org/10.1371/journal.pone.0323321.s002
(DOCX)
S3 Table. Subgroup analysis of AP in ICU at different time periods(28, 60, 90, and 360days).
https://doi.org/10.1371/journal.pone.0323321.s003
(DOCX)
Acknowledgments
We are grateful to acknowledge database administrator Xinrui Li for his support and guidance to this project.
References
- 1. Forsmark CE, Vege SS, Wilcox CM. Acute Pancreatitis. N Engl J Med. 2016;375(20):1972–81.
- 2. Sellers Z, MacIsaac D, Yu H, Dehghan M, Zhang K, Bensen R. Nationwide trends in acute and chronic pancreatitis among privately insured children and non-elderly adults in the United States, 2007-2014. Gastroenterology. 2018;155(2):469-78.e1.
- 3. Garg SK, Sarvepalli S, Campbell JP, Obaitan I, Singh D, Bazerbachi F, et al. Incidence, Admission Rates, and Predictors, and Economic Burden of Adult Emergency Visits for Acute Pancreatitis: Data From the National Emergency Department Sample, 2006 to 2012. J Clin Gastroenterol. 2019;53(3):220–5. pmid:29629907
- 4. Lankisch PG, Apte M, Banks PA. Acute pancreatitis. Lancet. 2015;386(9988):85–96. pmid:25616312
- 5. Mederos MA, Reber HA, Girgis MD. Acute pancreatitis: a review. JAMA. 2021;325(4):382–90.
- 6. Krishna SG, Kamboj AK, Hart PA, Hinton A, Conwell DL. The Changing Epidemiology of Acute Pancreatitis Hospitalizations: A Decade of Trends and the Impact of Chronic Pancreatitis. Pancreas. 2017;46(4):482–8. pmid:28196021
- 7. Wang Y, Gao S, Hong L, Hou T, Liu H, Li M. Prognostic impact of blood urea nitrogen to albumin ratio on patients with sepsis: a retrospective cohort study. Sci Rep. 2023;13(1):10013.
- 8. Alirezaei T, Hooshmand S, Irilouzadian R, Hajimoradi B, Montazeri S, Shayegh A. The role of blood urea nitrogen to serum albumin ratio in the prediction of severity and 30-day mortality in patients with COVID-19. Health Sci Rep. 2022;5(3):e606.
- 9. Zeng Z, Ke X, Gong S, Huang X, Liu Q, Huang X, et al. Blood urea nitrogen to serum albumin ratio: a good predictor of in-hospital and 90-day all-cause mortality in patients with acute exacerbations of chronic obstructive pulmonary disease. BMC Pulm Med. 2022;22(1):476. pmid:36522751
- 10. Ye L, Shi H, Wang X, Duan Q, Ge P, Shao Y. Elevated Blood Urea Nitrogen to Serum Albumin Ratio Is an Adverse Prognostic Predictor for Patients Undergoing Cardiac Surgery. Front Cardiovasc Med. 2022;9:888736. pmid:35600476
- 11. Zhang J, Zhong L, Min J, Wei Y, Ding L. Relationship between blood urea nitrogen to serum albumin ratio and short-term mortality among patients from the surgical intensive care unit: a population-based real-world study. BMC Anesthesiol. 2023;23(1):416. pmid:38114922
- 12. Johnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10(1):1. pmid:36596836
- 13. Pan L, Pan S, Lei B, Hong G, Chen K. Development of a novel nomogram incorporating red blood cell distribution width-albumin ratio for the prediction of 30-day mortality in acute pancreatitis patients. Emerg Med Int. 2022;2022:1573931.
- 14. Liu Q, Zheng H-L, Wu M-M, Wang Q-Z, Yan S-J, Wang M, et al. Association between lactate-to-albumin ratio and 28-days all-cause mortality in patients with acute pancreatitis: A retrospective analysis of the MIMIC-IV database. Front Immunol. 2022;13:1076121. pmid:36591285
- 15. Wang J, Li H, Luo H, Shi R, Chen S, Hu J, et al. Association between serum creatinine to albumin ratio and short- and long-term all-cause mortality in patients with acute pancreatitis admitted to the intensive care unit: a retrospective analysis based on the MIMIC-IV database. Front Immunol. 2024;15:1373371. pmid:38686375
- 16. Chen Y, Tang S, Wang Y. Prognostic Value of Glucose-to-Lymphocyte Ratio in Critically Ill Patients with Acute Pancreatitis. Int J Gen Med. 2021;14:5449–60. pmid:34526812
- 17. Nicholson JP, Wolmarans MR, Park GR. The role of albumin in critical illness. Br J Anaesth. 2000;85(4):599–610.
- 18. Holt ME, Ryall ME, Campbell AK. Albumin inhibits human polymorphonuclear leucocyte luminol-dependent chemiluminescence: evidence for oxygen radical scavenging. Br J Exp Pathol. 1984;65(2):231–41. pmid:6712882
- 19. Lapenna D, Ciofani G, Ucchino S, Pierdomenico SD, Cuccurullo C, Giamberardino MA, et al. Serum albumin and biomolecular oxidative damage of human atherosclerotic plaques. Clin Biochem. 2010;43(18):1458–60. pmid:20807525
- 20. Demling RH. Effect of plasma and interstitial protein content on tissue edema formation. Curr Stud Hematol Blood Transfus. 1986;53:36–52.
- 21. Vincent J-L, Dubois M-J, Navickis RJ, Wilkes MM. Hypoalbuminemia in acute illness: is there a rationale for intervention? A meta-analysis of cohort studies and controlled trials. Ann Surg. 2003;237(3):319–34. pmid:12616115
- 22. Kazory A. Emergence of blood urea nitrogen as a biomarker of neurohormonal activation in heart failure. Am J Cardiol. 2010;106(5):694–700. pmid:20723648
- 23. Cauthen CA, Lipinski MJ, Abbate A, Appleton D, Nusca A, Varma A, et al. Relation of blood urea nitrogen to long-term mortality in patients with heart failure. Am J Cardiol. 2008;101(11):1643–7. pmid:18489944
- 24. Nguyen KN, Chuang T-I, Wong L-T, Chan M-C, Chao W-C. Association between early blood urea nitrogen-to-albumin ratio and one-year post-hospital mortality in critically ill surgical patients: a propensity score-matched study. BMC Anesthesiol. 2023;23(1):247. pmid:37479965
- 25. Liu Y, Hu H, Li Z, Han Y, Chen F, Zhang M, et al. Association Between Pre-operative BUN and Post-operative 30-Day Mortality in Patients Undergoing Craniotomy for Tumors: Data From the ACS NSQIP Database. Front Neurol. 2022;13:926320. pmid:35928140
- 26. Faisst M, Wellner UF, Utzolino S, Hopt UT, Keck T. Elevated blood urea nitrogen is an independent risk factor of prolonged intensive care unit stay due to acute necrotizing pancreatitis. J Crit Care. 2010;25(1):105–11. pmid:19427764
- 27. Arnan M, Hsieh T, Yeboah J, Bertoni A, Burke G, Bahrainwala Z. Postoperative blood urea nitrogen is associated with stroke in cardiac surgical patients. Ann Thorac Surg. 2015;99(4):1314–20.
- 28. Yang Z, Dong L, Zhang Y, Yang C, Gou S, Li Y, et al. Prediction of Severe Acute Pancreatitis Using a Decision Tree Model Based on the Revised Atlanta Classification of Acute Pancreatitis. PLoS One. 2015;10(11):e0143486. pmid:26580397
- 29. Gharipour A, Razavi R, Gharipour M, Mukasa D. Lactate/albumin ratio: An early prognostic marker in critically ill patients. Am J Emerg Med. 2020;38(10):2088–95. pmid:33152585
- 30. Musch W, Verfaillie L, Decaux G. Age-related increase in plasma urea level and decrease in fractional urea excretion: clinical application in the syndrome of inappropriate secretion of antidiuretic hormone. Clin J Am Soc Nephrol. 2006;1(5):909–14. pmid:17699306
- 31. Corti MC, Guralnik JM, Salive ME, Sorkin JD. Serum albumin level and physical disability as predictors of mortality in older persons. JAMA. 1994;272(13):1036–42. pmid:8089886
- 32. Li M, Han S, Liang F, Hu C, Zhang B, Hou Q, et al. Machine Learning for Predicting Risk and Prognosis of Acute Kidney Disease in Critically Ill Elderly Patients During Hospitalization: Internet-Based and Interpretable Model Study. J Med Internet Res. 2024;26:e51354. pmid:38691403
- 33. Li X, Zheng R, Zhang T, Zeng Z, Li H, Liu J. Association between blood urea nitrogen and 30-day mortality in patients with sepsis: a retrospective analysis. Ann Palliat Med. 2021;10(11):11653–63. pmid:34872290
- 34. Cui SH, Liang CY, Hao YF. Analysis of risk factors affecting the prognosis of patients with sepsis and construction of nomogram prediction model. Eur Rev Med Pharmacol Sci. 2024;28(6):2409–18.
- 35. Wang H, Wu S, Pan D, Ning Y, Li Y, Feng C, et al. Comparison of different intensive care scoring systems and Glasgow Aneurysm score for aortic aneurysm in predicting 28-day mortality: a retrospective cohort study from MIMIC-IV database. BMC Cardiovasc Disord. 2024;24(1):513. pmid:39333879
- 36. Zhu Y, Zhang R, Ye X, Liu H, Wei J. SAPS III is superior to SOFA for predicting 28-day mortality in sepsis patients based on Sepsis 3.0 criteria. Int J Infect Dis. 2022;114:135–41.
- 37. Liu S, Zhao K, Shao C, Xu L, Cui X, Wang Y. Association between alkaline phosphatase to albumin ratio and mortality among patients with sepsis. Sci Rep. 2024;14(1):3170. pmid:38326383