F1000Research F1000Research 2046-1402 F1000 Research Limited London, UK 10.12688/f1000research.125896.1 Research Article Articles Comparative analysis of computed tomography severity indices in predicting the severity and clinical outcome in patients with acute pancreatitis

[version 1; peer review: 1 approved, 1 approved with reservations, 1 not approved]

 Parmar Geetanjali Formal Analysis Investigation Methodology Visualization Writing – Original Draft Preparation https://orcid.org/0000-0001-8865-6802 1 Noronha Griselda Philomena Writing – Review & Editing https://orcid.org/0000-0002-7645-3618 a 1 Poornima Vinaya Conceptualization 1 Department of Radiodiagnosis, Kasturba Medical College, Mangalore, Manipal Academy of Higher Education, Manipal, India., Mangalore, Karnataka, 575001, India griselda.pn@manipal.edu

No competing interests were disclosed.

 8 11 2022 2022 11 1272 6 10 2022 Copyright: © 2022 Parmar G et al. 2022 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background: Acute pancreatitis (AP) has unpredictable severity. Its management is based on initial assessment of disease severity. It ranges from mild interstitial to severe necrotic form; the latter is associated with poor prognosis. Contrast-enhanced computed tomography (CT) of the abdomen is the gold standard in early detection of pancreatic necrosis and in assessing the severity of AP. Two CT grading systems exist to assess the severity of AP: CT severity Index (CSI) and modified CSI (MCSI). This study compares the usefulness of these two systems in predicting the severity and clinical outcome in AP in comparison with Ranson’s criteria and clinical outcome parameters.

Methods: This is a prospective hospital-based screening study of 80 patients aged >12 years with clinical diagnosis of AP who underwent contrast-enhanced CT study of the abdomen. Comparative analysis between MCSI and CSI with Ranson’s criteria and clinical outcome parameters was assessed by Chi-Squared test.

Results: The accuracy of CSI and MCSI in predicting the requirement of critical care, superadded infection, multiple organ dysfunction syndrome (MODS) and requirement of intervention were 73.0%, 64.5%, 69.8% 60.9% and 77.2%, 76.0%, 74.4% & 56.6% respectively. Area under the curve for MCSI score was significantly higher (AUC: 0.861; 95% CI: 0.736-0.986) than CSI score (AUC:0.815;95% CI:0.749-0.941). MCSI and CSI showed significant correlation with Ranson’s criteria; however, MCSI correlation was better (r:0.53; p<0.01) than CSI (r:0.35;p:0.04).

Conclusion: CSI and MCSI are better predictors of severity, clinical outcome and mortality compared with Ranson’s criteria, with MCSI being more accurate and better predictor than CSI. The accuracy of MCSI is better than CSI for prediction of requirement of critical care, development of superadded infection and development of MODS in AP. However, CSI and MCSI have low accuracy in predicting intervention in AP.

 Acute Pancreatitis CT severity index Modified CT severity index clinical outcome parameters Ranson’s criteria hospital stay multisystem organ dysfunction syndrome sepsis. The author(s) declared that no grants were involved in supporting this work. Introduction

Acute pancreatitis (AP) is one of the most common causes of acute abdomen with unpredictable clinical course. Based on the severity, 80% of cases are mild and 20% of cases are severe which morphologically correlate with edematous and necrotizing forms of AP respectively. The mild form is self-limiting without causing major physiological insult. The severe form is life threatening and can lead to early or late multiple organ dysfunction syndrome (MODS) and superadded infection. 1 – 3

Contrast enhanced computed tomography (CT) of the abdomen is the gold standard 4 in identifying necrosis and fluid collections in AP which can aid in predicting disease severity and prognosis of the patient, thus guiding the management. Various studies suggest the evidence that severity can be better assessed by CT than the numerical grading systems due to direct visualization of necrosis and complications of AP on CT. 5 , 6 Although the CT severity index (CSI) shows good correlation with the severity of AP, few studies suggested few limitations. CSI doesn’t show good correlation with clinical outcome, mortality, need for surgical or percutaneous interventional procedures, MODS and superadded infection. 3 These shortcomings led to the modification and simplification of CSI by Mortele et al. 3 leading to the formation of modified CT severity index (MCSI). The present study is comparative analysis of MCSI and CSI with clinical outcome parameters and Ranson’s criteria in patients with AP.

 Methods

This is a prospective hospital-based screening study performed in the department of Radiodiagnosis affiliated to Kasturba Medical College (KMC), Mangalore (MLR), Manipal Academy of Higher Education (MAHE), Manipal, India on 80 patients with clinical diagnosis of AP who underwent contrast enhanced CT abdomen over a period of 2 years from September 2019 to September 2021. The study was performed after the approval from the Institutional Ethics Committee (IEC), KMC, MLR, MAHE with approval number of IEC KMC MLR 09-19/411.

Paediatric patients <12 years of age were excluded from the study, as Ranson’s criteria is not done in this group of patients in our hospital. Patients with poor imaging results due to poor compliance or motion artefacts were excluded. Patients without intravenous ( i.v.) contrast administration were also excluded. Patients with diagnosis of acute-on-chronic, recurrent and calcific pancreatitis and those that got discharged against medical advice or were lost to follow up were also excluded from the study. Patients with cardiac, renal & respiratory comorbidities were excluded from the study. Since the informed consent is routinely taken prior to every CT study and research data are obtained from the CT machine computer and patient case files with no direct interaction with the study participants, IEC, KMC, MLR waived off additional informed consent from the study participants for this research.

16-slice and 32-slice CT scanner machines were used to acquire 5-mm plain CT axial sections followed by the administration of 1.5–2.0 mL/kg body weight (80–100 mL) of non-ionic i.v. contrast through the automated injector. This was followed by around 1 mL/kg body weight (40–50 mL) of normal saline. The rate of injection for both contrast and saline administration was ~4 mL/s which was altered in accordance with haemodynamic status, body weight and size of the i.v. cannula. The images were acquired in the arterial and porto-venous phases at 6–8 and 35–45 seconds respectively in all cases by bolus tracking method which is described as follows. A locator was placed on the aorta at D12–L1 level and the contrast injection got automatically triggered via the automated injector once the aorta at this level showed optimum contrast opacification. Axial sections of 5 mm slice thickness were then reformatted to thin 0.6 mm axial, sagittal and coronal sections. The clinical and laboratory details of the patient were obtained from the CT requisition form and patient case file. This was followed by assessment of severity of acute pancreatitis using both CSI ( Tables 1a, 1b and 1c) and MCSI ( Tables 2a, 2b and 2c). Accordingly, severity of AP was graded as mild, moderate and severe based on the scores.

CT severity index (CSI) <sup> <xref ref-type="bibr" rid="ref7">7</xref> </sup> Grading of acute pancreatitis by CSI with allocation of points to each grade.
CT findings Points
GRADE A Normal pancreas 0
GRADE B Focal or diffuse enlargement of the pancreas including irregularity of gland contour, inhomogenous attenuation, dilatation of pancreatic duct and foci of small fluid collections within the gland, where there was no evidence of peri-pancreatic changes. 1
GRADE C Abnormalities of pancreas which were intrinsic associated with hazy streaky densities representing inflammation in the surrounding peri-pancreatic fat. 2
GRADE D A single ill-defined fluid collection (phlegmon). 3
GRADE E Two or multiple, ill-defined collections of fluid or evidence of gas within or surrounding to the pancreas. 4
Assessment of presence & extent of pancreatic necrosis in AP by CSI with allotment of points.
Percentage of necrosis (%) Points
Absent 0
<30 2
30–50 4
>50 6
Total points from CT grading of AP ( <xref ref-type="table" rid="T1">Table 1a</xref>) & assessment of pancreatic necrosis ( <xref ref-type="table" rid="T2">Table 1b</xref>) were combined to get CSI score with categorization of severity.
Severity of AP CSI score
Mild 0 to 3
Moderate 4 to 6
Severe 7 to 10
Modified CT Severity Index (MCSI) <sup> <xref ref-type="bibr" rid="ref7">7</xref> </sup> Grading of acute pancreatitis by MCSI with allocation of points to each grade.
CT findings Points
Normal pancreas 0
Intrinsic pancreatic abnormalities with or without inflammatory changes in peri-pancreatic fat 2
Pancreatic or peri-pancreatic fluid collection or peri-pancreatic fat necrosis 4
Assessment of presence & extent of pancreatic necrosis in AP by MCSI with addition of extra-pancreatic complications & allotment of points.
Percentage of necrosis (%) Points
Absent 0
<30 2
>30 4
Extra-pancreatic complications (one or more of pleural effusion, ascites, vascular complications or gastrointestinal tract involvement) 2
Total points from CT grading of AP ( <xref ref-type="table" rid="T4">Table 2a</xref>) & assessment of pancreatic necrosis with extra-pancreatic complications ( <xref ref-type="table" rid="T5">Table 2b</xref>) were combined to get MCSI score with categorization of severity.
Severity of AP MCSI score
Mild 0 to 2
Moderate 4 to 6
Severe 8 to 10
Ranson’s criteria <sup> <xref ref-type="bibr" rid="ref8">8</xref> </sup>

Wherever available Ranson’s criteria score was noted down from patient case file and the correspondence of both the CT indices were studied with respect to the Ranson’s criteria. Ranson’s criteria score consists of 11 prognostic parameters, out of which five parameters are assessed at the admission and six parameters are assessed during initial 48 hours of hospital stay ( Tables 3a and 3b).

Assessment of five prognostic parameters of Ranson’s criteria at admission.
Prognostic factors assessed at the time of admission
Age more than 55 years
WBC Count more than 16,000 cells/mm 3
Blood Glucose more than 200 mg/dL
Serum glutamic oxaloacetic transaminase (AST) more than 250 U/L
Serum Lactate dehydrogenase (LDH) more than 350 U/L
Assessment of remaining six parameters of Ranson’s criteria during the first 48 hours of hospital stay.
Prognostic factors assessed during initial 48 hours of hospital stay
Serum calcium <8.0 mg/dL (<2.0 mmol/L)
Haematocrit fall > 10%
Arterial oxygen tension ( PO 2) < 60 mmHg
Blood urea nitrogen increase by 5 mg/dL or more despite intravenous fluid hydration
Base deficit > 4 mEq/L
Sequestration of fluids >6 L
Ranson’s score interpretation

Ranson’s score of 0 or 1 suggests complications will not develop in AP and mortality is negligible. On the other hand, Ranson’s score of 3 or more predicts severe AP with possible mortality. 9 The mortality in AP is directly proportional to the Ranson’s criteria score ( Table 3c).

Shows percentage of mortality with respect to the Ranson’s criteria score. <sup> <xref ref-type="bibr" rid="ref10">10</xref> </sup>
Ranson’s criteria score Mortality (%)
0–2 0–3
3–4 15
5–6 40
7–11 100

The clinical outcome parameters 6 , 7 were noted down from all the patient case files and its association with CT severity indices were studied and are as follows:

The extent of hospital or intensive care unit (ICU) stay (greater than or equal to 15 days);

Requirement of critical care, (Arterial oxygen tension ( PO 2) <60 mmHg or requirement of ventilation, systolic blood pressure (BP) <90 mmHg);

Requirement for (surgical/percutaneous) intervention (like drainage and aspiration);

Evidence of infection, (combination of a fever more than 100°F and elevated WBC count greater than 15,000 cells/mm);

Existence of organ failure (Arterial PO 2 <60 mmHg or requirement of ventilation, serum creatinine of >3 mg/dL or urine output of <500 mL per 24 h and systolic BP of <90 mmHg); and

Death.

Outcome Variables that were studied are as follows:

Sensitivity, specificity, positive predictive value, negative predictive value and accuracy of CSI and MCSI with respect to clinical outcome parameters like mean hospital stay, requirement of critical care, superadded infection, MODS, requirement of intervention & mortality.

Concordance of CSI and MCSI with the score of Ranson’s criteria.

Statistical analysis

The data was collected on a pre-designed study proforma. Qualitative data was expressed as percentage and frequency. Chi-Squared test was used to assess the association among the qualitative variables. The level of significance was represented by p-value of less than 0.05. Screening efficacy was computed using standard formulae. Wherever necessary, the results were graphically represented. Pearson correlation was used to assess the magnitude and direction of association between CSI and MCSI with Ranson’s score. Receiver operating characteristics (ROC) curves were used to compare the role of CSI and MCSI in predicting the mortality in AP with r value from +1 to −1. The r value of +0.1 to +1, 0 and −0.1 to −1 was suggestive of positive, zero and negative correlation respectively. Area under the curve (AUC) between CSI and MCSI as predictor of mortality was analyzed. Statistical package for social sciences (SPSS) version 21.0 (RRID:SCR_002865) and Microsoft Excel 2010 (RRID:SCR_016137) were used for most of the analysis and graphical representation respectively.

 Results Demographics

The patients with AP in this study were more or less equally distributed across all the decades from 2nd to 6th decade with mean age of 44.41 years. There was clear male predominance of 77.5% with 22.5% female patients with male to female ratio of 3.5:1. The most common cause for acute pancreatitis was alcoholism (56.3%) followed by gall stones (28.8%).

 Severity grading on CT by MCSI and CSI

As per MCSI, more than half of patients (56%) with AP had mild disease, about one third of them (36.3%) had moderate disease and a small percentage (7.5%) had severe disease ( Figure 1a).

Bar diagram showing grading of AP by MCSI.

As per CSI, about half of patients (52.5%) with AP had moderate disease, about one fourth of them (26.3%) had mild disease and 21.3% had severe disease ( Figure 1b).

Bar diagram showing grading of AP by CSI. Association of MCSI and CSI score with clinical outcome parameters

Requirement of critical care

Based on MCSI score, all the patients with severe AP (100.0%) required critical care, 82.8% of moderate disease needed critical care and only one third of patients with mild disease (31.1%) needed intensive care with (p<0.01) ( Figure 2a). The overall sensitivity and specificity for prediction of requirement of critical care was 85.7% and 68.9% respectively with an accuracy of 77.2%.

Bar diagram showing association of MCSI score with requirement of critical care.

As per CSI score, most (95.2%) of moderate disease, about half (52.4%) of severe disease and small percentage (11.8%) of mild disease required critical care (p<0.01) ( Figure 2b). The overall sensitivity and specificity for prediction of critical care requirement was 66.7% and 88.2% respectively with an accuracy of 73%.

Bar diagram showing association of CSI score with requirement of critical care. Development of superadded infection

As per MCSI, the superadded infection was seen in 83%, 41% and 4% of severe, moderate and mild disease of AP respectively (p value<0.01) ( Figure 3a). The overall sensitivity and specificity were 49% & 96% respectively with an accuracy of 76% in predicting superadded infection in AP patients.

Bar diagram showing association of MCSI score with development of superadded infection.

Based on the CSI score, there was no (0.0%) superadded infection in mild disease, while it was present in slightly less than half (47.6%) of severe disease and about 21.4% in moderate disease (p<0.01) ( Figure 3b). Hence the overall specificity & sensitivity for prediction of presence of superadded infections was 100.0% and 30.2% respectively with accuracy of 64.5%.

Bar diagram showing association of CSI score with development of superadded infection. Development of multiple organ dysfunction syndrome (MODS)

As per MCSI, MODS developed in 15.6% of mild, 58.6% of moderate and 83.3% of severe AP (p<0.01) ( Figure 4a). Overall sensitivity & specificity for prediction of development of MODS were 62.9% and 84.4% respectively with an accuracy of 74.4% respectively.

Bar diagram showing association of MCSI score with development of MODS.

As per the CSI score, there was no (0.0%) development of MODS in mild disease. On the contrary, most (95.2%) of severe disease and 21.4% percentage of moderate disease developed MODS ( Figure 4b). The overall specificity & sensitivity for prediction of development of MODS was 100.0% and 40.0% respectively with an accuracy of ~69.8%.

Bar diagram showing association of CSI score with development of MODS. Requirement of intervention

As per MCSI, intervention was performed in 55.6% of mild, 65.5% of moderate & 83.3% of severe cases of AP (p<0.01) ( Figure 5a). The overall sensitivity & specificity for prediction of requirement of intervention was 68.6% and 44.4% respectively with an accuracy of 56.6%.

Bar diagram showing association of MCSI score with requirement of intervention.

As per CSI score, approximately three-quarters (76.2%) of patients with severe acute pancreatitis required intervention, 61.9% of patients with moderate disease and 41.2% of patients with mild disease required intervention (p<0.01) ( Figure 5b). The overall sensitivity and specificity for prediction of requirement of intervention by CSI was ~66.7% and ~58.8%, respectively, with an accuracy of ~60.9%.

Bar diagram showing association of CSI score with requirement of intervention. Comparison of screening efficacy of MCSI and CSI with clinical outcome parameters

MCSI score showed both good sensitivity and specificity for development of MODS, good sensitivity for prediction of requirement of critical care & intervention. MCSI showed good specificity for the development of superadded infection ( Figure 6a).

Shows screening efficacy of MCSI score with clinical outcome parameters.

CSI score showed high specificity for the development of MODS and superadded infection. The overall accuracy is better with MCSI score than CSI score for prediction of requirement of critical care, development of superadded infection & development of MODS. Both MCSI and CSI scores had low accuracy in predicting the requirement of intervention ( figure 6b).

Shows screening efficacy of CSI score with clinical outcome parameters. Association of MCSI and CSI score with mortality

As per MCSI score, mortality rate was 100.0% in AP, 17.2% in moderate disease and 2.2% in mild disease ( Figure 7a). The overall sensitivity and specificity for prediction of mortality was 91.7% and 64.7%, respectively, with an accuracy of 87.5%.

Bar diagram showing association of MCSI score with mortality.

As per CSI score, mortality rate was one-third (33.3%) and highest in severe grade of AP, followed by mild grade of AP (17.6%) with lowest mortality rate in moderate grade (4.8%) ( Figure 7b). The overall sensitivity & specificity for prediction of mortality was 58.5% and 79.4%, respectively, with an accuracy of 73.8%.

Bar diagram showing association of CSI score with mortality. Association of MCSI score with mean hospital stay

The mean hospital stay by MCSI was highest in moderate grade of AP with ~22 days as compared to approximately 12 & 13 days in mild & severe disease respectively (p<0.01) ( Table 4a).

Shows association of MCSI with mean hospital stay.
MCSI Score N Mean hospital stay (days)
Mild 45 12.27
Moderate 29 22.14
Severe 6 13.33
Total 80 15.93

The mean hospital stay as per CSI score was significantly higher in moderate and severe grade of acute pancreatitis corresponding to approximately 18 and 19 days respectively as opposed to approximately 8 days in mild disease (p<0.01) ( Table 4b).

Shows association of CSI with mean hospital stay.
CSI Score N Mean hospital stay (days)
Mild 17 7.53
Moderate 42 17.95
Severe 21 18.67
Total 80 15.93
Correlation analysis for MCSI & CSI scores with Ranson’s criteria

Ranson’s criteria score was available with 31 out of 80 patients (38.8%). There was significant correlation between Ranson’s criteria and both CT severity indices (CSI and MCSI) but the correlation was highly statistically significant and better with MCSI score (r=0.53; p<0.01) as compared to CSI score (r=0.35; p=0.04) ( Table 5, Figures 8a & 8b).

Shows Pearson correlation of MCSI & CSI scores with Ranson’s criteria score.
Pearson co-relation
Ranson’s criteria r-value p-value
CSI Score 0.35 0.040
MCSI Score 0.53 <0.01
Scatter plot between CSI score (x axis) with Ranson’s criteria score (y axis) which shows positive correlation with Pearson correlation coefficient (r value) of 0.35. Scatter plot between MCSI score (x-axis) with Ranson’s criteria score (y-axis) which shows positive correlation with Pearson correlation coefficient (r value) of 0.53. ROC Curve analysis of MCSI & CSI score for prediction of mortality

According to ROC Curve analysis, both CSI & MCSI scores were significant predictors of development of mortality in AP. However, area under curve was significantly higher for MCSI score (AUC 0.861; 95% CI 0.736–0.986) as compared to CSI score (AUC 0.815; 95% CI 0.749–0.941) ( Table 6 & Figure 9).

Shows area under the curve (AUC) analysis of CSI & MCSI in predicting mortality in AP.

Area under the curve
Test result variable(s) Area SE p-value Asymptotic 95% confidence interval
Lower bound Upper bound
CSI Score 0.815 0.049 <0.01 0.749 0.941
MCSI Score 0.861 0.064 <0.01 0.736 0.986
ROC curve analysis of CSI & MCSI scores shows both CSI (green coloured graph) & MCSI (purple coloured graph) as significant predictors of mortality in AP with MCSI score being better & more accurate than CSI score (yellow coloured graph is reference line). Discussion

Contrast enhanced computed tomography of the abdomen is the imaging modality of choice and is the gold standard in the diagnosis of AP. The necrotizing form of AP, though less common, if present, is associated with a myriad of life-threatening complications. Among all the diagnostic tests available, CT has the highest diagnostic accuracy in detecting pancreatic necrosis. 11

Steinberg et al. 12 in their study suggested the evidence of 80 to 90% of AP was due to cholelithiasis & chronic alcoholism. Our study suggests evidence of alcoholism as the most common etiological factor for AP (56.3%) followed by gall stones (28.8%). Similar evidence was suggested by Wongnai et al. 13 in their study on 90 patients of AP, where alcoholism and pancreatico-biliary ductal calculi were reported as aetiological factor in 60% and 18% patients respectively. In India, alcohol consumption is predominantly seen in males (male to female ratio of 24.3:1). 14 The suggestive evidence of alcohol abuse as the commonest aetiological factor of AP combined with the male predominance of alcohol consumption in India explains the male to female preponderance (3.5:1) in this study. Similar evidence was suggested by Dugernier T L et al. 15 and Balthazar EJ et al. 16

On the contrary, Raghuwanshi S et al. 17 suggested the evidence of most common aetiology for AP as cholecystolithiasis (42%) followed by alcoholism (38%) with remaining 20% aetiology for AP belonged to rest category which included idiopathic, trauma and drug induced cases (24%, 2% and 2% respectively). Casas et al. 18 in their study on 148 patients suggested cholelithiasis (57%) as the most common aetiological factor for AP followed by alcoholism (21%) with both together contributing to another 5% of AP patients. Bollen TL et al. 19 and Jauregui et al. 20 also suggested the evidence of cholelithiasis as the predominant aetiological factor for AP.

Severity of AP

This study is comparative analysis between MCSI and CSI grading systems in assessing severity and clinical outcome. Majority of the patients with AP belonged to mild category as per MCSI and moderate category as per CSI. This resulted in a small group of patients who had different category of severity by CSI and MCSI. The present study suggests MCSI to be more accurate predictor of severity than CSI as it predicted clinical outcome more accurately in those patients who were differently categorized in severity by CSI. This better prediction of severity and clinical outcome by MCSI in AP may be attributable to inclusion of extra-pancreatic complications of AP like ascites, pleural effusion, vascular complications and gastrointestinal complications in the assessment of MCSI which are not included in CSI. Kondekar S et al. 7 and Banday et al. 21 suggested partially opposing evidence from our study where majority of the patients by MCTSI belonged to mild category as per our study and the majority of the patients belonged to severe category as per CSI unlike our study.

 Clinical outcome parameters

Banday et al. 21 in their study suggested evidence of increasing mean duration of hospital stay with increasing severity by MCTSI score and concluded that the duration of mean hospital stay is directly proportional to severity grading by MCTSI system in acute pancreatitis.

Our study suggests mean hospital stay in AP by CSI score is significantly longer in moderate and severe disease as compared to mild disease (p<0.01) whereas the mean hospital stay by MCSI is significantly longer in moderate disease as compared to mild and severe disease (p<0.01). This can be attributed to the fact that mild cases were discharged relatively early from hospital in comparison to moderate category cases and very severe cases had higher mortality with lesser hospital stay.

Overall in the present study, MCSI score showed good sensitivity for prediction of requirement of critical care, development of MODS and requirement of intervention. MCSI showed good specificity for MODS and development of superadded infection. CSI showed high specificity for MODS and development of superadded infection. Overall accuracy of MCSI was better than CSI for prediction of requirement of critical care, development of superadded infection and development of MODS. Both scores showed lower accuracy with regard to requirement of intervention.

The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of MCSI in predicting severity according to the study by Bollen TL et al. 19 were 71%, 93%, 69% and 94%, respectively. This study suggested evidence of accurate correlation of clinical scoring systems with systemic complications & mortality in acute pancreatitis. The study also suggested evidence that radiological scoring system was more accurate in predicting the severity of acute pancreatitis, superadded infection and need for intervention than clinical scoring system. Among the two radiological scoring systems, the study suggested no evidence of significant differences between CSI and MCSI in predicting severity in acute pancreatitis.

Bollen et al. 22 suggested CSI showed better sensitivity, specificity, PPV and NPV than MCSI. Whereas Jauregui-Arrieta LK et al. 19 suggested different evidence where MCSI showed better sensitivity, specificity and PPV than CSI in severe AP and concluded that MCSI is better screening test than CSI in severe AP. Sharma et al. 23 performed suggested sensitivity and NPV is better with MCSI (98.6% and 90%, respectively) than CSI (87.3% and 57.1%, respectively) with similar PPV for both (~74%) and low specificity of 26.5% and 35.3% for MCSI and CSI, respectively.

 Ranson’s criteria

The present study shows significant correlation between Ranson’s criteria and both severity indices on CT (CSI and MCSI) but the correlation of MCSI with Ransons’ criteria is highly statistically significant which suggests MCSI as a better predictor of severity and clinical outcome than CSI.

On receiver operating characteristic (ROC) curve analysis, the present study suggests evidence that both CSI and MCSI are significant predictors of development of mortality in AP. However, the area under curve was significantly higher for MCSI score (AUC 0.861; 95% CI 0.736–0.986) as compared to CSI (AUC 0.815; 95% CI 0.749–0.941) which suggests MCSI as better predictor of mortality in AP than CSI.

Mangalanandan S et al. 24 suggested evidence of strong correlation between Ranson’s criteria and MCSI with mild and severe forms of AP showing 100% agreement with each other. But moderate category in MCSI Score had disagreeing results because Ranson’s criteria has only mild and severe categories due to which moderate category patients could not be studied. Their study suggested that MCSI (sensitivity of 93.33% and specificity of 54.17%) is more sensitive but less specific than Ranson’s criteria (sensitivity of 80% and specificity of 83.3%) in predicting actual outcome of AP. Although Chand P et al. 25 suggested evidence of lack of statistical significant difference between Ranson’s criteria and MCSI in evaluation of the outcome of AP with respect to the systemic complications, 7 there was statistically significant difference between MCSI and Ranson’s criteria with respect to local complications with increased incidence of local complications with higher Ranson’s criteria.

One of the drawbacks in the study was pediatric patients below the age of 12 years were not concluded as Ranson’s criteria is not done for them in our hospital. Also, larger sample size will reduce the margin of error in the study. Third drawback is the lack of availability of Ranson’s criteria score in 61.3% of the patients in the study due to usage of various other alternative clinical grading systems like Revised Atlanta Classification, Acute physiology and chronic health evaluation (APACHE) II & Bedside index of severity in acute pancreatitis (BISAP) by the treating clinician. These alternative clinical grading systems are affordable, quick & requires less effort in assessing the severity of AP than Ranson’s criteria.

 Conclusion

Both CSI and MCSI are better predictors of severity, clinical outcome and mortality than Ranson’s criteria in patients with AP with MCSI being more accurate & better predictor of the same than CSI. The accuracy of MCSI is better than CSI for prediction of requirement of critical care, development of superadded infection and development of MODS. Both CSI and MCSI scores have low accuracy with regard to requirement of intervention in AP patients.

 Data availability Underlying data

Mendeley: Underlying data for ‘Comparative analysis of computed tomography severity indices in predicting the severity and clinical outcome in patients with acute pancreatitis’, https://doi.org/10.17632/htkkzr9zbr.2. 26

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

 Acknowledgment

We would like to acknowledge Dr. Ashvini Kumar, Former Head of Department & Former Professor of Department of Radiodiagnosis, Kasturba Medical College, Mangalore, Manipal Academy of Higher Education, Manipal, India who provided insight & expertise that greatly assisted this research.

 References Frossard JL Steer ML Pastor CM : Acute pancreatitis. Lancet. 2008;371:143–152. 10.1016/S0140-6736(08)60107-5 Balthazar EJ : Acute Pancreatitis: Assessment of Severity with Clinical and CT Evaluation. Radiology. June 2002;223(3):603–613. 10.1148/radiol.2233010680 Mortele KJ Wiesner W Intriere L : A Modified CT Severity Index for Evaluating Acute Pancreatitis: Improved Correlation with Patient Outcome. AJR. 2004;183:1261–1265. 15505289 10.2214/ajr.183.5.1831261 Zhao K Adam SZ Keswani RN : Acute Pancreatitis: Revised Atlanta Classification and the Role of Cross-Sectional Imaging. AJR. 2015;205:W32–W41. 26102416 10.2214/AJR.14.14056 Leung TK Lee CM Lin SY : Balthazar computed tomography severity index is superior to Ranson criteria and APACHE II scoring system in predicting acute pancreatitis outcome. World J. Gastroenterol. 2005;11(38):6049–6052. 16273623 10.3748/wjg.v11.i38.6049 PMC4436733 Kumar AH Griwan MS : A comparison of APACHE II, BISAP, Ranson’s score and modified CTSI in predicting the severity of acute pancreatitis based on the 2012 revised Atlanta Classification. Gastroenterol. Rep (Oxf). 2018 May;6(2):127–131. 29780601 10.1093/gastro/gox029 Kondekar S Minne I : Assessment of acute pancreatitis using CT severity index and modified CT severity index: A tertiary care hospital based observational study. Int. J. Radiol. Diagn. Imaging. 2020;3(1):118–122. 10.33545/26644436.2020.v3.i1b.64 Abu-Eshy SA Abolfotouh MA Nawar E : Ranson's criteria for acute pancreatitis in high altitude: do they need to be modified? Saudi J. Gastroenterol. 2008 Jan;14(1):20–23. 19568489 10.4103/1319-3767.37797 Basit H Ruan GJ Mukherjee S : Ranson Criteria. [Updated 2021 Sep 28]. StatPearls. Treasure Island (FL): StatPearls Publishing;2022 Jan. Reference Source Bhat MS : SRB’s manual of surgery. 6th ed. India: Jaypee Brothers Medical Publishers (P) Ltd.;2019. Urooj T Shoukat S Bokhar I : Diagnostic accuracy of contrast enhanced computed tomography (CECT) in detection of necrosis in acute pancreatitis by taking surgical findings as gold standard. J. Pak. Med. Assoc. November 2020;70(11):1930–1933. Steinberg WM Chari ST Forsmark CE : Controversies in clinical pancreatology: management of acute idiopathic recurrent pancreatitis. Pancreas. 2003 Aug;27(2):103–117. 10.1097/00006676-200308000-00001 Wongnai A Mai WN : CT FINDINGS OF ACUTE PANCREATITIS IN MAHARAJ NAKORN CHIANG MAI HOSPITAL. Chiang Mai Med. J. 2007;46(2):45–53. Balasubramani K Paulson W Chellappan S : Sociodemographic Risk Factors of Alcohol Consumption in Indian Men and Women: Analysis of National Family Health Survey-4 (2015-16), a Nationally Representative Cross-Sectional Study Front. Public Health. August 2021;9:1–10. 10.3389/fpubh.2021.617311 Dugernier TL Laterre PF Wittebolr X : Compartmentalization of the inflammatory response during acute pancreatitis: correlation with local and systemic complications. Am. J. Respir. Crit. Care Med. 2003 Jul;168(2):148–157. 12851244 10.1164/rccm.2204019 Balthazar EJ Freeny PC VanSonnenberg E : Imaging and intervention in acute pancreatitis. Radiology. 1994 Nov;193(2):297–306. 10.1148/radiology.193.2.7972730 Raghuwanshi S Gupta R Vyas MM : CT Evaluation of Acute Pancreatitis and its Prognostic Correlation with CT Severity Index. J. Clin. Diagn. Res. 2016 Jun;10(6):TC06–TC11. 27504376 10.7860/JCDR/2016/19849.7934 Casas JD Díaz R Valderas G : Prognostic value of CT in the early assessment of patients with acute pancreatitis. AJR Am. J. Roentgenol. 2004 Mar;182(3):569–574. 14975947 10.2214/ajr.182.3.1820569 Bollen TL Singh VK Maurer R : Comparative evaluation of the modified CT severity index and CT severity index in assessing severity of acute pancreatitis. AJR Am. J. Roentgenol. 2011 Aug;197(2):386–392. 21785084 10.2214/AJR.09.4025 Jáuregui-Arrieta LK Alvarez-López F Cobián-Machuca H : Effectiveness of the modify tomographic severity index in patients with severe acute pancreatitis. Rev. Gastroenterol. Mex. 2008;73(3):144–148. 19671500 Banday IA Gattoo I Khan AM : Modified Computed Tomography Severity Index for Evaluation of Acute Pancreatitis and its Correlation with Clinical Outcome: A Tertiary Care Hospital Based Observational Study. J. Clin. Diagn. Res. 2015 Aug;9(8):TC01–TC05. 26436014 10.7860/JCDR/2015/14824.6368 Bollen TL Singh VK Maurer R : A comparative evaluation of radiologic and clinical scoring systems in the early prediction of severity in acute pancreatitis. Am. J. Gastroenterol. April 2012;107(4):612–619. 22186977 10.1038/ajg.2011.438 Sharma V Rana SS Sharma RK : A study of radiological scoring system evaluating extrapancreatic inflammation with conventional radiological and clinical scores in predicting outcomes in acute pancreatitis. Ann. Gastroenterol. 2015;28(3):399–404. 26129965 Mangalanandan S Thomas DA Benjamin G : Correlation of Modified Computed Tomography Severity Index with Ranson’s Criteria in Assessing Severity of Acute Pancreatitis. Int. J. Anat. Radiol. Surg. 2021 Jan;10(1):RO22–RO27. 10.7860/IJARS/2021/44831.2595 Chand P Singh R Singh DP : Evaluation of the Outcome of Acute Pancreatitis by Ranson’s Criteria and Modified CT Severity Index. International Journal of Contemporary Medicine Surgery and Radiology. 2017;2(2):58–61. Noronha G Parmar G Poornima V : “COMPARITIVE ANALYSIS OF CT SEVERITY INDICES IN PREDICTING THE SEVERITY & CLINICAL OUTCOME IN PATIENTS WITH ACUTE PANCREATITIS”, Mendeley [DATASET]. 2022;V2. 10.17632/htkkzr9zbr.2 10.5256/f1000research.138249.r206452 Reviewer response for version 1 Xiao Juan 1 Referee Guilin Medical University, Guangxi, China

Competing interests: No competing interests were disclosed.

 29 9 2023 Copyright: © 2023 Xiao J 2023 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

This study described the usefulness of modified CT severity index in the prediction of acute pancreatitis severity and outcome which is of clinical significance. However, there are still some issues which need to be addressed.

The authors should put the MSCI and SCI results compared to the same parameter together in one figure but not separately in Figure 1-5, 7.

The Ranson score result should be added in figure 6 and 9 in order to support your conclusion that MSCI is better than Ranscon score in outcome prediction.

Did the MSCI score correlate to the etiologies of acute pancreatitis? 

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

I cannot comment. A qualified statistician is required.

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Partly

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

Biochemistry, baic medicineļ¼Œacute pancreatitis

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Noronha Griselda Kasturba Medical College Hospital, India

Competing interests: No competing interests were disclosed.

 15 3 2024

Dear Reviewer,

Thank you for your detailed review of our article titled "Comparative Analysis of Computed Tomography Severity Indices in Predicting the Severity and Clinical Outcome in Patients with Acute Pancreatitis" submitted to the F1000 Journal. We appreciate your time and effort in providing constructive feedback to enhance the quality of our research.

We acknowledge your suggestions.

1) Based on your feedback, we will revise the figures accordingly to incorporate the MSCI and CSI results compared to the same parameter together in one figure but not separately in Figure 1-5, 7. We are committed to addressing all your concerns and making the necessary revisions to strengthen the validity and impact of our findings. Your valuable input will undoubtedly improve the clarity and reliability of our study, and we are grateful for your insightful suggestions.

2) Regarding the addition of the Ranson score in Figures 6 and 9 of our manuscript to better support the conclusion that the MCSI is superior to the Ranson score in predicting clinical outcomes in patients with acute pancreatitis, we apologize for any confusion or oversight on our part in not including the Ranson score in the mentioned figures. It would be great if you could suggest how to incorporate the Ranson score in figure 6 and 9 as we are unable to figure it out. I discussed with my statistician and he was unable to figure it out. Hence your input would be greatly appreciated on this. 

3) In our study we found that the most common etiology for acute pancreatitis was alcoholism followed by gall stones and third was idiopathic. we have not correlated MCSI with acute pancreatitis etiologies in our study.  

We will promptly update the manuscript as per your recommendations and hope that these revisions will further enhance the significance and relevance of our research in the field of acute pancreatitis management.

Thank you once again for your thorough review and for guiding us to refine our work for publication in F1000 Journal.

Sincerely,

Griselda Philomena Noronha

Corresponding Author

 10.5256/f1000research.138249.r206377 Reviewer response for version 1 Trna Jan 1 Referee Masaryk University, Brno, Czech Republic

Competing interests: No competing interests were disclosed.

 29 9 2023 Copyright: © 2023 Trna J 2023 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve

Interesting topic with interesting results showing that CT results can predict severity well. This is probably not a completely new finding, however comparison with Ranson scoring system and showing that CTSI is better is interesting since CT is reproducible and not subjective. There are many new scoring systems and showing that good old CTSI is good is in my opinion a very good piece of evidence.

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

Yes

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Yes

Are the conclusions drawn adequately supported by the results?

Yes

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

gastroenterology, endoscopy

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Noronha Griselda Kasturba Medical College Hospital, India

Competing interests: No competing interests were disclosed.

 15 3 2024

Thank you so much for your valuable comments and review report. Highly appreciate your feedback.

 10.5256/f1000research.138249.r182028 Reviewer response for version 1 Dawra Saurabh 1 Referee Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, Punjab and Haryana, India

Competing interests: No competing interests were disclosed.

 19 7 2023 Copyright: © 2023 Dawra S 2023 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation reject

Introduction:

I would recommend to change the first line that the disease has unpredicted severity. With the present level of research, we do have fair enough idea of predicted severity. The authors may rephrase the sentence.

While it is agreed that CTSI has it’s fair share of limitations, the same hasn’t been brought about clearly in the “Introduction“ section.

Please bring out clearly why have you chosen Hanson’s criteria as comparison to CTSI. What about Atlanta classification?

Methods:

How was pancreatitis diagnosed?

What is your institutional protocol of managing pancreatitis?

At what stage of disease/ day of illness was the CT done?

Do you perform all investigations as required to calculate Ranson’s criteria on all patients with AP?

What do you mean by critical care? What is your institute’s protocol in admitting the patients for critical care?

Discussion:

What is unique in your study? How is it different from other comparisons including meta analysis on the same topic?

What are your limitations?

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

No

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

No

Reviewer Expertise:

Pancreas, Liver

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.

Noronha Griselda Kasturba Medical College Hospital, India

Competing interests: No competing interests were disclosed.

 15 3 2024

Thank you for your insightful questions regarding our manuscript. We appreciate the opportunity to clarify these points and ensure a comprehensive understanding of our study protocols and findings. Below, please find detailed responses to each of your queries:

introduction: 

"I would recommend to change the first line that the disease has unpredicted severity. With the present level of research, we do have fair enough idea of predicted severity. The authors may rephrase the sentence"- Thank you so much for bringing out this to us. we will surely rephrase this sentence. 

"While it is agreed that CTSI has it’s fair share of limitations, the same hasn’t been brought about clearly in the “Introduction“ section" - Again thank you for your valuable feedback. we shall do this change. 

"Please bring out clearly why have you chosen Hanson’s criteria as comparison to CTSI. What about Atlanta classification?" 

a)The Ranson criteria was the first scoring system specifically designed to assess the severity and prognosis of acute pancreatitis & hence it is extensively studied and validated globally. Since its used since ages, it is a reliable tool.

b) Ranson’s score is directly correlated with patient outcomes which means it is a clear prognostic indicator. High score at admission and at 48hours later is associated with increased rate of mortality and morbidity which helps in stratifying patients according to the need for ICU admission and aggressive therapy.

c) Ranson criteria is simpler and more straightforward, making it easier to apply in clinical settings, especially where immediate decisions are required while other scores like APACHE II requires more amount of data & is relatively complex to calculate.

d) APACHE II and BISAP are non-specific to pancreatitis and are used in various critical care settings whereas Ranson’s criteria is specifically designed for acute pancreatitis.

e) Though BISAP score is simple & easy to use, it predicts based on the assessment within first 24 hours of admission. It is relatively inferior in predicting the complications or mortality which can happen later in the course of illness. Ranson’s score is relatively superior to BISAP in this context as it considers assessment both at admission & 48hours after admission.

f) APACHE II is more complex, need more data to calculate and it is not specific to acute pancreatitis as compared to Ransons score.  

why not Atlanta Classification? Atlanta classification is based on morphological assessment of acute pancreatitis severity on CT which usually takes places after atleast 48-72 hours. based on that we give CSI & MCSI scores.  Till then Ranson’s score will provide early prognostic information before detailed imaging findings are available.

Method: 

1. How was pancreatitis diagnosed?

The diagnosis of pancreatitis in our study followed a rigorous, multi-dimensional approach in line with established clinical guidelines. Initially, we considered the patient’s clinical presentation, including relevant history (e.g., alcohol consumption, gallstones), as well as symptoms like abdominal pain and vomiting. Objective findings such as ileus, fever, tachycardia, and hypotension were also critical. Crucially, we substantiated our clinical suspicion through laboratory evidence, particularly elevated lipase and amylase levels, which were at least threefold above the upper limit of normal. Finally, imaging played a pivotal role in our diagnostic algorithm. Contrast-Enhanced Computed Tomography (CECT) of the abdomen or Ultrasound was utilized to confirm the diagnosis, adhering to the principle that the clinician’s discretion, based on a synthesis of these data points, ultimately guided the diagnostic process.

 2. Institutional protocol for managing pancreatitis

At our institution, the Department of Surgery manages pancreatitis cases. The management protocol begins at the point of presentation, where patients with symptoms indicative of pancreatitis or those evaluated for abdominal pain are assessed by the attending surgeon. Our approach prioritizes conservative management, including bowel rest, fluid resuscitation, and the administration of IV antibiotics and analgesics. Supportive measures such as supplemental oxygen, inotropic or ventilatory support, and transfusions for DIC are employed based on individual patient needs. Admission decisions—whether to the ward for mild cases or the ICU for those with significant organ dysfunction or unstable vitals—are made in accordance with the severity of the clinical presentation. Follow-up and further management, including imaging with CECT abdomen, are guided by established criteria, such as Ranson's criteria, and the clinical judgment concerning suspected complications.

Reference :  (Banks PA, Freeman ML. Practice guidelines in acute pancreatitis. Am J Gastroenterol. 2006;101:2379–2400.)

 3. At what stage of disease was CT done?

The timing for deploying CECT in our cases was carefully considered, primarily focusing on patients exhibiting signs of complicated pancreatitis, as defined by the Atlanta criteria, or when clinical suspicion of complications arose. Adhering to best practice, we generally deferred CT imaging for at least 48-72 hours following the onset of symptoms. This delay allows the acute inflammatory processes to stabilize, thereby enhancing diagnostic accuracy and informing subsequent management decisions, particularly in our ICU patients who are closely monitored for any signs of worsening or complications.

Reference : Busireddy KK, AlObaidy M, Ramalho M, Kalubowila J, Baodong L, Santagostino I, Semelka RC. Pancreatitis-imaging approach. World J Gastrointest Pathophysiol. 2014 Aug 15;5(3):252-70. doi: 10.4291/wjgp.v5.i3.252. PMID: 25133027; PMCID: PMC4133524.)

4) Do you perform all investigations as required to calculate Ranson’s criteria on all patients with AP?

     The investigations for Ranson's criteria were not performed for all cases but based on the clinical condition as per the discretion of the attending surgeon. Complete blood counts, glucose, calcium, liver and renal function tests and blood gas was performed at baseline for all cases whereas the post 48 hour investigations were done selectively based on the clinical condition of the patient. Based on the predicted severity, a decision to perform a CECT scan abdomen was taken.

5. What do you mean by critical care? What is your institute’s protocol in admitting the patients for critical care?

We had a Surgical intensive care unit at our hospital with ventilators and routine intensive care support systems. Unfortunately, there was no step-up or high dependency unit (HDU) facility available. Any patient who was found to have any organ dysfunction or features of shock was admitted to the ICU for monitoring and treatment and started on appropriate treatment. They were provided with supplemental oxygen, Non invasive ventilation (NIV) or endotracheal intubation with ventilation as and when required and hemodynamic support with transfusions or inotropic and pressor supports along with routine conservative management.

Discussion: 

6) unique in our study :

There was no other study in the literature comparing both radiological severity indicators CSI & MCSI with Ranson's score alone in this combination at the time of conceptualization of this study. there were studies comparing CSI alone or MCSI alone with Ranson score or correlating radiological scores (CSI/MCSI) with multiple clinical scoring systems. 

7) Limitations of study : 

a) the patients who underwent CECT for acute pancreatitis but who were discharged against medial advice or lost for follow up could not be assessed and excluded from the study.

b)the study did not include patients below 12 years of age as Ranson's criteria is not done for paediatric patients in our hospital. 

c)smaller sample size which may increase the margin of error when compared to a study with a relatively larger sample size. 

d) Since no single scoring system can universally predict the patient outcome perfectly, it is a limitation of the study. 

e) Ranson's score needs 48hours to complete the score. CSI & MCSI is obtained after atleast 48-72hours of admission. So there is delay of obtaining scores and hence it has lesser sensitivity in very early stages of disease. 

We trust these responses address your queries comprehensively. Our team is committed to delivering research that meets the highest standards of clinical rigor and transparency. Should you require further details or clarification, we are at your disposal.