Predicting Hospitalization, Organ Dysfunction, and Mortality in Post-Endoscopic Retrograde Cholangiopancreatography Acute Pancreatitis: Are SIRS and qSOFA Reliable Tools?

: Background: Post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) has shown constant incidence throughout time, despite advances in endoscopic technology, devices, or personal skills of the operating endoscopists, with prevention and prediction of severity in PEP being constant concerns. Several prospective studies have investigated the role of systemic inﬂammatory response syndrome (SIRS) criteria or the quick Sequential Organ Failure Assessment (qSOFA) score in the PEP severity assessment. However, there are no clearly deﬁned tools for the prediction of PEP severity. Methods: A total of 403 patients were prospectively monitored 60 days after ERCP for the detection of PEP development. Consequently, we evaluated the lengths of stay, incidence of organic dysfunction, and mortality rates of these patients. The predictive power of the univariate model was evaluated by using the receiver operating characteristic curve and analyzing the area under the curve (AUC). Results: Incidence of PEP was similar to that reported in the majority of trials. The 60-day survival rate of PEP patients reached 82.8%. A qSOFA score ≥ 1 is a very good predictor for organ dysfunction (AUC 0.993, p < 0.0001). SIRS can also be considered a signiﬁcant predictor for organic dysfunctions in PEP patients (AUC 0.926, p < 0.0001). However, only qSOFA was found to signiﬁcantly predict mortality in PEP patients (AUC 0.885, p = 0.003), with SIRS criteria showing a much lower predictive power. Neither SIRS nor qSOFA showed any predictive value for the length of stay of PEP patients. Conclusion: Our study offers novel information about severity prediction in PEP patients. Both SIRS criteria and qSOFA showed good predictive value for organic dysfunction, mortality, and hospitalization.


Introduction
Incidence and severity of adverse events (AEs) after endoscopic retrograde cholangiopancreatography (ERCP) are undoubtfully different from those that may follow other therapeutic endoscopy procedures.Post-ERCP pancreatitis (PEP) has long been described as the most frequent post-procedural AE, resulting in relatively high morbidity and mortality rates.These are followed by increased lengths of hospital stay, thus triggering a heavy burden on healthcare-associated costs [1][2][3].Nevertheless, evaluation and quantification of such burden has been hard to achieve due to the ongoing controversy regarding definition of the AE itself, its follow-up protocols, and its severity assessment.
Currently, PEP is defined as a new or worsened abdominal pain combined with more than 3 times the normal value of amylase or lipase concentrations at more than 24 h after ERCP and a requirement of hospital admission or prolongation of a planned admission [4].The proposed definition of PEP has been introduced by Cotton et al. [5] and it has since been used in most trials and published papers.Based on this definition, the reported overall incidence of PEP ranges between 3% and 10%, from most of the systematic reviews and meta-analyses [2,3,6].The majority of PEP patients develop a mild case, with an overall mortality rate of under 1% [1,2].
Despite its relatively high incidence, there has yet been no specifically designed tool for the evaluation of PEP severity nor of its impact on the length of hospital stay and inpatient mortality rates.Although not specifically designed for PEP, the revised Atlanta Classification of Acute Pancreatitis Consensus [7] managed to stratify pancreatitis severity based on the presence and duration of organ failure rather than duration of hospitalization and may provide an alternative for assessing PEP severity and mortality.
Consequently, given such a potential gap in the workup of ERCP patients, our aim was to evaluate whether such parameters as the length of stay, incidence of organic dysfunction, or mortality could be predicted by standardized indicators like the systemic inflammatory response syndrome (SIRS) or the quick Sequential Organ Failure Assessment (qSOFA) score.Investigating SIRS follows several systemic parameters associated with inflammation including the heart and respiratory rates, body temperature, and white blood cells.On the other hand, qSOFA assessment consists of evaluating three components respiratory rate, change in mental status, and systolic blood pressure, which have been associated with a high risk for organ dysfunction.

Selection of Patients
A cohort of 403 patients was included in the study.All patients had received therapeutic ERCP procedures between the 1 January 2018 and the 31 August 2018, within the Institute of Gastroenterology and Hepatology at St. Spiridon Regional Emergency Hospital of Ias , i, Romania.General patient characteristics are shown in Table 1.All patients were prospectively monitored, starting at admission before ERCP until 60 days after the procedure.All patients included in the study underwent the therapeutic ERCP procedures to address bile duct obstructions.Patients with native papillae as well as those with previous papillotomy were included.Exclusion criteria were as follows: age below 18 years, indication of ERCP for pancreatic duct (PD) conditions, lack of patient availability for the 60 days follow-up, and impossibility of achieving all monitored parameters.Additionally, all patients with an altered qSOFA before ERCP were excluded for methodological purposes.All patients had signed an informed consent for both study enrollment and the ERCP procedure.The study was approved by the Ethical Committee of the Grigore T. Popa University of Medicine and Pharmacy of Ias , i, Romania.

Procedures and Devices
All procedures were performed within the same endoscopy suite.Conventional Olympus duodenoscopes (models TJF 160 and 180; Tokyo, Japan) were used for all procedures.Fluoroscopy was performed with a standard C-arm device (General Electrics, Frankfurt am Main, Germany).All procedures were attended by a radiology technician, and the total radiation exposure was registered for all patients (mGy/min) but was not included in the follow-up protocol.
Devices used for the ERCP procedures included the following: (1) triple lumen sphincterotomes V-System™ CleverCut3V (Olympus Europa, Hamburg, Germany); (2)  All of the ERCP procedures were carried out by a board-qualified ERCP endoscopist.A group of four endoscopists managed all included cases.Trainees were involved in patient preparation and follow-up.In patients where biliary clearance was not achieved, or with Bismuth type 2 to 4, with primary sclerosing cholangitis, or in immunocompromised cases, antibiotic prophylaxis was administered after local protocols and local antibiotic resistance, as recommended by current guidelines [8,9].All patients received intrarectal diclofenac or indomethacin for the prevention of PEP as required by local protocol.
Indications for ERCP for the studied cohort included various causes of bile duct obstruction: gallstone disease, benign and malignant strictures and extrinsic compression of bile ducts, bile duct leaks and post-operative injury, and hydatid disease.Insertion of biliary stents was performed for stricture management in patients with bile duct leaks and in cases of remnant bile duct stones and clearance failure.

Patient Evaluation before and during the Procedure
Pre-ERCP general evaluation of patients included the following parameters: indication for ERCP, personal history of ERCP procedures and the presence of in situ biliary stents, urgent vs elective, and curative vs palliative type of indication.The type of imagistic technique confirming the diagnosis (computed tomography, magnetic resonance imaging, direct cholangiography, or ultrasonography) was also included in the study.Laboratory workup included the following parameters: complete blood count, total and direct bilirubin (mg/dL), alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), alkaline phosphatase (U/L), C-reactive protein (CRP) (mg/dL), and international normalized ratio (INR).All patients were classified with the American Society of Anesthesia (ASA) score [10].Antibiotic prophylaxis, ASA score, and the type of anesthesia were also included as studied parameters.
Intra-procedural indicators included the following: anatomy of the major papilla [11], presence of duodenal diverticula, number of cannulation attempts, total time for cannulation, involuntary PD guidewire insertion or PD contract opacification, use of needle knife pre-cut papillotomy, fistulotomy or trans-pancreatic access sphincterotomy, and presence of a difficult cannulation or lack of achieving cannulation.All patients with difficult cannulation where there was guidewire access in the PD received prophylactic pancreatic stents (5 Fr, 3 cm to 5 cm).The definition of difficult cannulation was made after current recommendations of the European Society of Gastrointestinal Endoscopy (ESGE) [12], as follows: more than three cannulation attempts, more than 5 min of contact with the major papilla, and more than one involuntary PD guidewire insertion or PD contrast opacification.

Patient Post-ERCP Follow-Up
All patients had a minimum of 3 days of hospital admission, and a minimum of 24 h of in-hospital follow-up after each procedure.Clinical and biological workups were performed for all patients at 24 h after the ERCP procedures.The 24-h biological panels for all patients included complete blood count, serum lipase (mg/dL), CRP (mg/dL), and serum creatinine (mg/dL); natrium, potassium, chlorine, and bicarbonate levels (mmol/L); serum lactate (mg/dL), total and direct bilirubin (mg/dL), and glycaemia (mg/dL).When perforations were suspected, 24-h post-ERCP computed tomography scans were carried out.
The presence of post-ERCP SIRS and the qSOFA were monitored.The definition of SIRS included the presence of two or more of the following items at 24 h after the ERCP [13,14]: respiratory rate > 20/min; peripheral body temperature > 38 • C or <36 • C; heart rate > 90/min; and white blood cells > 12,000/mm 3 , or <4000/mm 3 , or >10% immature peripheral white cells.Calculation of qSOFA required the presence of the following parameters at 24 h post-ERCP [15]: systolic blood pressure ≤ 100 mmHg, respiratory rate > 22/min, or an altered mental status (Glasgow comma scale < 13).In patients with more than 24 h of post-ERCP hospitalization, monitored parameters included complete blood count and CRP levels (mg/dL) at the patient's release.
The survival rate was evaluated for all patients at 15 d and 60 d after the procedures.Consequently, given the follow-up protocol, the following parameters were monitored and included in the study: incidence of post-ERCP complications (24 to 48 h after ERCP), total hospitalization length (d), number and type of organic dysfunctions occurring within 24 to 48 h after ERCP, 15-day interval mortality, and 60-day interval mortality.
Organic dysfunctions have been defined for respiratory, hematological, hepatic, cardiovascular, neurological, and renal systems.Since arterial gas sampling is not routinely collected in our gastroenterology and endoscopy unit, oxygen arterial pressure (PaO 2 ) could not be monitored.Thus, oxygen saturation was measured (SpO 2 ) by pulse oximeter readings to calculate the SpO 2 /FiO 2 ratio, which was used to define respiratory dysfunction.Organic dysfunctions were therefore defined as follows: (a) respiratory system SpO 2 /FiO 2 below 500; (b) renal system: rise in serum creatinine ≥0.3 mg/dL within 48 h, or rise in serum creatinine ≥1.5 times baseline 48 h after ERCP, or urine output <0.5 mL/kg/h for six hours; (c) nervous system a decrease in Glasgow comma scale below or equal 13 after 24 h post-ERCP; (d) cardiovascular system: mean arterial pressure below 70 mmHg or new need of any dose of vasopressor infusion; and (e) hematology: platelet decrease below 150,000/mL after ERCP.Liver dysfunction was not monitored as such a parameter could be highly influenced by pre-existent bile duct obstruction, which serves as main indication for ERCP procedures.

Definition of Post-ERCP Complications
We defined PEP after the suggestion of the ESGE, as follows: detection of new or worsened abdominal pain combined with more than 3 times the normal value of lipase at 24 h after ERCP and the requirement of admission or prolongation of a planned admission.The severity of PEP was evaluated at 48 h to 72 h after detection but was not included in the prospective study report [4].
Other post-ERCP AEs were defined according to the 2010 lexicon of definitions proposed in 2010 by Cotton et al. [5] for the American Society of Gastrointestinal Endoscopy (ASGE).Post-ERCP bleeding was classified as either covert or overt bleeding.We defined post-ERCP bleeding as hematemesis and/or a melena or hemoglobin drop more than 2 g/dL following an ERCP procedure [16], with subsequently confirmed papillary bleeding or haemobilia.Severity of post-ERCP bleeding was monitored but was not included in the report for the present study.
Post-ERCP perforations were defined as the presence of gas or luminal contents outside of the gastrointestinal tract, as determined by computed tomography scans within 48 h after the procedure [16].Type of post-ERCP perforation was monitored but was not included in the report for the present study.Monitored cardiopulmonary AEs were as follows: (1) intraor post-procedural hypoxemia; hemoglobin oxygen saturation below 85%; and (2) arterial hypotension or hypertension, defined as either a blood pressure value below 90/50 mmHg or above 190/130 mmHg, or a change in value (down or up) to 20% [1,4,16].
Post-ERCP cholecystitis and cholangitis were defined according to the revised Tokyo Guidelines 2018.Post-ERCP cholecystitis was defined as right upper quadrant pain (or positive Murphy sign), systemic signs of inflammation, and imaging findings characteristic of acute cholecystitis, without any suggestive clinical or imaging findings prior to ERCP [17].Post-ERCP cholangitis was defined as the onset of fever and/or shaking chills and/or evidence of an inflammatory response combined with onset or worsening of cholestasis 24 h after ERCP [18].

Statistical Analysis and Report of Results
Statistical analysis was performed using SPSS v.25 software.Continuous variables are reported as mean values and standard deviation if the distribution was normal and as median and quartiles if the variable did not have a normal distribution.Comparisons were performed using the Student's t-test, ANOVA, and Kruskal-Wallis or Mann-Whitney U test for continuous variables.Homogeneity of variance across groups was evaluated using Levene's test.Correlations were tested using Pearson's test and the r correlation coefficient.
Qualitative variables are presented as their absolute (n) and relative (%) frequencies.Comparisons between groups were made using non-parametrical tests like M-L, Yates, or Pearson chi-square tests.Univariate and multivariate analyses of prognostic factors were performed using the logistic regression model.The predictive power of the univariate model was evaluated using the receiver operating characteristic (ROC) curve and analyzing the area under the curve (AUC).
Survival was estimated using Kaplan-Meier survival curves, which were then compared using Cox's F-test and the log-rank test (LRT).Statistical significance of tests is expressed as a p value, which when less than 0.05 was considered to indicate statistical significance.

Results
Patients included in the prospective cohort had been diagnosed with the following conditions: bile duct stones, choledochal cysts; cholangiocarcinoma, benign biliary strictures, post-operatory bile duct injury, chronic pancreatitis and pancreatic cancer, compressive pancreatic cystic lesions, tumors of the major papilla, and suspected type I sphincter of Oddi dysfunction.The latter could not be confirmed due to the lack of local availability of sphincter of Oddi manometry.
Within our cohort, we found no significant correlation between patients' background parameters and post-ERCP pancreatitis as seen in Table 1.
As shown in Table 2, overall complication rates slightly overarched 10%, of which the most frequent AEs were post-ERCP infections-acute cholangitis and cholecystitis (8.93%), followed by PEP (8.68%), post-sphincterotomy bleeding, and perforations occurring in a clearly lower frequency.Moreover, we evaluated the association between PEP and other concomitant adverse events, as stated in Table 3.Some patients developed more than one AE.Correlations between PEP, post-sphincterotomy bleeding, and post-ERCP infections are shown in Table 3.The highest significant correlation was registered between PEP and infections, with an incidence of 0.99%, followed by 0.5% of patients who developed PEP and post-sphincterotomy bleeding and 0.25% of patients who experienced post-ERCP infections and bleeding.Only three perforation cases occurred in the prospective cohort, one of which developed PEP, and another developed post-sphincterotomy bleeding.The vast majority of patients with post-ERCP AEs had only one type of complication.As shown in Table 3, the mean length of stay of the cohort was 13 days, being significantly lower than that of PEP patients, which reached 16 days (Z = 3.73, p = 0.0001).
The survival rate of patients with PEP was estimated using the Kaplan-Meier method after the 60 d of prospective follow-up and direct survival analysis.The life table analysis is shown in Table 4. Subsequently, the Kaplan-Meier curve was drawn (Figure 1), showing a statistically significant decline in the survival rate of patients with PEP when compared to the other patients in the prospective cohort (82.8% vs. 93.2%,p = 0.0387, LRT = −2.185,p = 0.0288).A statistically significant decline was shown in the survival rate of patients with PEP when compared to the prospective cohort (p = 0.0288).
Subsequently, we searched whether the presence of SIRS and qSOFA could predict organic dysfunctions, length of stay, and survival rates in patients with PEP.As shown in Table 5, the predictive power of such indicators was evaluated using ROC curves and study of the AUCs.A statistically significant decline was shown in the survival rate of patients with PEP when compared to the prospective cohort (p = 0.0288).
Subsequently, we searched whether the presence of SIRS and qSOFA could predict organic dysfunctions, length of stay, and survival rates in patients with PEP.As shown in Table 5, the predictive power of such indicators was evaluated using ROC curves and study of the AUCs.AUC analyses showed that in both PEP and non-PEP patients, a qSOFA score ≥ 1 was a very good predictor for organ dysfunction (AUC 0.993, p < 0.0001).Moreover, despite a lower predictive precision when compared to qSOFA, SIRS could also be considered a significant predictor for organic dysfunctions in PEP patients (AUC 0.926, p < 0.0001).However, only qSOFA was shown to significantly predict mortality in PEP patients (AUC 0.885, p = 0.003), with SIRS criteria showing a much lower prediction power.Neither SIRS nor qSOFA showed any predictive value for the length of stay of PEP patients.The data of ROC curve analyses for all parameters can be found in Figure 2.

Discussion
Given their proven applications as criteria for the definition of sepsis and sepsis-related organ dysfunctions, both SIRS and qSOFA could be useful markers in predicting the severity of post-ERCP AEs, especially of PEP.However, to date, only few prospective studies have investigated such an extended use of these criteria in order to identify possible alternative severity and prognostic markers of PEP.
Regarding PEP, large clinical trials have aimed to refine its definition and severity scaling but with no clear consensus, mainly due to the variability of hospital stay and discharge criteria [19], the exact timing for the measurement of pancreatic enzymes, and their diagnostic value [1,4,16,20].Pre-existing pain, organ dysfunction, or sepsis due to acute biliary pancreatitis or cholangitis further influence the definition and prediction of PEP severity [4,16,20].The Atlanta definition has not been embraced so far, probably because it requires pancreas imaging [10].Hence, identification of predictive tools for PEP severity seems mandatory.
Incidence of PEP in our prospective cohort is similar to that reported in most of the previous trials and meta-analyses [1,4,21].For the PEP severity assessment, practice guidelines of ASGE suggest grading severity after the length of hospital stay and the presence of local complications [1], while the ESGE recommends staging severity mainly after

Discussion
Given their proven applications as criteria for the definition of sepsis and sepsisrelated organ dysfunctions, both SIRS and qSOFA could be useful markers in predicting the severity of post-ERCP AEs, especially of PEP.However, to date, only few prospective studies have investigated such an extended use of these criteria in order to identify possible alternative severity and prognostic markers of PEP.
Regarding PEP, large clinical trials have aimed to refine its definition and severity scaling but with no clear consensus, mainly due to the variability of hospital stay and discharge criteria [19], the exact timing for the measurement of pancreatic enzymes, and their diagnostic value [1,4,16,20].Pre-existing pain, organ dysfunction, or sepsis due to acute biliary pancreatitis or cholangitis further influence the definition and prediction of PEP severity [4,16,20].The Atlanta definition has not been embraced so far, probably because it requires pancreas imaging [10].Hence, identification of predictive tools for PEP severity seems mandatory.
Incidence of PEP in our prospective cohort is similar to that reported in most of the previous trials and meta-analyses [1,4,21].For the PEP severity assessment, practice guidelines of ASGE suggest grading severity after the length of hospital stay and the presence of local complications [1], while the ESGE recommends staging severity mainly after the presence and length of organ dysfunction [4].Thus, we designed our research so that it could evaluate prediction of both length of stay and organ dysfunction.The cumulative 60-day mortality of PEP patients in our prospective cohort reached 17%, which is similar to figures presented in a recently published systematic review of PEP cases [3].However, such a mortality rate is also within the mortality range of patients with allcause acute pancreatitis [22,23].The only reason that could explain such high mortality is the complexity of cases-our center is a tertiary one with an emergency setting of admittance.To date, there is no complexity scoring system in patients with bile duct obstructions-as such conditions refer to a relatively large number of diseases: benign and malignant strictures, regular and difficult bile duct stones, and post-operative bile duct injuries-all these being either associated or not with acute cholangitis.The only tool to demonstrate the complexity of cases would be the ASA scaling-more than a half of patients were classified as ASA 2-4.Moreover, ASA class is significantly associated with the incidence of post-ERCP adverse events.Notwithstanding, although less reproducible, the overall complexity of cases is also sustained by the relatively high length of stay especially when referring to an emergency hospital.
Both SIRS and qSOFA were tested within our prospective cohort to predict organ dysfunction and mortality in PEP patients.Persistent SIRS criteria have previously been shown to predict a poor prognosis in post-ERCP cholangitis patients [24].Furthermore, persistent SIRS has been linked to post-ERCP perforations but with no additional value on the prediction of organic dysfunction or mortality in such cases [25].Despite lack of similar results within our cohort, a recent previous study suggests that SIRS may predict overall length of stay in PEP patients [26].
Organ dysfunctions should be screened for before and after digestive interventional procedures, irrespective of the type of procedure [27].qSOFA has been designed as a tool in predicting sepsis as an alternative to SIRS, which lacked specificity [14,28].Thus, it has been previously linked to intraabdominal sepsis [29], being considered an independent risk factor for severity and an indirect predictor for the length of stay in cases of acute cholangitis [30].Persistent organ dysfunction is associated with severe PEP [7], and multiorgan dysfunction seems to be a relatively frequent finding in post-ERCP AEs [24].
In ERCP patients, qSOFA predicts severity of acute biliary pancreatitis and cholangitis [31], being considered the earliest predictor of sepsis in patients with acute cholangitis presenting for ERCP procedures [32].So far, there are no data within the literature to show any correlation between qSOFA and the severity of PEP or any other post-ERCP AE.However, similar to our results, both SIRS and qSOFA seem to predict severity and mortality in all-cause pancreatitis irrespective of ERCP procedures [33,34].
Post-ERCP AEs increase the overall length of hospital stay and in-hospital mortality rates to over 1% [35].Nevertheless, both length of stay and in-hospital mortality have been linked to low caseloads and limited ERCP experience of specific centers [36].However, despite the AEs, ERCP procedures have been shown to decrease overall mortality rates of referred patients [35].Our results show no correlation and thus no predictive power of either SIRS or qSOFA over the length of hospital stay in PEP patients.This fact could be explained by the lack of locally implemented criteria for discharging patients, resulting in disparities regarding length of stay in patients regardless of admittance diagnosis.
One potential alternative predictor for PEP severity is the CRP level.The CRP level has previously been correlated with not only PEP severity [37], but also with all-cause length of stay in patients with post-ERCP AEs [38].However, it seems that CRP levels may be highly influenced by pre-ERCP conditions, such as acute cholangitis, acute biliary pancreatitis, or even uncomplicated bile duct obstructions, thus generating potential bias in interpreting their predictive value for post-ERCP PEP severity [39].Other studied markers for PEP severity have predictive value, although it has not yet been clearly defined [39,40].
Despite its prospective value, our study is limited by a number of factors that require attention.First, the cohort is comprised of patients referred to a tertiary emergency unit; thus, the severity and complexity of cases are predominantly high, resulting in higher rates of post-procedural complicated evolutions of patients.This calls into question how the observed parameters would predict severity of PEP in regular cases and in the absence of numerous comorbidities.Such comorbidities along with longstanding and complex bile duct obstructions could influence the predictive values of SIRS and qSOFA.Second, due to the lack of clearly established discharge criteria after post-ERCP AEs, length of hospital stay may have been biased, and thus, no predictive values of the studied parameters could be found.Moreover, the high rate of post-ERCP infections and the association between infectious AEs and PEP could have biased the true predictive potential of both SIRS and qSOFA when related solely to PEP.Subsequently, the lack of alternately studied scoring systems and biomarkers, such as those included in the revised Atlanta classification of acute pancreatitis, is among the most important limitations of our study, but their workup and follow-up in routine clinical practice is still challenging.

Conclusions
In conclusion, after observing a prospective cohort of 403 patients, we found that post-procedural SIRS criteria and a qSOFA score ≥ 1 may predict organ dysfunction and mortality in patients developing PEP, with qSOFA being superior to SIRS in predicting organic dysfunction as well as being the only predictor for mortality.As both SIRS and qSOFA have become diagnostic tools routinely used by clinicians, such data encourage routine post-procedural screening for SIRS criteria and qSOFA score calculation in ERCP patients.Moreover, the results of our study provide objective and reproducible data that may help clinicians stratify and individualize the risk for PEP, and thus refer patients for post-procedural intensive care.Such an approach adds to the current PEP prevention algorithms that have not yet included any prediction tool.Secondly, our study shows that, despite routine preventive measures, PEP patients are characterized by a significantly higher mortality rate, making the need for both severity and mortality predictors even more clear.Notwithstanding, implementing the use of such prediction tools helps not only management and follow-up of patients, but also offers a ground for offering better patient information and individualized care.
Our study offers novel information about severity prediction in PEP patients.Both SIRS criteria and qSOFA showed good predictive value for organic dysfunction, mortality, and hospitalization.Nevertheless, neither SIRS nor qSOFA showed any predictive value for the length of stay of PEP patients.

14 Figure 2 .
Figure 2. Receiver operating characteristic curves analyzing predictive effect of systemic inflammatory response syndrome criteria and the quick Sequential Organ Failure Assessment score on survival rates, organ dysfunction, and length of stay in post-endoscopic retrograde cholangiopancreatography pancreatitis patients.ERCP: endoscopic retrograde cholangiopancreatography; PEP: post-ERCP pancreatitis; ROC: receiver operating characteristic; qSOFA: quick Sequential Organ Failure Assessment; SIRS: systemic inflammatory response syndrome.

Figure 2 .
Figure 2. Receiver operating characteristic curves analyzing predictive effect of systemic inflammatory response syndrome criteria and the quick Sequential Organ Failure Assessment score on survival rates, organ dysfunction, and length of stay in post-endoscopic retrograde cholangiopancreatography pancreatitis patients.ERCP: endoscopic retrograde cholangiopancreatography; PEP: post-ERCP pancreatitis; ROC: receiver operating characteristic; qSOFA: quick Sequential Organ Failure Assessment; SIRS: systemic inflammatory response syndrome.

Table 1 .
The characteristics of studied population.

Table 2 .
Overall complication rates in the studied population.

Table 3 .
Evaluation of the association between PEP and bleeding and infection (a dependent variable: bleeding and infection).

Table 4 .
Life table evaluating survival rate of patients with post-endoscopic retrograde cholangiopancreatography pancreatitis.

Table 5 .
Comparative evaluation of the predictive values of the quick Sequential Organ Failure Assessment and systemic inflammatory response syndrome criteria for organ dysfunction, survival, and length of stay in post-endoscopic retrograde cholangiopancreatography pancreatitis patients.