A comparative analysis between Model for End-stage Liver Disease score (MELD), Modified Model for End-stage Liver Disease score (MELD-Na), and Child–Pugh score (CPS) in predicting complications among cirrhosis patients

Abstract

Aim

This study aims to compare the effectiveness of the model for end-stage liver disease (MELD), the modified model for end-stage liver disease including sodium (MELD-Na), and the Child–Pugh score (CPS) in predicting complications in patients with cirrhosis.

Methods

We conducted a prospective and observational study, analyzing 145 cirrhotic patients admitted to a tertiary care teaching hospital. The predictive accuracies of MELD, MELD-Na, and CPS for complications such as ascites, spontaneous bacterial peritonitis, hepatic encephalopathy, hepatorenal syndrome, varices, hepatocellular carcinoma, and hepatic hydrothorax were assessed using receiver operating characteristic (ROC) curves.

Results

The CPS was more accurate in predicting ascites (AUC = 0.753), while MELD and MELD-Na scores more reliably predicted hepatic encephalopathy and hepatorenal syndrome (AUCs from 0.734 to 0.834). None of the scoring systems was effective in predicting varices or hepatic hydrothorax, with AUCs approaching 0.5.

Conclusion

The CPS remains a robust predictor for ascites, and the MELD scores, particularly when sodium is included, are better for predicting hepatorenal syndrome and HCC. The limited utility of these scores for predicting varices and hepatic hydrothorax underscores the necessity for developing more sophisticated models or utilizing additional clinical parameters in prognostication.

Introduction

Cirrhosis, the end stage of chronic liver disease (CLD), continues to be a major public health issue. In 2021, cirrhosis and other chronic liver diseases caused about 1.4 million deaths and 46 million disability-adjusted life years (DALYs) globally, highlighting the enormous burden that these diseases place on health systems [1]. Although age-standardized death and DALY rates have decreased in the last decade, the absolute number of incident cases has increased, with the majority being due to the rising incidence of metabolic dysfunction-associated steatohepatitis (MASH) [1].

The etiological landscape of cirrhosis has changed, with the rise of nonalcoholic fatty liver disease (NAFLD) and alcohol-related liver disease as top causes, overtaking viral hepatitis in most parts of the world [2]. This shift initiates a need to re-evaluate prognostic tools, as we need to confirm that they remain valid in heterogeneous patient cohort.

Prognostication in cirrhosis is important for management decisions including when and if to use intervention or transplant. The Child–Pugh score (CPS), which was created in the 1970 s, has been a widespread use methodology for assessing the severity of disease and predicting outcomes. The subjective elements in CPS, such as encephalopathy and ascites grading, might undermine its accuracy. The model for end-stage liver disease (MELD) score, and its sodium-adjusted variant MELD-Na, provides more objective measures and has been implemented for liver transplant prioritization. Recent evidence indicates that MELD-Na can have higher prognostic accuracy in certain patients, such as those with decompensated cirrhosis [3].

Notwithstanding the common use of these scoring models, their relative ability to predict individual complications of cirrhosis—e.g., ascites, hepatic encephalopathy, variceal hemorrhage, spontaneous bacterial peritonitis (SBP), hepatorenal syndrome (HRS), and hepatocellular carcinoma (HCC)—is a subject of ongoing research. Knowledge about the predictive value of CPS, MELD, and MELD-Na for such complications is vital to optimize the management and outcome of the patient.

This research seeks to perform a comparative evaluation of CPS, MELD, and MELD-Na scores for the prediction of the abovementioned complications in cirrhosis patients. Through this analysis, we strive to clarify the advantages and limitations of every model, providing knowledge about their relative uses in clinical management of cirrhosis and its complications and clearing the way for future progress in prognostic modeling.

Materials and methods

It is a prospective observational study conducted during 2020–2021 including a total of 145 CLD patients from a tertiary care hospital. The study was approved by the Institutional Ethics Committee with approval number CSP/20/OCT/86/169. The study did not receive any financial support. All patients included were above the age of 18 years admitted with cirrhosis irrespective of the etiology, inclusive of the patients admitted in wards and ICU after obtaining their informed consent. Patients aged less than 18 years, acute fulminant liver failure cases, and congestive hepatomegaly cases were excluded from the study population. Their demographic, clinical, and laboratory data were collected and entered into a pre-structured proforma. Investigations on the day of admission, day 3, and day 5 of admission were collected to see for the trend in the values and correlate them with their clinical condition and the MELD score. MELD scores were calculated using the Mayo Clinic MELD score calculator.

Statistical analysis

The study’s analysis was conducted using SPSS software version 23. Continuous variables were expressed as mean and standard deviation, whereas categorical variables were represented as frequencies and percentages. Statistical differences in categorical data were derived using the chi-square test or Fisher’s exact test. The area under the curve (AUC) was determined through receiver operating curves (ROCs) for the various scores to assess their predictive accuracy for different cirrhosis-related complications. A p-value of less than 0.05 was considered significant for statistical comparisons.

Results

The study assessed the clinical characteristics and outcomes of 145 patients diagnosed with cirrhosis. This study included 125 males (86.2%) and 17 females (11.7%). Mean age of the study participants was 55.8 ± 12.8 years. The etiology of cirrhosis was varied with alcoholism (50.3%) and NASH (26.2%) as the leading causes followed by hepatitis B (14.5%) and hepatitis C (4.8%) followed by idiopathic etiology (11.7%). Symptoms at the time of presentation in frequency of occurrence were abdominal distension (78%), swelling of the feet (33.1%), abdominal pain (28.3), melena (17.2%), reduced urine output (15.2%), hematemesis (14.5%), jaundice (13.8%), altered sensorium (11.7%), and hematochezia (4.1%).

The mean MELD score of study population was 18.3 ± 8.9, 28 patients had scores < 10, 70 patients had scores 11–20, 32 patients had scores 21–30, 9 patients had scores 31–40, and 6 patients had scores above 40. Similarly, the mean MELD-Na was at 23.1 ± 12.2, 5 patients had scores < 10, 66 patients had scores 11–20, 48 patients had scores 21–30, 20 patients had scores 31–40, and 6 patients had scores above 40. Fifteen patients came under Class A of modified Child–Pugh score, 66 under Class B, and 64 under Class C. The presence of complications of cirrhosis like ascites, varices, spontaneous bacterial peritonitis (SBP), hepatic encephalopathy, hepatorenal syndrome, coagulopathy, hepatocellular carcinoma (HCC), and hepatic hydrothorax was documented across various scores of MELD, MELD-Na, and the three classes of Child–Pugh scores as shown in Tables 1, 2, and 3. The occurrence of these complications was further statistically analyzed which showed significant correlation for all the complications except for varices, HCC, and hepatic hydrothorax. A comparison in the predictability of each of these complications by the three scores, viz, MELD, MELD-Na, and Child–Pugh score, have been further derived with the help of receiver operating curves (ROC), assessing the area under the curve. The ROCs of the various complications as predicted by each of the scoring system are shown in Figs. 1, 2, 3, 4, 5, 6, 7, and 8. Table 4 shows the area under curve (AUC) and the p-value for the predictive ability of complications by the scores. The ROC curve analysis for ascites revealed the Child–Pugh score significantly outperformed the MELD and MELD-Na scores. The predictive ability for varices and hepatic hydrothorax was poor for all the three scores, whereas HCC was better predicted by MELD-Na.

Table 1. Complications in different MELD score categories

Parameter	n	MELD score categories					p-value
		≤ 10 (n = 28)	11 to 20 (n = 70)	21 to 30 (n = 32)	31 to 40 (n = 9)	> 40 (n = 6)
Ascites	96	17 (60.7)	39 (55.7)	27 (84.4)	7 (77.8)	6 (100.0)	0.016
Varices	85	18 (64.3)	45 (64.3)	14 (43.8)	4 (44.4)	4 (66.7)	0.275
SBP	26	2 (7.1)	10 (14.3)	10 (31.3)	1 (11.1)	3 (50.0)	0.011
Hepatic encephalopathy	46	1 (3.6)	20 (28.6)	15 (46.9)	5 (55.6)	5 (83.3)	< 0.001
Hepatorenal syndrome	45	1 (3.6)	14 (20.0)	16 (50.0)	8 (88.9)	6 (100.0)	< 0.001
Coagulopathy	97	3 (10.7)	48 (68.6)	31 (96.9)	9 (100.0)	6 (100.0)	< 0.001
HCC	13	3 (10.7)	2 (2.9)	7 (21.9)	1 (11.1)	0	0.033
Hepatic hydrothorax	15	3 (10.7)	9 (12.9)	2 (6.3)	1 (11.1)	0	0.780

Table 2. Complications stratified by MELD-Na score categories

Parameter	n	MELD-Na score categories					p-value
		≤ 10 (n = 5)	11 to 20 (n = 66)	21 to 30 (n = 48)	31 to 40 (n = 20)	> 40 (n = 6)
Ascites	96	2 (40.0)	34 (51.5)	37 (77.1)	17 (85.0)	6 (100.0)	0.002
Varices	85	2 (40.0)	45 (68.2)	26 (54.2)	8 (40.0)	4 (66.7)	0.158
SBP	26	0	7 (10.6)	10 (20.8)	6 (30.0)	3 (50.0)	0.041
Hepatic encephalopathy	46	0	12 (18.2)	17 (35.4)	12 (60.0)	5 (83.3)	< 0.001
Hepatorenal syndrome	45	0	8 (12.1)	16 (33.3)	15 (75.0)	6 (100.0)	< 0.001
Coagulopathy	97	0	29 (43.9)	42 (87.5)	20 (100.0)	6 (100.0)	< 0.001
HCC	13	0	4 (6.1)	5 (10.4)	4 (20.0)	0	0.301
Hepatic hydrothorax	15	2 (40.0)	5 (7.6)	7 (14.6)	1 (5.0)	0	0.111

Table 3. Complications in Child–Pugh score categories

Parameter	n	CPS category			p-value
		A (n = 15)	B (n = 66)	C (n = 64)
Ascites	96	3 (20.0)	40 (60.6)	53 (82.2)	< 0.001
Varices	85	10 (66.7)	38 (57.6)	37 (57.8)	0.800
SBP	26	0	8 (12.1)	18 (28.1)	0.010
Hepatic encephalopathy	46	2 (13.3)	11 (16.7)	33 (51.6)	< 0.001
Hepatorenal syndrome	45	3 (20.0)	13 (19.7)	29 (45.3)	0.004
Coagulopathy	97	2 (13.3)	33 (50.0)	62 (96.9)	< 0.001
HCC	13	2 (13.3)	3 (4.5)	8 (12.5)	0.233
Hepatic hydrothorax	15	1 (6.7)	9 (13.5)	5 (7.8)	0.489

[See PDF for image]

Fig. 1

ROC curve for prediction of ascites by three scores

[See PDF for image]

Fig. 2

ROC curve for prediction of spontaneous bacterial peritonitis by three scores

[See PDF for image]

Fig. 3

ROC curve for prediction of hepatic encephalopathy by three scores

[See PDF for image]

Fig. 4

ROC curve for prediction of hepatorenal syndrome by three scores

[See PDF for image]

Fig. 5

ROC curve for prediction of varices by three scores

[See PDF for image]

Fig. 6

ROC curve for prediction of coagulation disorder by three scores

[See PDF for image]

Fig. 7

ROC curve for prediction of HCC by three scores

[See PDF for image]

Fig. 8

ROC curve for prediction of hepatic hydrothorax by three scores

Table 4. The AUC and p-value values suggestive of the predictive performance of the scores for complications

Complications	CPS		MELD		MELD-Na
	AUC	p-value	AUC	p-value	AUC	p-value
Ascites	0.753	< 0.001	0.657	0.002	0.680	< 0.001
SBP	0.754	< 0.001	0.678	0.005	0.692	0.003
Hepatic encephalopathy	0.774	< 0.001	0.734	< 0.001	0.734	< 0.001
Hepatorenal syndrome	0.717	< 0.001	0.829	< 0.001	0.834	< 0.001
Varices	0.473	0.569	0.453	0.552	0.442	0.542
Coagulopathy	0.883	< 0.001	0.931	< 0.001	0.853	< 0.001
HCC	0.561	0.471	0.604	0.217	0.700	0.017
Hepatic hydrothorax	0.502	0.981	0.436	0.417	0.450	0.524

Discussion

Prognostic models and scoring systems in cirrhosis

Cirrhosis represents end-stage chronic liver disease characterized by progressive hepatic fibrosis, portal hypertension, and the development of complications that can be potentially life threatening. PSGI predicts morbidity from liver disease-related complications and mortality; prognostic assessment is critical in both guiding treatment, referring need for transplant, and predicting mortality. In recent decades, numerous scoring systems have been proposed to classify disease severity and predict outcome in cirrhotic patients [4].

The Child–Pugh score (CPS), first presented in the 1970 s, was one of the first popular scoring systems for cirrhosis. It contains both clinical (ascites and encephalopathy) and laboratory parameters (bilirubin, albumin, and prothrombin time), making it possible to stratify into three classes (A to C) [5]. Although old, CPS has disadvantages such as the subjectivity of grading of ascites and encephalopathy and its ceiling effect resulting from a low score range [6].

In an effort to bypass these limitations, the MELD score was created. MELD applies solely objective laboratory values—serum bilirubin, creatinine, and INR—to estimate 3-month mortality [7]. Its implementation in liver transplant allocation has transformed prioritization by emphasizing disease severity over waiting time [8].

Realizing that hyponatremia highly correlates with poor outcomes in cirrhosis, the MELD-Na score was subsequently introduced by adding serum sodium to the initial MELD equation [9]. MELD-Na has proved to be better at predicting short-term mortality and is currently extensively practiced in clinical settings and liver transplant allocation [10].

While these scores are used in routine practice, their relative ability to predict individual complications—e.g., hepatic encephalopathy, SBP, or HCC—is the subject of ongoing investigation [4]. This research attempts to fill that void by measuring the discriminatory performance of MELD, MELD-Na, and CPS in the prediction of significant cirrhotic complications.

Predictive value for ascites, SBP, and hepatic encephalopathy

Cirrhosis complications pose a significant burden of morbidity and mortality, necessitating predictive indices that not only provide an estimate of the overall prognosis but also can predict individual events. In this study, we examined the ability of the Child–Pugh score (CPS), the MELD, and the MELD-Na scores to predict common complications, such as the development of ascites, spontaneous bacterial peritonitis (SBP), and hepatic encephalopathy (HE).

Ascites, the most common complication of decompensated cirrhosis, is due primarily to portal hypertension and sodium retention. In the results of our study, the CPS was superior to both MELD and MELD-Na for predicting the development of ascites. This finding is similar to other studies that have shown that the CPS can more accurately reflect a patient’s volume status and synthetic function by incorporating clinical parameters such as ascites and hypoalbuminemia [11, 12].

For SBP, a high-risk infectious event in ascitic fluid, both CPS and MELD-Na had predictive ability for SBP, with CPS having a slightly better area under the curve (AUC). This may be because CPS is sensitive to hypoalbuminemia and coagulopathy, potentially due to indicators of poor immune status and also factors that increase the risk of infection [13]. Also, by adding sodium to the MELD-Na, it enhances predictive ability because hyponatremia is a known risk factor for at least the development of SBP [14].

Hepatic encephalopathy is a neural-psychiatric illness characterized by impaired ammonia excretion and compromised integrity of the blood–brain barrier. Our own analysis indicated that MELD and MELD-Na were much better predictive scores for HE than CPS. This agrees with emerging studies that found a higher MELD score, more specifically the potential decreased value in conjunction with incremental failure of kidney and liver function, particularly correlated to encephalopathy [15, 16].

These findings reinforce the point that there is not necessarily one gold standard score; instead, they all have their strengths based on the individual complication being measured; thus by individualizing prognostic tools to symptom presentation, it can enhance early intervention and risk stratification.

Prediction of HRS, variceal bleeding, coagulopathy, and HCC

Hepatorenal syndrome (HRS) is a critical renal dysfunction in cirrhosis, maintained by splanchnic vasodilation and reduced renal perfusion. In our population, MELD-Na was best at predicting HRS, closely seconded by MELD, and with CPS lagging behind [17]. This could be because MELD-Na incorporates the inclusion of creatinine and sodium in its score, which both detect renal dysfunction early on as well as circulatory failure [18]. Recent multicenter data have demonstrated that each 5-point increase in MELD-Na is correlated with doubling of the risk for HRS [19].

Variceal bleeding, the lethal complication of portal hypertension, was poorly predicted by all three scores (AUCs ~ 0.5). This is in keeping with evidence that endoscopic findings and portal pressures are superior to global liver function scores in predicting varices and variceal bleeding [20]. Adjunctive measures—transient elastography and platelet-to-spleen ratio—are emerging as the better noninvasive predictors of variceal development and bleeding [21].

Coagulopathy, as defined by elevated INR, was most accurately predicted by MELD (which already contains INR), with somewhat inferior performance by MELD-Na and CPS [22]. This is not unexpected as INR is already a component of the MELD. However, INR alone cannot portray the complex interaction of pro- and anticoagulants in cirrhosis; viscoelastic testing is increasingly advocated for bleeding risk stratification [23].

Finally, HCC risk prediction remains challenging with baseline scores. In this study, MELD-Na was moderately predictive (AUC ~ 0.70), and this was superior to MELD and CPS [24]. Nonetheless, established HCC risk models (e.g., GALAD, aMAP) incorporating demographic, tumor, and biomarker data have superior discrimination and would complement MELD-Na in surveillance practices [25].

Limitations of prognostic scores and influencing factors

While clinicians commonly use MELD, MELD-Na, and Child–Pugh prognostic scores, each has its limitations, which could challenge reliability in certain circumstances. Knowledge of these limitations is important to use the scores correctly and not misjudge the severity of the disease.

A key limitation of the Child–Pugh score (CPS) is its inclusion of subjective parameters, such as the severity of ascites and hepatic encephalopathy, whose application can be impacted by a physician’s clinical judgement [6]. CPS can also be limited in providing a dynamic assessment of liver function over time because it has a fixed ceiling (Class A, B, or C) and nominal score categories [5]. The laboratory variables in CPS like INR and serum albumin can be influenced by other factors relevant to infection and nutritional status, which will affect the validity of these lab variables included in CPS [26].

In contrast, even though the MELD score is more objective than the Child–Pugh score, it can still be influenced by certain conditions. For example, cholestatic disease may increase bilirubin, thereby increasing the MELD score even though it does not affect prognosis [27]. Also, the impact of malnutrition, infections, or muscle wasting may increase serum creatinine without the patient having renal dysfunction and thus inflate MELD [28]. Although INR measurements have become more standardized, variability in results across different labs is still a confounding factor [29].

MELD-Na scoring may also miss a true prognostic representation if a patient is experiencing dilutional hyponatremia from diuretics or fluid overload. Similarly, sodium changes from acute illness and/or medical interventions can cause temporary fluctuations in MELD-Na that do not reflect changes in liver function [30].

Further, many factors beyond the liver—like sarcopenia, systemic inflammation, portal vein thrombosis, or comorbidities (i.e., diabetes, cardiovascular disease)—can majorly affect outcomes and are not factored into the scoring systems [31, 32]. Thus, a more holistic assessment of the patient’s overall health is warranted.

Emerging models and future directions in cirrhosis prognostication

With the changing clinical context of cirrhosis, driven by etiologic shifts, aging populations, and therapeutic advancements, there is an increasing agreement that conventional scoring systems, as important as they are, no longer adequately capture the dynamic and multifaceted nature of liver disease. This has prompted the creation of novel prognostic tools that leverage larger datasets, such as imaging, biomarkers, and machine learning algorithms [33].

Machine learning (ML) and artificial intelligence (AI) are leading this revolution. Predictive models constructed with decision trees, random forests, and neural networks have provided higher accuracy than standard scores in predicting mortality, hospital readmission, and onset of complications in patients with cirrhosis [34]. A new model combining serum ammonia, albumin, sodium, and ascites status outperformed MELD-Na and CPS in predicting 1-year transplant-free survival following TIPS placement [35].

Concurrently, composite scores such as the CLIF-C ACLF score and ALBI grade have come into use to overcome shortcomings of conventional models, particularly in hepatocellular carcinoma (HCC) and acute-on-chronic liver failure (ACLF) [36, 37]. They incorporate organ failure, systemic inflammation, or tumor markers, providing more personalized risk predictions.

There is also growing interest in functional measures, such as frailty and sarcopenia measures with CT scanning, grip strength, or chair-stand tests, which have been independently predictive of mortality, hospitalization, and transplant success [38, 39]. These can be implemented in clinical practice with minimal expense or disruption.

Finally, personalized prediction, integrating objective scoring models with patient-level variables—comorbidities, nutrition, quality of life, and inflammatory status—is set to characterize cirrhosis management in the future. Multimodal data platforms, real-time analytics, and electronic health record (EHR)-driven predictive alerts will also augment clinical decision-making [40]. These technologies hold the promise to change the paradigm toward precision hepatology, enhancing outcomes through proactive, individualized care.

Conclusion

The study’s conclusions draw on the critical observation that while MELD and MELD-Na scores are better suited for predicting hepatorenal syndrome and HCC, CPS demonstrated superior predictive accuracy for ascites and SBP, significantly outperforming the other two, and this finding is critical for clinicians focusing on ascites and SBP management in cirrhotic patients. All three scores showed limitations in predicting varices and hepatic hydrothorax, which cannot be overlooked, indicating the necessity for a broader clinical assessment and possibly additional diagnostic tools for these patients. The cirrhosis has a complex prognostic landscape, with no single model providing comprehensive predictive power across all complications. This necessitates a multidimensional approach to the assessment of cirrhotic patients, potentially involving combinations of scores or the development of new models that integrate additional clinical parameters. The significant variability in score performance also underlines the importance of personalized medicine, where prognostic tools must be tailored to individual patient profiles to optimize outcomes.

Authors’ contributions

AN—collected raw data, wrote the initial draft of manuscript and prepared figures SM—analyzed the data and prepared tables, vetted the figures, wrote the draft of manuscript and prepared the final manuscript AN and SM reviewed the manuscript.

Funding

None.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The study has been carried out after obtaining approval by the Institutional Ethics Committee with approval number CSP/20/OCT/86/169, and the study participants were enrolled in the study after getting their written informed consent for participation and future publishing of the study results.

Competing interests

The authors declare no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Tham, EKJ; Tan, DJH; Danpanichkul, P; Ng, CH; Syn, N; Koh, B et al. The global burden of cirrhosis and other chronic liver diseases in 2021. Liver Int; 2025; 45, 31:CAS:528:DC%2BB2MXjsFOjsr4%3D [DOI: https://dx.doi.org/10.1111/liv.70001] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39927433][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11808647]e70001.

2. Huang, DQ; Terrault, NA; Tacke, F; Gluud, LL; Arrese, M; Bugianesi, E et al. Global epidemiology of cirrhosis — aetiology, trends and predictions. Nat Rev Gastroenterol Hepatol; 2023; 20, 6 pp. 388-398. [DOI: https://dx.doi.org/10.1038/s41575-023-00759-2] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36977794][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10043867]

3. Wu, SL; Zheng, YX; Tian, ZW; Chen, MS; Tan, HZ. Scoring systems for prediction of mortality in decompensated liver cirrhosis: a meta-analysis of test accuracy. World J Clin Cases; 2018; 6, 15 pp. 995-1006. [DOI: https://dx.doi.org/10.12998/wjcc.v6.i15.995] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30568954][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6288518]

4. Moon, AM; Singal, AG; Tapper, EB. Contemporary epidemiology of chronic liver disease and cirrhosis. Clin Gastroenterol Hepatol; 2020; 18, 12 pp. 2650-2666. [DOI: https://dx.doi.org/10.1016/j.cgh.2019.07.060] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31401364]

5. Transection of the oesophagus for bleeding oesophageal varices | British Journal of Surgery | Oxford Academic. Available from: https://academic.oup.com/bjs/article-abstract/60/8/646/6188292?redirectedFrom=fulltext&login=false. Cited 2025 May 24.

6. Peng, Y; Qi, X; Guo, X. Child-Pugh versus MELD score for the assessment of prognosis in liver cirrhosis: a systematic review and meta-analysis of observational studies. Medicine (Baltimore); 2016; 95, 8 [DOI: https://dx.doi.org/10.1097/MD.0000000000002877] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26937922]e2877.

7. Kamath, PS; Wiesner, RH; Malinchoc, M; Kremers, W; Therneau, TM; Kosberg, CL et al. A model to predict survival in patients with end–stage liver disease. Hepatology; 2001; 33, 2 464.1:STN:280:DC%2BD3M7kvVOqug%3D%3D [DOI: https://dx.doi.org/10.1053/jhep.2001.22172] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11172350]

8. Freeman, RBJ; Wiesner, RH; Harper, A; McDiarmid, SV; Lake, J; Edwards, E et al. The new liver allocation system: moving toward evidence-based transplantation policy. Liver Transpl; 2002; 8, 9 851. [DOI: https://dx.doi.org/10.1053/jlts.2002.35927] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/12200791]

9. Biggins, SW; Kim, WR; Terrault, NA; Saab, S; Balan, V; Schiano, T et al. Evidence-based incorporation of serum sodium concentration into MELD. Gastroenterology; 2006; 130, 6 pp. 1652-1660. [DOI: https://dx.doi.org/10.1053/j.gastro.2006.02.010] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16697729]

10. Kim, WR; Biggins, SW; Kremers, WK; Wiesner, RH; Kamath, PS; Benson, JT et al. Hyponatremia and mortality among patients on the liver-transplant waiting list. N Engl J Med; 2008; 359, 10 pp. 1018-1026.1:CAS:528:DC%2BD1cXhtVymurbF [DOI: https://dx.doi.org/10.1056/NEJMoa0801209] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18768945][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4374557]

11. Angeli, P; Bernardi, M; Villanueva, C; Francoz, C; Mookerjee, RP; Trebicka, J et al. EASL clinical practice guidelines for the management of patients with decompensated cirrhosis. J Hepatol; 2018; 69, 2 pp. 406-460. [DOI: https://dx.doi.org/10.1016/j.jhep.2018.03.024]

12. Berzigotti, A. Non-invasive evaluation of portal hypertension using ultrasound elastography. J Hepatol; 2017; 67, 2 pp. 399-411. [DOI: https://dx.doi.org/10.1016/j.jhep.2017.02.003] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28223101]

13. Poca, M; Alvarado-Tapias, E; Concepción, M; Pérez-Cameo, C; Cañete, N; Gich, I et al. Predictive model of mortality in patients with spontaneous bacterial peritonitis. Aliment Pharmacol Ther; 2016; 44, 6 pp. 629-637.1:CAS:528:DC%2BC28XhtlajurzM [DOI: https://dx.doi.org/10.1111/apt.13745] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27464682]

14. Kinami, S; Nakamura, N; Tomita, Y; Miyata, T; Fujita, H; Ueda, N et al. Precision surgical approach with lymph-node dissection in early gastric cancer. World J Gastroenterol; 2019; 25, 14 pp. 1640-1652. [DOI: https://dx.doi.org/10.3748/wjg.v25.i14.1640] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31011251][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6465935]

15. Tapper EB, Zhao L, Nikirk S, Baki J, Parikh ND, Lok AS, et al. Incidence and bedside predictors of the first episode of.. : Official journal of the American College of Gastroenterology | ACG. Available from: https://journals.lww.com/ajg/abstract/2020/12000/incidence_and_bedside_predictors_of_the_first.21.aspx. Cited 2025 May 24.

16. Stewart, CA; Malinchoc, M; Kim, RW; Kamath, PS. Hepatic encephalopathy as a predictor of survival in patients with end-stage liver disease. Liver Transpl; 2007; 13, 10 1366. [DOI: https://dx.doi.org/10.1002/lt.21129] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17520742]

17. Angeli, P; Ginès, P; Wong, F; Bernardi, M; Boyer, TD; Gerbes, A et al. Diagnosis and management of acute kidney injury in patients with cirrhosis: revised consensus recommendations of the International Club of Ascites. J Hepatol; 2015; 62, 4 pp. 968-974. [DOI: https://dx.doi.org/10.1016/j.jhep.2014.12.029] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25638527]

18. Wong, F; Nadim, MK; Kellum, JA; Salerno, F; Bellomo, R; Gerbes, A et al. Working party proposal for a revised classification system of renal dysfunction in patients with cirrhosis. Gut; 2011; 60, 5 pp. 702-709. [DOI: https://dx.doi.org/10.1136/gut.2010.236133] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21325171]

19. Sasso, R; Abou Yassine, A; Deeb, L. Predictors of development of hepatorenal syndrome in hospitalized cirrhotic patients with acute kidney injury. J Clin Med; 2021; 10, 23 5621.1:CAS:528:DC%2BB3MXislGrsb3J [DOI: https://dx.doi.org/10.3390/jcm10235621] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34884323][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8658275]

20. Horvatits, T; Mahmud, N; Serper, M; Seiz, O; Reher, D; Drolz, A et al. MELD-lactate predicts poor outcome in variceal bleeding in cirrhosis. Dig Dis Sci; 2023; 68, 3 pp. 1042-1050.1:CAS:528:DC%2BB38XivV2lt7vE [DOI: https://dx.doi.org/10.1007/s10620-022-07744-w] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36376577]

21. Marasco, G; Colecchia, A; Colli, A; Ravaioli, F; Casazza, G; Bacchi Reggiani, ML et al. Reply to correspondence concerning: “role of liver and spleen stiffness in predicting the recurrence of hepatocellular carcinoma after resection”. J Hepatol; 2019; 70, 4 pp. 809-810. [DOI: https://dx.doi.org/10.1016/j.jhep.2018.12.025] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30661766]

22. Kim WR, Mannalithara A, Heimbach JK, Kamath PS, Asrani SK, Biggins SW et al (2021) MELD 3.0: the model for end-stage liver disease updated for the modern era. Gastroenterology 161(6):1887–1895.e4.

23. Intagliata, NM; Davitkov, P; Allen, AM; Falck-Ytter, YT; Stine, JG. AGA technical review on coagulation in cirrhosis. Gastroenterology; 2021; 161, 5 pp. 1630-1656. [DOI: https://dx.doi.org/10.1053/j.gastro.2021.09.004] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34579937]

24. Yoo SK, Kim JH, Choe WH, Kwon SY (2025) Comparison of mortality prediction between the model for end-stage liver disease-3.0 (MELD-3.0) and the model for end-stage liver disease-lactate (MELD-La) in Korean patients with liver cirrhosis. Medicina (Mex) 61(3):494.

25. GALAD score for hepatocellular carcinoma detection in comparison with liver ultrasound and proposal of GALADUS score | Cancer Epidemiology, Biomarkers & Prevention | American Association for Cancer Research. Available from: https://aacrjournals.org/cebp/article/28/3/531/71819/GALAD-Score-for-Hepatocellular-Carcinoma-Detection. Cited 2025 May 24.

26. Montagnese, S; Bajaj, JS. Impact of hepatic encephalopathy in cirrhosis on quality-of-life issues. Drugs; 2019; 79, 1 pp. 11-16. [DOI: https://dx.doi.org/10.1007/s40265-018-1019-y] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30706419][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6416233]

27. Mukherjee, T; Nataraju, KT; Rakesh, BM; Dandothi, SD. Comparison of model for end stage liver disease-Na score and model for end stage liver disease score in predicting in-hospital mortality in patients with end stage liver disease: an observational study. Int J Adv Med; 2019; 6, 3 pp. 932-939. [DOI: https://dx.doi.org/10.18203/2349-3933.ijam20192267]

28. Muscle wasting is associated with mortality in patients with cirrhosis - clinical gastroenterology and hepatology. Available from: https://www.cghjournal.org/article/S1542-3565(11)00916-5/fulltext. Cited 2025 May 24.

29. Shikdar S, Vashisht R, Zubair M, Bhattacharya PT (2025) International normalized ratio: assessment, monitoring, and clinical implications. In: StatPearls. Treasure Island (FL): StatPearls Publishing; Available from: http://www.ncbi.nlm.nih.gov/books/NBK507707/. Cited 2025 May 24.

30. Rondon-Berrios, H; Velez, JCQ. Hyponatremia in cirrhosis. Clin Liver Dis; 2022; 26, 2 pp. 149-164. [DOI: https://dx.doi.org/10.1016/j.cld.2022.01.001] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35487602][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9060324]

31. Ferguson Toll, J; Solà, E; Perez, MA; Piano, S; Cheng, A; Subramanian, AK et al. Infections in decompensated cirrhosis: pathophysiology, management, and research agenda. Hepatol Commun; 2024; 8, 10 [DOI: https://dx.doi.org/10.1097/HC9.0000000000000539] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39365139][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11458171]e0539.

32. Ii, ERW; Satapathy, SK. Sarcopenia in the cirrhotic patient: current knowledge and future directions. J Clin Exp Hepatol; 2023; 13, 1 pp. 162-177. [DOI: https://dx.doi.org/10.1016/j.jceh.2022.06.005]

33. Balsano, C; Burra, P; Duvoux, C; Alisi, A; Piscaglia, F; Gerussi, A et al. Artificial intelligence and liver: opportunities and barriers. Dig Liver Dis; 2023; 55, 11 pp. 1455-1461. [DOI: https://dx.doi.org/10.1016/j.dld.2023.08.048] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37718227]

34. Yang B, Lu H, Ran Y (2024) Advancing non-alcoholic fatty liver disease prediction: a comprehensive machine learning approach integrating SHAP interpretability and multi-cohort validation. Front Endocrinol 15. Available from: https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2024.1450317/full. Cited 2025 May 24.

35. Development and validation of a machine learning-based model to predict survival in patients with cirrhosis after transjugular intrahepatic portosystemic shunt - eClinicalMedicine. Available from: https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(24)00580-7/fulltext. Cited 2025 May 24.

36. Toyoda H, Johnson PJ (2022) The ALBI score: from liver function in patients with HCC to a general measure of liver function. JHEP Rep 4(10). Available from: https://www.jhep-reports.eu/article/S2589-5559(22)00129-X/fulltext. Cited 2025 May 24.

37. Niewiński G, Morawiec S, Janik MK, Grąt M, Graczyńska A, Zieniewicz K et al (2020) Acute-on-chronic liver failure: the role of prognostic scores in a single-center experience. Med Sci Monit 26:0–0.

38. Elsheikh, M; Sabagh, AE; Mohamed, IB; Bhongade, M; Hassan, MM; Jalal, PK. Frailty in end-stage liver disease: understanding pathophysiology, tools for assessment, and strategies for management. World J Gastroenterol; 2023; 29, 46 pp. 6028-6048.1:CAS:528:DC%2BB2cXhsFels7k%3D [DOI: https://dx.doi.org/10.3748/wjg.v29.i46.6028] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38130738][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10731159]

39. Zhou, D; Zhang, D; Zeng, C; Zhang, L; Gao, X; Wang, X. Impact of sarcopenia on the survival of patients undergoing liver transplantation for decompensated liver cirrhosis. J Cachexia Sarcopenia Muscle; 2023; 14, 6 pp. 2602-2612. [DOI: https://dx.doi.org/10.1002/jcsm.13334] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37735907][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10751414]

40. Schattenberg, JM; Chalasani, N; Alkhouri, N. Artificial intelligence applications in hepatology. Clin Gastroenterol Hepatol; 2023; 21, 8 pp. 2015-2025. [DOI: https://dx.doi.org/10.1016/j.cgh.2023.04.007] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37088460]