- Email: [email protected]
A multicenter, prospective evaluation of quality of care and mortality in Japan based on the Surviving Sepsis Campaign guidelines
J Infect Chemother 20 (2014) 115e120
Contents lists available at ScienceDirect
Journal of Infection and Chemotherapy journal homepage: http://www.elsevier.com/locate/jic
Original article
A multicenter, prospective evaluation of quality of care and mortality in Japan based on the Surviving Sepsis Campaign guidelines Seitaro Fujishima, MD *, Satoshi Gando, MD, FCCM, Daizoh Saitoh, MD, Toshihiko Mayumi, MD, Shigeki Kushimoto, MD, Shin-ichiro Shiraishi, MD, Hiroshi Ogura, MD, Kiyotsugu Takuma, MD, Joji Kotani, MD, Hiroto Ikeda, MD, Norio Yamashita, MD, Koichiro Suzuki, MD, Ryosuke Tsuruta, MD, Naoshi Takeyama, MD, Tsunetoshi Araki, MD, Yasushi Suzuki, MD, Yasuo Miki, MD, Yoshihiro Yamaguchi, MD, Naoki Aikawa, MD, FACS, Japanese Association for Acute Medicine Sepsis Registry (JAAM SR) Study Group1 Department of Emergency & Critical Care Medicine, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 March 2013 Received in revised form 29 August 2013 Accepted 6 September 2013
To elucidate the standard Surviving Sepsis Campaign (SSC) guidelines-based quality of care and mortality related to severe sepsis in Japan, we conducted a multicenter, prospective, observational study using a new web-based database between June 1, 2010, and December 31, 2011. A total of 1104 patients with severe sepsis were enrolled from 39 Japanese emergency and critical care centers. All-cause hospital mortality was 29.3% in patients with severe sepsis and 40.7% in patients with septic shock. Pulmonary, renal, hepatic, and hematological dysfunctions were associated with significantly higher mortality, and hematological dysfunction, especially coagulopathy, was associated with the highest odds ratio for mortality. Compliance with severe sepsis bundles in our study was generally low compared with that in a previous international sepsis registry study, and glycemic control was associated with lowest odds ratio for mortality. Despite higher complication rates of multiple organ dysfunction syndrome and low compliance with severe sepsis bundles on the whole, mortality in our study was similar to that in the international sepsis registry study. From these results, we concluded that our prospective multicenter study was successful in evaluating SSC guidelines-based standard quality of care and mortality related to severe sepsis in Japan. Although mortality in Japan was equivalent to that reported worldwide in the above-mentioned international sepsis registry study, compliance with severe sepsis bundles was low. Thus, there is scope for improvement in the initial treatment of severe sepsis and septic shock in Japanese emergency and critical care centers. Ó 2013, Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
keywords: Surviving Sepsis Campaign guidelines Septic shock Quality indicator Intensive care unit Emergency department Japanese association for acute medicine
1. Introduction Sepsis is prevalent among hospitalized patients worldwide. In fact, severe sepsis and septic shock are the major cause of admission and mortality in intensive care units (ICUs). However, mortality rates vary among countries and regions, depending on medical resources available. Japan has government health insurance and emergency medical systems but is also at the forefront of * Corresponding author. Tel.: þ81 3 3225 1323; fax: þ81 3 3353 2232. E-mail address: [email protected] (S. Fujishima). 1 See in Appendix.
countries with an aging population; therefore, morbidity and mortality statistics may not be comparable with those of other countries in which sepsis epidemiology has been demonstrated on a multicenter basis. A Japanese retrospective study examined incidence in a single emergency room and found that 4.5% of patients met the criteria for sepsis, 2.5% for severe sepsis, and 2.1% for septic shock [1]. In another retrospective study using Japan Nosocomial Infection Surveillances (JANIS), the incidence of sepsis was reported to be 2.1% among patients admitted to ICUs [2]; however, this value is very low compared with that observed in previous multicenter cohort studies, and it does not appear to be generalizable to nationwide ICUs accepting patients from emergency
1341-321X/$ e see front matter Ó 2013, Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jiac.2013.09.003
116
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
rooms, including our facilities. Prospective, multicenter-based, epidemiological studies of sepsis should have been conducted in Japan in order to advance the development of institutional and governmental strategies aimed at tackling this critical illness. In response to an increasing demand for a systematic approach to sepsis, the Society of Critical Care Medicine, the European Society of Intensive Care Medicine, and the International Sepsis Forum launched the Surviving Sepsis Campaign (SSC) in the hope of decreasing mortality [3]. Updated versions of guidelines developed by SSC were published in 2008 and 2013, in which the Japanese Association for Acute Medicine (JAAM) evolved as an organization comprising various professional societies. In accordance with SSC, a set of interventions, referred to as severe sepsis bundles, was introduced to change bedside behavior, and chart review database software was freely distributed in several languages to evaluate the quality of care in each hospital and to register patients worldwide [4]. However, because this software was not available in a Japanese version, almost no Japanese patients were included in the international sepsis registry; therefore, quality of care standards and mortality associated with severe sepsis and septic shock remain poorly clarified in Japan. The present study aimed to elucidate the epidemiology of severe sepsis and examine the SSC guidelines-based standard quality of care related to severe sepsis in Japan. To the best of our knowledge, this is the first report examining severe sepsis in Japan on the basis of a multicenter observational study. 2. Patients and methods This multicenter, prospective, observational study was registered at the University Hospital Medical Information Network Clinical Trial Registry (UMIN-CTR ID: UMIN000008195). Patients fulfilling the original and revised definition of severe sepsis or septic shock according to the American College of Chest Physicians/Society of Critical Care Medicine consensus conference were registered [5,6]. JAAM established a sepsis registry (SR) interim committee in 2007. The committee developed a web-based database in collaboration with the Internet Data and Information Center for Medical Research, a division of the University Hospital Medical Information Network. Since the JAAM SR database used in this study was equipped with functions equivalent to the SSC chart review database, it was capable of collecting the data necessary for evaluating quality indicators at individual institutes and sending anonymous patient data to the SSC office [4,7]. Data collected in this study were exactly the same as data collected in the international sepsis registry database and thus did not include information regarding parameters such as age, gender, comorbidities, Acute Physiology and Chronic Health Evaluation (APACHE) II score, Sequential Organ Failure Assessment (SOFA) score, or treatments other than severe sepsis bundles. Data regarding blood glucose, median BG values and times of hypoglycemic episodes during the initial 24 h were collected. This secure database is accessible through the Internet. Quality indicators were programmed to be calculated automatically as soon as data input from patients was completed, and included Q1: blood lactate measurement; Q2: blood culture; Q3: early antibiotic use; Q4: fluid resuscitation and vasopressors; Q5: achievement of CVP > 8 mmHg; Q6: achievement of ScvO2 > 70%; Q7: all resuscitation bundles (Q1eQ6); Q8: following glucocorticoid policy; Q9: following activated protein C (APC) policy (not evaluated in this study); Q10: glycemic control; Q11: low inspiratory plateau pressure; and Q12: all management bundles (Q8, 10, 11). In addition to the 18 core JAAM SR committee institutes, we recruited JAAM board-approved hospitals that could register patients in the database. All institutes and hospitals obtained approval from their ethics committees and recruited patients for 365
sequential days. Multiple organ dysfunction syndrome (MODS) was defined according to the SSC severe sepsis screening table [4]. Incidence of individual organ dysfunction in MODS, quality indicators, and their correlation with all-cause hospital mortality were evaluated. To evaluate the effect of organ dysfunction in MODS and compliance with individual severe sepsis bundles on mortality, univariate analysis was used. We also applied multivariate logistic regression analysis to MODS and two sets of quality indicators were applied to the same denominator, i.e., Q1, Q2, Q3, and Q10 in patients with severe sepsis and Q1, Q2, Q3, Q4, Q5, Q6, Q8, and Q10 in patients with septic shock. Simultaneous test procedures were used for analysis of MODS, and the stepwise backward elimination method (likelihood ratio) was used for evaluation of MODS subcategories and quality indicators. To compare values between two groups, Pearson’s chi-square test was used. To evaluate the effect of the total number of dysfunctional organs and systems on mortality, univariate analysis was used. Survival between two groups was compared using KaplaneMeier survival curves with a log-rank significance test. All statistical analyses were performed using IBM SPSS statistics (V20) software (International Business Machines Corp., NY, USA), and the level of significance was set at P < 0.05. 3. Results Using the JAAM SR database, 1104 patients with severe sepsis from 39 hospitals were registered between June 1, 2010, and December 31, 2011 (information of participated sites is shown in a supplementary table). Overall sepsis-related hospital mortality was 29.3%. Patients with septic shock had a significantly higher mortality (n ¼ 484, 40.7%) than those without (n ¼ 620, 20.3%, P < 0.0005). When our results were compared with the 30.8% mortality rate of 509 patients included in the final quarter of the SSC international sepsis registry study, the difference was not significant according to Pearson’s chi-square test (P ¼ 0.517) [7]. Table 1 shows patient demographics in our study. The emergency department was the most common route of admission, followed by the general ward. Pneumonia or empyema was the most common type of infection, followed by acute abdominal infection, urinary tract infection, and skin/soft tissue infection. Among MODS, hemodynamic dysfunction was the most frequent, followed by hematological, renal, pulmonary, and hepatic dysfunction. The average number of dysfunctional organs and systems reported in our study was 2.16 1.06 (mean standard deviation), which was significantly higher than that (1.94 1.01) reported in the international sepsis registry study, according to chi-square test mean scores with modified ridit scores (P < 0.0001). The incidence of individual organ dysfunction in our study was equivalent to or greater than that reported in previous studies, indicating a similar or higher severity of illness. We next analyzed the effects of MODS on mortality (Table 2). Existence of pulmonary, renal, hepatic, and hematological dysfunction was associated with significantly higher mortality. The mortality in patients with one, two, three, four, and five MODS was 19.7%, 28.0%, 34.7%, 48.0%, and 53.6%, respectively, and each additional MODS was associated with an approximately 1.5-fold increase in mortality (Table 2). We also assessed the effect of the following MODS subcategory abnormalities on mortality: hypotension, hyperlactatemia, acute lung injury, thrombocytopenia, and coagulopathy. Hypotension, thrombocytopenia, and coagulopathy were associated with higher mortality. We further performed multivariate logistic regression analysis. When the effects of individual organ dysfunction and system in MODS were analyzed, pulmonary, renal, and hematological dysfunction were extracted as significant parameters (Table 3A), and when MODS subcategory abnormalities in addition to renal and hepatic dysfunction were
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120 Table 1 Patient demographics. Patient number Route of admission Emergency department 671 General ward 316 ICU 117 Type of infection Pneumonia/empyema 436 Acute abdominal infection 268 Urinary tract infection 160 Skin/soft tissue inflammation 110 Meningitis 21 Blood stream catheter infection 20 Bone/joint infection 20 Wound infection 13 Endocarditis 12 Other infection 44 Baseline organ dysfunction (MODS) Hemodynamic 928 Pulmonary 374 Renal 412 Hepatic 184 Hematological 483 Number of dysfunctional organs and systems 1 361 2 364 3 248 4 102 5 28
Percentage 60.8 28.6 10.6 39.5 24.3 14.5 10.0 1.9 1.8 1.8 1.2 1.1 4.0 84.1 33.9 37.3 16.7 43.8 32.7 33.0 22.5 9.2 2.5
ICU: intensive care unit, Hemodynamic: systolic blood pressure <90 mmHg or mean blood pressure <65 mmHg or systolic blood pressure decrease >40 mmHg from baseline or blood lactate >2 mmol/L, Pulmonary: bilateral infiltration on chest X-ray with PaO2/FIO2 <300 or oxygenation required to maintain SpO2 >90%, Renal: Serum creatinine >2 mg/dl, Hepatic: serum bilirubin >2 mg/dl, Hematological: peripheral blood platelet counts <100,000/ml or prothrombin time-international normalized ratio >1.5 or activated partial thromboplastin time >60 s.
analyzed, acute lung injury, renal dysfunction, thrombocytopenia, and coagulopathy were extracted as significant parameters (Table 3B). In Table 4, the severe sepsis quality indicators in our study are listed alongside those from the international sepsis registry study [7]. Among these, the rate of Q3, Q4, Q5, Q6, Q7, and Q8 were lower, Table 2 Effects of MODS on mortality, by univariate analysis. Individual MODS
OR
95% CI
P value
Hemodynamic Hypotension Hyperlactatemia Pulmonary Acute lung injury Renal Hepatic Hematological Thrombocytopenia Coagulopathy Each Additional MODS
1.245 1.389 1.218 1.323 1.270 1.847 1.485 2.022 1.730 2.197 1.506
0.863e1.796 1.064e1.814 0.924e1.605 1.010e1.734 0.959e1.680 1.418e2.406 1.065e2.072 1.555e2.629 1.321e2.266 1.640e2.942 1.332e1.702
0.241 0.016 0.163 0.042 0.095 <0.001 0.020 <0.001 <0.001 <0.001 <0.001
OR: odds ratio, 95% CI: 95% confidence interval, MODS: multiple organ dysfunction syndrome, Hemodynamic: systolic blood pressure <90 mmHg or mean blood pressure <65 mmHg or systolic blood pressure decrease >40 mmHg from baseline or blood lactate >2 mmol/L, Hypotension: systolic blood pressure <90 mmHg or mean blood pressure <65 mmHg or systolic blood pressure decrease >40 mmHg from baseline, Hyperlactatemia: blood lactate >2 mmol/L, Pulmonary: bilateral infiltration on chest X-ray with PaO2/FIO2 <300 or oxygen required to maintain SpO2 >90%, Acute lung injury: bilateral infiltration on chest X-ray with PaO2/FIO2 <300, Renal: Serum creatinine >2 mg/dl, Hepatic: serum bilirubin >2 mg/dl, Hematological: peripheral blood platelet counts <100,000/ml or prothrombin timeinternational normalized ratio >1.5 or activated partial thromboplastin time >60 s, Thrombocytopenia: peripheral blood platelet counts <100,000/ml, Coagulopathy: prothrombin time-international normalized ratio >1.5 or activated partial thromboplastin time >60 s.
117
Table 3 Effects of MODS on mortality, by multivariate logistic regression analysis. A) Effects of MODS. B) Effects of MODS subcategory abnormalities in addition to renal and hepatic dysfunctions. A. MODS Hemodynamic Pulmonary Renal Hepatic Hematological
OR 1.229 1.457 1.628 1.152 1.836
95% CI 0.847e1.783 1.103e1.925 1.237e2.143 0.808e1.642 1.386e2.431
P value 0.279 0.008 0.001 0.435 <0.001
OR 1.359 1.626 1.378 1.912
95% CI 1.019e1.813 1.234e2.143 1.034e1.838 1.410e2.593
P value 0.038 0.001 0.029 <0.001
B. Subcategory abnormalities Acute lung injury Renal Thrombocytopenia Coagulopathy
OR: odds ratio, 95% CI: 95% confidence interval, MODS: multiple organ dysfunction syndrome, five subcategory abnormalities: hypotension, hyperlactatemia, acute lung injury, thrombocytopenia, and coagulopathy. Simultaneous test procedure was used in Table A, and stepwise backward-elimination method (likelihood ratio) in Table B. For details, please see Table 2.
whereas that of Q10 was equivalent and that for Q11 and Q12 was higher than that reported in the international sepsis registry study. We then analyzed whether compliance with each severe sepsis bundle could influence outcome. Univariate analysis indicated that compliance with Q3, Q4, Q6, Q7, Q8, Q10, and Q12 was associated with significantly lower mortality (Table 5). When analysis was restricted to patients with septic shock, Q6, Q8, and Q12 were found to be quality indicators affecting outcome (data not shown). When KaplaneMeier analysis was applied, the results were similar: compliance with Q4, Q6, Q8, Q10, and Q12 was confirmed as being associated with significantly better survival (Fig. 1). We further applied multivariate logistic regression analysis to elucidate the association of mortality with two sets of quality indicators (Table 6). Among patients with severe sepsis, Q10 was associated with
Table 4 Severe sepsis bundles quality indicator: Japan (JAAM) vs. international. Quality indicators Q1: Blood lactate measurement (severe sepsis) Q2: Blood culture (severe sepsis) Q3: Early antibiotic use (severe sepsis) Q4: Fluid resuscitation and vasopressors (hypotension or lactate > 4 mmol/L) Q5: Achievement of CVP over 8 mmHg (septic shock or lactate >4 mmol/L) Q6: Achievement of ScvO2 > 70% (septic shock or lactate >4 mmol/L) Q7: All resuscitation bundles (Q1eQ6) (severe sepsis) Q8: Following glucocorticoid policy (septic shock) Low-dose glucocorticoid use (septic shock) Q9: Following activated protein C (APC) policy (septic shock) Q10: Glycemic control (severe sepsis) Q11: Low inspiratory plateau pressure (requiring mechanical ventilation) Q12: All management bundles (Q8, 10, 11) (severe sepsis) Q13: Sepsis-related mortality
Japan (JAAM)
Internationala (final quarter)
987
81.4%
78.7%
784 692 791
81.4% 48.8% 47.2%
78.3% 67.9% 77.0%
667
21.0%
38.0%
667
10.3%
24.3%
534
11.2%
21.5%
484
36.2%
73.9%
484
18.2%
NA
NA
53.5%
674 703
51.3% 94.2%
56.8% 83.8%
675
39.7%
25.5%
29.3
34.8
Patient number
0
1104
JAAM: Japanese Association for Acute Medicine, CVP: central venous pressure, ScvO2: central venous oxygen saturation, NA: not available. Parentheses indicate target patients to whom each quality indicator of the severe sepsis bundle was applied to calculate the quality indicator. a Adopted from Ref. [7].
118
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
Table 5 Quality indicators vs. hospital mortality, by univariate analysis. Quality indicator (Q)
OR
Q1. Blood lactate measurement (severe sepsis) Q2. Blood culture (severe sepsis) Q3. Early antibiotic use (severe sepsis) Q4. Fluid resuscitation & vasopressors (hypotension or lactate >4 mmol/L) Q5: Achievement of central venous pressure over 8 mmHg (septic shock or lactate > 4 mmol/L) Q6: Achievement of central venous oxygen saturation>70% (severe sepsis) Q7. All resuscitation bundles (severe sepsis) Q8. Following glucocorticoid policy (septic shock) Low dose glucocorticoid use (septic shock) Q10. Glycemic control (severe sepsis) Q11. Low inspiratory plateau pressure (requiring mechanical ventilation) Q12. All management bundles (severe sepsis)
0.850 0.602e1.201
95% CI
0.358
0.897 0.605e1.330 0.677 0.483e0.949 0.633 0.470e0.852
0.588 0.023 0.003
0.679 0.454e1.014
0.059
0.483 0.270e0.865
0.014
0.453 0.630 1.199 0.614 0.992
0.035 0.019 0.446 0.004 0.981
0.217e0.945 0.429e0.927 0.752e1.911 0.440e0.856 0.515e1.190
P value
0.447 0.312e0.640 <0.001
OR: odds ratio, 95% CI: 95% confidence interval. Parentheses indicate target patients to whom each quality indicator of a severe sepsis bundle was applied to calculate the quality indicator.
mortality, but the difference was not significant. In patients with septic shock, Q8 was extracted as significant. To further analyze the efficacy of Q10, we examined the effect of hypoglycemia on mortality using univariate analysis. Blood glucose control <150 mg/dL, regardless of hypoglycemic episodes, did not affect the outcome (OR: 0.987, 95% CI: 0.689e1.415, P ¼ 0.943). In addition, in patients whose median blood glucose was controlled <150 mg/dL (n ¼ 408), the number of hypoglycemic episodes significantly correlated with outcome (OR: 1.519, 95% CI: 1.168e1.976, P ¼ 0.002). 4. Discussion Between June 1, 2010, and December 31, 2011, 1104 patients with severe sepsis and septic shock from 39 Japanese hospitals were registered using a new web-based database equivalent to the SSC chart review database. Compliance with severe sepsis bundles was generally low in our study compared with that in the international sepsis registry study, and glycemic control was associated with the lowest OR for mortality. Despite higher complication rates of MODS and low compliance with severe sepsis bundles on the whole, mortality was equivalent to that in the international sepsis registry study. To the best of our knowledge, this study is the first prospective multicenter study in Japan to achieve success in evaluating the standard quality of care on the basis of SSC guidelines. In our study, patients with four MODS, namely pulmonary, renal, hepatic, and hematological, were associated with significantly higher mortality. In addition, mortality increased in proportion to the total number of dysfunctional organs and systems. Among these MODS, hematological dysfunction, especially coagulopathy, was associated with the highest OR for mortality by both univariate and multivariate logistic regression analyses. In contrast, a previous large French epidemiological study by Annane et al. indicated that acute lung injury/acute respiratory distress syndrome and renal failure significantly affected patient outcome; however, they did not mention hematological dysfunction [8]. In a Spanish multicenter study, neurological, respiratory, renal, and hematological dysfunction, as assessed by the organ dysfunction system score, was associated with a significantly higher mortality rate. Furthermore, OR for hematological dysfunction was the highest [9]. These results, including those of the present study, may indicate that hematological dysfunction could be critical in determining the outcome of severe sepsis and septic shock. Although recombinant human APC is
no longer available, the attenuation of abnormal coagulopathy might still be a potential approach for treating sepsis. Of late, numerous SSC-based, worldwide multicenter studies have been published with mortality ranging from 27.0% to 44.5% [7,10e15]. In our study, mortality and the average number of dysfunctional organs were 29.3% and 2.2, respectively, with 11.2% compliance for all sepsis resuscitation bundles and 39.7% compliance for all sepsis management bundles, with the exception of adequate use of APC. Levy et al. reported the results of the international sepsis registry collected from 252 qualifying sites. Original and MODS number-adjusted mortality was 30.8% and 35.4%, respectively, with the average number of dysfunctional organs being 1.9; in addition, sepsis resuscitation and management bundle compliance was 31.3% and 36.1%, respectively, for the final quarter [7]. When comparing compliance with individual sepsis resuscitation and management bundles in our study with that reported in the international sepsis registry study, compliance with antibiotic use, fluid resuscitation and vasopressors, central venous pressure achievement, central venous oxygen saturation achievement, and steroid policy was low, but achievement of all these bundles was associated with better outcomes in our study. The reason for the discrepancy observed between fairly low mortality and low bundle compliance in the present study is unclear. The discrepancy might be attributed to a difference in the case mix of patients, including comorbidities and ages of patients between the 2 studies, even if severity of illness as assessed by the number of dysfunctional organs were adjusted. In addition, another prospective observational study was conducted in parallel to accumulate more precise clinical information from selected JAAM board-approved hospitals. Other factors that improved the outcome may be revealed by this study. In our study, compliance with five severe sepsis bundles, namely early antibiotic use, fluid resuscitation and vasopressors, achievement of central venous oxygen saturation >70%, following glucocorticoid policy, and glycemic control, was associated with better outcomes. Of these specific bundles, evidence of effectiveness has been accumulated for early antibiotic use [16e18] and fluid resuscitation and vasopressors [19]. In contrast, the efficacy of the other three bundles remains controversial. However, compliance with glycemic control was associated with the lowest OR for mortality in our study according to univariate analysis along with that noted in the international sepsis registry study [7]. Although significance was not achieved by multivariate analysis, we speculate that this was because of our small patient number and missing data. In the international sepsis registry study and the present study, a median blood glucose level maintained <150 mg/dL and its lowest value being higher than the lower limit of normal was considered good compliance. Although previous randomized controlled trials have shown no efficacy or adverse effects associated with tight glycemic control in critically ill patients [20e22], the effectiveness of glycemic control might have been canceled out by the disadvantages of hypoglycemia and variations in blood glucose levels in these studies. Also, the efficacy of intermediate glycemic control has not been determined yet [23e25]. In fact, we could not establish the efficacy of controlling blood glucose <150 mg/dL when patients with more than one episodes of hypoglycemia were included. In patients whose blood glucose was controlled <150 mg/dL, the number of hypoglycemic episodes significantly correlated with mortality. Therefore, we believe that the efficacy of intermediate glycemic control, in combination with a technique for avoiding hypoglycemia and blood glucose variation, needs to be examined in future studies. The present study had several limitations. First, although we demonstrated significant bundle factors associated with improved clinical outcomes in multivariate analysis, other possible risk factors affecting clinical outcomes such as Acute Physiology and Chronic Health Evaluation II score, old age, and comorbidity were
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
119
Fig. 1. KaplaneMeier Analysis of Compliance with Severe Sepsis Bundles. There are significant differences in survival between the groups with (black line) or without (gray line) fluid resuscitation and vasopressors (Q4), achievement of central venous oxygen saturation over 70% (Q6), following glucocorticoid policy (Q8), achievement of glycemic control (Q10), as shown in the graphs along with all management bundles (Q12).
not included in the analysis. In addition, the efficacy of commonly used treatments in Japan such as administration of immunoglobulin and continuous renal replacement therapy for non-renal indication, which may have had an impact on the results, were not evaluated in the study. Second, we included patients who met inclusion criteria equivalent to those in the SSC chart review database. However, the number of patients recruited from some hospitals was low, and we could not rule out the possibility of selection bias associated with inclusion at these facilities. Furthermore, 44% of patients were recruited from hospitals participating in this study under their own initiative, and thus, these may not Table 6 Quality indicators vs. hospital mortality in severe sepsis and septic shock by multivariate analysis. Quality indicator (Q) Severe sepsis (n ¼ 389) Q10. Glycemic control Septic shock (n ¼ 176) Q8. Following glucocorticoid policy
OR
95% CI
P value
0.727
0.468e1.127
0.154
0.468
0.253e0.867
0.016
OR: odds ratio, 95% CI: 95% confidence interval. Since target patients were not equal among quality indicators, two sets of quality indicators were applied to the same denominator (i.e., Q1, Q2, Q3, and Q10 in patients with severe sepsis, and Q1, Q2, Q3, Q4, Q5, Q6, Q8, and Q10 in patients with septic shock).
represent the average Japanese hospital, which could be another source of selection bias. Third, our study was prospective but observational and used ORs to evaluate the correlation between outcome and MODS or compliance with severe sepsis bundles, which limited its merits as compared with randomized controlled studies. Fourth, our study recruited patients during a relatively short period and thus could not evaluate quality improvement. Despite these limitations, this multicenter study, the first of its kind in Japan, was successful in evaluating standard quality of care and mortality related to severe sepsis. In conclusion, we were successful in evaluating the SSC guideline-based standard quality of care for severe sepsis and septic shock patients in Japan in a prospective multicenter study. Although mortality related with severe sepsis was determined to be equivalent to that reported worldwide in the above-mentioned international sepsis registry study, despite low compliance with severe sepsis bundles, further improvement in the quality of care may result in even better patient outcomes.
Acknowledgments This study was funded and supported by the Japanese Association for Acute Medicine (JAAM), and in part by a Health Labor
120
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
Sciences Research Grant (H22-Shinko-Ippan-013). Part of the study was presented at the 39th and 40th JAAM annual conferences and at the 2012 American Thoracic Society International Conference. We thank Dr. Takayuki Abe (Center for Clinical Research, School of Medicine, Keio University) for help with the statistical analyses, and Mr. Kiyohiko Sato (Japanese Association for Acute Medicine) for administrative work.
Conflict of interest None.
Appendix Japanese Association for Acute Medicine Sepsis Registry (JAAMSR) Study Group: Hideharu Hagiya, MD, Emergency Unit and Critical Care Center, Tsuyama Central Hospital; Yoshihiro Tada, MD, Emergency Department, Hiroshima Prefectural Hospital; Akiko Oshiro, MD, Division of Emergency Medicine, Teine Keijinkai Hospital; Akihiro Noda, MD, Division of Emergency and Critical Care Medicine, Department of Acute Medicine, Nihon University School of Medicine; Satoru Futami, MD, Department of Emergency Medicine and Critical Care, Cancer Institute Hospital, Japanese Foundation for Cancer Research; Shunsuke Otani, MD, Department of Emergency and Critical Care Medicine, Kimitsu Chuo Hospital; Takeshi Takahashi, MD, Department of Emergency and Critical Care Medicine, Kumamoto Medical Center, National Hospital Organization; Minoru Nakano, MD, Advanced Medical Emergency and Critical Care Center, Japanese Red Cross Maebashi Hospital; Takuya Oda, MD, Department of General Internal Medicine, Iizuka Hospital; Yuichiro Sakamoto, MD, Department of Emergency and Critical Care Medicine, Saga University Hospital, Shigeaki Inoue, MD, Department of Emergency and Critical Care Medicine, Tokai University School of Medicine; Kenji Kubo, MD, Department of Emergency and Critical Care Medicine/Department of Infectious Diseases, Japanese Red Cross Society Wakayama Medical Center; Takeshi Kasai, MD, Department of Emergency and Trauma Center, Kameda Medical Center; Nobuyuki Saito, MD, Department of Emergency and Critical Care Medicine, Chiba-Hokusoh Hospital, Nippon Medical School; Masamune Kuno, MD, Department of Emergency and Critical Care Medicine, Tama Nagayama Hospital Nippon Medical School; Hideaki Anan, MD, Emergency Medical Center, Fujisawa City Hospital; Hiroshi Moro, MD, Advanced Disaster Medical and Emergency Critical Care Center, Niigata University Medical and Dental Hospital; Toshiaki Mochizuki, MD, Department of Emergency and Disaster Medicine, Hamamatsu University School of Medicine; Yasushi Takasaki, MD, Nan-yo Emergency Critical Care Center, Uwajima City Hospital; Hideki Tokuyama, MD, Department of Emergency and Critical Care Medicine, Komaki City Hospital; Katsuyuki Maruyama, MD, Department of Emergency and Critical Care Medicine, Kinki University School of Medicine; Hiroshi Morita, MD, Department of Emergency Medicine, Osaka Medical College Hospital; Yoshimi Kitagawa, MD, Emergency Department, Nagoya Ekisaikai Hospital; Hiroshi Chiba, MD, Center for Emergency and Critical Care Medicine, Hachinohe City Hospital.
Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jiac.2013.09.003.
References [1] Sun D, Aikawa N. The natural history of the systemic inflammatory response syndrome and the evaluation of SIRS criteria as a predictor of severity in patients hospitalized through emergency services. Keio J Med 1999;48:28e37. [2] Suka M, Yoshida K, Takezawa J. Incidence and outcome of sepsis in Japanese intensive care units: the Japanese nosocomial infection surveillance system. Environ Health Preventive Med 2006;11:298e303. [3] Dellinger RP, Carlet JM, Masur H, Gerlach H, Calandra T, Cohen J, et al. Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Intensive Care Med 2004;30:536e55. [4] Surviving Sepsis Campaign Website. Available from: www.survivingsepsis.org. [5] Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS international sepsis definitions conference. Intensive Care Med 2003;29:530e8. [6] Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992;101: 1644e55. [7] Levy MM, Dellinger RP, Townsend SR, Linde-Zwirble WT, Marshall JC, Bion J, et al. The Surviving Sepsis Campaign: results of an international guidelinebased performance improvement program targeting severe sepsis. Crit Care Med 2010;38:367e74. [8] Annane D, Aegerter P, Jars-Guincestre MC, Guidet B, Network CU- R. Current epidemiology of septic shock: the CUB-Rea Network. Am J Respir Crit Care Med 2003;168:165e72. [9] Blanco J, Muriel-Bombin A, Sagredo V, Taboada F, Gandia F, Tamayo L, et al. Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study. Crit Care 2008;12:R158. [10] Tromp M, Tjan DH, van Zanten AR, Gielen-Wijffels SE, Goekoop GJ, van den Boogaard M, et al. The effects of implementation of the Surviving Sepsis Campaign in the Netherlands. Neth J Med 2011;69:292e8. [11] Phua J, Koh Y, Du B, Tang YQ, Divatia JV, Tan CC, et al. Management of severe sepsis in patients admitted to Asian intensive care units: prospective cohort study. BMJ 2011;342:d3245. [12] Lefrant JY, Muller L, Raillard A, Jung B, Beaudroit L, Favier L, et al. Reduction of the severe sepsis or septic shock associated mortality by reinforcement of the recommendations bundle: a multicenter study. Ann Fr Anesth Reanim 2010;29:621e8. [13] Castellanos-Ortega A, Suberviola B, Garcia-Astudillo LA, Holanda MS, Ortiz F, Llorca J, et al. Impact of the Surviving Sepsis Campaign protocols on hospital length of stay and mortality in septic shock patients: results of a three-year follow-up quasi-experimental study. Crit Care Med 2010;38:1036e43. [14] Cardoso T, Carneiro AH, Ribeiro O, Teixeira-Pinto A, Costa-Pereira A. Reducing mortality in severe sepsis with the implementation of a core 6-hour bundle: results from the Portuguese community-acquired sepsis study (SACiUCI study). Crit Care 2010;14:R83. [15] Ferrer R, Artigas A, Levy MM, Blanco J, Gonzalez-Diaz G, Garnacho-Montero J, et al. Improvement in process of care and outcome after a multicenter severe sepsis educational program in Spain. JAMA 2008;299:2294e303. [16] Puskarich MA, Trzeciak S, Shapiro NI, Arnold RC, Horton JM, Studnek JR, et al. Association between timing of antibiotic administration and mortality from septic shock in patients treated with a quantitative resuscitation protocol. Crit Care Med 2011;39:2066e71. [17] Gaieski DF, Mikkelsen ME, Band RA, Pines JM, Massone R, Furia FF, et al. Impact of time to antibiotics on survival in patients with severe sepsis or septic shock in whom early goal-directed therapy was initiated in the emergency department. Crit Care Med 2010;38:1045e53. [18] Ferrer R, Artigas A, Suarez D, Palencia E, Levy MM, Arenzana A, et al. Effectiveness of treatments for severe sepsis: a prospective, multicenter, observational study. Am J Respir Crit Care Med 2009;180:861e6. [19] Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, et al. Early goaldirected therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368e77. [20] Investigators N-SS, Finfer S, Chittock DR, Su SY, Blair D, Foster D, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283e97. [21] Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 2008;358:125e39. [22] Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449e61. [23] Investigators N-SS, Finfer S, Liu B, Chittock DR, Norton R, Myburgh JA, et al. Hypoglycemia and risk of death in critically ill patients. N Engl J Med 2012;367:1108e18. [24] Hermanides J, Bosman RJ, Vriesendorp TM, Dotsch R, Rosendaal FR, Zandstra DF, et al. Hypoglycemia is associated with intensive care unit mortality. Crit Care Med 2010;38(6):1430e4. [25] Ali NA, O’Brien Jr JM, Dungan K, Phillips G, Marsh CB, Lemeshow S, et al. Glucose variability and mortality in patients with sepsis. Crit Care Med 2008;36:2316e21.
Contents lists available at ScienceDirect
Journal of Infection and Chemotherapy journal homepage: http://www.elsevier.com/locate/jic
Original article
A multicenter, prospective evaluation of quality of care and mortality in Japan based on the Surviving Sepsis Campaign guidelines Seitaro Fujishima, MD *, Satoshi Gando, MD, FCCM, Daizoh Saitoh, MD, Toshihiko Mayumi, MD, Shigeki Kushimoto, MD, Shin-ichiro Shiraishi, MD, Hiroshi Ogura, MD, Kiyotsugu Takuma, MD, Joji Kotani, MD, Hiroto Ikeda, MD, Norio Yamashita, MD, Koichiro Suzuki, MD, Ryosuke Tsuruta, MD, Naoshi Takeyama, MD, Tsunetoshi Araki, MD, Yasushi Suzuki, MD, Yasuo Miki, MD, Yoshihiro Yamaguchi, MD, Naoki Aikawa, MD, FACS, Japanese Association for Acute Medicine Sepsis Registry (JAAM SR) Study Group1 Department of Emergency & Critical Care Medicine, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 March 2013 Received in revised form 29 August 2013 Accepted 6 September 2013
To elucidate the standard Surviving Sepsis Campaign (SSC) guidelines-based quality of care and mortality related to severe sepsis in Japan, we conducted a multicenter, prospective, observational study using a new web-based database between June 1, 2010, and December 31, 2011. A total of 1104 patients with severe sepsis were enrolled from 39 Japanese emergency and critical care centers. All-cause hospital mortality was 29.3% in patients with severe sepsis and 40.7% in patients with septic shock. Pulmonary, renal, hepatic, and hematological dysfunctions were associated with significantly higher mortality, and hematological dysfunction, especially coagulopathy, was associated with the highest odds ratio for mortality. Compliance with severe sepsis bundles in our study was generally low compared with that in a previous international sepsis registry study, and glycemic control was associated with lowest odds ratio for mortality. Despite higher complication rates of multiple organ dysfunction syndrome and low compliance with severe sepsis bundles on the whole, mortality in our study was similar to that in the international sepsis registry study. From these results, we concluded that our prospective multicenter study was successful in evaluating SSC guidelines-based standard quality of care and mortality related to severe sepsis in Japan. Although mortality in Japan was equivalent to that reported worldwide in the above-mentioned international sepsis registry study, compliance with severe sepsis bundles was low. Thus, there is scope for improvement in the initial treatment of severe sepsis and septic shock in Japanese emergency and critical care centers. Ó 2013, Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
keywords: Surviving Sepsis Campaign guidelines Septic shock Quality indicator Intensive care unit Emergency department Japanese association for acute medicine
1. Introduction Sepsis is prevalent among hospitalized patients worldwide. In fact, severe sepsis and septic shock are the major cause of admission and mortality in intensive care units (ICUs). However, mortality rates vary among countries and regions, depending on medical resources available. Japan has government health insurance and emergency medical systems but is also at the forefront of * Corresponding author. Tel.: þ81 3 3225 1323; fax: þ81 3 3353 2232. E-mail address: [email protected] (S. Fujishima). 1 See in Appendix.
countries with an aging population; therefore, morbidity and mortality statistics may not be comparable with those of other countries in which sepsis epidemiology has been demonstrated on a multicenter basis. A Japanese retrospective study examined incidence in a single emergency room and found that 4.5% of patients met the criteria for sepsis, 2.5% for severe sepsis, and 2.1% for septic shock [1]. In another retrospective study using Japan Nosocomial Infection Surveillances (JANIS), the incidence of sepsis was reported to be 2.1% among patients admitted to ICUs [2]; however, this value is very low compared with that observed in previous multicenter cohort studies, and it does not appear to be generalizable to nationwide ICUs accepting patients from emergency
1341-321X/$ e see front matter Ó 2013, Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jiac.2013.09.003
116
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
rooms, including our facilities. Prospective, multicenter-based, epidemiological studies of sepsis should have been conducted in Japan in order to advance the development of institutional and governmental strategies aimed at tackling this critical illness. In response to an increasing demand for a systematic approach to sepsis, the Society of Critical Care Medicine, the European Society of Intensive Care Medicine, and the International Sepsis Forum launched the Surviving Sepsis Campaign (SSC) in the hope of decreasing mortality [3]. Updated versions of guidelines developed by SSC were published in 2008 and 2013, in which the Japanese Association for Acute Medicine (JAAM) evolved as an organization comprising various professional societies. In accordance with SSC, a set of interventions, referred to as severe sepsis bundles, was introduced to change bedside behavior, and chart review database software was freely distributed in several languages to evaluate the quality of care in each hospital and to register patients worldwide [4]. However, because this software was not available in a Japanese version, almost no Japanese patients were included in the international sepsis registry; therefore, quality of care standards and mortality associated with severe sepsis and septic shock remain poorly clarified in Japan. The present study aimed to elucidate the epidemiology of severe sepsis and examine the SSC guidelines-based standard quality of care related to severe sepsis in Japan. To the best of our knowledge, this is the first report examining severe sepsis in Japan on the basis of a multicenter observational study. 2. Patients and methods This multicenter, prospective, observational study was registered at the University Hospital Medical Information Network Clinical Trial Registry (UMIN-CTR ID: UMIN000008195). Patients fulfilling the original and revised definition of severe sepsis or septic shock according to the American College of Chest Physicians/Society of Critical Care Medicine consensus conference were registered [5,6]. JAAM established a sepsis registry (SR) interim committee in 2007. The committee developed a web-based database in collaboration with the Internet Data and Information Center for Medical Research, a division of the University Hospital Medical Information Network. Since the JAAM SR database used in this study was equipped with functions equivalent to the SSC chart review database, it was capable of collecting the data necessary for evaluating quality indicators at individual institutes and sending anonymous patient data to the SSC office [4,7]. Data collected in this study were exactly the same as data collected in the international sepsis registry database and thus did not include information regarding parameters such as age, gender, comorbidities, Acute Physiology and Chronic Health Evaluation (APACHE) II score, Sequential Organ Failure Assessment (SOFA) score, or treatments other than severe sepsis bundles. Data regarding blood glucose, median BG values and times of hypoglycemic episodes during the initial 24 h were collected. This secure database is accessible through the Internet. Quality indicators were programmed to be calculated automatically as soon as data input from patients was completed, and included Q1: blood lactate measurement; Q2: blood culture; Q3: early antibiotic use; Q4: fluid resuscitation and vasopressors; Q5: achievement of CVP > 8 mmHg; Q6: achievement of ScvO2 > 70%; Q7: all resuscitation bundles (Q1eQ6); Q8: following glucocorticoid policy; Q9: following activated protein C (APC) policy (not evaluated in this study); Q10: glycemic control; Q11: low inspiratory plateau pressure; and Q12: all management bundles (Q8, 10, 11). In addition to the 18 core JAAM SR committee institutes, we recruited JAAM board-approved hospitals that could register patients in the database. All institutes and hospitals obtained approval from their ethics committees and recruited patients for 365
sequential days. Multiple organ dysfunction syndrome (MODS) was defined according to the SSC severe sepsis screening table [4]. Incidence of individual organ dysfunction in MODS, quality indicators, and their correlation with all-cause hospital mortality were evaluated. To evaluate the effect of organ dysfunction in MODS and compliance with individual severe sepsis bundles on mortality, univariate analysis was used. We also applied multivariate logistic regression analysis to MODS and two sets of quality indicators were applied to the same denominator, i.e., Q1, Q2, Q3, and Q10 in patients with severe sepsis and Q1, Q2, Q3, Q4, Q5, Q6, Q8, and Q10 in patients with septic shock. Simultaneous test procedures were used for analysis of MODS, and the stepwise backward elimination method (likelihood ratio) was used for evaluation of MODS subcategories and quality indicators. To compare values between two groups, Pearson’s chi-square test was used. To evaluate the effect of the total number of dysfunctional organs and systems on mortality, univariate analysis was used. Survival between two groups was compared using KaplaneMeier survival curves with a log-rank significance test. All statistical analyses were performed using IBM SPSS statistics (V20) software (International Business Machines Corp., NY, USA), and the level of significance was set at P < 0.05. 3. Results Using the JAAM SR database, 1104 patients with severe sepsis from 39 hospitals were registered between June 1, 2010, and December 31, 2011 (information of participated sites is shown in a supplementary table). Overall sepsis-related hospital mortality was 29.3%. Patients with septic shock had a significantly higher mortality (n ¼ 484, 40.7%) than those without (n ¼ 620, 20.3%, P < 0.0005). When our results were compared with the 30.8% mortality rate of 509 patients included in the final quarter of the SSC international sepsis registry study, the difference was not significant according to Pearson’s chi-square test (P ¼ 0.517) [7]. Table 1 shows patient demographics in our study. The emergency department was the most common route of admission, followed by the general ward. Pneumonia or empyema was the most common type of infection, followed by acute abdominal infection, urinary tract infection, and skin/soft tissue infection. Among MODS, hemodynamic dysfunction was the most frequent, followed by hematological, renal, pulmonary, and hepatic dysfunction. The average number of dysfunctional organs and systems reported in our study was 2.16 1.06 (mean standard deviation), which was significantly higher than that (1.94 1.01) reported in the international sepsis registry study, according to chi-square test mean scores with modified ridit scores (P < 0.0001). The incidence of individual organ dysfunction in our study was equivalent to or greater than that reported in previous studies, indicating a similar or higher severity of illness. We next analyzed the effects of MODS on mortality (Table 2). Existence of pulmonary, renal, hepatic, and hematological dysfunction was associated with significantly higher mortality. The mortality in patients with one, two, three, four, and five MODS was 19.7%, 28.0%, 34.7%, 48.0%, and 53.6%, respectively, and each additional MODS was associated with an approximately 1.5-fold increase in mortality (Table 2). We also assessed the effect of the following MODS subcategory abnormalities on mortality: hypotension, hyperlactatemia, acute lung injury, thrombocytopenia, and coagulopathy. Hypotension, thrombocytopenia, and coagulopathy were associated with higher mortality. We further performed multivariate logistic regression analysis. When the effects of individual organ dysfunction and system in MODS were analyzed, pulmonary, renal, and hematological dysfunction were extracted as significant parameters (Table 3A), and when MODS subcategory abnormalities in addition to renal and hepatic dysfunction were
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120 Table 1 Patient demographics. Patient number Route of admission Emergency department 671 General ward 316 ICU 117 Type of infection Pneumonia/empyema 436 Acute abdominal infection 268 Urinary tract infection 160 Skin/soft tissue inflammation 110 Meningitis 21 Blood stream catheter infection 20 Bone/joint infection 20 Wound infection 13 Endocarditis 12 Other infection 44 Baseline organ dysfunction (MODS) Hemodynamic 928 Pulmonary 374 Renal 412 Hepatic 184 Hematological 483 Number of dysfunctional organs and systems 1 361 2 364 3 248 4 102 5 28
Percentage 60.8 28.6 10.6 39.5 24.3 14.5 10.0 1.9 1.8 1.8 1.2 1.1 4.0 84.1 33.9 37.3 16.7 43.8 32.7 33.0 22.5 9.2 2.5
ICU: intensive care unit, Hemodynamic: systolic blood pressure <90 mmHg or mean blood pressure <65 mmHg or systolic blood pressure decrease >40 mmHg from baseline or blood lactate >2 mmol/L, Pulmonary: bilateral infiltration on chest X-ray with PaO2/FIO2 <300 or oxygenation required to maintain SpO2 >90%, Renal: Serum creatinine >2 mg/dl, Hepatic: serum bilirubin >2 mg/dl, Hematological: peripheral blood platelet counts <100,000/ml or prothrombin time-international normalized ratio >1.5 or activated partial thromboplastin time >60 s.
analyzed, acute lung injury, renal dysfunction, thrombocytopenia, and coagulopathy were extracted as significant parameters (Table 3B). In Table 4, the severe sepsis quality indicators in our study are listed alongside those from the international sepsis registry study [7]. Among these, the rate of Q3, Q4, Q5, Q6, Q7, and Q8 were lower, Table 2 Effects of MODS on mortality, by univariate analysis. Individual MODS
OR
95% CI
P value
Hemodynamic Hypotension Hyperlactatemia Pulmonary Acute lung injury Renal Hepatic Hematological Thrombocytopenia Coagulopathy Each Additional MODS
1.245 1.389 1.218 1.323 1.270 1.847 1.485 2.022 1.730 2.197 1.506
0.863e1.796 1.064e1.814 0.924e1.605 1.010e1.734 0.959e1.680 1.418e2.406 1.065e2.072 1.555e2.629 1.321e2.266 1.640e2.942 1.332e1.702
0.241 0.016 0.163 0.042 0.095 <0.001 0.020 <0.001 <0.001 <0.001 <0.001
OR: odds ratio, 95% CI: 95% confidence interval, MODS: multiple organ dysfunction syndrome, Hemodynamic: systolic blood pressure <90 mmHg or mean blood pressure <65 mmHg or systolic blood pressure decrease >40 mmHg from baseline or blood lactate >2 mmol/L, Hypotension: systolic blood pressure <90 mmHg or mean blood pressure <65 mmHg or systolic blood pressure decrease >40 mmHg from baseline, Hyperlactatemia: blood lactate >2 mmol/L, Pulmonary: bilateral infiltration on chest X-ray with PaO2/FIO2 <300 or oxygen required to maintain SpO2 >90%, Acute lung injury: bilateral infiltration on chest X-ray with PaO2/FIO2 <300, Renal: Serum creatinine >2 mg/dl, Hepatic: serum bilirubin >2 mg/dl, Hematological: peripheral blood platelet counts <100,000/ml or prothrombin timeinternational normalized ratio >1.5 or activated partial thromboplastin time >60 s, Thrombocytopenia: peripheral blood platelet counts <100,000/ml, Coagulopathy: prothrombin time-international normalized ratio >1.5 or activated partial thromboplastin time >60 s.
117
Table 3 Effects of MODS on mortality, by multivariate logistic regression analysis. A) Effects of MODS. B) Effects of MODS subcategory abnormalities in addition to renal and hepatic dysfunctions. A. MODS Hemodynamic Pulmonary Renal Hepatic Hematological
OR 1.229 1.457 1.628 1.152 1.836
95% CI 0.847e1.783 1.103e1.925 1.237e2.143 0.808e1.642 1.386e2.431
P value 0.279 0.008 0.001 0.435 <0.001
OR 1.359 1.626 1.378 1.912
95% CI 1.019e1.813 1.234e2.143 1.034e1.838 1.410e2.593
P value 0.038 0.001 0.029 <0.001
B. Subcategory abnormalities Acute lung injury Renal Thrombocytopenia Coagulopathy
OR: odds ratio, 95% CI: 95% confidence interval, MODS: multiple organ dysfunction syndrome, five subcategory abnormalities: hypotension, hyperlactatemia, acute lung injury, thrombocytopenia, and coagulopathy. Simultaneous test procedure was used in Table A, and stepwise backward-elimination method (likelihood ratio) in Table B. For details, please see Table 2.
whereas that of Q10 was equivalent and that for Q11 and Q12 was higher than that reported in the international sepsis registry study. We then analyzed whether compliance with each severe sepsis bundle could influence outcome. Univariate analysis indicated that compliance with Q3, Q4, Q6, Q7, Q8, Q10, and Q12 was associated with significantly lower mortality (Table 5). When analysis was restricted to patients with septic shock, Q6, Q8, and Q12 were found to be quality indicators affecting outcome (data not shown). When KaplaneMeier analysis was applied, the results were similar: compliance with Q4, Q6, Q8, Q10, and Q12 was confirmed as being associated with significantly better survival (Fig. 1). We further applied multivariate logistic regression analysis to elucidate the association of mortality with two sets of quality indicators (Table 6). Among patients with severe sepsis, Q10 was associated with
Table 4 Severe sepsis bundles quality indicator: Japan (JAAM) vs. international. Quality indicators Q1: Blood lactate measurement (severe sepsis) Q2: Blood culture (severe sepsis) Q3: Early antibiotic use (severe sepsis) Q4: Fluid resuscitation and vasopressors (hypotension or lactate > 4 mmol/L) Q5: Achievement of CVP over 8 mmHg (septic shock or lactate >4 mmol/L) Q6: Achievement of ScvO2 > 70% (septic shock or lactate >4 mmol/L) Q7: All resuscitation bundles (Q1eQ6) (severe sepsis) Q8: Following glucocorticoid policy (septic shock) Low-dose glucocorticoid use (septic shock) Q9: Following activated protein C (APC) policy (septic shock) Q10: Glycemic control (severe sepsis) Q11: Low inspiratory plateau pressure (requiring mechanical ventilation) Q12: All management bundles (Q8, 10, 11) (severe sepsis) Q13: Sepsis-related mortality
Japan (JAAM)
Internationala (final quarter)
987
81.4%
78.7%
784 692 791
81.4% 48.8% 47.2%
78.3% 67.9% 77.0%
667
21.0%
38.0%
667
10.3%
24.3%
534
11.2%
21.5%
484
36.2%
73.9%
484
18.2%
NA
NA
53.5%
674 703
51.3% 94.2%
56.8% 83.8%
675
39.7%
25.5%
29.3
34.8
Patient number
0
1104
JAAM: Japanese Association for Acute Medicine, CVP: central venous pressure, ScvO2: central venous oxygen saturation, NA: not available. Parentheses indicate target patients to whom each quality indicator of the severe sepsis bundle was applied to calculate the quality indicator. a Adopted from Ref. [7].
118
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
Table 5 Quality indicators vs. hospital mortality, by univariate analysis. Quality indicator (Q)
OR
Q1. Blood lactate measurement (severe sepsis) Q2. Blood culture (severe sepsis) Q3. Early antibiotic use (severe sepsis) Q4. Fluid resuscitation & vasopressors (hypotension or lactate >4 mmol/L) Q5: Achievement of central venous pressure over 8 mmHg (septic shock or lactate > 4 mmol/L) Q6: Achievement of central venous oxygen saturation>70% (severe sepsis) Q7. All resuscitation bundles (severe sepsis) Q8. Following glucocorticoid policy (septic shock) Low dose glucocorticoid use (septic shock) Q10. Glycemic control (severe sepsis) Q11. Low inspiratory plateau pressure (requiring mechanical ventilation) Q12. All management bundles (severe sepsis)
0.850 0.602e1.201
95% CI
0.358
0.897 0.605e1.330 0.677 0.483e0.949 0.633 0.470e0.852
0.588 0.023 0.003
0.679 0.454e1.014
0.059
0.483 0.270e0.865
0.014
0.453 0.630 1.199 0.614 0.992
0.035 0.019 0.446 0.004 0.981
0.217e0.945 0.429e0.927 0.752e1.911 0.440e0.856 0.515e1.190
P value
0.447 0.312e0.640 <0.001
OR: odds ratio, 95% CI: 95% confidence interval. Parentheses indicate target patients to whom each quality indicator of a severe sepsis bundle was applied to calculate the quality indicator.
mortality, but the difference was not significant. In patients with septic shock, Q8 was extracted as significant. To further analyze the efficacy of Q10, we examined the effect of hypoglycemia on mortality using univariate analysis. Blood glucose control <150 mg/dL, regardless of hypoglycemic episodes, did not affect the outcome (OR: 0.987, 95% CI: 0.689e1.415, P ¼ 0.943). In addition, in patients whose median blood glucose was controlled <150 mg/dL (n ¼ 408), the number of hypoglycemic episodes significantly correlated with outcome (OR: 1.519, 95% CI: 1.168e1.976, P ¼ 0.002). 4. Discussion Between June 1, 2010, and December 31, 2011, 1104 patients with severe sepsis and septic shock from 39 Japanese hospitals were registered using a new web-based database equivalent to the SSC chart review database. Compliance with severe sepsis bundles was generally low in our study compared with that in the international sepsis registry study, and glycemic control was associated with the lowest OR for mortality. Despite higher complication rates of MODS and low compliance with severe sepsis bundles on the whole, mortality was equivalent to that in the international sepsis registry study. To the best of our knowledge, this study is the first prospective multicenter study in Japan to achieve success in evaluating the standard quality of care on the basis of SSC guidelines. In our study, patients with four MODS, namely pulmonary, renal, hepatic, and hematological, were associated with significantly higher mortality. In addition, mortality increased in proportion to the total number of dysfunctional organs and systems. Among these MODS, hematological dysfunction, especially coagulopathy, was associated with the highest OR for mortality by both univariate and multivariate logistic regression analyses. In contrast, a previous large French epidemiological study by Annane et al. indicated that acute lung injury/acute respiratory distress syndrome and renal failure significantly affected patient outcome; however, they did not mention hematological dysfunction [8]. In a Spanish multicenter study, neurological, respiratory, renal, and hematological dysfunction, as assessed by the organ dysfunction system score, was associated with a significantly higher mortality rate. Furthermore, OR for hematological dysfunction was the highest [9]. These results, including those of the present study, may indicate that hematological dysfunction could be critical in determining the outcome of severe sepsis and septic shock. Although recombinant human APC is
no longer available, the attenuation of abnormal coagulopathy might still be a potential approach for treating sepsis. Of late, numerous SSC-based, worldwide multicenter studies have been published with mortality ranging from 27.0% to 44.5% [7,10e15]. In our study, mortality and the average number of dysfunctional organs were 29.3% and 2.2, respectively, with 11.2% compliance for all sepsis resuscitation bundles and 39.7% compliance for all sepsis management bundles, with the exception of adequate use of APC. Levy et al. reported the results of the international sepsis registry collected from 252 qualifying sites. Original and MODS number-adjusted mortality was 30.8% and 35.4%, respectively, with the average number of dysfunctional organs being 1.9; in addition, sepsis resuscitation and management bundle compliance was 31.3% and 36.1%, respectively, for the final quarter [7]. When comparing compliance with individual sepsis resuscitation and management bundles in our study with that reported in the international sepsis registry study, compliance with antibiotic use, fluid resuscitation and vasopressors, central venous pressure achievement, central venous oxygen saturation achievement, and steroid policy was low, but achievement of all these bundles was associated with better outcomes in our study. The reason for the discrepancy observed between fairly low mortality and low bundle compliance in the present study is unclear. The discrepancy might be attributed to a difference in the case mix of patients, including comorbidities and ages of patients between the 2 studies, even if severity of illness as assessed by the number of dysfunctional organs were adjusted. In addition, another prospective observational study was conducted in parallel to accumulate more precise clinical information from selected JAAM board-approved hospitals. Other factors that improved the outcome may be revealed by this study. In our study, compliance with five severe sepsis bundles, namely early antibiotic use, fluid resuscitation and vasopressors, achievement of central venous oxygen saturation >70%, following glucocorticoid policy, and glycemic control, was associated with better outcomes. Of these specific bundles, evidence of effectiveness has been accumulated for early antibiotic use [16e18] and fluid resuscitation and vasopressors [19]. In contrast, the efficacy of the other three bundles remains controversial. However, compliance with glycemic control was associated with the lowest OR for mortality in our study according to univariate analysis along with that noted in the international sepsis registry study [7]. Although significance was not achieved by multivariate analysis, we speculate that this was because of our small patient number and missing data. In the international sepsis registry study and the present study, a median blood glucose level maintained <150 mg/dL and its lowest value being higher than the lower limit of normal was considered good compliance. Although previous randomized controlled trials have shown no efficacy or adverse effects associated with tight glycemic control in critically ill patients [20e22], the effectiveness of glycemic control might have been canceled out by the disadvantages of hypoglycemia and variations in blood glucose levels in these studies. Also, the efficacy of intermediate glycemic control has not been determined yet [23e25]. In fact, we could not establish the efficacy of controlling blood glucose <150 mg/dL when patients with more than one episodes of hypoglycemia were included. In patients whose blood glucose was controlled <150 mg/dL, the number of hypoglycemic episodes significantly correlated with mortality. Therefore, we believe that the efficacy of intermediate glycemic control, in combination with a technique for avoiding hypoglycemia and blood glucose variation, needs to be examined in future studies. The present study had several limitations. First, although we demonstrated significant bundle factors associated with improved clinical outcomes in multivariate analysis, other possible risk factors affecting clinical outcomes such as Acute Physiology and Chronic Health Evaluation II score, old age, and comorbidity were
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
119
Fig. 1. KaplaneMeier Analysis of Compliance with Severe Sepsis Bundles. There are significant differences in survival between the groups with (black line) or without (gray line) fluid resuscitation and vasopressors (Q4), achievement of central venous oxygen saturation over 70% (Q6), following glucocorticoid policy (Q8), achievement of glycemic control (Q10), as shown in the graphs along with all management bundles (Q12).
not included in the analysis. In addition, the efficacy of commonly used treatments in Japan such as administration of immunoglobulin and continuous renal replacement therapy for non-renal indication, which may have had an impact on the results, were not evaluated in the study. Second, we included patients who met inclusion criteria equivalent to those in the SSC chart review database. However, the number of patients recruited from some hospitals was low, and we could not rule out the possibility of selection bias associated with inclusion at these facilities. Furthermore, 44% of patients were recruited from hospitals participating in this study under their own initiative, and thus, these may not Table 6 Quality indicators vs. hospital mortality in severe sepsis and septic shock by multivariate analysis. Quality indicator (Q) Severe sepsis (n ¼ 389) Q10. Glycemic control Septic shock (n ¼ 176) Q8. Following glucocorticoid policy
OR
95% CI
P value
0.727
0.468e1.127
0.154
0.468
0.253e0.867
0.016
OR: odds ratio, 95% CI: 95% confidence interval. Since target patients were not equal among quality indicators, two sets of quality indicators were applied to the same denominator (i.e., Q1, Q2, Q3, and Q10 in patients with severe sepsis, and Q1, Q2, Q3, Q4, Q5, Q6, Q8, and Q10 in patients with septic shock).
represent the average Japanese hospital, which could be another source of selection bias. Third, our study was prospective but observational and used ORs to evaluate the correlation between outcome and MODS or compliance with severe sepsis bundles, which limited its merits as compared with randomized controlled studies. Fourth, our study recruited patients during a relatively short period and thus could not evaluate quality improvement. Despite these limitations, this multicenter study, the first of its kind in Japan, was successful in evaluating standard quality of care and mortality related to severe sepsis. In conclusion, we were successful in evaluating the SSC guideline-based standard quality of care for severe sepsis and septic shock patients in Japan in a prospective multicenter study. Although mortality related with severe sepsis was determined to be equivalent to that reported worldwide in the above-mentioned international sepsis registry study, despite low compliance with severe sepsis bundles, further improvement in the quality of care may result in even better patient outcomes.
Acknowledgments This study was funded and supported by the Japanese Association for Acute Medicine (JAAM), and in part by a Health Labor
120
S. Fujishima et al. / J Infect Chemother 20 (2014) 115e120
Sciences Research Grant (H22-Shinko-Ippan-013). Part of the study was presented at the 39th and 40th JAAM annual conferences and at the 2012 American Thoracic Society International Conference. We thank Dr. Takayuki Abe (Center for Clinical Research, School of Medicine, Keio University) for help with the statistical analyses, and Mr. Kiyohiko Sato (Japanese Association for Acute Medicine) for administrative work.
Conflict of interest None.
Appendix Japanese Association for Acute Medicine Sepsis Registry (JAAMSR) Study Group: Hideharu Hagiya, MD, Emergency Unit and Critical Care Center, Tsuyama Central Hospital; Yoshihiro Tada, MD, Emergency Department, Hiroshima Prefectural Hospital; Akiko Oshiro, MD, Division of Emergency Medicine, Teine Keijinkai Hospital; Akihiro Noda, MD, Division of Emergency and Critical Care Medicine, Department of Acute Medicine, Nihon University School of Medicine; Satoru Futami, MD, Department of Emergency Medicine and Critical Care, Cancer Institute Hospital, Japanese Foundation for Cancer Research; Shunsuke Otani, MD, Department of Emergency and Critical Care Medicine, Kimitsu Chuo Hospital; Takeshi Takahashi, MD, Department of Emergency and Critical Care Medicine, Kumamoto Medical Center, National Hospital Organization; Minoru Nakano, MD, Advanced Medical Emergency and Critical Care Center, Japanese Red Cross Maebashi Hospital; Takuya Oda, MD, Department of General Internal Medicine, Iizuka Hospital; Yuichiro Sakamoto, MD, Department of Emergency and Critical Care Medicine, Saga University Hospital, Shigeaki Inoue, MD, Department of Emergency and Critical Care Medicine, Tokai University School of Medicine; Kenji Kubo, MD, Department of Emergency and Critical Care Medicine/Department of Infectious Diseases, Japanese Red Cross Society Wakayama Medical Center; Takeshi Kasai, MD, Department of Emergency and Trauma Center, Kameda Medical Center; Nobuyuki Saito, MD, Department of Emergency and Critical Care Medicine, Chiba-Hokusoh Hospital, Nippon Medical School; Masamune Kuno, MD, Department of Emergency and Critical Care Medicine, Tama Nagayama Hospital Nippon Medical School; Hideaki Anan, MD, Emergency Medical Center, Fujisawa City Hospital; Hiroshi Moro, MD, Advanced Disaster Medical and Emergency Critical Care Center, Niigata University Medical and Dental Hospital; Toshiaki Mochizuki, MD, Department of Emergency and Disaster Medicine, Hamamatsu University School of Medicine; Yasushi Takasaki, MD, Nan-yo Emergency Critical Care Center, Uwajima City Hospital; Hideki Tokuyama, MD, Department of Emergency and Critical Care Medicine, Komaki City Hospital; Katsuyuki Maruyama, MD, Department of Emergency and Critical Care Medicine, Kinki University School of Medicine; Hiroshi Morita, MD, Department of Emergency Medicine, Osaka Medical College Hospital; Yoshimi Kitagawa, MD, Emergency Department, Nagoya Ekisaikai Hospital; Hiroshi Chiba, MD, Center for Emergency and Critical Care Medicine, Hachinohe City Hospital.
Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jiac.2013.09.003.
References [1] Sun D, Aikawa N. The natural history of the systemic inflammatory response syndrome and the evaluation of SIRS criteria as a predictor of severity in patients hospitalized through emergency services. Keio J Med 1999;48:28e37. [2] Suka M, Yoshida K, Takezawa J. Incidence and outcome of sepsis in Japanese intensive care units: the Japanese nosocomial infection surveillance system. Environ Health Preventive Med 2006;11:298e303. [3] Dellinger RP, Carlet JM, Masur H, Gerlach H, Calandra T, Cohen J, et al. Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Intensive Care Med 2004;30:536e55. [4] Surviving Sepsis Campaign Website. Available from: www.survivingsepsis.org. [5] Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS international sepsis definitions conference. Intensive Care Med 2003;29:530e8. [6] Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992;101: 1644e55. [7] Levy MM, Dellinger RP, Townsend SR, Linde-Zwirble WT, Marshall JC, Bion J, et al. The Surviving Sepsis Campaign: results of an international guidelinebased performance improvement program targeting severe sepsis. Crit Care Med 2010;38:367e74. [8] Annane D, Aegerter P, Jars-Guincestre MC, Guidet B, Network CU- R. Current epidemiology of septic shock: the CUB-Rea Network. Am J Respir Crit Care Med 2003;168:165e72. [9] Blanco J, Muriel-Bombin A, Sagredo V, Taboada F, Gandia F, Tamayo L, et al. Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study. Crit Care 2008;12:R158. [10] Tromp M, Tjan DH, van Zanten AR, Gielen-Wijffels SE, Goekoop GJ, van den Boogaard M, et al. The effects of implementation of the Surviving Sepsis Campaign in the Netherlands. Neth J Med 2011;69:292e8. [11] Phua J, Koh Y, Du B, Tang YQ, Divatia JV, Tan CC, et al. Management of severe sepsis in patients admitted to Asian intensive care units: prospective cohort study. BMJ 2011;342:d3245. [12] Lefrant JY, Muller L, Raillard A, Jung B, Beaudroit L, Favier L, et al. Reduction of the severe sepsis or septic shock associated mortality by reinforcement of the recommendations bundle: a multicenter study. Ann Fr Anesth Reanim 2010;29:621e8. [13] Castellanos-Ortega A, Suberviola B, Garcia-Astudillo LA, Holanda MS, Ortiz F, Llorca J, et al. Impact of the Surviving Sepsis Campaign protocols on hospital length of stay and mortality in septic shock patients: results of a three-year follow-up quasi-experimental study. Crit Care Med 2010;38:1036e43. [14] Cardoso T, Carneiro AH, Ribeiro O, Teixeira-Pinto A, Costa-Pereira A. Reducing mortality in severe sepsis with the implementation of a core 6-hour bundle: results from the Portuguese community-acquired sepsis study (SACiUCI study). Crit Care 2010;14:R83. [15] Ferrer R, Artigas A, Levy MM, Blanco J, Gonzalez-Diaz G, Garnacho-Montero J, et al. Improvement in process of care and outcome after a multicenter severe sepsis educational program in Spain. JAMA 2008;299:2294e303. [16] Puskarich MA, Trzeciak S, Shapiro NI, Arnold RC, Horton JM, Studnek JR, et al. Association between timing of antibiotic administration and mortality from septic shock in patients treated with a quantitative resuscitation protocol. Crit Care Med 2011;39:2066e71. [17] Gaieski DF, Mikkelsen ME, Band RA, Pines JM, Massone R, Furia FF, et al. Impact of time to antibiotics on survival in patients with severe sepsis or septic shock in whom early goal-directed therapy was initiated in the emergency department. Crit Care Med 2010;38:1045e53. [18] Ferrer R, Artigas A, Suarez D, Palencia E, Levy MM, Arenzana A, et al. Effectiveness of treatments for severe sepsis: a prospective, multicenter, observational study. Am J Respir Crit Care Med 2009;180:861e6. [19] Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, et al. Early goaldirected therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368e77. [20] Investigators N-SS, Finfer S, Chittock DR, Su SY, Blair D, Foster D, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283e97. [21] Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 2008;358:125e39. [22] Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449e61. [23] Investigators N-SS, Finfer S, Liu B, Chittock DR, Norton R, Myburgh JA, et al. Hypoglycemia and risk of death in critically ill patients. N Engl J Med 2012;367:1108e18. [24] Hermanides J, Bosman RJ, Vriesendorp TM, Dotsch R, Rosendaal FR, Zandstra DF, et al. Hypoglycemia is associated with intensive care unit mortality. Crit Care Med 2010;38(6):1430e4. [25] Ali NA, O’Brien Jr JM, Dungan K, Phillips G, Marsh CB, Lemeshow S, et al. Glucose variability and mortality in patients with sepsis. Crit Care Med 2008;36:2316e21.