Documente Academic
Documente Profesional
Documente Cultură
analysis. The primary outcome was in-hospital mortality. Secon- institutions in Japan, including 25 tertiary academic medical
dary outcomes included the incidence of pancreatic infection and centers, and community hospitals with intensive care units.
the need for surgical intervention. Inclusion criteria were age 18 years or older, and the diag-
Results: We analyzed 1097 patients, and the mean fluid volume nosis of SAP, from January 1, 2009 to December 31, 2013.
administered was 5618 ± 3018 mL (mean ± SD), with 708 and 389 Acute pancreatitis was diagnosed if the patient presented
patients stratified into the fluid <6000 mL and fluid ≥ 6000 mL with at least 2 of the following 3 features: (1) acute pain and
groups, respectively. Overall in-hospital mortality was 12.3%. The tenderness in the upper abdomen; (2) elevated pancreatic
fluid ≥ 6000 mL group had significantly higher mortality than the enzyme levels in the blood and urine; or (3) findings of acute
fluid <6000 mL group (univariable analysis, 15.9% vs. 10.3%; pancreatitis as detected by ultrasonography, computed
P < 0.05). In multivariable logistic regression analysis, admin- tomography (CT) scan, or magnetic resonance imaging. The
istration of ≥ 6000 mL of fluid within the first 24 hours was sig- diagnosis of SAP was based on criteria of the Japanese
nificantly associated with reduced mortality (odds ratio, 0.58;
P < 0.05). No significant association was found between the
Ministry of Health, Labour and Welfare (Japanese Severity
administered fluid volume and pancreatic infection, or between the Score)10 (Table 1). We report the present study in accord-
volume administered and the need for surgical intervention. ance with the Strengthening Reporting of Observational
studies in Epidemiology guidelines.16
Conclusions: In patients with SAP, administration of a large fluid
volume within the first 24 hours is associated with decreased
mortality. Data Collection
Data were collected retrospectively, including patient
Key Words: severe acute pancreatitis, fluid resuscitation, mortality, age, gender, etiology of acute pancreatitis, acute physiology
pancreatic infection, surgical intervention and chronic health evaluation (APACHE) II score, CT
(J Clin Gastroenterol 2018;00:000–000) severity index (CTSI), prognostic factors, and CT grade of
the Japanese Severity Score,10 Charlson index and severity
2 | www.jcge.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
J Clin Gastroenterol Volume 00, Number 00, ’’ 2018 Effect of Fluid Therapy in Severe Acute Pancreatitis
grade of the revised Atlanta classification.17 Acute pan- covariables included age, gender, etiology, APACHE II
creatitis was classified into 4 categories according to the score, Charlson index, CTSI, prognostic factors of the
etiology (alcohol, gallstones, idiopathic, and others).18 Japanese Severity Score,10 enteral nutrition within first
Treatments analyzed included administered fluid vol- 48 hours, CRAI of protease inhibitors, administration of
ume within the first 24 hours after diagnosis, administration prophylactic antibiotics, mechanical ventilation, and renal
of enteral nutrition within the first 48 hours after diagnosis, replacement therapy. As a sensitivity analysis, we conducted
administration of prophylactic antibiotics, continuous an identical analysis for the subgroup of patients with SAP
regional arterial infusion (CRAI) of protease inhibitors, diagnosed according to the revised Atlanta classification.17
mechanical ventilation, and renal replacement therapy. We performed another subgroup analysis for patients
Fluids were defined only as intravenously infused fluids. who required open surgery, a percutaneous approach or an
Oral intake, enteral nutrition, and other nonintravenous endoscopic approach. We additionally evaluated the
fluid intakes were excluded from analysis. association between fluid resuscitation and mortality in
multivariable analysis for each of the 3 interventions (open
Outcomes surgery, percutaneous approach, and endoscopic approach).
The primary outcome was in-hospital mortality. Sec- Each factor was included in multivariable analysis. A
ondary outcomes included the incidence of pancreatic 2-sided P < 0.05 was considered statistically significant. All
infection and the need for surgical intervention. We defined statistical analysis was performed with JMP (r) statistical
pancreatic infection as the presence of bacteria in a blood software version 11.2 (SAS Institute Inc., Cary, NC).
culture or local culture obtained by percutaneous, image-
guided, or endoscopic fine-needle aspiration, or the presence RESULTS
of extraluminal gas in the pancreatic and/or peripancreatic
tissues on contrast-enhanced CT scan. Surgical interventions Patient Characteristics
included percutaneous, endoscopic, laparoscopic, or lapa- A total of 1159 patients with SAP were enrolled. In total,
rotomy drainage or necrosectomy for infected acute necrotic 41 patients who did not undergo contrast-enhanced CT scan
collections or walled-off necrosis, interventional radiologic, and 21 patients who had at least 1 missing data point for
or endoscopic treatment for bleeding, etc. variables used in the multivariable analysis were excluded,
leaving 1097 patients for analysis (Fig. 1). The mean volume of
Statistical Analysis fluid administered was 5618 ± 3038 mL (mean ± SD) within
On the basis of the recommendation that initial fluid the first 24 hours (Fig. 2). In total, 708 patients were classified
administration should be at least 250 mL/h,8 we divided the into the fluid <6000 mL group, and 389 patients into the fluid
cohort into 2 groups according to the fluid volume admin- ≥ 6000 mL group (Fig. 1). Table 2 shows the clinical charac-
istered within the first 24 hours: <6000 and ≥ 6000 mL. teristics of the study patients. The fluid ≥ 6000 mL group was
Differences between the 2 groups were analyzed by the significantly younger and had a higher proportion of males
Student t test or the Wilcoxon rank-sum tests for continuous than the fluid <6000 mL group. There were significant differ-
variables and the χ2 tests for categorical variables. ences between the 2 groups with regard to the etiology of SAP.
Multivariable logistic regression analysis was used to The fluid ≥ 6000 mL group had a significantly higher severity
evaluate the independent effect of fluid volume on clinical as showed by the APACHE II scores (P < 0.001), prognostic
outcomes in patients with SAP. In multivariable analysis, factors of the Japanese Severity Score10 (P < 0.001), CTSI
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved. www.jcge.com |3
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
Yamashita et al J Clin Gastroenterol Volume 00, Number 00, ’’ 2018
FIGURE 2. Administered fluid volume within the first 24 hours. The mean fluid volume administered within the first 24 hours to all
patients was 5618 ± 3038 mL (mean ± SD).
4 | www.jcge.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
J Clin Gastroenterol Volume 00, Number 00, ’’ 2018 Effect of Fluid Therapy in Severe Acute Pancreatitis
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved. www.jcge.com |5
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
Yamashita et al J Clin Gastroenterol Volume 00, Number 00, ’’ 2018
such as pancreatic infection or necrosis, were not affected by multivariable analysis. Third, we included patients with
the fluid volume administered but rather were affected by the SAP diagnosed by the Japanese Severity Score,10 severity
severity of the disease itself. There is no significant association assessment criteria that have been validated only in Japa-
between administered fluid volume and mortality in subgroup nese patients. Therefore, we performed subgroup analysis
analysis for patients who required open surgery, percutaneous for patients with SAP based on the revised Atlanta
approach, or endoscopic approach. Neither fluid volume nor classification17 and found consistent associations between
any of the 3 interventions have a significant association with total volume of fluid administered and outcomes. Fourth,
mortality, possibly due to the small sample size with insuffi- the type of fluids administered were unknown in our cohort.
cient statistical power. The Japanese guideline recommends balanced electrolyte
The volume of fluid administered in this study was solutions, such as lactated Ringer’s solution, for fluid
greater than in previous studies.21–24 The present study resuscitation.10 Further studies will be needed to identify the
includes only patients with SAP and a majority of study optimal fluid for administration to patients with SAP.
institutions were considered to have followed the guidelines Finally, the volume of fluid resuscitation given after the first
recommendations. The American College of Gastro- 24 hours following admission was unclear, and may affect
enterology guidelines recommend aggressive fluid resusci- the outcomes.
tation defined as 250 to 500 mL/h for all patients unless In conclusion, the administration of a large volume of
cardiovascular or renal comorbidities exist9 and also state fluid within the first 24 hours of the onset of SAP is asso-
that “early aggressive intravenous hydration is most bene- ciated with lower mortality. Large volume fluid resuscita-
ficial during the first 12 to 24 hours, and may have little tion may be associated with an improved prognosis in
benefit beyond this time period.”9 Although these recom- patients with SAP.
mendations do not define a daily target dose, the daily
amount by calculation (total 6000 to 12,000 mL/24 h) is
comparable to the fluid volume administered in this study. ACKNOWLEDGMENTS
The American Gastroenterological Association technical The authors acknowledge for Kazuichi Okazaki,
review states that “in severe acute pancreatitis, fluid needs of Tsuyoshi Takeda, Seiya Suzuki, Jun Kataoka, Tomohiro
5 L or more daily are not uncommon.”11 The Japanese Adachi, Wataru Shinomiya, Shin Namiki, Sakue Masuda,
guideline for the treatment of patients with acute pan- Tomoaki Hashida, Naoki Shinyama, Hitoshi Yamamura,
creatitis also recommends rapid fluid resuscitation (150 to Takashi Moriya, Kunihiro Shirai, Kazuo Inui, Satoshi
600 mL/h) to treat shock and dehydration.10 Although the Yamamoto, Kyoji Oe, Takashi Muraki, Tetsuya Ito, Junichi
fluid ≥ 6000 mL group had higher severity than the fluid Sakagami, Hiroaki Yasuda, Yoshinori Azumi, Masayuki
<6000 mL group, the fluid ≥ 6000 mL group had a lower Kamochi, Keiji Nagata, Nobuyuki Saito, Mizuki Sato, Kyohei
risk of mortality when adjusted for severity. Although the Miyamoto, Koji Saito, Kazunori Takeda, Motohiro Sekino,
universal cut-off value (eg, 6000 mL) for all patients cannot Tomoki Furuya, Yoshimoto Seki, Tetsuya Mine, Youhei
be determined, there is a group of patients who require a Kawashima, Naoyuki Matsuda, Masato Inaba, Mineji Hay-
large volume infusion to improve outcomes. The results may akawa, Takuyo Misumi and Yuki Takeda with the support of
indicate that large volume resuscitation is a critical part of the data collection at 44 institutions (Osaka Saiseikai Senri
the management of patients with severe pancreatitis. Hospital, Hiroshima City Hiroshima Citizens Hospital,
Mao et al13 reported that aggressive fluid resuscitation Kansai Medical University Hirakata Hospital, The University
increased mortality and complications including respiratory of Tokyo Hospital, Iizuka Hospital, Japanese Red Cross
failure, abdominal compartment syndrome, and sepsis. Musashino Hospital, Tokyo Metropolitan Tama Medical
However, aggressive fluid resuscitation in the study by Mao Center, Japanese Red Cross Maebashi Hospital, Shonan
and colleagues represented the rate of fluid administration Kamakura General Hospital, Showa University Hospital, Nihon
rather the total volume in the early stage. The fluid infusion University Hospital, Saiseikai Kumamoto Hospital, Fukuyama
rate in the aggressive group was 10 to 15 mL/kg/h and in City Hospital, Jichi Medical University Hospital, Chiba Uni-
the control group was 5 to 10 mL/kg/h.13 The total mean versity Hospital, Osaka City University Hospital, Tohoku
volume of fluid from admission to the first day was similar University Hospital, Nihon University Itabashi Hospital,
comparing the aggressively resuscitated group (9535 mL: Gifu University Hospital, Second Teaching Hospital, Fujita
crystalloids 6855 mL, colloids 2680 mL) and the control Health University, Asahi General Hospital, Shinshu University
group (8387 mL: crystalloids 5841 mL, colloids 2546 mL).13 Hospital, National Hospital Organization Nagasaki Medical
Thus, the comparison in the Mao et al study13 (rate) was Center, University Hospital, Kyoto Prefectural University of
different from that in the present study (total volume). As Medicine, Mie University Hospital, Hospital of the University of
the data in the present study are limited only to the volume Occupational and Environmental Health, Nippon Medical
of fluid, we could not define an optimal rate of fluid School Chiba Hokusoh Hospital, Jichi Medical University,
infusion. Saitama Medical Center, Wakayama Medical University Hos-
This study has several acknowledged limitations. This pital, Tokyo Metropolitan Bokutoh Hospital, Jikei University
is a multicenter retrospective study in which protocols for School of Medicine, Saka General Hospital, National Hospital
fluid resuscitation were not consistent among participating Organization Sendai Medical Center, Nagasaki University
institutions. However, all participating institutions are Hospital, Keio University School of Medicine, Japanese Red
believed to have followed the Japanese guidelines for the Cross Akita Hospital, Ibaraki Prefectural Central Hospital,
management of patients with acute pancreatitis suggesting Tokai University Hospital, Nagoya University Hospital, Hok-
early fluid administration,10 and had sufficient experience in kaido University Hospital, National Cancer Center, Akita City
treating patients with SAP. Second, the retrospective nature Hospital, Kobe University Hospital, Tokyo Rosai Hospital).
of the study precludes adjustments for undefined con- The authors thank the Japanese Society of Education for
founders. Because of the large study population, we Physicians and Trainees in Intensive Care, and the Japanese
were able to adjust for other confounding factors by Society of Intensive Care Medicine.
6 | www.jcge.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
J Clin Gastroenterol Volume 00, Number 00, ’’ 2018 Effect of Fluid Therapy in Severe Acute Pancreatitis
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved. www.jcge.com |7
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.