American Journal of Emergency Medicine 32 (2014) 1481–1484
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Performance assessment of the Simplified Acute Physiology Score II, the Acute Physiology and Chronic Health Evaluation II score, and the Sequential Organ Failure Assessment score in predicting the outcomes of adult patients with hepatic portal venous gas in the ED Chen-June Seak, MD a,b,c, Chip-Jin Ng, MD a,b,⁎, David Hung-Tsang Yen, MD, PhD c,d, Yon-Cheong Wong, MD b,e, Kuang-Hung Hsu, PhD f,⁎⁎, Joanna Chen-Yeen Seak, MD g, Chen-Ken Seak, MBBS g a
Department of Emergency Medicine, Lin-Kou Medical Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan College of Medicine, Chang Gung University, Taoyuan, Taiwan Institute of Emergency and Critical Care Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan d Department of Emergency Medicine, Taipei Veterans General Hospital, Taipei, Taiwan e Division of Emergency and Critical Care Radiology, Department of Medical Imaging and Intervention, Lin-Kou Medical Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan f Laboratory for Epidemiology, Department of Health Care Management, and Healthy Aging Research Center, Chang Gung University, Taoyuan, Taiwan g Sarawak General Hospital, Kuching, Sarawak, Malaysia b c
a r t i c l e
i n f o
Article history: Received 4 July 2014 Received in revised form 2 September 2014 Accepted 9 September 2014
a b s t r a c t Objective: This study aims to evaluate the performance of Simplified Acute Physiology Score II (SAPS II), the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and the Sequential Organ Failure Assessment (SOFA) score for predicting illness severity and the mortality of adult hepatic portal venous gas (HPVG) patients presenting to the emergency department (ED). This will assist emergency physicians in risk stratification. Methods: Data for 48 adult HPVG patients who visited our ED between December 2009 and December 2013 were analyzed. The SAPS II, APACHE II score, and SOFA score were calculated based on the worst laboratory values in the ED. The probability of death was calculated for each patient based on these scores. The ability of the SAPS II, APACHE II score, and SOFA score to predict group mortality was assessed by using receiver operating characteristic curve analysis and calibration analysis. Results: The sensitivity, specificity, and accuracy were 92.6%,71.4%, and 83.3%, respectively, for the SAPS II method; 77.8%, 81%, and 79.2%, respectively, for the APACHE II scoring system, and 77.8%, 76.2%, and 79.2%, respectively, for the SOFA score. In the receiver operating characteristic curve analysis, the areas under the curve for the SAPS II, APACHE II scoring system, and SOFA score were 0.910, 0.878, and 0.809, respectively. Conclusion: This is one of the largest series performed in a population of adult HPVG patients in the ED. The results from the present study showed that SAPS II is easier and more quickly calculated than the APACHE II and more superior in predicting the mortality of ED adult HPVG patients than the SOFA. We recommend that the SAPS II be used for outcome prediction and risk stratification in adult HPVG patients in the ED. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Hepatic portal venous gas (HPVG) was first reported in 1955 by Wolfe and Evans [1]. In the past, the diagnosis of HPVG was challenging, as massive amounts of HPVG are required to be visible on plain films. However, the discovery of HPVG incidence has increased over the past ⁎ Correspondence to: CJ Ng, Department of Emergency Medicine, Lin-Kou Medical Center, Chang Gung Memorial Hospital, No. 5, Fusing St, Gueishan Township, Taoyuan County, 333 Taiwan. Tel.: +886 3 3281200x2505; fax: +886 3 3287715. ⁎⁎ Correspondence to: KH Hsu, Laboratory for Epidemiology, Department of Health Care Management, and Healthy Aging Research Center, Chang Gung University, No. 259, WenHwa 1st Rd, Gueishan Township, Taoyuan County, 333 Taiwan. Tel.: + 886 3 2118800x5486; fax: +886 3 2118138. E-mail addresses: [email protected] (C.-J. Ng), [email protected] (K.-H. Hsu). http://dx.doi.org/10.1016/j.ajem.2014.09.011 0735-6757/© 2014 Elsevier Inc. All rights reserved.
60 years due to improved diagnostic imaging and increased incidental findings detected by modern high-resolution computed tomography (CT) [2,3]. Historically, this rare radiologic finding has been associated with lethal surgical conditions [2] and high mortality rates ranging from 29% to 90% [4-7]. However, gradually over the years, some patients have survived after being treated conservatively depending on the etiology of the underlying disease [2]. Hence, risk stratification in the emergency department (ED) is particularly important to assist emergency physicians (EPs) in identifying critically ill patients with HPVG and informing family members about treatment options. However, to our knowledge, there is no established method which can precisely and easily predict the outcomes of adult patients with HPVG in the ED. The increasing patient volumes in the ED and limited bed availability in the intensive care unit (ICU) lead to the ED being used as the first
1482
C-J. Seak et al. / American Journal of Emergency Medicine 32 (2014) 1481–1484
contact for primary care. Consequently, the provision of critical care in the ED has increased [8-10]. Although early and aggressive intervention in critically ill patients in the ED is becoming increasingly important, adequate critical care in the ED is often difficult to provide and sometimes overlooked in the busy ED setting [8]. To improve critical care in the ED, the use of predictive scoring systems is considered potentially valuable for EPs for early disease recognition and prognostication of outcome to identify high-risk patients and adopt the appropriate intensive care intervention [8,11]. Existing ICU scoring systems were thus applied with some success [8,9,12]. Among the ICU scoring systems, the Simplified Acute Physiology Score II (SAPS II), the Acute Physiology and Chronic Health Evaluation II (APACHE II), and the Sequential Organ Failure Assessment (SOFA) scoring systems have been widely used for prognosticating outcomes [8,13,14]. However, there is limited dearth of literature describing the application of these scoring systems in HPVG patients despite their high mortality rate, and no studies have investigated the relationship between prognosis of HPVG patients and the SOFA and SAPS II scores at the time of ED presentation. In our previous study [15], we found that shock (odds ratio, 17.02; 95% confidence interval, 3.3686.22) and pneumatosis intestinalis (odds ratio, 5.14; 95% confidence interval, 1.03-25.67) were the significant predictors of mortality in adult patients with HPVG in the ED after the results were adjusted for age and sex. However, using the existing scoring systems to predict patient outcomes will be more feasible and less invasive and thus favorable in clinical practice. In this study, we further aimed to evaluate the performance of the SAPS II, APACHE II, and SOFA scoring systems in predicting illness severity and the mortality of adult HPVG patients presenting to ED due to the high HPVG mortality rate. This may help in risk stratification and permits earlier detection of adult HPVG patients in the ED who will benefit most from early and aggressive therapeutic intervention.
2.3. Etiology If surgery was implemented, intraoperative findings were integrated as the reference standard, and histopathology reports were evaluated to confirm the underlying disease. Clinical presentations and courses with discharge diagnoses and CT images were considered as the reference standard in patients who did not receive surgical interventions. 2.4. Statistical analysis Numerical and categorical variables are displayed as the means ± SD, and frequencies are shown as percentages (%). Univariate analyses were applied to examine the association between predictors and mortality, whereas categorical and numerical variables were analyzed with a χ 2 test and 2-sample t test, respectively. A logistic regression analysis was performed to develop predictive models between scoring systems and mortality. The probability of death was calculated based on the predictive models using the logit formula: P¼
1 1 þ exp ½−ðβ0 þ β1 X 1 Þ ðβ0 : Intercept; β1 : Parameter estimate of score; X 1 : ScoreÞ
A 2-sample t test was applied to compare the differences in death probability between survivors and nonsurvivors. Area under the receiver operating characteristic curve (AUROC) analysis was used to compare the predictability of mortality among scoring systems. The HosmerLemeshow statistic was used to measure model fitness of risk prediction. In addition, the sensitivity, specificity, and accuracy rates were calculated based on the optimal cutoff point derived from the AUROC analysis.
2. Materials and methods
3. Results
2.1. Design and setting
A total of 48 patients who met the inclusion criteria (admitted to ED with acute abdominal pain and subsequent CT findings of HPVG) were identified over 4 years. Among the 48 patients, 32 patients had ischemic bowel disease; 16, with nonischemic bowel disease. In the group of patients with nonischemic bowel disease, 7 patients had intra-abdominal abscesses, 3 had biliary tract infections, 4 had colitis, 1 had gastric ulcer, and 1 had appendicitis. All scoring systems were found to be significantly different between survivors and nonsurvivors. The Glascow Coma Scale was 8.96 and 14.19, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. The SAPS II score was 67.00 and 37.62, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. The APACHE II score was 29.19 and 16.76, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. The SOFA score was 8.37 and 3.52, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. All other patient characteristics were not significantly different between nonsurvivors and survivors (Table 1). Based on the predictive model with the scoring systems using logistic regression analysis, the probability of death was calculated and compared between survivors and nonsurvivors. The probability of death in the model with the SAPS II score was found to be 0.269 and 0.790 (P b .0001) in survivors and nonsurvivors, respectively. The probability of death in the model with the APACHE II score was 0.336 and 0.739 (P b .0001) in survivors and nonsurvivors, respectively. The probability of death in the model with the SOFA score was 0.402 and 0.688 (P b .0001) in survivors and nonsurvivors, respectively (Table 2). The AUROC analysis demonstrated the predictability, in descending order, of the SAPS II, APACHE II, and SOFA scores as 0.9101, 0.8783, and 0.8086, respectively. The Hosmer-Lemeshow statistic P value of SAPS II, APACHE II, and SOFA scores are 0.709, 0.532, and 0.163, respectively (Figure). The highest accuracy rate was found in the model with
This is a retrospective data longitudinal analysis conducted at the ED of a tertiary medical center from December 2009 to December 2013. The ED of the tertiary care medical center has approximately 12000 visits per month. Patients older than 18 years who were admitted to the ED presented with acute abdominal pain and who then had undergone an abdominal CT scan with the findings of HPVG within this period were recruited. Initial keywords searched include portal venous gas, portal gas (air), and hepatic gas (air). The study was conducted following the institutional review board approval issued by the local ethical committee. 2.2. Data collection A total of 48 of 52 patients with HPVG were selected via a search engine examining all radiology data stored in the hospital emergency radiological database. Two patients were excluded from the study as they were younger than 18 years. Two other patients were excluded due to incomplete laboratory results, which may affect the result of the scoring systems. The parameters, including medical records with age; sex; clinical presentation; laboratory findings; radiographic imaging; relevant data; and the SAPS II, APACHE II, and SOFA scores, were assessed carefully. The worst laboratory values observed during the stay in the ED were recorded. All of the physiological scores were computed using these data. All unenhanced and contrast-enhanced CT scans were performed by using a helical CT scanner with 64-slice technology (Aquilion TSX-101A; Toshiba Medical Systems, Tokyo, Japan). All CT scans were performed and stored in our computer workflow management system (Centricity RIS; GE Healthcare, Milwaukee, WI). All of the patients in the study had a contrast-enhanced CT scan. The images of the CT scans from all of the patients in the study were reviewed, and the presence of HPVG was confirmed by a board-certified radiologist with 20 years of experience in abdominal imaging.
C-J. Seak et al. / American Journal of Emergency Medicine 32 (2014) 1481–1484
1483
Table 1 Comparison of findings between survivors and nonsurvivors Characteristic
Age Sex Male Female Body temperature (°C) Pulse rate (/min) Respiratory Rate (/min) Glascow Coma Scale⁎ SAPS II score⁎ APACHE II score⁎ SOFA score⁎
Total
Nonsurvivors
Survivors
n = 48
n = 27
n = 21
P
69 ± 15.89
72 ± 13.25
65.14 ±18.37
19 (39.58) 29 (60.42) 36.61 ± 1.26 101.56 ± 24.89 20.58 ± 4.78 11.25 ± 4.56 54.15 ± 21.97 23.75 ± 10.26 6.25 ± 4.60
9 (47.37) 18 (62.07) 36.51 ± 1.24 98.78 ± 25.77 21.33 ± 5.16 8.96 ± 4.82 67.00 ± 18.27 29.19 ± 8.73 8.37 ± 4.46
10 (52.63) 11 (37.93) 36.74 ± 1.29 105.14 ± 23.84 19.62 ± 4.17 14.19 ± 1.57 37.62 ± 13.81 16.76 ± 7.54 3.52 ± 3.16
.140 .480
.538 .385 .221 b.0001 b.0001 b.0001 b.0001
⁎ Indicates a statistically significant difference between the survivors and nonsurvivors.
4. Discussion Our study is the one of the largest analyses conducted in a population of adult patients with HPVG in the ED. In this study, we compared the performance of the SAPS II, APACHE II, and SOFA scores at the time of ED presentation of adult patients with HPVG in predicting their mortality. The SAPS II score is an appropriate index of severity for adult patients presenting to the ED with HPVG, as demonstrated in our study. Originally, the SAPS II score was designed to measure illness severity in patients older than 18 years who were admitted to the ICU. The score is calculated using the worst value from 17 variables for the initial 24-hour stay in the ICU [13]. Higher severity scores indicate more severe illness. Compared with the APACHE II and SOFA scores, the SAPS II score was found to have the highest sensitivity (92.6%) in predicting mortality in our patient population (Table 3). With a cutoff value of 50, 92.6% of our patients' mortality was accurately predicted. The high sensitivity and accuracy (83.3%) of the SAPS II scoring system may help EPs in prescreening adult HPVG patients presenting to the ED with high mortality rates. The SAPS II scoring systems may assist EPs and surgeons as it enables them to predict the mortality rate and to discuss the treatment options with the patients' families. Compared with the APACHE II score, the SAPS II score requires only a limited number of parameters and takes less time to calculate, and it is also easier to use. Clinical and laboratory data included in the SAPS II score are routinely collected in the ED. Le Gall et al [13] reported that collecting the data necessary to calculate the SAPS II score is very simple and quick and would take less than 5 minutes per patient. In our institute, the values were obtained from the ED database using a computerized system, and the results were obtained within 5 minutes per patient as reported by Le Gall et al. These advantages are particularly effective and important for EPs in clinical emergency settings and may help EPs to easily calculate mortality rates and to guarantee appropriate therapeutic approach in crowded EDs. The SAPS II score is reported to provide Table 2 The probability of death predicted by the SAPS II, APACHE II, and SOFA scoring systems Assessment tools
Survivors
Nonsurvivors
P
SAPS IIa APACHE IIb SOFAc
0.269 ± 0.282 0.336 ± 0.252 0.402 ± 0.220
0.790 ± 0.230 0.739 ± 0.238 0.688 ± 0.240
b.001 b.001 b.001
a b c
1 P ¼ 1þexp ½−ð−7:42þ0:15SAPS IIÞ. 1 P ¼ 1þexp ½−ð−4:4þ0:21APACHE IIÞ. 1 P ¼ 1þexp ½−ð−1:61þ0:34SOFA Þ.
an estimation of the risk of death without having to specify a primary diagnosis [13]. This is especially useful and practical for EPs in the prediction of the mortality rates of patients with HPVG in the ED because, in most situations, HPVG is revealed as a radiologic sign with no definite preliminary diagnosis [3,16,17]. Thus, the SAPS II score can be used in HPVG patients who are admitted to the ED even while their underlying etiology is not yet identified. Our study found that the SAPS II score estimates mortality risks with satisfactory accuracy in patients with HPVG in the ED. The SAPS II score had previously been reported to have excellent discrimination and calibration [18]. The assessment of illness severity with the SAPS II score had been demonstrated to be reliable not only in ICUs but also in intermediate care unit settings [19]. As a result from overcrowding, our EDs sometimes also function as intermediate care units, handling monitoring needs, and providing adequate treatment for critically ill patients at a lower cost. This may explain why the SAPS II score was predictive of mortality in our database with satisfactory discriminative power and calibration. The APACHE II score has been shown to predict outcomes accurately, not only in a variety of medical illnesses but also among surgical ICU patients [8]. The APACHE II score was reported to be a prognostic predictor in patients with ischemic bowel disease [20] and patients who develop acute abdominal complications in a medical ICU [21]. However, it was not superior to the SAPS score in our population. In addition, the APACHE II score is relatively complex compared with the SAPS II and
1.0
0.8
Sensitivity
the SAPS II score (83.3%), compared with models with the APACHE II (79.2%) or SOFA (79.2%) scores. The model with the SAPS II score had the highest sensitivity (92.6%) and the lowest specificity (71.4%) among the 3 scoring systems (Table 3).
0.6
0.4
0.2
0.0 0.0
0.2
0.4
0.6
0.8
1.0
1-Specificity
Figure. Receiver operating curves for predicting death according to the SAPS II, APACHE II, and SOFA scoring systems.
1484
C-J. Seak et al. / American Journal of Emergency Medicine 32 (2014) 1481–1484
Table 3 Sensitivities, specificities, and accuracy rates of the SAPS II, APACHE II, and SOFA scoring systems for predicting mortality a
Diagnostic values
SAPS II
Sensitivity Specificity Accuracy rates
25/27 (92.6%) 15/21 (71.4%) 40/48 (83.3%)
a
APACHE II
a
21/27 (77.8%) 17/21 (81.0%) 38/48 (79.2%)
SOFA
a
21/27 (77.8%) 16/21 (76.2%) 38/48 (79.2%)
The optimal cutoff point of SAPS II, APACHE II, and SOFA is 50, 21, and 5, respectively.
SOFA scores. Nevertheless, the APACHE II score may serve as a supplement tool for EPs in communications with surgeons. On the other hand, the APACHE II score may also be used by surgeons to discuss the risks and benefits of operative interventions with family members. The SOFA score is a well-known objective, simple, easily applied, and repeated scoring system. It is also predictive of mortality and/or morbidity in a variety of critically ill patients [14,22]. Despite the SOFA score being easier for EPs to calculate, its predictability of the mortality in our population was not as good as the APACHE II and SAPS II scoring systems. Our study found that the mortality of patients with HPVG can be precisely predicted using any of the 3 score systems upon ED presentation. Although all of these scores can be used as an objective parameter to identify HPVG patients who need close monitoring in the ED, the SAPS II score has the best overall performance compared with the 2 other scoring systems. It is also easier and faster to calculate than the APACHE II score. Moreover, the SAPS II score has strong calibration and discrimination power. These scores could be accurate and useful tools for EPs in the stratification of risk among adult HPVG patients presenting to the ED. This research is limited by the single-center retrospective nature of the study. Wider prospective applications of these scores should be conducted at multiple centers for further confirmation of the findings of this study. Although our study is one of the largest among adult ED patients with HPVG, a larger sample size is warranted for future confirmation of these findings.
5. Conclusion Our study is the first evaluation of the performance of the SAPS II and SOFA scores in predicting mortality of adult patients with HPVG in the ED. We also contributed to the validity of the APACHE II score in predicting the outcomes of adult patients with HPVG in the ED. This study found that these 3 predictive scoring systems provide potentially valuable prognostic information in evaluating the severity and predicting the mortality of adult HPVG patients presenting in the ED. The SAPS II score is less complex and more readily calculated than the APACHE II score and superior in prediction compared with the SOFA score. We recommend that the SAPS II score be used for outcome prediction and risk stratification in adult patients with HPVG in the ED.
References [1] Wolfe JN, Evans WA. Gas in the portal veins of the liver in infants; a roentgenographic demonstration with postmortem anatomical correlation. Am J Roentgenol Radium Ther Nucl Med 1955;74(3):486–8. [2] Monneuse O, Pilleul F, Barth X, Gruner L, Allaouchiche B, Valette PJ, et al. Portal venous gas detected on computed tomography in emergency situations: surgery is still necessary. World J Surg 2007;31(5):1065–71. [3] Abboud B, El Hachem J, Yazbeck T, Doumit C. Hepatic portal venous gas: physiopathology, etiology, prognosis and treatment. World J Gastroenterol 2009;15(29):3585–90. [4] Faberman RS, Mayo-Smith WW. Outcome of 17 patients with portal venous gas detected by CT. Am J Roentgenol 1997;169(6):1535–8. [5] Sellner F, Sobhian B, Baur M, Sellner S, Horvath B, Mostegel M, et al. Intermittent hepatic portal vein gas complicating diverticulitis—a case report and literature review. Int J Colorectal Dis 2007;22(11):1395–9. [6] Muscari F, Suc B, Lagarrigue J. Hepatic portal venous gas: is it always a sign of severity and surgical emergency? Chirurgie 1999;124(1):69–72. [7] Leibman PR, Patten MT, Manny J, Benfield JR, Hechtman HB. Hepatic-portal venous gas in adults: etiology, pathophysiology and clinical significance. Ann Surg 1978;187(3):281–7. [8] Hargrove J, Nguyen HB. Bench-to-bedside review: outcome predictions for critically ill patients in the emergency department. Crit Care Med 2005;9(4):376–83. [9] Nguyen HB, Rivers EP, Havstad S, Knoblich B, Ressler JA, Muzzin AM, et al. Critical care in the emergency department: a physiologic assessment and outcome evaluation. Acad Emerg Med 2000;7(12):1354–61. [10] O'Connor G, Geary U, Moriarty J. Critical care in the emergency department. Eur J Emerg Med 2009;16(6):296–300. [11] Nguyen HB, Banta JE, Cho TW, Van Ginkel C, Burroughs K, Wittlake WA, et al. Mortality predictions using current physiologic scoring systems in patients meeting criteria for early goal-directed therapy and the severe sepsis resuscitation bundle. Shock 2008;30(1):23–8. [12] Jones AE, Fitch MT, Kline JA. Operational performance of validated physiologic scoring systems for predicting in-hospital mortality among critically ill emergency department patients. Crit Care Med 2005;33(5):974–8. [13] Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993;270(24):2957–63. [14] Vincent JL, de Mendonça A, Cantraine F, Moreno R, Takala J, Suter PM, et al. Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on “sepsisrelated problems” of the European Society of Intensive Care Medicine. Crit Care Med 1998;26(11):1793–800. [15] Seak CJ, Hsu KH, Wong YC, Ng CJ, Yen DH, Seak JC, et al. The prognostic factors of adult patients with hepatic portal venous gas in the ED. Am J Emerg Med 2014;32 (9):972–5. [16] Kinoshita H, Shinozaki M, Tanimura H, Umemoto Y, Sakaguchi S, Takifuji K, et al. Clinical features and management of hepatic portal venous gas: four case reports and cumulative review of the literature. Arch Surg 2001;136(2):1410–4. [17] Shah PA, Cunningham SC, Morgan TA, Daly BD. Hepatic gas: widening spectrum of causes detected at CT and US in the interventional era. Radiographics 2011;31(5): 1403–13. [18] Castella X, Artigas A, Bion J, Kari A. A comparison of severity of illness scoring systems for intensive care unit patients: results of a multicenter, multinational study. The European/North American Severity Study Group. Crit Care Med 1995;23(8):1327–35. [19] Auriant I, Vinatier I, Thaler F, Tourneur M, Loirat P. Simplified acute physiology score II for measuring severity of illness in intermediate care units. Crit Care Med 1998;26 (8):1368–71. [20] Hsu HP, Shan YS, Hsieh YH, Sy ED, Lin PW. Impact of etiologic factors and APACHE II and POSSUM scores in management and clinical outcome of acute intestinal ischemic disorders after surgical treatment. World J Surg 2006;30(12):2152–62. [21] Gajic O, Urrutia LE, Sewani H, Schroeder DR, Cullinane DC, Peters SG. Acute abdomen in the medical intensive care unit. Crit Care Med 2002;30(6):1187–90. [22] Jones AE, Trzeciak S, Kline JA. The Sequential Organ Failure Assessment score for predicting outcome in patients with severe sepsis and evidence of hypoperfusion at the time of emergency department presentation. Crit Care Med 2009;37(5): 1649–54.
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Performance assessment of the Simplified Acute Physiology Score II, the Acute Physiology and Chronic Health Evaluation II score, and the Sequential Organ Failure Assessment score in predicting the outcomes of adult patients with hepatic portal venous gas in the ED Chen-June Seak, MD a,b,c, Chip-Jin Ng, MD a,b,⁎, David Hung-Tsang Yen, MD, PhD c,d, Yon-Cheong Wong, MD b,e, Kuang-Hung Hsu, PhD f,⁎⁎, Joanna Chen-Yeen Seak, MD g, Chen-Ken Seak, MBBS g a
Department of Emergency Medicine, Lin-Kou Medical Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan College of Medicine, Chang Gung University, Taoyuan, Taiwan Institute of Emergency and Critical Care Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan d Department of Emergency Medicine, Taipei Veterans General Hospital, Taipei, Taiwan e Division of Emergency and Critical Care Radiology, Department of Medical Imaging and Intervention, Lin-Kou Medical Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan f Laboratory for Epidemiology, Department of Health Care Management, and Healthy Aging Research Center, Chang Gung University, Taoyuan, Taiwan g Sarawak General Hospital, Kuching, Sarawak, Malaysia b c
a r t i c l e
i n f o
Article history: Received 4 July 2014 Received in revised form 2 September 2014 Accepted 9 September 2014
a b s t r a c t Objective: This study aims to evaluate the performance of Simplified Acute Physiology Score II (SAPS II), the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and the Sequential Organ Failure Assessment (SOFA) score for predicting illness severity and the mortality of adult hepatic portal venous gas (HPVG) patients presenting to the emergency department (ED). This will assist emergency physicians in risk stratification. Methods: Data for 48 adult HPVG patients who visited our ED between December 2009 and December 2013 were analyzed. The SAPS II, APACHE II score, and SOFA score were calculated based on the worst laboratory values in the ED. The probability of death was calculated for each patient based on these scores. The ability of the SAPS II, APACHE II score, and SOFA score to predict group mortality was assessed by using receiver operating characteristic curve analysis and calibration analysis. Results: The sensitivity, specificity, and accuracy were 92.6%,71.4%, and 83.3%, respectively, for the SAPS II method; 77.8%, 81%, and 79.2%, respectively, for the APACHE II scoring system, and 77.8%, 76.2%, and 79.2%, respectively, for the SOFA score. In the receiver operating characteristic curve analysis, the areas under the curve for the SAPS II, APACHE II scoring system, and SOFA score were 0.910, 0.878, and 0.809, respectively. Conclusion: This is one of the largest series performed in a population of adult HPVG patients in the ED. The results from the present study showed that SAPS II is easier and more quickly calculated than the APACHE II and more superior in predicting the mortality of ED adult HPVG patients than the SOFA. We recommend that the SAPS II be used for outcome prediction and risk stratification in adult HPVG patients in the ED. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Hepatic portal venous gas (HPVG) was first reported in 1955 by Wolfe and Evans [1]. In the past, the diagnosis of HPVG was challenging, as massive amounts of HPVG are required to be visible on plain films. However, the discovery of HPVG incidence has increased over the past ⁎ Correspondence to: CJ Ng, Department of Emergency Medicine, Lin-Kou Medical Center, Chang Gung Memorial Hospital, No. 5, Fusing St, Gueishan Township, Taoyuan County, 333 Taiwan. Tel.: +886 3 3281200x2505; fax: +886 3 3287715. ⁎⁎ Correspondence to: KH Hsu, Laboratory for Epidemiology, Department of Health Care Management, and Healthy Aging Research Center, Chang Gung University, No. 259, WenHwa 1st Rd, Gueishan Township, Taoyuan County, 333 Taiwan. Tel.: + 886 3 2118800x5486; fax: +886 3 2118138. E-mail addresses: [email protected] (C.-J. Ng), [email protected] (K.-H. Hsu). http://dx.doi.org/10.1016/j.ajem.2014.09.011 0735-6757/© 2014 Elsevier Inc. All rights reserved.
60 years due to improved diagnostic imaging and increased incidental findings detected by modern high-resolution computed tomography (CT) [2,3]. Historically, this rare radiologic finding has been associated with lethal surgical conditions [2] and high mortality rates ranging from 29% to 90% [4-7]. However, gradually over the years, some patients have survived after being treated conservatively depending on the etiology of the underlying disease [2]. Hence, risk stratification in the emergency department (ED) is particularly important to assist emergency physicians (EPs) in identifying critically ill patients with HPVG and informing family members about treatment options. However, to our knowledge, there is no established method which can precisely and easily predict the outcomes of adult patients with HPVG in the ED. The increasing patient volumes in the ED and limited bed availability in the intensive care unit (ICU) lead to the ED being used as the first
1482
C-J. Seak et al. / American Journal of Emergency Medicine 32 (2014) 1481–1484
contact for primary care. Consequently, the provision of critical care in the ED has increased [8-10]. Although early and aggressive intervention in critically ill patients in the ED is becoming increasingly important, adequate critical care in the ED is often difficult to provide and sometimes overlooked in the busy ED setting [8]. To improve critical care in the ED, the use of predictive scoring systems is considered potentially valuable for EPs for early disease recognition and prognostication of outcome to identify high-risk patients and adopt the appropriate intensive care intervention [8,11]. Existing ICU scoring systems were thus applied with some success [8,9,12]. Among the ICU scoring systems, the Simplified Acute Physiology Score II (SAPS II), the Acute Physiology and Chronic Health Evaluation II (APACHE II), and the Sequential Organ Failure Assessment (SOFA) scoring systems have been widely used for prognosticating outcomes [8,13,14]. However, there is limited dearth of literature describing the application of these scoring systems in HPVG patients despite their high mortality rate, and no studies have investigated the relationship between prognosis of HPVG patients and the SOFA and SAPS II scores at the time of ED presentation. In our previous study [15], we found that shock (odds ratio, 17.02; 95% confidence interval, 3.3686.22) and pneumatosis intestinalis (odds ratio, 5.14; 95% confidence interval, 1.03-25.67) were the significant predictors of mortality in adult patients with HPVG in the ED after the results were adjusted for age and sex. However, using the existing scoring systems to predict patient outcomes will be more feasible and less invasive and thus favorable in clinical practice. In this study, we further aimed to evaluate the performance of the SAPS II, APACHE II, and SOFA scoring systems in predicting illness severity and the mortality of adult HPVG patients presenting to ED due to the high HPVG mortality rate. This may help in risk stratification and permits earlier detection of adult HPVG patients in the ED who will benefit most from early and aggressive therapeutic intervention.
2.3. Etiology If surgery was implemented, intraoperative findings were integrated as the reference standard, and histopathology reports were evaluated to confirm the underlying disease. Clinical presentations and courses with discharge diagnoses and CT images were considered as the reference standard in patients who did not receive surgical interventions. 2.4. Statistical analysis Numerical and categorical variables are displayed as the means ± SD, and frequencies are shown as percentages (%). Univariate analyses were applied to examine the association between predictors and mortality, whereas categorical and numerical variables were analyzed with a χ 2 test and 2-sample t test, respectively. A logistic regression analysis was performed to develop predictive models between scoring systems and mortality. The probability of death was calculated based on the predictive models using the logit formula: P¼
1 1 þ exp ½−ðβ0 þ β1 X 1 Þ ðβ0 : Intercept; β1 : Parameter estimate of score; X 1 : ScoreÞ
A 2-sample t test was applied to compare the differences in death probability between survivors and nonsurvivors. Area under the receiver operating characteristic curve (AUROC) analysis was used to compare the predictability of mortality among scoring systems. The HosmerLemeshow statistic was used to measure model fitness of risk prediction. In addition, the sensitivity, specificity, and accuracy rates were calculated based on the optimal cutoff point derived from the AUROC analysis.
2. Materials and methods
3. Results
2.1. Design and setting
A total of 48 patients who met the inclusion criteria (admitted to ED with acute abdominal pain and subsequent CT findings of HPVG) were identified over 4 years. Among the 48 patients, 32 patients had ischemic bowel disease; 16, with nonischemic bowel disease. In the group of patients with nonischemic bowel disease, 7 patients had intra-abdominal abscesses, 3 had biliary tract infections, 4 had colitis, 1 had gastric ulcer, and 1 had appendicitis. All scoring systems were found to be significantly different between survivors and nonsurvivors. The Glascow Coma Scale was 8.96 and 14.19, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. The SAPS II score was 67.00 and 37.62, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. The APACHE II score was 29.19 and 16.76, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. The SOFA score was 8.37 and 3.52, a statistically significant difference (P b .0001), in nonsurvivors and survivors, respectively. All other patient characteristics were not significantly different between nonsurvivors and survivors (Table 1). Based on the predictive model with the scoring systems using logistic regression analysis, the probability of death was calculated and compared between survivors and nonsurvivors. The probability of death in the model with the SAPS II score was found to be 0.269 and 0.790 (P b .0001) in survivors and nonsurvivors, respectively. The probability of death in the model with the APACHE II score was 0.336 and 0.739 (P b .0001) in survivors and nonsurvivors, respectively. The probability of death in the model with the SOFA score was 0.402 and 0.688 (P b .0001) in survivors and nonsurvivors, respectively (Table 2). The AUROC analysis demonstrated the predictability, in descending order, of the SAPS II, APACHE II, and SOFA scores as 0.9101, 0.8783, and 0.8086, respectively. The Hosmer-Lemeshow statistic P value of SAPS II, APACHE II, and SOFA scores are 0.709, 0.532, and 0.163, respectively (Figure). The highest accuracy rate was found in the model with
This is a retrospective data longitudinal analysis conducted at the ED of a tertiary medical center from December 2009 to December 2013. The ED of the tertiary care medical center has approximately 12000 visits per month. Patients older than 18 years who were admitted to the ED presented with acute abdominal pain and who then had undergone an abdominal CT scan with the findings of HPVG within this period were recruited. Initial keywords searched include portal venous gas, portal gas (air), and hepatic gas (air). The study was conducted following the institutional review board approval issued by the local ethical committee. 2.2. Data collection A total of 48 of 52 patients with HPVG were selected via a search engine examining all radiology data stored in the hospital emergency radiological database. Two patients were excluded from the study as they were younger than 18 years. Two other patients were excluded due to incomplete laboratory results, which may affect the result of the scoring systems. The parameters, including medical records with age; sex; clinical presentation; laboratory findings; radiographic imaging; relevant data; and the SAPS II, APACHE II, and SOFA scores, were assessed carefully. The worst laboratory values observed during the stay in the ED were recorded. All of the physiological scores were computed using these data. All unenhanced and contrast-enhanced CT scans were performed by using a helical CT scanner with 64-slice technology (Aquilion TSX-101A; Toshiba Medical Systems, Tokyo, Japan). All CT scans were performed and stored in our computer workflow management system (Centricity RIS; GE Healthcare, Milwaukee, WI). All of the patients in the study had a contrast-enhanced CT scan. The images of the CT scans from all of the patients in the study were reviewed, and the presence of HPVG was confirmed by a board-certified radiologist with 20 years of experience in abdominal imaging.
C-J. Seak et al. / American Journal of Emergency Medicine 32 (2014) 1481–1484
1483
Table 1 Comparison of findings between survivors and nonsurvivors Characteristic
Age Sex Male Female Body temperature (°C) Pulse rate (/min) Respiratory Rate (/min) Glascow Coma Scale⁎ SAPS II score⁎ APACHE II score⁎ SOFA score⁎
Total
Nonsurvivors
Survivors
n = 48
n = 27
n = 21
P
69 ± 15.89
72 ± 13.25
65.14 ±18.37
19 (39.58) 29 (60.42) 36.61 ± 1.26 101.56 ± 24.89 20.58 ± 4.78 11.25 ± 4.56 54.15 ± 21.97 23.75 ± 10.26 6.25 ± 4.60
9 (47.37) 18 (62.07) 36.51 ± 1.24 98.78 ± 25.77 21.33 ± 5.16 8.96 ± 4.82 67.00 ± 18.27 29.19 ± 8.73 8.37 ± 4.46
10 (52.63) 11 (37.93) 36.74 ± 1.29 105.14 ± 23.84 19.62 ± 4.17 14.19 ± 1.57 37.62 ± 13.81 16.76 ± 7.54 3.52 ± 3.16
.140 .480
.538 .385 .221 b.0001 b.0001 b.0001 b.0001
⁎ Indicates a statistically significant difference between the survivors and nonsurvivors.
4. Discussion Our study is the one of the largest analyses conducted in a population of adult patients with HPVG in the ED. In this study, we compared the performance of the SAPS II, APACHE II, and SOFA scores at the time of ED presentation of adult patients with HPVG in predicting their mortality. The SAPS II score is an appropriate index of severity for adult patients presenting to the ED with HPVG, as demonstrated in our study. Originally, the SAPS II score was designed to measure illness severity in patients older than 18 years who were admitted to the ICU. The score is calculated using the worst value from 17 variables for the initial 24-hour stay in the ICU [13]. Higher severity scores indicate more severe illness. Compared with the APACHE II and SOFA scores, the SAPS II score was found to have the highest sensitivity (92.6%) in predicting mortality in our patient population (Table 3). With a cutoff value of 50, 92.6% of our patients' mortality was accurately predicted. The high sensitivity and accuracy (83.3%) of the SAPS II scoring system may help EPs in prescreening adult HPVG patients presenting to the ED with high mortality rates. The SAPS II scoring systems may assist EPs and surgeons as it enables them to predict the mortality rate and to discuss the treatment options with the patients' families. Compared with the APACHE II score, the SAPS II score requires only a limited number of parameters and takes less time to calculate, and it is also easier to use. Clinical and laboratory data included in the SAPS II score are routinely collected in the ED. Le Gall et al [13] reported that collecting the data necessary to calculate the SAPS II score is very simple and quick and would take less than 5 minutes per patient. In our institute, the values were obtained from the ED database using a computerized system, and the results were obtained within 5 minutes per patient as reported by Le Gall et al. These advantages are particularly effective and important for EPs in clinical emergency settings and may help EPs to easily calculate mortality rates and to guarantee appropriate therapeutic approach in crowded EDs. The SAPS II score is reported to provide Table 2 The probability of death predicted by the SAPS II, APACHE II, and SOFA scoring systems Assessment tools
Survivors
Nonsurvivors
P
SAPS IIa APACHE IIb SOFAc
0.269 ± 0.282 0.336 ± 0.252 0.402 ± 0.220
0.790 ± 0.230 0.739 ± 0.238 0.688 ± 0.240
b.001 b.001 b.001
a b c
1 P ¼ 1þexp ½−ð−7:42þ0:15SAPS IIÞ. 1 P ¼ 1þexp ½−ð−4:4þ0:21APACHE IIÞ. 1 P ¼ 1þexp ½−ð−1:61þ0:34SOFA Þ.
an estimation of the risk of death without having to specify a primary diagnosis [13]. This is especially useful and practical for EPs in the prediction of the mortality rates of patients with HPVG in the ED because, in most situations, HPVG is revealed as a radiologic sign with no definite preliminary diagnosis [3,16,17]. Thus, the SAPS II score can be used in HPVG patients who are admitted to the ED even while their underlying etiology is not yet identified. Our study found that the SAPS II score estimates mortality risks with satisfactory accuracy in patients with HPVG in the ED. The SAPS II score had previously been reported to have excellent discrimination and calibration [18]. The assessment of illness severity with the SAPS II score had been demonstrated to be reliable not only in ICUs but also in intermediate care unit settings [19]. As a result from overcrowding, our EDs sometimes also function as intermediate care units, handling monitoring needs, and providing adequate treatment for critically ill patients at a lower cost. This may explain why the SAPS II score was predictive of mortality in our database with satisfactory discriminative power and calibration. The APACHE II score has been shown to predict outcomes accurately, not only in a variety of medical illnesses but also among surgical ICU patients [8]. The APACHE II score was reported to be a prognostic predictor in patients with ischemic bowel disease [20] and patients who develop acute abdominal complications in a medical ICU [21]. However, it was not superior to the SAPS score in our population. In addition, the APACHE II score is relatively complex compared with the SAPS II and
1.0
0.8
Sensitivity
the SAPS II score (83.3%), compared with models with the APACHE II (79.2%) or SOFA (79.2%) scores. The model with the SAPS II score had the highest sensitivity (92.6%) and the lowest specificity (71.4%) among the 3 scoring systems (Table 3).
0.6
0.4
0.2
0.0 0.0
0.2
0.4
0.6
0.8
1.0
1-Specificity
Figure. Receiver operating curves for predicting death according to the SAPS II, APACHE II, and SOFA scoring systems.
1484
C-J. Seak et al. / American Journal of Emergency Medicine 32 (2014) 1481–1484
Table 3 Sensitivities, specificities, and accuracy rates of the SAPS II, APACHE II, and SOFA scoring systems for predicting mortality a
Diagnostic values
SAPS II
Sensitivity Specificity Accuracy rates
25/27 (92.6%) 15/21 (71.4%) 40/48 (83.3%)
a
APACHE II
a
21/27 (77.8%) 17/21 (81.0%) 38/48 (79.2%)
SOFA
a
21/27 (77.8%) 16/21 (76.2%) 38/48 (79.2%)
The optimal cutoff point of SAPS II, APACHE II, and SOFA is 50, 21, and 5, respectively.
SOFA scores. Nevertheless, the APACHE II score may serve as a supplement tool for EPs in communications with surgeons. On the other hand, the APACHE II score may also be used by surgeons to discuss the risks and benefits of operative interventions with family members. The SOFA score is a well-known objective, simple, easily applied, and repeated scoring system. It is also predictive of mortality and/or morbidity in a variety of critically ill patients [14,22]. Despite the SOFA score being easier for EPs to calculate, its predictability of the mortality in our population was not as good as the APACHE II and SAPS II scoring systems. Our study found that the mortality of patients with HPVG can be precisely predicted using any of the 3 score systems upon ED presentation. Although all of these scores can be used as an objective parameter to identify HPVG patients who need close monitoring in the ED, the SAPS II score has the best overall performance compared with the 2 other scoring systems. It is also easier and faster to calculate than the APACHE II score. Moreover, the SAPS II score has strong calibration and discrimination power. These scores could be accurate and useful tools for EPs in the stratification of risk among adult HPVG patients presenting to the ED. This research is limited by the single-center retrospective nature of the study. Wider prospective applications of these scores should be conducted at multiple centers for further confirmation of the findings of this study. Although our study is one of the largest among adult ED patients with HPVG, a larger sample size is warranted for future confirmation of these findings.
5. Conclusion Our study is the first evaluation of the performance of the SAPS II and SOFA scores in predicting mortality of adult patients with HPVG in the ED. We also contributed to the validity of the APACHE II score in predicting the outcomes of adult patients with HPVG in the ED. This study found that these 3 predictive scoring systems provide potentially valuable prognostic information in evaluating the severity and predicting the mortality of adult HPVG patients presenting in the ED. The SAPS II score is less complex and more readily calculated than the APACHE II score and superior in prediction compared with the SOFA score. We recommend that the SAPS II score be used for outcome prediction and risk stratification in adult patients with HPVG in the ED.
References [1] Wolfe JN, Evans WA. Gas in the portal veins of the liver in infants; a roentgenographic demonstration with postmortem anatomical correlation. Am J Roentgenol Radium Ther Nucl Med 1955;74(3):486–8. [2] Monneuse O, Pilleul F, Barth X, Gruner L, Allaouchiche B, Valette PJ, et al. Portal venous gas detected on computed tomography in emergency situations: surgery is still necessary. World J Surg 2007;31(5):1065–71. [3] Abboud B, El Hachem J, Yazbeck T, Doumit C. Hepatic portal venous gas: physiopathology, etiology, prognosis and treatment. World J Gastroenterol 2009;15(29):3585–90. [4] Faberman RS, Mayo-Smith WW. Outcome of 17 patients with portal venous gas detected by CT. Am J Roentgenol 1997;169(6):1535–8. [5] Sellner F, Sobhian B, Baur M, Sellner S, Horvath B, Mostegel M, et al. Intermittent hepatic portal vein gas complicating diverticulitis—a case report and literature review. Int J Colorectal Dis 2007;22(11):1395–9. [6] Muscari F, Suc B, Lagarrigue J. Hepatic portal venous gas: is it always a sign of severity and surgical emergency? Chirurgie 1999;124(1):69–72. [7] Leibman PR, Patten MT, Manny J, Benfield JR, Hechtman HB. Hepatic-portal venous gas in adults: etiology, pathophysiology and clinical significance. Ann Surg 1978;187(3):281–7. [8] Hargrove J, Nguyen HB. Bench-to-bedside review: outcome predictions for critically ill patients in the emergency department. Crit Care Med 2005;9(4):376–83. [9] Nguyen HB, Rivers EP, Havstad S, Knoblich B, Ressler JA, Muzzin AM, et al. Critical care in the emergency department: a physiologic assessment and outcome evaluation. Acad Emerg Med 2000;7(12):1354–61. [10] O'Connor G, Geary U, Moriarty J. Critical care in the emergency department. Eur J Emerg Med 2009;16(6):296–300. [11] Nguyen HB, Banta JE, Cho TW, Van Ginkel C, Burroughs K, Wittlake WA, et al. Mortality predictions using current physiologic scoring systems in patients meeting criteria for early goal-directed therapy and the severe sepsis resuscitation bundle. Shock 2008;30(1):23–8. [12] Jones AE, Fitch MT, Kline JA. Operational performance of validated physiologic scoring systems for predicting in-hospital mortality among critically ill emergency department patients. Crit Care Med 2005;33(5):974–8. [13] Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993;270(24):2957–63. [14] Vincent JL, de Mendonça A, Cantraine F, Moreno R, Takala J, Suter PM, et al. Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on “sepsisrelated problems” of the European Society of Intensive Care Medicine. Crit Care Med 1998;26(11):1793–800. [15] Seak CJ, Hsu KH, Wong YC, Ng CJ, Yen DH, Seak JC, et al. The prognostic factors of adult patients with hepatic portal venous gas in the ED. Am J Emerg Med 2014;32 (9):972–5. [16] Kinoshita H, Shinozaki M, Tanimura H, Umemoto Y, Sakaguchi S, Takifuji K, et al. Clinical features and management of hepatic portal venous gas: four case reports and cumulative review of the literature. Arch Surg 2001;136(2):1410–4. [17] Shah PA, Cunningham SC, Morgan TA, Daly BD. Hepatic gas: widening spectrum of causes detected at CT and US in the interventional era. Radiographics 2011;31(5): 1403–13. [18] Castella X, Artigas A, Bion J, Kari A. A comparison of severity of illness scoring systems for intensive care unit patients: results of a multicenter, multinational study. The European/North American Severity Study Group. Crit Care Med 1995;23(8):1327–35. [19] Auriant I, Vinatier I, Thaler F, Tourneur M, Loirat P. Simplified acute physiology score II for measuring severity of illness in intermediate care units. Crit Care Med 1998;26 (8):1368–71. [20] Hsu HP, Shan YS, Hsieh YH, Sy ED, Lin PW. Impact of etiologic factors and APACHE II and POSSUM scores in management and clinical outcome of acute intestinal ischemic disorders after surgical treatment. World J Surg 2006;30(12):2152–62. [21] Gajic O, Urrutia LE, Sewani H, Schroeder DR, Cullinane DC, Peters SG. Acute abdomen in the medical intensive care unit. Crit Care Med 2002;30(6):1187–90. [22] Jones AE, Trzeciak S, Kline JA. The Sequential Organ Failure Assessment score for predicting outcome in patients with severe sepsis and evidence of hypoperfusion at the time of emergency department presentation. Crit Care Med 2009;37(5): 1649–54.
