Scandinavian Journal of Gastroenterology
ISSN: 0036-5521 (Print) 1502-7708 (Online) Journal homepage: http://www.tandfonline.com/loi/igas20
Persistent SIRS and acute fluid collections are associated with increased CT scanning in acute interstitial pancreatitis Ayesha Kamal, Mahya Faghih, Robert A. Moran, Elham Afghani, Amitasha Sinha, Nasim Parsa, Martin A. Makary, Atif Zaheer, Elliot K. Fishman, Mouen A. Khashab, Anthony N. Kalloo & Vikesh K. Singh To cite this article: Ayesha Kamal, Mahya Faghih, Robert A. Moran, Elham Afghani, Amitasha Sinha, Nasim Parsa, Martin A. Makary, Atif Zaheer, Elliot K. Fishman, Mouen A. Khashab, Anthony N. Kalloo & Vikesh K. Singh (2017): Persistent SIRS and acute fluid collections are associated with increased CT scanning in acute interstitial pancreatitis, Scandinavian Journal of Gastroenterology, DOI: 10.1080/00365521.2017.1383510 To link to this article: http://dx.doi.org/10.1080/00365521.2017.1383510
Published online: 11 Oct 2017.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=igas20 Download by: [UNIVERSITY OF ADELAIDE LIBRARIES]
Date: 12 October 2017, At: 05:23
SCANDINAVIAN JOURNAL OF GASTROENTEROLOGY, 2017 https://doi.org/10.1080/00365521.2017.1383510
ORIGINAL ARTICLE
Persistent SIRS and acute fluid collections are associated with increased CT scanning in acute interstitial pancreatitis
Downloaded by [UNIVERSITY OF ADELAIDE LIBRARIES] at 05:23 12 October 2017
Ayesha Kamala, Mahya Faghiha, Robert A. Morana, Elham Afghania, Amitasha Sinhaa, Nasim Parsaa, Martin A. Makaryb, Atif Zaheerc,d, Elliot K. Fishmand, Mouen A. Khashaba, Anthony N. Kallooa,c and Vikesh K. Singha,c a Division of Gastroenterology, Johns Hopkins Medical Institutions, Baltimore, MD, USA; bDivision of Surgical Oncology, Department of Surgery, Johns Hopkins Medical Institutions, Baltimore, MD, USA; cPancreatitis Center, Johns Hopkins Medical Institutions, Baltimore, MD, USA; dDepartment of Radiology, Johns Hopkins Medical Institutions, Baltimore, MD, USA
ABSTRACT
ARTICLE HISTORY
Background: The use of computed tomography (CT) in acute pancreatitis (AP) continues to increase in parallel with the increasing use of diagnostic imaging in clinical medicine. Aim: To determine the factors associated with obtaining >1 CT scan in acute interstitial pancreatitis (AIP). Methods: Demographic and clinical data of all adult patients admitted between 1/2010 and 1/2015 with AP (AP) were evaluated. Only patients with a CT severity index (CTSI) 3 on a CT obtained within 48 h of presentation were included. Results: A total of 229 patients were included, of whom 206 (90%) had a single CT and 23 (10%) had >1 CT during the first week of hospitalization. Patients undergoing >1 CT had significantly higher rates of acute fluid collection (AFC), persistent SIRS, opioid use 4 days, and persistent organ failure compared to those undergoing 1 CT (p < .05 for all). On multivariable analysis, only persistent SIRS (OR ¼ 3.6, 95% CI 1.4–9.6, p ¼ .01) and an AFC on initial CT (OR ¼ 3.5, 95% CI 1.4–9, p ¼ .009) were independently associated with obtaining >1 CT. Conclusion: An AFC on initial CT and persistent SIRS are associated with increased CT imaging in AIP patients. However, these additional CT scans did not change clinical management.
Received 25 July 2017 Revised 13 September 2017 Accepted 15 September 2017
Introduction Acute pancreatitis (AP) is the third leading cause of hospitalization in the United States among gastrointestinal diseases, resulting in 275,170 hospitalizations at a cost of nearly 2 billion dollars in 2012 [1]. Approximately 90% of patients with AP have acute interstitial pancreatitis (AIP) that is typically mild in severity with full clinical recovery commonly seen within 1 week [2,3]. Severe acute pancreatitis (SAP), which is typically defined as persistent organ failure (POF) and/or infected pancreatic necrosis, occurs in about 5–10% of AP patients and is associated with increased mortality, morbidity and length of hospital stay [3–8]. The primary indications for obtaining a contrast enhanced CT (CECT) in AP are to determine if there is an alternative diagnosis in a patient presenting with abdominal pain or to detect local complication(s) during hospitalization [9,10]. Despite this, the use of CT imaging in AP continues to increase in parallel with the increasing use of diagnostic imaging in medical practice but has not been shown to
KEYWORDS
Acute pancreatitis; CT scans; persistent SIRS; acute fluid collection
result in improved outcomes [11]. Half of all patients with AP undergo an abdominal CT during hospitalization, with the majority obtained within 4 days of presentation, a time period which is often too early to detect local complication(s) [12,13]. The excessive use of CT imaging is particularly notable in acute necrotizing pancreatitis (ANP), where patients undergo a median of 5 CTs during the course of their hospitalization to monitor the progress of disease but only 20% of these scans are associated with a change in management [6]. Although early CT imaging is common, the frequency and factors associated with repeat CT imaging during the first week of hospitalization are not well characterized in patients with AIP. The primary aim of this study was to determine the factors associated with obtaining more than one CT during the first week of hospitalization in patients with AIP. The secondary aim was to determine if additional CT imaging during the first week of hospitalization leads to detection of evolving pancreatic necrosis and/or changes in clinical management.
CONTACT Vikesh K. Singh [email protected] Johns Hopkins Hospital, Division of Gastroenterology, 1830 E. Monument Street, Room 436, Baltimore, MD 21205, USA Both authors contributed equally to this work The preliminary data for this study was presented as a poster at the 44th annual meeting of the American Pancreatic Association on October 31, 2013 in Miami, Florida. ß 2017 Johns Hopkins University. Published by Informa UK Limited, trading as Taylor & Francis Group
Downloaded by [UNIVERSITY OF ADELAIDE LIBRARIES] at 05:23 12 October 2017
2
A. KAMAL ET AL.
Materials and methods
Inclusion/exclusion criteria
This study was approved by the Johns Hopkins Institutional Review Board for Human Research and complied with Health Insurance Portability and Accountability Act (HIPPA) regulations. Acute pancreatitis (AP) was defined as the presence of two or more of the following: (1) characteristic abdominal pain; (2) serum amylase and/or lipase three times the upper limit of normal and (3) CECT imaging demonstrating changes consistent with AP [3]. Acute interstitial pancreatitis (AIP) was defined as a CTSI score 3 within the 48 h of admission. Acute necrotizing pancreatitis (ANP) was defined by findings of parenchymal non-enhancement on contrast enhanced CT scan according to the revised Atlanta classification [3]. Chronic pancreatitis (CP) was defined by abdominal imaging demonstrating calcifications and/or a markedly irregular and dilated main pancreatic duct 5 mm [14]. An acute fluid collection (AFC) was defined as a collection with homogenous enhancement on CT associated with AIP [3]. A bedside index of severity in acute pancreatitis (BISAP) score of 3 in the first 24 hours was used to define predicted SAP [13]. Systemic inflammatory response syndrome (SIRS) was defined as two or more of the following: (1) temperature >38 C or 20 breaths/minutes or PaCO2 90 beats/min and (4) WBC 10% immature bands. Persistent SIRS was defined as presence of SIRS for >48 hours [13]. Organ failure (OF) was defined as a score 2 in the renal, cardiovascular, and pulmonary systems using the modified Marshall scoring system [3]. Persistent organ failure (POF) was defined as the presence of OF for >48 hours [3]. Comorbidity was quantified using the age-adjusted Charlson comorbidity index. Mortality was defined as death during hospitalization [15]. Opioid use was defined as the median number of days the patient received intravenous opioid analgesics including meperidine, morphine, hydromorphone, fentanyl, and buprenorphine. A change in clinical management was defined changes in the type of antibiotic(s), type or rate of intravenous fluid administration, diet, opioid use or performance of a surgical, endoscopic, or radiographic intervention. This information was obtained from paper as well as electronic records.
Patient who meet all three of the following inclusion criteria were included for analysis: 1) Age 18 years; 2) Documented AP based on the aforementioned definition; and 3) CT scan obtained within the first 48 hours of admission. The exclusion criteria for the study and the reason for exclusion (in parentheses) are as follows: (1) Transferred patients (insufficient and/or unreliable clinical data from their initial presentation as well as the potential bias of including patients with increased disease severity); (2) Chronic pancreatitis (lower incidence of SAP [16]); (3) Post-ERCP pancreatitis (PEP) (imaging is not required for the diagnosis of PEP as per consensus criteria and repeat imaging in these patients may be obtained for the evaluation of other suspected post-procedural complications, e.g., duodenal perforation [17]); (3) CT severity index (CTSI) 4 on the initial CT performed within 48 hours (difficult to distinguish between AIP and ANP due to a heterogeneous appearing pancreas [8]); and (4) Contraindication to receiving intravenous contrast (contrastenhanced imaging required to distinguish ANP from AIP [18].
Study design and population This is a retrospective study of all adult patients admitted with a diagnosis of AP from January 2010 to January 2015. Patients with AP were initially identified using the International Classification of Diseases, Ninth Revision, and Clinical Modification (ICD-9-CM) code. Demographics, laboratory and clinical data regarding severity including LOS (length of stay), POF, ICU admission and mortality was collected on all AIP patients during the course of hospitalization.
Statistical analysis Continuous and categorical data were compared between groups using the two-sample Student’s t-test and Pearson’s chi-squared test, respectively. A cutoff of 4 days of opioid use was calculated using logistic regression ROC analysis (the sensitivity and specificity of 4 days was 54.2% and 69.8%, respectively). Univariable and multivariable logistic regression analysis was performed to evaluate the factors associated with obtaining >1 CT in AIP. Variables chosen for the multivariable analysis included acute fluid collection, persistent SIRS, persistent organ failure and opioid use 4 days. They were selected on an a priori basis with the recognition that certain variables are more clinically relevant. Alcoholic AP is a known confounder for the presence of SIRS [19] and an ICU stay is a known confounder for the presence of POF [20], thus they were not incorporated into the multivariable model. The results are presented as estimated odds ratios (OR) with respective 95% confidence intervals (95% CI) and p values. A two-sided p-value of 1 CT obtained within a week of initial presentation. On univariable analysis, patients who underwent >1 CT scan had a significantly longer mean length of stay (18.4 ± 33.2 vs. 4.4 ± 4.4 days, p < .0001) and ICU stay during admission (23 ± 43.2 vs. 4.5 ± 4.9 days, p < .0001) than those who underwent 1 CT scan. Opioid use 4 days (p ¼ .02), BISAP >3 (p ¼ .004), day one SIRS (p ¼ .04), persistent SIRS (p < .001), acute fluid collection (p ¼ .03), POF (p ¼ .01) and alcohol use (p ¼ .03) were more commonly found in patients who underwent >1 CT scan as compared to those who underwent 1 CT scan. However, on multivariable analysis, only persistent SIRS (OR 3.6, 95% CI 1.4–9.6) and acute fluid collection (OR 3.5, 95% CI 1.4–9.0) were independent predictors of obtaining >1 CT scan within a week of initial hospitalization (Table 3). There were no changes in the clinical management, including choice of antibiotics, fluid administration, diet and opioid use, following the additional CT scan(s). No patient
that met the study inclusion criteria developed ANP during hospitalization.
Discussion The present study has shown that persistent SIRS and AFC are independent factors associated with repeat CT imaging in patients hospitalized with AIP. However, repeat CT imaging was not associated with a change in clinical management or a subsequent evolution from AIP to ANP. Clinicians may be compelled to obtain additional CT scans for AIP patients with persistent SIRS and/or an AFC on the premise that these patients might be developing ANP or other local complications. However, the simple presence of ANP does not mandate a change in management and local infection of pancreatic necrosis as well as vascular complications are extremely rare events in the first week of hospitalization [12].
4
A. KAMAL ET AL.
Obtaining a CECT 3–10 days after presentation in patients with AIP has been shown to change the diagnosis to ANP in approximately 10% [10,21,22]. In the present study, where patients with ANP who would have had a CTSI score of >3
Downloaded by [UNIVERSITY OF ADELAIDE LIBRARIES] at 05:23 12 October 2017
Table 1. Demographic characteristics, markers of severity and interventions in the study population (n ¼ 229). Demographics Median Age [Q1, Q3] Male Race Black White Other Etiology Alcoholic Biliary Idiopathic Other Age adjusted CCI Pain onset prior to presentation (days) Measures of severity at admission or 25 mg/dl HCT 44% Day 1 SIRS Markers of severity during hospitalization Persistent SIRS Persistent organ failure Acute fluid collection Length of stay (days) ICU admission Median length of stay in ICU Mortality Intervention Duration of opioid use (days)
N (%) 48 [38,58] 128 (55) 139 (60.7) 71 (31.2) 19 (8.10) 125 (54.6) 24 (10.5) 58 (24.5) 26 (11.4) 2 [0,3] 2 [1,4] 11 (4.80) 0 (0) 51 (22.3) 117 (51.1) 65 (28.4) 8 (3.5) 50 (21.8) 3 (2,6) 12 (5.4) 3.5 [3,9] 1 (0.4) 2 [1,4]
Values are presented as number (%), median [Q1, Q3] and mean ± standard deviation.
were excluded, no patient was found to have ANP on repeat imaging during their hospitalization. Thus, patients with a CTSI score of 3 within 48 hours of presentation are very unlikely to be subsequently diagnosed with ANP on repeat imaging. Local infection is rare in the first week of hospitalization and, when this occurs, it is associated exclusively with ANP. There is no data to support local infection of AFCs occurring in the first week of hospitalization. Despite this, the current study suggests that treating clinicians were compelled to evaluate for local infection with CECT imaging based on persistent SIRS and AFC. Ultimately, no patient developed infection in the present study and no change in management was initiated based on the findings of these repeat CTs. There was a single death (0.4%) in our study, in comparison to a mortality rate of 1–3% in previously published studies of patients with AIP [2,4]. The likely explanation for this difference in mortality is the inclusion of patients with a CTSI ¼ 4 in the other studies as these patients will often be found to have pancreatic necrosis on subsequent imaging. There are consequences to excessive CT imaging. The average abdominopelvic CT scan has been reported to Table3. Multivariable Analysis showing predictors of >1 CT in AIP. Variables
OR (95% CI)
Acute fluid collection Persistent SIRS Persistent organ failure Opioid use 4 days
3.5 3.6 1.9 1.8
p Value
(1.4–9) (1.4–9.6) (0.4–10.4) (0.7–4.6)
.009 .01 .4 .2
OR: Odds Ratio; CI: Confidence Interval; SIRS: systemic inflammatory response syndrome.
Table 2. Comparison of risk factors, measures of severity and intervention use among patients who underwent 1 CT and >1 CT during their first week of admission. Demographics
1 CT (N ¼ 206)
>1 CT (N ¼ 23)
p value
47.8 ± 16.3 17 (73.9)
.9 .4
18 (75) 4 (16.7)
.13 .10
18 (75) 1 (5.3) 2 (8.2) 3 (12.5) 2 (0–3) 1.9 (3.1)
.03 .29 .05 .85 .50 .1
4 (16.7) 0 (0) 43 (20.9) 17 (73.9)
.004 0 .16 .04
14 (58.3) 3 (12.5) 11 (45.8) 18.4 ± 33.2 6 (25) 23 ± 43.2 1
.001 .01 .003
ISSN: 0036-5521 (Print) 1502-7708 (Online) Journal homepage: http://www.tandfonline.com/loi/igas20
Persistent SIRS and acute fluid collections are associated with increased CT scanning in acute interstitial pancreatitis Ayesha Kamal, Mahya Faghih, Robert A. Moran, Elham Afghani, Amitasha Sinha, Nasim Parsa, Martin A. Makary, Atif Zaheer, Elliot K. Fishman, Mouen A. Khashab, Anthony N. Kalloo & Vikesh K. Singh To cite this article: Ayesha Kamal, Mahya Faghih, Robert A. Moran, Elham Afghani, Amitasha Sinha, Nasim Parsa, Martin A. Makary, Atif Zaheer, Elliot K. Fishman, Mouen A. Khashab, Anthony N. Kalloo & Vikesh K. Singh (2017): Persistent SIRS and acute fluid collections are associated with increased CT scanning in acute interstitial pancreatitis, Scandinavian Journal of Gastroenterology, DOI: 10.1080/00365521.2017.1383510 To link to this article: http://dx.doi.org/10.1080/00365521.2017.1383510
Published online: 11 Oct 2017.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=igas20 Download by: [UNIVERSITY OF ADELAIDE LIBRARIES]
Date: 12 October 2017, At: 05:23
SCANDINAVIAN JOURNAL OF GASTROENTEROLOGY, 2017 https://doi.org/10.1080/00365521.2017.1383510
ORIGINAL ARTICLE
Persistent SIRS and acute fluid collections are associated with increased CT scanning in acute interstitial pancreatitis
Downloaded by [UNIVERSITY OF ADELAIDE LIBRARIES] at 05:23 12 October 2017
Ayesha Kamala, Mahya Faghiha, Robert A. Morana, Elham Afghania, Amitasha Sinhaa, Nasim Parsaa, Martin A. Makaryb, Atif Zaheerc,d, Elliot K. Fishmand, Mouen A. Khashaba, Anthony N. Kallooa,c and Vikesh K. Singha,c a Division of Gastroenterology, Johns Hopkins Medical Institutions, Baltimore, MD, USA; bDivision of Surgical Oncology, Department of Surgery, Johns Hopkins Medical Institutions, Baltimore, MD, USA; cPancreatitis Center, Johns Hopkins Medical Institutions, Baltimore, MD, USA; dDepartment of Radiology, Johns Hopkins Medical Institutions, Baltimore, MD, USA
ABSTRACT
ARTICLE HISTORY
Background: The use of computed tomography (CT) in acute pancreatitis (AP) continues to increase in parallel with the increasing use of diagnostic imaging in clinical medicine. Aim: To determine the factors associated with obtaining >1 CT scan in acute interstitial pancreatitis (AIP). Methods: Demographic and clinical data of all adult patients admitted between 1/2010 and 1/2015 with AP (AP) were evaluated. Only patients with a CT severity index (CTSI) 3 on a CT obtained within 48 h of presentation were included. Results: A total of 229 patients were included, of whom 206 (90%) had a single CT and 23 (10%) had >1 CT during the first week of hospitalization. Patients undergoing >1 CT had significantly higher rates of acute fluid collection (AFC), persistent SIRS, opioid use 4 days, and persistent organ failure compared to those undergoing 1 CT (p < .05 for all). On multivariable analysis, only persistent SIRS (OR ¼ 3.6, 95% CI 1.4–9.6, p ¼ .01) and an AFC on initial CT (OR ¼ 3.5, 95% CI 1.4–9, p ¼ .009) were independently associated with obtaining >1 CT. Conclusion: An AFC on initial CT and persistent SIRS are associated with increased CT imaging in AIP patients. However, these additional CT scans did not change clinical management.
Received 25 July 2017 Revised 13 September 2017 Accepted 15 September 2017
Introduction Acute pancreatitis (AP) is the third leading cause of hospitalization in the United States among gastrointestinal diseases, resulting in 275,170 hospitalizations at a cost of nearly 2 billion dollars in 2012 [1]. Approximately 90% of patients with AP have acute interstitial pancreatitis (AIP) that is typically mild in severity with full clinical recovery commonly seen within 1 week [2,3]. Severe acute pancreatitis (SAP), which is typically defined as persistent organ failure (POF) and/or infected pancreatic necrosis, occurs in about 5–10% of AP patients and is associated with increased mortality, morbidity and length of hospital stay [3–8]. The primary indications for obtaining a contrast enhanced CT (CECT) in AP are to determine if there is an alternative diagnosis in a patient presenting with abdominal pain or to detect local complication(s) during hospitalization [9,10]. Despite this, the use of CT imaging in AP continues to increase in parallel with the increasing use of diagnostic imaging in medical practice but has not been shown to
KEYWORDS
Acute pancreatitis; CT scans; persistent SIRS; acute fluid collection
result in improved outcomes [11]. Half of all patients with AP undergo an abdominal CT during hospitalization, with the majority obtained within 4 days of presentation, a time period which is often too early to detect local complication(s) [12,13]. The excessive use of CT imaging is particularly notable in acute necrotizing pancreatitis (ANP), where patients undergo a median of 5 CTs during the course of their hospitalization to monitor the progress of disease but only 20% of these scans are associated with a change in management [6]. Although early CT imaging is common, the frequency and factors associated with repeat CT imaging during the first week of hospitalization are not well characterized in patients with AIP. The primary aim of this study was to determine the factors associated with obtaining more than one CT during the first week of hospitalization in patients with AIP. The secondary aim was to determine if additional CT imaging during the first week of hospitalization leads to detection of evolving pancreatic necrosis and/or changes in clinical management.
CONTACT Vikesh K. Singh [email protected] Johns Hopkins Hospital, Division of Gastroenterology, 1830 E. Monument Street, Room 436, Baltimore, MD 21205, USA Both authors contributed equally to this work The preliminary data for this study was presented as a poster at the 44th annual meeting of the American Pancreatic Association on October 31, 2013 in Miami, Florida. ß 2017 Johns Hopkins University. Published by Informa UK Limited, trading as Taylor & Francis Group
Downloaded by [UNIVERSITY OF ADELAIDE LIBRARIES] at 05:23 12 October 2017
2
A. KAMAL ET AL.
Materials and methods
Inclusion/exclusion criteria
This study was approved by the Johns Hopkins Institutional Review Board for Human Research and complied with Health Insurance Portability and Accountability Act (HIPPA) regulations. Acute pancreatitis (AP) was defined as the presence of two or more of the following: (1) characteristic abdominal pain; (2) serum amylase and/or lipase three times the upper limit of normal and (3) CECT imaging demonstrating changes consistent with AP [3]. Acute interstitial pancreatitis (AIP) was defined as a CTSI score 3 within the 48 h of admission. Acute necrotizing pancreatitis (ANP) was defined by findings of parenchymal non-enhancement on contrast enhanced CT scan according to the revised Atlanta classification [3]. Chronic pancreatitis (CP) was defined by abdominal imaging demonstrating calcifications and/or a markedly irregular and dilated main pancreatic duct 5 mm [14]. An acute fluid collection (AFC) was defined as a collection with homogenous enhancement on CT associated with AIP [3]. A bedside index of severity in acute pancreatitis (BISAP) score of 3 in the first 24 hours was used to define predicted SAP [13]. Systemic inflammatory response syndrome (SIRS) was defined as two or more of the following: (1) temperature >38 C or 20 breaths/minutes or PaCO2 90 beats/min and (4) WBC 10% immature bands. Persistent SIRS was defined as presence of SIRS for >48 hours [13]. Organ failure (OF) was defined as a score 2 in the renal, cardiovascular, and pulmonary systems using the modified Marshall scoring system [3]. Persistent organ failure (POF) was defined as the presence of OF for >48 hours [3]. Comorbidity was quantified using the age-adjusted Charlson comorbidity index. Mortality was defined as death during hospitalization [15]. Opioid use was defined as the median number of days the patient received intravenous opioid analgesics including meperidine, morphine, hydromorphone, fentanyl, and buprenorphine. A change in clinical management was defined changes in the type of antibiotic(s), type or rate of intravenous fluid administration, diet, opioid use or performance of a surgical, endoscopic, or radiographic intervention. This information was obtained from paper as well as electronic records.
Patient who meet all three of the following inclusion criteria were included for analysis: 1) Age 18 years; 2) Documented AP based on the aforementioned definition; and 3) CT scan obtained within the first 48 hours of admission. The exclusion criteria for the study and the reason for exclusion (in parentheses) are as follows: (1) Transferred patients (insufficient and/or unreliable clinical data from their initial presentation as well as the potential bias of including patients with increased disease severity); (2) Chronic pancreatitis (lower incidence of SAP [16]); (3) Post-ERCP pancreatitis (PEP) (imaging is not required for the diagnosis of PEP as per consensus criteria and repeat imaging in these patients may be obtained for the evaluation of other suspected post-procedural complications, e.g., duodenal perforation [17]); (3) CT severity index (CTSI) 4 on the initial CT performed within 48 hours (difficult to distinguish between AIP and ANP due to a heterogeneous appearing pancreas [8]); and (4) Contraindication to receiving intravenous contrast (contrastenhanced imaging required to distinguish ANP from AIP [18].
Study design and population This is a retrospective study of all adult patients admitted with a diagnosis of AP from January 2010 to January 2015. Patients with AP were initially identified using the International Classification of Diseases, Ninth Revision, and Clinical Modification (ICD-9-CM) code. Demographics, laboratory and clinical data regarding severity including LOS (length of stay), POF, ICU admission and mortality was collected on all AIP patients during the course of hospitalization.
Statistical analysis Continuous and categorical data were compared between groups using the two-sample Student’s t-test and Pearson’s chi-squared test, respectively. A cutoff of 4 days of opioid use was calculated using logistic regression ROC analysis (the sensitivity and specificity of 4 days was 54.2% and 69.8%, respectively). Univariable and multivariable logistic regression analysis was performed to evaluate the factors associated with obtaining >1 CT in AIP. Variables chosen for the multivariable analysis included acute fluid collection, persistent SIRS, persistent organ failure and opioid use 4 days. They were selected on an a priori basis with the recognition that certain variables are more clinically relevant. Alcoholic AP is a known confounder for the presence of SIRS [19] and an ICU stay is a known confounder for the presence of POF [20], thus they were not incorporated into the multivariable model. The results are presented as estimated odds ratios (OR) with respective 95% confidence intervals (95% CI) and p values. A two-sided p-value of 1 CT obtained within a week of initial presentation. On univariable analysis, patients who underwent >1 CT scan had a significantly longer mean length of stay (18.4 ± 33.2 vs. 4.4 ± 4.4 days, p < .0001) and ICU stay during admission (23 ± 43.2 vs. 4.5 ± 4.9 days, p < .0001) than those who underwent 1 CT scan. Opioid use 4 days (p ¼ .02), BISAP >3 (p ¼ .004), day one SIRS (p ¼ .04), persistent SIRS (p < .001), acute fluid collection (p ¼ .03), POF (p ¼ .01) and alcohol use (p ¼ .03) were more commonly found in patients who underwent >1 CT scan as compared to those who underwent 1 CT scan. However, on multivariable analysis, only persistent SIRS (OR 3.6, 95% CI 1.4–9.6) and acute fluid collection (OR 3.5, 95% CI 1.4–9.0) were independent predictors of obtaining >1 CT scan within a week of initial hospitalization (Table 3). There were no changes in the clinical management, including choice of antibiotics, fluid administration, diet and opioid use, following the additional CT scan(s). No patient
that met the study inclusion criteria developed ANP during hospitalization.
Discussion The present study has shown that persistent SIRS and AFC are independent factors associated with repeat CT imaging in patients hospitalized with AIP. However, repeat CT imaging was not associated with a change in clinical management or a subsequent evolution from AIP to ANP. Clinicians may be compelled to obtain additional CT scans for AIP patients with persistent SIRS and/or an AFC on the premise that these patients might be developing ANP or other local complications. However, the simple presence of ANP does not mandate a change in management and local infection of pancreatic necrosis as well as vascular complications are extremely rare events in the first week of hospitalization [12].
4
A. KAMAL ET AL.
Obtaining a CECT 3–10 days after presentation in patients with AIP has been shown to change the diagnosis to ANP in approximately 10% [10,21,22]. In the present study, where patients with ANP who would have had a CTSI score of >3
Downloaded by [UNIVERSITY OF ADELAIDE LIBRARIES] at 05:23 12 October 2017
Table 1. Demographic characteristics, markers of severity and interventions in the study population (n ¼ 229). Demographics Median Age [Q1, Q3] Male Race Black White Other Etiology Alcoholic Biliary Idiopathic Other Age adjusted CCI Pain onset prior to presentation (days) Measures of severity at admission or 25 mg/dl HCT 44% Day 1 SIRS Markers of severity during hospitalization Persistent SIRS Persistent organ failure Acute fluid collection Length of stay (days) ICU admission Median length of stay in ICU Mortality Intervention Duration of opioid use (days)
N (%) 48 [38,58] 128 (55) 139 (60.7) 71 (31.2) 19 (8.10) 125 (54.6) 24 (10.5) 58 (24.5) 26 (11.4) 2 [0,3] 2 [1,4] 11 (4.80) 0 (0) 51 (22.3) 117 (51.1) 65 (28.4) 8 (3.5) 50 (21.8) 3 (2,6) 12 (5.4) 3.5 [3,9] 1 (0.4) 2 [1,4]
Values are presented as number (%), median [Q1, Q3] and mean ± standard deviation.
were excluded, no patient was found to have ANP on repeat imaging during their hospitalization. Thus, patients with a CTSI score of 3 within 48 hours of presentation are very unlikely to be subsequently diagnosed with ANP on repeat imaging. Local infection is rare in the first week of hospitalization and, when this occurs, it is associated exclusively with ANP. There is no data to support local infection of AFCs occurring in the first week of hospitalization. Despite this, the current study suggests that treating clinicians were compelled to evaluate for local infection with CECT imaging based on persistent SIRS and AFC. Ultimately, no patient developed infection in the present study and no change in management was initiated based on the findings of these repeat CTs. There was a single death (0.4%) in our study, in comparison to a mortality rate of 1–3% in previously published studies of patients with AIP [2,4]. The likely explanation for this difference in mortality is the inclusion of patients with a CTSI ¼ 4 in the other studies as these patients will often be found to have pancreatic necrosis on subsequent imaging. There are consequences to excessive CT imaging. The average abdominopelvic CT scan has been reported to Table3. Multivariable Analysis showing predictors of >1 CT in AIP. Variables
OR (95% CI)
Acute fluid collection Persistent SIRS Persistent organ failure Opioid use 4 days
3.5 3.6 1.9 1.8
p Value
(1.4–9) (1.4–9.6) (0.4–10.4) (0.7–4.6)
.009 .01 .4 .2
OR: Odds Ratio; CI: Confidence Interval; SIRS: systemic inflammatory response syndrome.
Table 2. Comparison of risk factors, measures of severity and intervention use among patients who underwent 1 CT and >1 CT during their first week of admission. Demographics
1 CT (N ¼ 206)
>1 CT (N ¼ 23)
p value
47.8 ± 16.3 17 (73.9)
.9 .4
18 (75) 4 (16.7)
.13 .10
18 (75) 1 (5.3) 2 (8.2) 3 (12.5) 2 (0–3) 1.9 (3.1)
.03 .29 .05 .85 .50 .1
4 (16.7) 0 (0) 43 (20.9) 17 (73.9)
.004 0 .16 .04
14 (58.3) 3 (12.5) 11 (45.8) 18.4 ± 33.2 6 (25) 23 ± 43.2 1
.001 .01 .003
