Introduction

Acute pancreatitis (AP) is one of the most common hospitalization diagnoses in the Gastroenterology Departments, and its incidence has been reported increasing. The course of acute pancreatitis is mostly self-limiting, and the patients could recover within several days. However, about 20% of the patients will eventually develop into severe acute pancreatitis (SAP), accompanied by systemic inflammatory response syndrome and persistent multiple organ failure, and the mortality rate is as high as 30% to 50%1,2,3.

Therefore, early identification and classification of acute pancreatitis severity is of great significance to take measures to prevent severe acute pancreatitis, which will improve the prognosis of the patients. There are many studied biomarkers, clinical scores and radiological signs for prediction severity and mortality of acute pancreatitis. Yet, none of them is the gold standard1,2,3,4,5.

The purpose of the current work was to assess different available scores to early identification patients with SAP. These scores included modified Ranson criteria (mRC) score at admission, mRC score at 48 h after admission, Acute Physiology and Chronic Health Examination-II score that consists of the following parameters (APACHE-II), Bedside Index of Severity in Acute Pancreatitis (BISAP) Score, Harmless Acute Pancreatitis (HAP) score, Computed tomography severity index (CTSI), neutrophil/lymphocytes ratio (NLR) and platelets/lymphocytes ratio (PLR).

Methods

Study setting and design

Between October 2018 and October 2022, a prospective hospital-based study was conducted at Tropical Medicine and Gastroenterology Department to enroll all patients with suspected acute pancreatitis.

Patients’ characteristics and criteria of selection

During the study period, all patients with AP were enrolled. AP was diagnosed by fulfilling two out of three criteria: (1) abdominal pain consistent with pancreatitis, (2) a serum amylase or lipase three or more times the upper limit of normal, and (3) findings consistent with pancreatitis on cross-sectional abdominal imaging; abdominal computed tomography (CT) or magnetic resonance image (MRI)6,7. Imaging was completed within 6 h of admission, ensuring consistency between inclusion criteria and data collection.

Any patient with one or more of the following criteria was excluded; age less than 18 years old, previous history of AP (prior AP may alter pancreatic morphology and inflammatory response, potentially confounding severity assessment and score performance), confirmed diagnosis of corona virus infectious disease-19 based on polymerase chain reaction, and/or pancreatic cancer.

Ethical consideration

The study was conducted according to the principles of the Declaration of Helsinki and was done after obtaining approval by the Medical Ethics Committee of the Faculty of Medicine at Assiut University (IRB: 12009/2018). We confirmed that all methods were performed in accordance with the relevant guidelines and regulations.

Also, a written informed consent was obtained from all participants before being enrolled in the study. The study was registered on www.Clinicaltrials.gov Identifier: NCT 03601325.

Data collection

Enrolled patients were clinically evaluated with recording the following data; age, sex, comorbidities, laboratory blood tests as complete blood count (NLR and PLR were recorded), liver function tests, blood urea nitrogen and serum creatinine. All patients underwent abdominal ultrasound and multi-slice computed tomography in the first 6 h of admission.

Grading of severity of acute pancreatitis

Based on revised Atlanta criteria (RAC); the severity of acute pancreatitis is primarily determined by the presence and duration of organ failure. Organ failure is assessed using the modified Marshall scoring system and may involve the respiratory, cardiovascular, or renal systems8;

  • Mild acute pancreatitis is characterized by the absence of organ failure and local or systemic complications.

  • Moderately severe acute pancreatitis is defined by the presence of transient organ failure that resolves within 48 h and/or local or systemic complications without persistent organ failure.

  • Severe acute pancreatitis (SAP) is diagnosed when persistent organ failure lasting longer than 48 h affects one or more organ systems.

The duration of organ failure is a critical determinant of prognosis, as persistent organ failure is associated with increased morbidity and mortality. This classification provides a standardized framework for risk stratification and guides clinical decision-making in both research and practice.

In the current study we divided the patients into either severe group (included patients with severe AP based on RAC) or non-severe group (included patients with mild and moderately severe AP based on RAC).

Severity scoring systems

  • Modified Ranson criteria for acute pancreatitis include five parameters assessed at admission: age greater than 70 years, white blood cell count exceeding 18,000 cells/mm³, blood glucose above 220 mg/dL, serum lactate dehydrogenase (LDH) greater than 400 IU/L, and serum aspartate aminotransferase (AST) greater than 250 IU/L. After 48 h, five additional parameters are evaluated: hematocrit fall greater than 10%, blood urea nitrogen (BUN) rise of 5 mg/dL or more, serum calcium below 8 mg/dL, arterial partial pressure of oxygen (PaO₂) less than 60 mmHg, and base deficit exceeding 4 mEq/L9.

  • The Acute Physiology and Chronic Health Evaluation II (APACHE II) score was calculated within the first 24 h of admission using standard physiological, laboratory, age, and chronic health variables10,

  • The Bedside Index for Severity in Acute Pancreatitis (BISAP) score was assessed at admission based on blood urea nitrogen level, impaired mental status, systemic inflammatory response syndrome, age > 60 years, and presence of pleural effusion11,

  • The Harmless Acute Pancreatitis (HAP) score was calculated at presentation using the absence of rebound tenderness or guarding, normal hematocrit, and normal serum creatinine12.

  • Inflammatory indices were derived from admission complete blood counts, including the neutrophil-to-lymphocyte ratio (NLR), calculated by dividing the absolute neutrophil count by the absolute lymphocyte count, and the platelet-to-lymphocyte ratio (PLR), calculated by dividing the platelet count by the absolute lymphocyte count.

  • Radiological severity was assessed using the Computed Tomography Severity Index (CTSI) based on contrast-enhanced CT findings, which incorporates the degree of pancreatic inflammation and the extent of pancreatic necrosis. CT imaging was performed according to institutional protocol, and CTSI was calculated by experienced radiologists blinded to clinical outcomes. Severity of AP was assigned according to the total points: 0–3, mild; 4–6, moderately severe; 7–10, severe13.

Statistical analysis

The collected data were coded, processed and analyzed using the SPSS (Statistical Package for Social Sciences) version 20 for Windows® (IBM SPSS Inc, Chicago, IL, USA). Continuous variables are presented as mean and standard deviation (SD), and compared by Student’s t test. While nominal data are expressed as frequency (percentage) and were compared by Chi2 test.

Multivariate logistic regression analysis was performed to identify independent predictors of SAP. The area under the curve (AUC) of receiver operator characteristics (ROC) curve was used to evaluate the discriminative ability of different predictors for predicting SAP. A P value < 0.05 was considered significant.

Results

Table 1 reveals that both groups of patients based on severity of AP had insignificant differences as regard different characteristics with except of significantly higher mean age among patients with severe AP (59.11 ± 13.22 vs. 44.50 ± 9.45 (years); p < 0.001). Main etiologies of AP were gall stones (60% vs. 705) and post-endoscopic retrograde cholangiopancreatography (16.5% vs. 10%). Three patients with non-SAP and one patient with SAP had hypertriglyceridemia induced AP. Etiology wasn’t known in up to 21% of non-SAP group and 19% of SAP group.

Table 1 Characteristics of studied patients based on severity of AP.
Full size table

Table 2 shows a significantly higher NLR (6.09 ± 2.45 vs. 2.34 ± 0.56; p < 0.001), PLR (212.12 ± 45.09 vs. 90.45 ± 18.90; p < 0.001) and red cell distribution width (16.11 ± 2.11 vs. 11.45 ± 1.65; p < 0.001) among patients with severe AP. Also, patients with SAP have significantly lower serum albumin (33.09 ± 5.09 vs. 37.22 ± 5.11 (mg/dl); p = 0.01).

Table 2 Laboratory data among the studied patients based on severity of AP.
Full size table

Table 3 reveals patients with SAP have significantly longer stay within the hospital (18.99 ± 4.44 vs. 9.11 ± 2.11 (days); p < 0.001). Also, SAP group had significantly higher complications (3% vs. 0; p = 0.04) and mortality (12% vs. 1.5%; p < 0.001).

Table 3 Scoring of severity and outcome in studied patients based on severity of AP.
Full size table

Predictors of SAP were NLR, PLR, mRC (after 48 h), HAP score, BISAP score and severe CTSI. With ROC curve, we find that CTSI and BISAP score has the best diagnostic accuracy (100% and 96.4%, respectively) followed by HAP score that had 83% overall accuracy (Tables 4 and 5; Fig. 1). acc.

Table 5 summarizes the diagnostic performance of the evaluated predictors for severe acute pancreatitis. The mRC after 48 h demonstrated limited discriminatory ability, with an accuracy of 62.2% and an AUC of 0.63 at a cutoff value of ≥ 3, and was not statistically significant. Inflammatory indices showed moderate predictive performance; the NLR achieved an accuracy of 79% with an AUC of 0.78 at a cutoff of ≥ 2.50, while the PLR demonstrated an accuracy of 65.9% and an AUC of 0.67 at a cutoff of > 180.90. Both markers were statistically significant.

Among clinical scoring systems, the BISAP score showed excellent predictive performance, with an accuracy of 96.4% and an AUC of 0.94 at a cutoff of ≥ 1. The HAP score also demonstrated good discrimination, achieving an accuracy of 83% and an AUC of 0.84 at the same cutoff. The highest diagnostic performance was observed for the CT Severity Index, which demonstrated perfect sensitivity, specificity, overall accuracy, and an AUC of 1.00 at a cutoff value of ≥ 7.

Table 4 Predictors of severity among patients with acute pancreatitis.
Full size table
Table 5 Accuracy of different predictors of severe acute pancreatitis.
Full size table
Fig. 1
Fig. 1
Full size image

Accuracy of different predictors of severe acute pancreatitis. NLR neutrophil/ lymphocytes ratio, PLR platelets/ lymphocytes ratio, mRC modified ranson’s criteria, HAP score harmless acute pancreatitis score, BISAP bedside index of severity in acute pancreatitis score, CTSI computed tomography severity index.

Discussion

This study was designed to evaluate eight scores in prediction of SAP. The main findings of this study included; (1) Based on RAC; 200 (66.7%) and 100 (33.3%) patients had non-severe and severe AP and a total of 15 (5%) patients were deteriorated and died and (2) another finding of our study was that predictors of SAP were NLR, PLR, mRC (after 48 h), HAP score and BISAP score. With ROC curve, we find that CTSI and BISAP scores have the best diagnostic accuracy for prediction of SAP.

Approximately one-third of our cohort presented with severe acute pancreatitis (SAP). Consistent with this observation, prior evidence indicates that nearly two-thirds of individuals with acute pancreatitis experience a mild or non-severe course, with clinical symptoms typically resolving within a few days14.

El-Masry et al. stated that SAP occurred in 46.3% of patients15. The observed variation in the reported frequency of SAP may be explained by differences in sample size, study populations, and potential selection bias. However, the principal factor underlying this variability is likely the diagnostic criteria applied, as most studies utilized RAC, whereas others relied on Ranson’s criteria.

Our study revealed that higher SAP incidence related to older age and higher comorbidity burden in the severe group. Older age and a higher comorbidity burden were associated with an increased incidence of severe acute pancreatitis in the present study. Advanced age is known to reduce physiological reserve and increase vulnerability to systemic inflammation and organ failure. Comorbid conditions may further exacerbate disease severity and complicate clinical course. These findings are consistent with previous studies identifying age and comorbidities as key risk factors for severe acute pancreatitis. Early risk stratification in these high-risk patients is therefore essential16,17,18.

Our study also, reported the most frequent etiology of AP was secondary to gall stones followed by post-endoscopic retrograde cholangiopancreatography. This was consistent with many previous literatures19. Regarding the absence of alcohol-related acute pancreatitis in our cohort, we agree that this is an important point that warrants clarification. The study was conducted in a predominantly Muslim population, where alcohol consumption is culturally and religiously prohibited and therefore markedly uncommon20,21. Consequently, alcohol-related acute pancreatitis is rarely encountered in our clinical practice, which explains its absence in the present study.

In the current study, we found that patients with SAP had significantly lower serum albumin. Hypoalbuminemia observed in patients with severe acute pancreatitis reflects the intensity of the systemic inflammatory response and is an important marker of disease severity. During acute pancreatitis, increased capillary permeability driven by inflammatory mediators leads to transcapillary albumin leakage, while reduced hepatic synthesis and increased catabolism further contribute to low serum albumin levels22.

Hypoalbuminemia has been associated with intravascular volume depletion, impaired tissue perfusion, and a higher risk of persistent organ failure. Previous studies have demonstrated that lower albumin levels are independently associated with increased complications, longer hospital stay, and higher mortality in acute pancreatitis, underscoring its value as a readily available prognostic indicator of adverse outcomes23,24.

With logistic regression analysis we concluded that predictors of SAP were NLR, PLR, mRC (after 48 h), HAP score, BISAP score and severe CTSI. With ROC curve, we find that CTSI and BISAP score has the best diagnostic accuracy (100% and 96.4%, respectively) followed by HAP score that had 83% overall accuracy. A variety of scoring systems based on clinical features, laboratory markers, and imaging characteristics have been described over the years to predict severe AP25,26,27,28,29,30,31.

Peripheral blood indices as NLR, PLR and red cell distribution width (RDW) have been widely reported as markers for rapid assessment of AP patients32. It was observed that elevated peripheral blood NLR and PLR demonstrated high sensitivity, specificity, and predictive value for identifying SAP, showing good correlation with the BISAP score. AUC for NLR, PLR, and RDW in predicting severe pancreatitis based on BISAP were 0.574, 0.521, and 0.722, respectively33.

In another study; BISAP score, NLR combined with BISAP score (BN score), PLR combined with BISAP score (BP score), and NLR, PLR combined with BISAP score (BNP score) were compared respectively to predict SAP. NLR and PLR were positively correlated with BISAP, APACHE Ⅱ and Ranson score. The AUC of BISAP score, BN score, BP score and BNP score for predicting SAP in HTGP were 0.865 (95% CI: 0.787–0.943), 0.925 (95% CI: 0.869–0.981),0.930 (95% CI: 0.885–0.987), and 0.936 (95% CI: 0.874–0.986)34.

HAP and BISAP were mainly created to evaluate patients with non-severe course of AP without indications for intensive management and expensive imaging procedures35,36. A meta-analysis of 30 studies which contained the data of 5,988 AP cases. The AUC for the prediction of severity of AP were 0.80 for CTSI; 0.79, for BISAP; 0.73 for CRP level; 0.81 for Ranson score and 0.80 for APACHE II score25.

Gupta et al. said that HAP and BISAP had 65% and 93% overall accuracy for prediction of SAP, respectively37. Borges et al. established that BISAP had better accuracy than HAP in prediction of SAP38.

In 1990 Balthazar et al. established the computed tomography severity index (CTSI), which can be used to predict the severity and mortality of patients with AP by grading pancreatic inflammation and necrosis and giving a numeric score13. Alberti et al. said that CTSI had fairly accuracy with AUC was 0.7439. Also, a meta-analysis reported that CTSI had 81% sensitivity with 19.1 OR for prediction of SAP25.

Another report utilized machine learning algorithms to construct predictive models for assessing the severity of AP and to identify key predictors influencing model performance. In that study, the CTSI demonstrated an overall diagnostic accuracy of 90% for SAP, with an AUC of 0.90640.

The superior accuracy of the CTSI observed in this study can be attributed to its ability to directly assess the underlying morphological and pathophysiological changes of acute pancreatitis. Unlike clinical or laboratory-based scores, CTSI incorporates imaging findings of pancreatic inflammation, peripancreatic fluid collections, and the extent of pancreatic necrosis, which are central determinants of disease severity and the development of organ failure39.

BISAP and HAP scores offer rapid bedside assessment using readily available clinical and laboratory data, making them particularly useful for early risk stratification at admission. The NLR and PLR are inexpensive, easily calculated hematological markers that can be repeatedly assessed to reflect the inflammatory response during early disease course41.

In contrast, the CT Severity Index provides the highest diagnostic accuracy by directly assessing pancreatic and peripancreatic pathology; however, its use is resource-dependent, requiring contrast-enhanced imaging, specialized expertise, and appropriate timing. This comparison highlights that integrating simple bedside scores and laboratory markers with selective use of CTSI may offer a balanced and pragmatic approach to severity assessment in acute pancreatitis42.

Limitations

The current study had some limitations; (1) it was conducted in single center, (2) no long term follow up of those patients after discharge from hospital to assess the frequency of possible local complications as pseudocysts and, (3) CTSI was calculated using contrast-enhanced CT performed within the first 6 h of admission, acknowledging that this timing may underestimate pancreatic necrosis, noting that delayed CT (≥ 48 h) may improve sensitivity for necrosis assessment.

And yet, our study had many strength points included large sample size (300 patients) that was considerable to obtain valid results. Also, we discussed several scores as NLR, PLR, RDW, HAP, BISAP, APACHE-II and CTSI in prediction severity of AP.

Conclusion

Acute pancreatitis is mild in the majority of cases and is associated with a favorable prognosis; however, a subset of patients develops severe disease requiring meticulous clinical monitoring and management. Accurate early risk stratification is therefore essential. Clinical, laboratory, and radiological parameters contribute to validate scoring systems for severity assessment.

In the present study, the BISAP score is recommended for early bedside risk stratification, while the CT Severity Index serves as a complementary tool when imaging is available to enhance diagnostic accuracy. Future research should focus on multicenter prospective validation, standardization of imaging timing, integration of clinical scores with machine-learning approaches, and evaluation of long-term outcomes and complications to further refine severity prediction and clinical decision-making in acute pancreatitis.