Validation of the efficacy of the prognostic factor score in the Japanese severity criteria for severe acute pancreatitis: A large multicenter study

Abstract

Background

The Japanese severity criteria for acute pancreatitis (AP), which consist of a prognostic factor score and contrast-enhanced computed tomography grade, have been widely used in Japan.

Objective

This large multicenter retrospective study was conducted to validate the predictive value of the prognostic factor score for mortality and complications in severe AP patients in comparison to the Acute Physiology and Chronic Health Evaluation II (APACHE II) score.

Methods

Data of 1159 patients diagnosed with severe AP according to the Japanese severity criteria for AP were retrospectively collected in 44 institutions.

Results

The area under the curve (AUC) for the receiver-operating characteristic curve of the prognostic factor score for predicting mortality was 0.78 (95% confidence interval (CI), 0.74–0.82), whereas the AUC for the APACHE II score was 0.80 (95% CI, 0.76–0.83), respectively. There were no significant differences in the AUC for predicting mortality between two scoring systems. The AUCs of the prognostic factor scores for predicting the need for mechanical ventilation, the development of pancreatic infection, and severe AP according to the revised Atlanta classification were 0.84 (95% CI, 0.81–0.86), 0.73 (95% CI, 0.69–0.77), and 0.83 (95% CI, 0.81–0.86), respectively, which were significantly greater than the AUCs for the APACHE II score; 0.81 (95% CI, 0.78–0.83) for the need for mechanical ventilation (p = 0.03), 0.68 (95% CI, 0.63–0.72) for the development of pancreatic infection (p = 0.02), and 0.80 (95% CI, 0.77–0.82) for severe AP according to the revised Atlanta classification (p = 0.01).

Conclusion

The prognostic factor score has an equivalent ability for predicting mortality compared with the APACHE II score. Regarding the ability for predicting the development of severe complications during the clinical course of AP, the prognostic factor score may be superior to the APACHE II score.

Keywords

Acute pancreatitis diagnosis severity criteria prognostic factor score mortality

Introduction

Acute pancreatitis (AP) is an inflammatory disease resulting from the self-digestion of the pancreas due to the activation of pancreatic enzymes. The overall prevalence of AP ranges from 36–78 cases per 100,000 people annually.^1–4 The clinical course of AP varies substantially, from a mild disease treated conservatively to a severe disease resulting in multiple organ failure, infected pancreatic necrosis, and mortality.^5–7 Recently, the overall mortality for AP has decreased partly due to the widespread use of guidelines for the management of AP.^8,9 Nonetheless, severe AP is still associated with high mortality, ranging from 3–24%.^10–13 Early identification of patients with severe AP who are at high risk for complications and unfavorable outcomes is likely to further decrease mortality.

The Ministry of Health, Labour, and Welfare of Japan has developed a set of criteria for assessing the severity of AP (Japanese severity criteria for AP), which is used for clinical evaluations in Japan. The Japanese severity criteria for AP consist of two parts: the prognostic factor score and the contrast-enhanced computed tomography (CT) grade evaluation. The severity of AP is determined according to the prognostic factor score or the CT grade, or a combination of these (Table 1).¹⁴ The prognostic factor score is designed to allow the assessment of AP severity during the early stages of hospital admission, using physiologic and laboratory data that are readily available in most local hospitals. To the best of our knowledge, although the validity of the prognostic factor score has been compared with other established scoring systems at a single center in three previous studies, there are no published studies conducted in a multicenter context.^15–17

Table 1.

The Japanese severity criteria for acute pancreatitis of the Japanese Ministry of Health, Labour and Welfare.

Severe acute pancreatitis: prognostic factor≥3 or CT grade≥2
Prognostic factors (1 point for each factor)
1. Base excess ≤–3 mEq/l or shock (systolic blood pressure <80 mm Hg)
2. PaO₂ ≤60 mm Hg (room air) or ventilatory failure (ventilator management is needed)
3. BUN ≥40 mg/dl (or Cr ≥2.0 mg/dl) or oliguria (daily urine output <400 ml even after IV fluid resuscitation)
4. LDH ≥2 times of upper limit of normal
5. Platelet count ≤100,000/mm³
6. Serum Ca≤7.5 mg/dl
7. CRP≥15 mg/dl
8. Number of positive measures in SIRS criteria≥3
9. Age≥70 years
CT grade by contrast-enhanced CT
1. Extrapancreatic progression of inflammation
Anterior pararenal space	0 point
Root of mesocolon	1 point
Beyond lower pole of kidney	2 points
2. Hypoenhanced lesion of the pancreas
The pancreas is conveniently divided into three segments (head, body, and tail).
Localized in each segment or only surrounding the pancreas	0 point
Covers 2 segments	1 point
Occupies entire 2 segments or more	2 points
1 + 2 = Total scores
Total score = 0 or 1	Grade 1
Total score = 2	Grade 2
Total score = 3 or more	Grade 3

BUN: blood urea nitrogen; CRP: C-reactive protein; CT: computed tomography; LDH: lactate dehydrogenase; PaO₂: partial pressure of oxygen in blood; SIRS: systemic inflammatory response syndrome.

Measures in SIRS diagnostic criteria: (a) temperature >38℃ or <36℃; (b) heart rate >90 beats/min; (c) ventilatory rate >20 breaths/min or PaCO₂ <32 torr; (d) white blood cell (WBC) count>12,000 cells/mm³ <4000 cells/mm³, or >10% immature (band) forms.

The aim of this large multi-center retrospective study was to validate the efficacy of the prognostic factor score for predicting mortality and severe complications in patients with severe AP by comparing to the Acute Physiology and Chronic Health Evaluation II (APACHE II) score,¹⁸ which is the globally used severity assessment method.

Materials and methods

Study design

This retrospective multi-center cohort study was performed at 44 institutions in Japan. Twenty-five of the institutions were tertiary academic medical centers, and the other institutions were community hospitals with intensive care units. In each participating institution, all consecutive patients classified as severe AP within 72 h after admission according to the Japanese severity criteria for AP between 1 June 2009–31 December 2013 were enrolled.

Patient data, including demographic, clinical, laboratory, and radiological information, were retrospectively collected. All patients were followed up until hospital discharge.

For the purpose of this study, the total prognostic factor score and the APACHE II score were calculated for all enrolled patients at the diagnosis of severe AP. Based on calculated scores and clinical outcomes of all patients, we assessed the ability of the prognostic factor score to predict hospital mortality, the development of severe complications, the need for surgical interventions, and severe AP according to the revised Atlanta classification¹⁹ by comparing with the APACHE II score.

This study was a post-hoc analysis using data from a retrospective multicenter cohort study to evaluate the effectiveness of continuous regional arterial infusion of protease inhibitors for patients with severe AP.²⁰ This study was registered at the University Hospital Medical Information Network Clinical Trials Registry (registry number: 000012220), and approved by the Institutional Review Board or the Medical Ethics Committee at each institution. The study was conducted and reported in accordance with the Strengthening Reporting of Observational Studies in Epidemiology guidelines.²¹

Definitions

AP was diagnosed if the patient presented with at least two of the following three features: (a) acute abdominal pain and tenderness in the upper abdomen; (b) elevated levels of pancreatic enzymes in the blood and urine; or (c) abdominal findings of acute pancreatitis as detected by ultrasonography, CT, or magnetic resonance imaging.¹⁴ Severe AP, based on the Japanese severity criteria, was diagnosed when the total prognostic factor score was ≥3 or the contrast-enhanced CT grade was ≥2 (Table 1). When patients were diagnosed with mild pancreatitis at admission, we repeatedly evaluated the severity during the early stages of AP under the Japanese guideline.¹⁴

Severe complications during hospitalization included pulmonary failure, renal failure, and pancreatic infection. Pulmonary failure and renal failure were defined as a need for mechanical ventilation and for hemodialysis, respectively. Pancreatic infection was defined as the presence of bacteria based on blood culture or local culture obtained by percutaneous, image-guided, or endoscopic fine-needle aspiration, or the presence of extra-luminal gas in the pancreatic and/or peripancreatic tissues on contrast-enhanced CT.¹⁹ Surgical interventions included percutaneous, endoscopic, laparoscopic, or laparotomy drainage, necrosectomy for infected acute necrotic collection, or walled-off necrosis, interventional radiology, or endoscopic treatment for bleeding.

Statistical analysis

The relationships between the prognostic factor score and hospital mortality were determined by univariate logistic regression analysis. To evaluate the predictive accuracy of scorings systems for each outcome, receiver-operating characteristic (ROC) curves were generated, and the area under the curve (AUC) with 95% confidence interval (CI) was calculated. The Japanese prognostic factor score was compared with the APACHE II score by Delong's test.²² A p-value < 0.05 was considered statistically significant. All analyses were conducted using SPSS (r) Version 23.0 (IBM Corp, Armonk, New York, USA).

Results

Patient characteristics

A total of 1159 patients diagnosed with severe AP were included in this study. The demographic and clinical information, AP severity, treatments, and patient outcomes are shown in Table 2. The median length of hospital stay duration was 24 days (interquartile range, 14–42).

Table 2.

Characteristics of severe acute pancreatitis patients analyzed in this study.

	All patients (n = 1159)
Age, mean (SD), years	59.0 (17.4)
Men, no. (%)	780 (67.3)
Etiologic factor
Alcohol, no. (%)	456 (39.3)
Cholelithiasis, no. (%)	241 (20.8)
Idiopathic, no. (%)	240 (20.7)
Post-ERCP, no. (%)	110 (9.49)
Hyperlipidemia, no. (%)	26 (2.24)
Miscellaneous, no. (%)	86 (7.42)
Prognosis factor score, mean (SD)	3.09 (2.25)
CECT grade
Grade 1, no. (%)	143(12.3)
Grade 2, no. (%)	706 (60.9)
Grade 3, no. (%)	285 (22.9)
CECT was not performed	25 (2.16)
APACHE II score, mean (SD)	13.2 (7.92)
Revised Atlanta classification, no. (%)
Mild acute pancreatitis	327 (28.2)
Moderately severe acute pancreatitis	425 (36.7)
Severe acute pancreatitis	406 (35.0)
Unknown	1 (0.09)
Treatments
Mechanical ventilation, no. (%)	358 (30.9)
Renal replacement therapy, no. (%)	174 (15.0)
Outcome
Hospital mortality, no. (%)	150 (12.9)
Pancreatic infection, no. (%)	143 (12.3)
Surgical intervention, no. (%)	179 (15.4)

APACHE: acute physiology and chronic health evaluation; CECT: contrast enhanced CT; ERCP: endoscopic retrograde cholangiopancreatography; SD: standard deviation.

Regarding the primary outcome, 150 patients (12.9%) died during hospitalization. During the clinical course of AP, 358 patients (30.9%) required mechanical ventilation due to pulmonary failure. Hemodialysis for acute renal failure was performed in 174 patients (15.0%) without pre-existing renal disease. Pancreatic infection occurred in 143 patients (12.3%), and surgical interventions were required in 179 patients (15.4%).

Relationship between the prognostic factor score and the revised Atlanta classification

Table 3 shows the relationship between the prognostic factor score and the revised Atlanta classification.¹⁹ As the prognostic factor score increased, the prevalence of severe AP according to the revised Atlanta classification increased.

Table 3.

Relationship between prognostic factor score and revised Atlanta classification.

Prognostic factor score	Revised Atlanta classification			Total
Prognostic factor score	Mild AP	Moderately severe AP	Severe AP	Total
0	85 (55.5%)	60 (38.2%)	8 (5.2%)	153
1	83 (47.4%)	79 (45.1%)	13 (7.4%)	175
2	78 (39.6%)	80 (40.6%)	39 (19.8%)	197
3	47 (26.6%)	80 (45.2%)	50 (28.3%)	177
4	22 (14.6%)	66 (43.7%)	63 (41.7%)	151
5	7 (6.5%)	31 (29.0%)	69 (64.5%)	107
6	4 (4.3%)	19 (20.2%)	71 (75.5%)	94
7	1 (1.6%)	5 (8.2%)	55 (90.2%)	61
8	0 (0%)	4 (11.8%)	30 (88.2%)	34
9	0 (0%)	1 (11.1%)	8 (88.9%)	9
Total	327	425	406	1158

AP: acute pancreatitis.

Mortality according to the prognostic factor score

The mortality according to the prognostic factor score is shown in Figure 1. The mortality for each prognostic factor score was as follows: 1.3% (2/153) for score = 0, 4% (7/176) for score = 1, 7.1% (14/197) for score = 2, 7.9% (14/177) for score = 3, 13.3% (20/151) for score = 4, 15% (16/107) for score = 5, 34% (32/94) for score = 6, 37.7% (23/61) for score = 7, 44.1% (15/34) for score = 8, and 77.8% (7/9) for score = 9. A significant relationship was observed between the prognostic factor score and the hospital mortality rate by univariate logistic regression analysis (p < 0.001). The odds ratio was 1.59 (95% CI, 1.46–1.74).

Figure 1.

Mortality rate according to prognostic factor score.

The ability of the prognostic factor score, APACHE II score, and combination score (prognostic factor score plus APACHE II score) to predict clinical outcomes

Figure 2 and Table 4 show ROC curves and AUCs with 95% CI, respectively. For predicting hospital mortality, the AUCs of the prognostic factor score were not significantly different from AUCs of APACHE II score (0.78 vs 0.80, p = 0.30), the need for renal replacement therapy (0.83 vs 0.82, p = 0.49) and surgical interventions (0.69 vs 0.66, p = 0.16). The AUCs of the prognostic factor score were significantly higher than the AUCs of APACHE II score for predicting the need for mechanical ventilation (0.84 vs 0.81, p = 0.03), development of pancreatic infection (0.73 vs. 0.68, p = 0.02), and severe AP according to the revised Atlanta classification (0.83 vs 0.80, p = 0.01).

Figure 2.

Receiver operator characteristic curves of different severity assessment methods for prediction of severe outcomes.

Table 4.

Area under the receiver operating characteristic curves of prognostic factor score and Acute Physiology and Chronic Health Evaluation (APACHE) II score for prediction of severe outcome.

	Prognostic factor score	APACHE II	p Value
Hospital mortality	0.78 (0.74–0.82)	0.80 (0.76–0.83)	0.30
Mechanical ventilator	0.84 (0.81–0.86)	0.81 (0.78–0.83)	0.03
Renal replacement therapy	0.83 (0.80–0.86)	0.82 (0.79–0.85)	0.49
Pancreatic infection	0.73 (0.69–0.77)	0.68 (0.63–0.72)	0.02
Surgical intervention	0.69 (0.65–0.73)	0.66 (0.62–0.70)	0.16
SAP (revised Atlanta classification)	0.83 (0.81–0.86)	0.80 (0.77–0.82)	0.01

SAP: severe acute pancreatitis.

The 95% confidence interval is shown in parentheses.

The AUC of the combination score (the prognostic factor score plus APACHE II score) was significantly higher than that of either the prognostic factor score or APACHE II score for predicting hospital mortality (0.81 vs 0.78, p = 0.02; 0.81 vs 0.80, p < 0.001, respectively). Similarly, the AUCs of the combination score for predicting other severe outcomes were significantly higher than AUCs of the APACHE II score alone, but were not significantly different from AUCs of the prognostic factor score alone (Supplementary Material, Tables 1 and 2).

In subgroup analysis investigating etiological differences, the AUCs of the prognostic factor score for predicting hospital mortality were 0.76 in the alcohol group, 0.79 in the cholelithiasis group, 0.78 in the idiopathic group, and 0.78 in the other etiologies group.

Sensitivity, specificity, positive likelihood ratio, and negatives likelihood ratio of the prognostic factor score and APACHE II score

From the ROC curve for predicting hospital mortality, prognostic factor score ≥ 4 and APACHE II score ≥14 were selected as best cut-off values to identify patients who were at risk for mortality, respectively. Using these cut-off values, sensitivity, specificity, positive likelihood ratio, and negatives likelihood ratio of the prognostic factor score and APACHE II score were calculated (Table 5). The prognostic factor score and APACHE II score had the similar accuracy for predicting hospital mortality.

Table 5.

Sensitivity, specificity, positive predictive value, positive likelihood rate, and negative likelihood rate of the prognostic factor score and Acute Physiology and Chronic Health Evaluation (APACHE) II score for prediction of hospital mortality using best cut-off values.

	Sensitivity	Specificity	Positive LR	Negative LR
Prognostic factore score (cut-off value ≥4)	0.75 (0.68–0.82)	0.66 (0.63–0.69)	2.22 (1.96–2.51)	0.37 (0.28–0.49)
APACHE II (cut-off value ≥14)	0.83 (0.77–0.89)	0.65 (0.62–0.68)	2.34 (2.09–2.61)	0.27 (0.19–0.38)

LR: likelihood rate.

The 95% confidence interval is shown in parentheses.

Discussion

In this large multicenter study, the ability to predict hospital mortality and severe complications using the prognostic factor score in the Japanese severity criteria for AP was comparable to the APACHE II score. The prognostic factor score has the advantage of assessment simplicity compared with the APACHE II score, and therefore, appears to be a more practical triage tool during the early stages of AP.

Treatments for severe AP, including close monitoring of vital signs, fasting, fluid resuscitation, and analgesic administration should be initiated immediately after establishing the diagnosis. It is also extremely important to correctly assess AP severity to determine the need for early transfer to specialty care centers,^7,8,14,23 because intensive care based on rigorous physiological monitoring and specific therapeutic procedures are associated with improved survival of severe AP patients. For this reason, the Japanese severity criteria were developed to manage patients with AP during the early stage of admission and used for clinical evaluations and public financial support in Japan. In the current study, we evaluated the accuracy of the prognostic factor score in the Japanese severity criteria for predicting disease prognosis.

The prognostic factor score consists of nine clinical and biochemical parameters that can be easily obtained in most local hospitals. In this scoring system, each parameter is assessed using a binary score (positive: 1; negative: 0), and the sum of all positive parameters is represented as the total score using a scale of 0–9.¹⁴

In the current study, we demonstrated that a higher prognostic factor score was strongly correlated with higher mortality. These results provide strong evidence that the prognostic factor score can be used to stratify AP patients accurately according to their risk of mortality. To further evaluate the ability of the prognostic factor score to predict mortality, the AUC for the prognostic factor score was compared with that for the APACHE II score. The prognostic factor score had a reasonably high AUC of 0.78, which was comparable to the AUCs for the APACHE II score (0.80). In previous studies, the AUCs for the prognostic factor score ranged from 0.798–0.822, and the AUC for the prognostic factor score was higher than the Ranson’s score, Glasgow score, and APACHE II score.^16,24 From these results, the prognostic factor score can be considered non-inferior to international prognostic evaluation scores for predicting mortality.

The development of organ failure during the early phase of AP is associated with an increased risk of mortality.^19,25,26 In addition, infected pancreatic necrosis during the late phase of AP and undergoing surgical interventions are associated with a significant mortality risk.²⁷ Therefore, it is also meaningful to evaluate the accuracy of the severity assessment scores for predicting the occurrence of these fatal conditions in addition to mortality. In the present study, organ failure included pulmonary failure and renal insufficiency. The AUCs of the prognostic factor score for predicting the need for mechanical ventilation, renal replacement therapy due to renal failure, pancreatic infection, and surgical interventions were 0.84, 0.83, 0.73, and 0.69, respectively, which were comparable to or partially greater than the AUC for the APACHE II score. In a previous cohort study as well, the prognostic factor score was found to be the best scoring system for the prediction of persistent organ failure at 48 h of hospitalization among existing scoring systems.¹⁵

The APACHE II score, which was originally designed to assess the severity of patients with acute illness admitted to intensive care units in the 1970s, is still the most widely used scoring system for severity assessment. It has been selected as a reference standard in numerous studies to evaluate proposed prognostic scoring systems or to identify individual risk factors for severe outcomes.¹⁸ However, the APACHE II score requires the evaluation of 14 parameters with weighting based on the degree of deviation from reference values using the worst data during initial 24 h after admission. Therefore, the scoring system needed 24 h for the final determination of AP severity and is complex and difficult to use in clinical practice. Furthermore, the scoring system is not specific to AP and includes some parameters that are not directly associated with the severity of AP, such as serum sodium and potassium. In a previous study using a large sample of population-based data, only 2.2% of included patients with AP had complete data for the APACHE II classification.²⁸ The ideal classification system for identifying patients at high risk of mortality should be easy to calculate, consisting of readily available parameters, and accurate.^29,30 According to our results, the ability of the prognostic factor score for predicting mortality was comparable to the APACHE II score despite fewer and simpler evaluation parameters. To the best of our knowledge, this was the first large multicenter study in which the prognostic factor score was compared with another scoring system.^15–17

For prediction of mortality, the combination score (prognostic factor score plus APACHE II score) was significantly better than the prognostic factor score alone, while for prediction of other severe outcomes, the combination score was not significantly different from prognostic factor score alone. For practical purposes, the combination of prognostic factor score and APACHE II score is cumbersome to use due to increased number of assessment parameter, and therefore its utility will be limited.

In our study with missing data of contrast-enhanced CT in 41 patients, we avoided a direct comparison of the AUCs of prognostic factor score with the AUCs of contrast-enhanced CT grade. Contrast-enhanced CT is an accurate and effective method to evaluate extra-pancreatic extensions of inflammation and pancreatic necrosis. In particular, the identification of pancreatic necrosis on contrast-enhanced CT during an early stage of AP has been reported to be an accurate predictive indicator for severe outcomes.^31,32 In the Japanese guidelines for the management of AP, however, the classification of the contrast-enhanced CT grade is not mandatory for the severity assessment in an early phase of AP, but rather, the severity assessment on the basis of the prognostic factor score is recommended.³³ Therefore, we conducted this retrospective study with focusing on evaluating the predictive value of prognostic factor score.

This study had several limitations. Firstly, our study subject was limited to severe AP diagnosed by the Japanese severity criteria. This was because our primary goal was to determine the ability of the prognostic factor score for predicting hospital mortality, while the mortality rate of mild AP was extremely row.⁴ Therefore, it was reasonable to assess the efficacy of the prognostic factor score in the severe AP population. Secondly, although patients were treated according to Japanese guidelines for the management of AP,¹⁴ there may be differences in the treatment strategies of AP and the indication for surgical or endoscopic interventions of pancreatic infection depending on participating institutions, which might have affected patient outcomes. Thirdly, prognosis factor scores were calculated by individual institutions, not by the data center. However, since data collection and severity stratification in the prognosis factors were quite simple as mentioned above, the data quality was considered to be adequate for the analyses. Fourthly, in this retrospective study, it was impossible to evaluate predictive ability for the development of cardiovascular complications because we did not collect clinical data on the assessment of cardiovascular complications according to the modified Marshall scoring system used in the revised Atlanta classification.³⁴

In conclusion, the prognostic factor score has an equivalent ability for predicting mortality compared with the APACHE II score. Regarding the ability for predicting the development of severe complications during the clinical course of AP, the prognostic factor score may be superior to the APACHE II score. The prognostic factor score is designed to allow clinicians to conduct evaluations based on readily available parameters and, therefore, appears to be a more practical tool than the APACHE II score.

Footnotes

Acknowledgments

The authors would like to acknowledge Yukiko Masuda, Natsuko Tokuhira, Tsuyoshi Takeda, Seiya Suzuki, Jun Kataoka, Tomohiro Adachi, Shin Namiki, Sakue Masuda, Tomoaki Hashida, Naoki Shinyama, Hitoshi Yamamura, Takashi Moriya, Kunihiro Shirai, Kazuo Inui, Takashi Muraki, Junichi Sakagami, Hiroaki Yasuda, Yoshinori Azumi, Masayuki Kamochi, Keiji Nagata, Nobuyuki Saito, Mizuki Sato, Kyohei Miyamoto, Mioko Kobayashi, Koji Saito, Shinjiro Saito, Junko Izai, Kazunori Takeda, Motohiro Sekino, Tomoki Furuya, Yoshimoto Seki, Tetsuya Mine, Youhei Kawashima, Naoyuki Matsuda, Masato Inaba, Mineji Hayakawa, Takuyo Misumi and Yuki Takeda with the support of the data collection at 44 institutions: Osaka Saiseikai Senri Hospital, Hiroshima City Hiroshima Citizens Hospital, Kansai Medical University Hirakata Hospital, The University of Tokyo Hospital, Iizuka Hospital, Japanese Red Cross Musashino Hospital, Tokyo Metropolitan Tama Medical Center, Japanese Redcross Maebashi Hospital, Shonan Kamakura General Hospital, Showa University Hospital, Nihon University Hospital, Saiseikai Kumamoto Hospital, Fukuyama City Hospital, Jichi Medical University Hospital, Chiba University Hospital, Osaka City University Hospital, Tohoku University Hospital, Nihon University Itabashi Hospital, Gifu University Hospital, Second Teaching Hospital, Fujita Health University, Asahi General Hospital, Shinshu University Hospital, National Hospital Organization Nagasaki Medical Center, University Hospital, Kyoto Prefectural University of Medicine, Mie University Hospital, Hospital of the University of Occupational and Environmental Health, Nippon Medical School Chiba Hokusoh Hospital, Jichi Medical University Saitama Medical Center, Wakayama Medical University Hospital, Tokyo Metropolitan Bokutoh Hospital, Jikei University School of Medicine, Saka General Hospital, National Hospital Organization Sendai Medical Center, Nagasaki University Hospital, Keio University School of Medicine, Japanese Red Cross Akita Hospital, Ibaraki Prefectural Central Hospital, Tokai University Hospital, Nagoya University Hospital, Hokkaido University Hospital, National Cancer Center, Akita City Hospital, Kobe University Hospital and Tokyo Rosai Hospital. They also wish to thank the Japanese Society of Education for Physicians and Trainees in Intensive Care, and the Japanese Society of Intensive Care Medicine.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Corfield

Cooper

Williamson

. Acute pancreatitis: A lethal disease of increasing incidence. Gut 1985; 26: 724–729.

Roberts

Akbari

Thorne

. The incidence of acute pancreatitis: Impact of social deprivation, alcohol consumption, seasonal and demographic factors. Aliment Pharmacol Ther 2013; 38: 539–548.

Fagenholz

Fernandez-del Castillo

Harris

. Direct medical costs of acute pancreatitis hospitalizations in the United States. Pancreas 2007; 35: 302–307.

Hamada

Masamune

Kikuta

. Nationwide epidemiological survey of acute pancreatitis in Japan. Pancreas 2014; 43: 1244–1248.

Wada

Takada

Hirata

. Treatment strategy for acute pancreatitis. J Hepatobiliary Pancreat Sci 2010; 17: 79–86.

Makhija

Kingsnorth

. Cytokine storm in acute pancreatitis. J Hepatobiliary Pancreat Surg 2002; 9: 401–410.

Werner

Feuerbach

Uhl

. Management of acute pancreatitis: From surgery to interventional intensive care. Gut 2005; 54: 426–436.

Working Group IAP/APA Acute Pancreatitis Guidelines. IAP/APA evidence-based guidelines for the management of acute pancreatitis. Pancreatology 2013; 13: e1–e15.

Tenner

Baillie

DeWitt

. American College of Gastroenterology guideline: Management of acute pancreatitis. Am J Gastroenterol 2013; 108: 1400–1415.

10.

Mitchell

RMS

Byrne

Baillie

. Pancreatitis. Lancet 2003; 361: 1447–1455.

11.

Forsmark CE and Baillie J. AGA Institute technical review on acute pancreatitis. Gastroenterology 2007; 132: 2022–2044.

12.

Tsujino

Isayama

Komatsu

. Risk factors for pancreatitis in patients with common bile duct stones managed by endoscopic papillary balloon dilation. Am J Gastroenterol 2005; 100: 38–42.

13.

Mayerle

Hlouschek

Lerch

. Current management of acute pancreatitis. Nat Clin Pract Gastroenterol Hepatol 2005; 2: 473–483.

14.

Yokoe

Takada

Mayumi

. Japanese guidelines for the management of acute pancreatitis: Japanese guidelines 2015. J Hepatobiliary Pancreat Sci 2015; 22: 405–432.

15.

Mounzer

Langmead

. Comparison of existing clinical scoring systems to predict persistent organ failure in patients with acute pancreatitis. Gastroenterology 2012; 142: 1476–1482; quiz e15–e16.

16.

Ueda

Takeyama

Yasuda

. Utility of the new Japanese severity score and indications for special therapies in acute pancreatitis. J Gastroenterol 2009; 44: 453–459.

17.

Senol

Gundogdu

Ozkan

. External validation of the new Japanese severity score in Turkish patients with acute pancreatitis. Pancreas 2014; 43: 487–488.

18.

Knaus

Draper

Wagner

. APACHE II: A severity of disease classification system. Crit Care Med 1985; 13: 818–829.

19.

Banks

Bollen

Dervenis

. Classification of acute pancreatitis–2012: Revision of the Atlanta classification and definitions by international consensus. Gut 2013; 62: 102–111.

20.

Horibe M, Sasaki M, Sanui M, et al. Continuous regional arterial infusion of protease inhibitors has no efficacy in the treatment of severe acute pancreatitis: A retrospective multi-center cohort study. Pancreas. In Press.

21.

Von Elm

Altman

Egger

. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Int J Surg 2014; 12: 1495–1499.

22.

DeLong

Clarke-Pearson

. Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach. Biometrics 1988; 44: 837–845.

23.

Working Party of the British Society of Gastroenterology; Association of Surgeons of Great Britain and Ireland; Pancreatic Society of Great Britain and Ireland; Association of Upper GI Surgeons of Great Britain and Ireland. UK guidelines for the management of acute pancreatitis. Gut 2005; 54 Suppl 3: iii1–iii9.

24.

Hamada

Yasunaga

Nakai

. Japanese severity score for acute pancreatitis well predicts in-hospital mortality: A nationwide survey of 17,901 cases. J Gastroenterol 2013; 48: 1384–1391.

25.

Petrov

Shanbhag

Chakraborty

. Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology 2010; 139: 813–820.

26.

Yang

Chen

Phillips

. Predictors of severe and critical acute pancreatitis: A systematic review. Dig Liver Dis 2014; 46: 446–451.

27.

Mofidi

Lee

Madhavan

. Prognostic factors in patients undergoing surgery for severe necrotizing pancreatitis. World J Surg 2007; 31: 2002–2007.

28.

Johannes

Sun

. The early prediction of mortality in acute pancreatitis: A large population-based study. Gut 2008; 57: 1698–1703.

29.

Papachristou

Muddana

Yadav

. Comparison of BISAP, Ranson's, APACHE-II, and CTSI scores in predicting organ failure, complications, and mortality in acute pancreatitis. Am J Gastroenterol 2010; 105: 435–441; quiz 442.

30.

Brown

James-Stevenson

Dyson

. The Panc 3 score: A rapid and accurate test for predicting severity on presentation in acute pancreatitis. J Clin Gastroenterol 2007; 41: 855–858.

31.

Casas

Diaz

Valderas

. Prognostic value of CT in the early assessment of patients with acute pancreatitis. AJR Am J Roentgenol 2004; 182: 569–574.

32.

Hirota

Satoh

Kikuta

. Early detection of low enhanced pancreatic parenchyma by contrast-enhanced computed tomography predicts poor prognosis of patients with acute pancreatitis. Pancreas 2012; 41: 1099–1104.

33.

Takeda

Yokoe

Takada

. Assessment of severity of acute pancreatitis according to new prognostic factors and CT grading. J Hepatobiliary Pancreat Sci 2010; 17: 37–44.

34.

Marshall

Cook

Christou

. Multiple organ dysfunction score: A reliable descriptor of a complex clinical outcome. Crit Care Med 1995; 23: 1638–1652.