Dynamic Nomogram for predicting the severity of acute pancreatitis in adults and Decision Curve analysis: a single-center retrospective study

doi:10.21203/rs.3.rs-2219176/v1

Download PDF

Research Article

Dynamic Nomogram for predicting the severity of acute pancreatitis in adults and Decision Curve analysis: a single-center retrospective study

https://doi.org/10.21203/rs.3.rs-2219176/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective

Acute pancreatitis (AP) is a common and unpredictable disease. Severity stratification and prognostic prediction play an important role in reducing the mortality of AP and improving its prognosis in the early stage. We aimed to establish a clinical prediction model by logistic regression analysis visualized by nomogram.

Methods

A total of 497 patients with AP in the Nanfang Hospital, Southern Medical University between January 2015 and January 2018 were retrospectively collected. 497 patients were randomly (ratio = 7:3) divided into the training cohort (N = 347) and the validation cohort (N = 147), and both cohorts were divided into non-severe acute pancreatitis (non-SAP) and SAP groups. Univariate and multivariate logistic regression analysis were used to identify the factors associated with the severity of AP. Based on multivariate logistic regression analysis, nomogram was established and validated. The performance, discrimination, and calibration of nomogram were conducted. Decision curve analysis was used to evaluate the net benefits and clinical usefulness of the prediction model.

Result

Multivariate logistic regression analysis showed that alcohol use, hyperlipidemia, gender, hypertension, ionized serum calcium and serum albumin were independent risk factors for SAP. The individualized nomogram showed good discrimination both in the training cohort (area under the receiver operating characteristic curve [AUC], 0.782) and in the validation cohort (AUC, 0.764) with good calibration. Decision curve analysis demonstrated that in terms of clinical usefulness, the nomogram was found to have some clinical values and can be used in clinical practice.

Conclusion

The proposed nomogram based on easy-to-obtain features are of high efficacy in predicting SAP patients, which may facilitate clinical decision-making.

Severe acute pancreatitis

Nomogram

Calibration curve

Decision curve

Acute pancreatitis (AP) is one of the main diseases of the total length of hospital stay, and causes high hospitalization costs, and is also one of the main causes of hospital mortality [1]. AP is divided into mild AP (MAP), moderate severe AP (MSAP) and severe AP (SAP) according to the Revised Atlanta Classification in 2012 [2]. MAP is characterized by controllable symptoms such as abdominal pain, gastrointestinal symptoms and pancreatic edema (no organ failure or systemic or local complications), comprising most AP cases. In addition to these clinical manifestations, MSAP presents with transient organ failure (TOF) that can recover spontaneously within 48 hours. Fortunately, the prognosis of hospitalized MAP/MSAP patients is generally good and mortality remains relatively low. However, AP can worsen into SAP. The death rate of patients with SAP is approximately half within two weeks of the onset [3, 4, 5]. The primary cause of early mortality and morbidity in SAP patients is organ failure caused by systemic inflammation. Persistent single or multiple organ failure is the characteristic of SAP, and the mortality rate among most SAP patients with persistent organ failure (POF) is significantly high [5, 6].

Therefore, early assessment of AP severity, especially the prediction of SAP, combined with appropriate clinical intervention and treatment, will greatly improve the prognosis of SAP patients. There are various scoring systems that can be used to evaluate the severity of AP in clinic, including Ranson score, the bedside index for severity in AP (BISAP) score and Acute Physiology and Chronic Health Examination (APACHE)-II score. However, the above scoring system is complex and requires extensive data collection. Some scoring systems, such as APACHE- II scoring system, were not originally used for AP evaluation, so they were not sensitive and practical enough [7]. In this study, clinical evaluation of AP sensitive indicators has important value for AP classification. In our study, factors which extracted from the above scoring systems and laboratory tests were selected, and the clinical data within 24h of onset of AP patients admitted between January 2015 and January 2018 were analyzed to establish an early prediction model of possibility for SAP.

To our knowledge, there are few relative studies on predicting SAP patients and no clinical tools or systems to quickly identify SAP patients. And the purpose of this study is to establish a prediction nomogram model to identify SAP patients and provide evidence for early clinical prevention and treatment of SAP.

Data source and Patient selection

This study was conducted in the Nanfang Hospital, Southern Medical University. All patients’ information was recorded in the electronic database of the Nanfang Hospital, Southern Medical University. The electronic database of the Nanfang Hospital contains the patient's diagnosis, biochemical results, related examination results and the corresponding course of disease are also included.

We extracted the data of patients with AP from January 2015 to January 2018 from the database, and screened patients according to the following criteria. Patient inclusion criteria were having 18 to 60 years of age and a diagnosis of AP. Patients with (1) age below 18 or older than 60 years; (2) Being patients; (3) Malignant pancreatic tumor patients; (4) Immune system diseases patients, such as systemic lupus erythematosus patients; and (5) Patients with incomplete information (e.g. lack of laboratory indicators or CT, B ultrasound, etc.).

Figure 1 showed that included and excluded population in our study. Thus, this study finally collected 497 patients in in AP between January 2015 and January 2018 and they were randomly divided into the training cohort (n = 349) and the validation cohort (n = 147) by 7:3 prior to data analysis.

Relevant Definitions

The diagnosis of AP was based on the revised Atlanta criteria in 2012 [2]. Patients who met the following criteria were diagnosed as AP: (1) Clinical presentation consistent with AP; (2) serum amylase or lipase was greater than the normal upper limit of three times; (3) Abdominal imaging findings were consistent with acute pancreatitis. MAP refers to pancreatitis without organ dysfunction or systemic complications; MSAP was diagnosed in AP patients with TOF (OF < 48h), previous complication deterioration or local complications (including acute peripancreatic effusion, pseudocyst, acute necrotic effusion, broken wall necrosis); Persistent organ failure (POF) (OF ≧ 48h) was diagnosed as SAP.

Clinical Factors Collection

Baseline clinical data, including age, gender, history of hypertension, history of diabetes, and etiology, were obtained from medical records. Laboratory indicators of data, including sodium ion (Na⁺), calcium ion (Ca²⁺), serum potassium ion (K⁺), chloride ion (Cl⁻), phosphorus ions (P), white blood cells count (WBC), neutrophils count (Neu), platelets count (Plt), serum amylase (Amy), alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), serum creatinine (SCr), lymphocytes count (Lym), serum lipase (Lip), albumin (Alb), platelet to lymphocyte ratio (PLR) and neutrophil to lymphocyte ratio (NLR) were obtained within 24h of onset.

Nomogram Establishment And Validation

Univariate and multivariate logistic regression analyses were performed in the training cohort to analyze the significant features for predicting SAP. The multivariate logistic regression analysis included factors significant at the 0.05 level in the univariate analysis. The multiple variables were considered significant in multivariate logistic regression analysis when p values were less than 0.05. The clinical nomogram incorporated only the independent clinical risk factors. Evaluations and comparisons between the training cohort and validation cohort of discrimination, calibration, and clinical utility of the nomogram were performed. In the training and validation cohorts, receiver operating characteristic curves (ROCs) were plotted to assess the performance of the nomogram in discriminating SAP from non-SAP patients. The area under the curve (AUC) was used to quantify discrimination. Calibration curves were drawn to explore the predictive accuracy of the individual nomogram. The net benefits at threshold probabilities were calculated by performing decision curve analysis (DCA) to determine the nomograms clinical usefulness in the training and validation cohorts.

Statistical Analysis

All selected parameters were analyzed using SPSS 24.0 (Chicago, IL) and R language software (version 4.2.0). Categorical variables were compared using the χ² test, and continuous variables were compared using the t test for variables with a normal distribution or the Mann-Whitney U test for variables with an abnormal or unknown distribution. The concentrated trends of normal distribution and non-normal distribution are represented by median and quartile, respectively. The normal distribution features are expressed as x ± s, and the difference between groups is tested by two independent samples t test. The Rank sum test was used to compare non-normal distribution data, and the chi-square test was used for comparing between groups, with p < 0.05 as statistically significant. In addition, univariate and multivariate logistic regression analyses were used to test the above selected related factors to identify relative factors. Test α = 0.05, when P < 0.05 was considered statistically significant. Finally, R language (version 4.2.0) was used to include statistically significant variables and draw the prediction model nomogram. The Concordance index (C-index) was used to verify its performance through cohorts. ROCs, Calibration curves, and Decision curves analysis were performed to evaluate the performance of the prediction model by using R software package. The “Hmisc” package was used to plot the nomogram. The “pROC” package was used to plot the ROCs and measure the AUC. The “CalibrationCurves” package was used for plotting the calibration curves. The “DecisionCurve” package was used to perform DCA.

Patient characteristics

The flowchart of the study was shown in Fig. 1. There were 1026 patients primarily diagnosed with AP in our dataset from January 2015 to January 2018. 529 patients were excluded for the following reasons: age < 18 years (n = 63); age > 60 years (n = 186); AP in pregnant patients (n = 14); having malignant pancreatic tumor (n = 84); having immune system diseases patients (n = 56); and incomplete information (n = 126). The training cohort consisted of 348 AP patients (including 94 patients with SAP and 254 patients with non-SAP), and the validation cohort comprised 149 patients (including 30 patients with SAP and 119 patients with non-SAP). Table 1 showed the main baseline characteristics of patients in the training cohort and validation cohort, there was no significant difference between the two cohorts (P > 0.05). The incidence of SAP was comparable between the two cohorts (27.0% vs. 20.1%; P = 0.131). Based on training cohort, there were no significant differences in age, other etiologies, K⁺, Cl⁻, P, WBC, Lym, Plt, Amy, Lip, ALT, AST, SCr, BUN, or PLR between the SAP group and the non-SAP group. Na⁺, Ca²⁺, Alb, NLR in the SAP group patients were lower than those in the non-SAP group (P < 0.05) (Table 2). Significant differences were also observed in the aspect of etiology, gender, history of hypertension, history of diabetes, Neu between these two groups (P < 0.05) (Table 2).

Table 1

Clinical characteristics of patients with AP in training and validation cohorts
Variables	Training cohort (n = 348)	Validation cohort (n = 149)	P-value
Age, years	41.9 ± 10.6	42.73 ± 9.97	0.490
Gender			0.626
Male	243 (69.8%)	108 (72.5%)
Female	105 (30.2%)	41 (27.5%)
Diabetes			0.920
Yes	59 (17%)	24 (16.1%)
No	289 (83%)	125 (83.9%)
Hypertension			0.258
Yes	47 (13.5%)	14 (9.4%)
No	301 (86.5%)	135 (90.6%)
Etiology			0.708
Gallstone	214 (61.5%)	93 (62.4%)
Hyperlipemia	33 (9.5%)	18 (12.1%)
Alcohol use	42 (12.1%)	14 (9.4%)
Other	59 (17%)	24 (16.1%)
Laboratory examination (24 h), mean ± SD
Na⁺, mmol/L	138.77 ± 4.86	138.01 ± 4.68	0.401
Ca²⁺, mmol/L	2.17 ± 0.24	2.21 ± 0.23	0.142
K⁺, mmol/L	3.98 ± 0.52	4.02 ± 0.48	0.227
Cl⁻, mmol/L	104.32 ± 48.37	100.96 ± 5.60	0.270
P, mmol/L	1.08 ± 0.47	1.76 ± 8.32	0.914
WBC, ×10⁹/L	11.5 ± 15.96	11 ± 6.21	0.655
Neu, ×10⁹/L	9.27 ± 9.62	29.22 ± 248.09	0.863
Lym, ×10⁹/L	1.96 ± 4.29	2.2 ± 4.81	0.163
Plt, ×10⁹/L	223.82 ± 91.25	234.7 ± 107.61	0.403
Amy, U/L	363.86 ± 817.54	382.26 ± 654.02	0.708
Lip, U/L	579.38 ± 932.48	867.75 ± 2568.32	0.427
Alb, g/L	36.94 ± 6.48	37.58 ± 6.53	0.220
ALT, U/L	115.04 ± 540.46	68.75 ± 93.00	0.154
AST, U/L	89.42 ± 418.53	47.35 ± 50.54	0.238
SCr, µmol/L	88.83 ± 83.52	85.1 ± 70.59	0.302
BUN, mmol/L	6.57 ± 15.36	5.22 ± 3.90	0.152
PLR	185.57 ± 141.81	172.49 ± 98.27	0.812
NLR	8.06 ± 8.40	21.25 ± 172.23	0.409
Na⁺, sodium ions; Ca²⁺, calcium ions; K⁺, serum potassium ions; Cl⁻, chloride ions; P, phosphorus ions; WBC, white blood cells; Neu, neutrophils; Plt, platelets; Amy, serum amylase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; SCr, serum creatinine; Lym, lymphocytes; Lip, serum lipase; Alb, albumin; PLR, platelet-to-lymphocyte ratio; NLR, neutrophil-to-lymphocyte ratio.

Table 2

Baseline characteristics of patients in the training cohort.
Variables	Non-SAP group (n = 254)	SAP group (n = 94)	P-value
Age, years	42.26 ± 10.8	40.94 ± 10.0	0.302
Gender			0.626
Male	166 (65.4%)	77 (81.9%)
Female	88 (34.6%)	17 (18.1%)
Diabetes			0.004
Yes	34 (13.4%)	25 (26.6%)
No	220 (86.6%)	69 (83.0%)
Hypertension			0.003
Yes	26 (10.2%)	21 (22.3%)
No	228 (89.8%)	73 (77.7%)
Etiology
Gallstone	171 (67.3%)	43 (45.7%)	0.000
Hyperlipemia	15 (5.9%)	18 (19.1%)	0.000
Alcohol use	22 (8.7%)	20 (21.3%)	0.001
Other	46 (18.1%)	13 (13.8%)	0.345
Laboratory examination (24 h), mean ± SD
Na⁺, mmol/L	139.15 ± 4.37	137.0 ± 5.80	0.006
Ca²⁺, mmol/L	2.20 ± 0.22	2.10 ± 0.25	0.000
K⁺, mmol/L	3.90 ± 0.51	4.00 ± 0.55	0.376
Cl⁻, mmol/L	102.0 ± 56.45	100.2 ± 6.72	0.334
P, mmol/L	1.10 ± 0.34	0.80 ± 0.70	0.070
WBC, ×10⁹/L	8.70 ± 18.18	11.10 ± 7.02	0.513
Neu, ×10⁹/L	6.60 ± 7.49	9.50 ± 13.42	0.005
Lym, ×10⁹/L	1.40 ± 4.93	1.15 ± 1.42	0.174
Plt, ×10⁹/L	208.0 ± 85.87	211.5 ± 104.68	0.526
Amy, U/L	134.90 ± 616.19	136.0 ± 1196.19	0.613
Lip, U/L	183.50 ± 1031.84	226.50 ± 615.30	0.230
Alb, g/L	38.60 ± 6.45	34.10 ± 5.69	0.000
ALT, U/L	34.0 ± 119.56	23.0 ± 1023.0	0.511
AST, U/L	29.6 ± 100.55	31.0 ± 787.06	0.430
SCr, µmol/L	70.0 ± 72.29	71.0 ± 106.65	0.066
BUN, mmol/L	4.30 ± 4.44	5.60 ± 28.58	0.096
PLR	146.50 ± 148.59	172.95 ± 119.61	0.075
NLR	4.40 ± 8.50	8.40 ± 7.76	0.005
Na⁺, sodium ions; Ca²⁺, calcium ions; K⁺, serum potassium ions; Cl⁻, chloride ions; P, phosphorus ions; WBC, white blood cells; Neu, neutrophils; Plt, platelets; Amy, serum amylase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; SCr, serum creatinine; Lym, lymphocytes; Lip, serum lipase; Alb, albumin; PLR, platelet-to-lymphocyte ratio; NLR, neutrophil-to-lymphocyte ratio.

Development Of Predicting Nomogram

Table 3 shows the results of the univariate and multivariate logistic regression analyses for predicting SAP. Hyperlipidemia, alcohol use, gender, history of hypertension, Ca²⁺, and Alb were the independent risk factors associated with SAP (P < 0.05). Thus, based on the multivariate logistic regression analysis of the training cohort, the novel nomogram was developed with hyperlipidemia, alcohol using, gender, hypertension, Ca²⁺, and Alb (Fig. 2). Ca²⁺ showed the most extended scale and the most significant effect on predicting severity of AP, and low level of Alb acquired high score on scale.

Table 3

Logistic regression analysis
Variables	Univariate logistic analysis		Multivariate logistic analysis
Variables	P-value	OR (95%CI)	P-value	OR (95%CI)
Etiology
Gallstone	0.000	0.409 (0.252–0.663)	0.490	1.361 (0.567–3.263)
Hyperlipemia	0.000	3.774 (1.815–7.848)	0.001	7.365 (2.305–23.534)
Alcohol use	0.002	2.850 (1.474–5.513)	0.039	2.995 (1.059–8.470)
Gender	0.003	2.401 (1.337–4.312)	0.007	2.840 (1.325–6.085)
Hypertension	0.004	2.523 (1.340–4.749)	0.007	3.051 (1.363–6.831)
Diabetes	0.004	2.344 (1.309–4.199)	0.239	1.560 (0.744–3.271)
Na⁺	0.002	0.915 (0.865–0.968)	0.061	0.945 (0.891–1.003)
Ca²⁺	0.000	0.032 (0.010–0.106)	0.004	0.116 (0.027–0.503)
Neu	0.002	1.046 (1.016–1.076)	0.046	1.031 (1.000-1.062)
Alb	0.000	0.908 (0.873–0.945)	0.011	0.933 (0.885–0.984)
NLR	0.009	1.038 (1.010–1.068)	0.636	1.009 (0.971–1.049)
Na⁺, sodium ions; Ca²⁺, calcium ions; Neu, neutrophils; Alb, albumin; NLR, neutrophil-to-lymphocyte ratio.

Validation Of Individual Nomogram

Through Bootstrap analysis, validation on this individual nomogram showed high internal validation concordance index (C-index) of 0.782 (95% CI, 0.727–0.837), and high external validation C-index of 0.803 (95% CI, 0.713–0.893). The area under the ROC curves (AUC) of the predictive nomogram was 0.782 (95% CI, 0.727–0.837) in the training cohort (Fig. 3A), and 0.764 (95% CI, 0.663–0.865) in the validation cohort (Fig. 3B), respectively. Both cohorts showed the good performance on C-index and ROCs. The calibration curves of training and validation cohorts are shown in Fig. 4A, 4B. The x-axis represents the nomogram-predicted probability and y-axis represents the actual probability of SAP. The ideal prediction would correspond to the 45° black dashed line. The blue dashed line represents the entire cohort (n(A) = 348, n(B) = 149), and the solid red line is bias-corrected by bootstrapping (B = 1000repetitions), indicating observed nomogram performance. And both cohorts have a good performance.

The clinical usefulness of the individual nomogram was evaluated by plotting DCA (Fig. 5A, 5B). Based on the two data cohorts, the nomogram model showed a positive net benefit at risk thresholds between 0% and 70%.

In recent years, with the increasing prevalence of AP year by year [8, 9, 10], the early, rapid and effective screening of SAP high-risk patients poses a great challenge to clinical treatment. In clinical practice, the classification and prognosis of AP are mainly evaluated based on the scoring system. Therefore, it is of great significance to establish a simple prediction tool for clinical practice.

Our analysis of the collected AP cases showed that six indicators could predict the probability of SAP, and men are more likely to develop SAP. In previous studies, the association between gender and SAP was not clear. However, previous studies demonstrated that acute biliary pancreatitis was significantly more common in female than in male patients, in contrast, alcoholic AP is more likely to occur in male patients [11, 12]. To our knowledge, through different channels, alcohol can directly damage the pancreas, and even lead to pancreatic secretion dysfunction by different mechanisms [15]. Although patients who have never consumed alcohol can develop pancreatitis, alcohol appears to increase the sensitivity of the pancreas to other etiologic factors, and the risk of pancreatitis undoubtedly increases with alcohol consumption. Previous studies have shown that alcohol use is the single most common cause of chronic pancreatitis (its attributable risk is about 40%), and after gallstones, is the second-most common cause of AP [13, 14, 15]. This can explain why male patients are more likely to develop SAP than female patients.

Previous studies have shown that gallstone and alcohol are the most common etiological factors in China, and gallstone ranks first among the known etiological factors [1, 10, 11]. Consistently, in our study, gallstone was also the main etiological factor of AP, responsible for 61.8% of all factors (Table 1). However, hyperlipidemia was also one of the significant etiological factors for AP in our study, and the incidence of SAP caused by hyperlipidemia was significantly higher than that caused by gallstone. Zhu et al. [11] analyzed 3260 hospitalized patients between 2005 and2012 from AP database and found that there was a significantly higher incidence of SAP in the hyperlipidemia group than cholelithiasis (18.2% vs 14.0%, P = 0.022) and hyperlipidemia-induced pancreatitis leads to poor prognosis. Similarly, in our study, hyperlipidemia-induced pancreatitis in the SAP group accounted for an important part and one of the major causes of AP.

In our study, it can be found that the probability of SAP in patients with hypertension was higher than that in patients without hypertension. Hypertension is often accompanied by hypercholesterolemia, and hypercholesterolemia and hypertension have synergistic pro-oxidant, pro-inflammatory, and prothrombotic effects [16, 17, 18]. Therefore, hypertension, with or without hyperlipidemia, increases the risk of AP, aggravates the inflammatory process of AP, and even leads to the development of SAP.

It can be seen from the data that calcium ions had significant differences between the non-SAP group and the SAP group in this study. Serum calcium in the non-SAP group was significantly higher than that in the SAP group. Another study found that the serum calcium level in non-SAP patients is higher than that on admission, showing that serum calcium ion may be a potential independent predictor of AP severity [19, 20]. Obviously, the Ca²⁺ level accounts for a large proportion in our model nomogram, and the lower Ca²⁺ level indicates a higher risk of SAP.

To our best knowledge, there are few studies on the association between AP and albumin, and their correlation is still unclear. Hypoalbuminemia is associated with inflammation, and low serum albumin levels caused by inflammation-induced capillary leakage or disease-related anorexia are associated with poor outcomes [22, 23]. Moreover, low serum albumin was proved to be predictive tool for inpatient mortality [21, 22, 23]. Through multivariate logistic regression, it was demonstrated that serum albumin could be used as an independent predictor of SAP. Nevertheless, previous studies have shown that albumin not only has good diagnostic accuracy for SAP, but also has high predictive ability for in-hospital mortality in patients with AP [24, 25, 26]. It can be seen from the individual analysis that AP patients were more likely to develop SAP and poor prognosis with the decrease of serum albumin level. In summary, albumin can be a potential predictor of SAP.

In this study, we developed and validated a machine learning-based model that incorporates the clinical features and etiologies in patients with AP. The nomogram demonstrated favorable discrimination in both training cohort (AUC, 0.782) and validation cohort (AUC, 0.764) and good calibration, which confirms that the training-validation model achieved excellent predictive efficacy. DCA indicated the clinical usefulness of the prediction model. By establishing the nomogram prediction model, a simple prediction tool is provided. The probability of SAP in AP can be routinely estimated in clinical practice within 24 hours of admission through a series of convenient and accessible laboratory tests. The above indicators are easy-to-obtain indicators, and the possibility of SAP can be obtained within 1 minute by using this nomogram (Fig. 2). From this perspective, clinicians can flexibly carry out early fluid resuscitation and close monitoring according to the early predictive value of AP patients, and even quickly transfer patients to hospitals with better pancreatic care or large medical institutions.

In this study, individual nomogram performed well in both training and the validation cohorts, and it also showed good calibration in both training and validation cohorts. However, there are some limitations to our analysis. The main limitation of this study may be that it is not a prospective but a single-center retrospective study and the selection bias cannot be ignored. Secondly, the sample size of this study is still limited, so an external validation cohort is necessary for our study verify the model nomogram. Finally, although the nomogram showed good performance in calibration and clinical usefulness, AUCs in the training (AUC = 0.782) and validation (AUC = 0.764) cohorts were unsatisfactory. In the next step, we will focus on multi-center validation and select more accurate features to develop a deep learning model to obtain high-level evidence with clinical application.

Six indicators (Gender, hyperlipidemia, alcohol use, hypertension, Ca²⁺, and Alb) were selected to establish the nomogram for predicting SAP. ROCs, calibration curve, and decision curve analysis suggest the potential clinical application value and clinical practicability of this nomogram.

Ethics approval

The Human Investigation Committee of Nanfang Hospital, Southern Medical University approved this study (Ethics approval number: NFEC-2021-433). The data are anonymous, and the requirement for informed consent was therefore waived.

Consent for publication

Not applicable

Availability of Data and Materials

The data that support the findings of this study are available from the Nanfang hospital but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the Nanfang hospital.

Conflict of interest statement

The authors declare that they have no competing interests

Funding

The work was supported by National Natural Science Foundation of China (82170655) & Clinical Research Program of Nanfang Hospital, Southern Medical University (2018CR046) & College Students’ Innovative Entrepreneurial Training Plan Program (S202012121097).

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Hao Xu, Xing-da Xu, Wang-jun Yang, Wang Xiao, Xiao-lou Zhang, Si-yun Zhang, Qun He, Xiao-kang Zhang and Jian-ping Qian. The original manuscript was written by Hao Xu and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Acknowledgements:

The authors would like to express their gratitude to EditSprings (https://www.editsprings.cn ) for the expert linguistic services provided.

Lankisch PG, Apte M, Banks PA. Acute pancreatitis. The Lancet. 2015;386(9988):85-96.
Banks PA, Bollen TL, Dervenis C, et al. Classification of acute pancreatitis--2012: revision of the Atlanta classification and definitions by international consensus. Gut. 2013;62(1):102-111.
Swaroop VS. Severe Acute Pancreatitis. JAMA. 2004;291(23):2865.
Fu CY. Timing of mortality in severe acute pancreatitis: Experience from 643 patients. World J Gastroenterol. 2007;13(13):1966.
Portelli M, Jones CD. Severe acute pancreatitis: pathogenesis, diagnosis and surgical management. Hepatobiliary Pancreat Dis Int. 2017;16(2):155-159.
Garg PK, Singh VP. Organ Failure Due to Systemic Injury in Acute Pancreatitis. Gastroenterology. 2019;156(7):2008-2023.
Papachristou GI, Muddana V, Yadav D, et al. 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(2):435-441; quiz 442.
Oskarsson V, Hosseini S, Discacciati A, et al. Rising incidence of acute pancreatitis in Sweden: National estimates and trends between 1990 and 2013. United Eur Gastroenterol J. 2020;8(4):472-480.
Bai Y, Liu Y, Jia L, et al. Severe acute pancreatitis in China: etiology and mortality in 1976 patients. Pancreas. 2007;35(3):232-237.
Bai Y, Jia L, Wang B, et al. Acute pancreatitis in the Guangdong Province, China. Digestion. 2007;75(2-3):74-79.
Zhu Y, Pan X, Zeng H, et al. A Study on the Etiology, Severity, and Mortality of 3260 Patients With Acute Pancreatitis According to the Revised Atlanta Classification in Jiangxi, China Over an 8-Year Period. Pancreas. 2017;46(4):504-509.
Pu W, Luo G, Chen T, et al. A 5-Year Retrospective Cohort Study: Epidemiology, Etiology, Severity, and Outcomes of Acute Pancreatitis. Pancreas. 2020;49(9):1161-1167.
Criddle DN. The role of fat and alcohol in acute pancreatitis: A dangerous liaison. Pancreatol Off J Int Assoc Pancreatol IAP Al. 2015;15(4 Suppl):S6-S12.
Yadav D, Whitcomb DC. The role of alcohol and smoking in pancreatitis. Nat Rev Gastroenterol Hepatol. 2010;7(3):131-145.
Yadav D, Lowenfels AB. The epidemiology of pancreatitis and pancreatic cancer. Gastroenterology. 2013;144(6):1252-1261.
Ivanovic B, Tadic M. Hypercholesterolemia and Hypertension: Two Sides of the Same Coin. Am J Cardiovasc Drugs Drugs Devices Interv. 2015;15(6):403-414.
Ettehad D, Emdin CA, Kiran A, et al. Blood pressure lowering for prevention of cardiovascular disease and death: a systematic review and meta-analysis. Lancet Lond Engl. 2016;387(10022):957-967.
Ruixing Y, Jinzhen W, Weixiong L, Yuming C, Dezhai Y, Shangling P. The environmental and genetic evidence for the association of hyperlipidemia and hypertension. J Hypertens. 2009;27(2):251-258.
Peng T, Peng X, Huang M, et al. Serum calcium as an indicator of persistent organ failure in acute pancreatitis. Am J Emerg Med. 2017;35(7):978-982.
Yuan S, Giovannucci EL, Larsson SC. Gallstone disease, diabetes, calcium, triglycerides, smoking and alcohol consumption and pancreatitis risk: Mendelian randomization study. NPJ Genomic Med. 2021;6(1):27.
Soeters PB, Wolfe RR, Shenkin A. Hypoalbuminemia: Pathogenesis and Clinical Significance. JPEN J Parenter Enteral Nutr. 2019;43(2):181-193.
Jellinge ME, Henriksen DP, Hallas P, Brabrand M. Hypoalbuminemia is a strong predictor of 30-day all-cause mortality in acutely admitted medical patients: a prospective, observational, cohort study. PloS One. 2014;9(8):e105983.
Ferguson RP, O’Connor P, Crabtree B, Batchelor A, Mitchell J, Coppola D. Serum albumin and prealbumin as predictors of clinical outcomes of hospitalized elderly nursing home residents. J Am Geriatr Soc. 1993;41(5):545-549.
Xu X, Ai F, Huang M. Deceased serum bilirubin and albumin levels in the assessment of severity and mortality in patients with acute pancreatitis. Int J Med Sci. 2020;17(17):2685-2695.
Hong W, Lin S, Zippi M, et al. Serum Albumin Is Independently Associated with Persistent Organ Failure in Acute Pancreatitis. Can J Gastroenterol Hepatol. 2017;2017:5297143.
Wang X, Cui Z, Li H, et al. Nosocomial mortality and early prediction of patients with severe acute pancreatitis. J Gastroenterol Hepatol. 2010;25(8):1386-1393.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Dynamic Nomogram for predicting the severity of acute pancreatitis in adults and Decision Curve analysis: a single-center retrospective study

Status:

Version 1

Abstract

Objective

Methods

Result

Conclusion

Figures

Introduction

Materials And Methods

Data source and Patient selection

Relevant Definitions

Clinical Factors Collection

Nomogram Establishment And Validation

Statistical Analysis

Results

Patient characteristics

Development Of Predicting Nomogram

Validation Of Individual Nomogram

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1