Emergency. 2018; 6 (1): e4 OR I G I N A L RE S E A RC H Pro-BNP versus MEDS Score in Determining the Prognosis of Sepsis Patients; a Diagnostic Accuracy Study Majid Shojaee1, Saeed Safari2, Anita Sabzghabaei2, Mostafa Alavi-Moghaddam1, Ali Arhami Dolatabadi1, Hamid Kariman1, Soheil Soltani1∗ 1. Emergency Department, Imam Hossein Hospital, Shahid Beheshti University of Medical Science, Tehran, Iran. 2. Emergency Department, Shohadaye Tajrish Hospital, Shahid Beheshti University of Medical Science, Tehran, Iran. Received: November 2017; Accepted: December 2017; Published online: 15 January 2018 Abstract: Introduction: Pro-brain natriuretic peptide (Pro-BNP) can act as an independent predictor of mortality in septic patients. This study aimed to compare the diagnostic accuracy of pro-BNP and Mortality in Emergency Depart- ment Sepsis (MEDS) score in this regard. Methods: This cross-sectional study was conducted on > 14 years old sepsis patients of an emergency department (ED), during 2 years. The level of Pro-BNP and MEDS score were measured for all eligible patients and considering one month mortality as reference, screening performance characteristics of the two tests were compared using SPSS 21 and STATS 11. Results: 121 patients with the mean age of 75.87±11.82 years were studied (55.4% male). 85 (70.25%) patients had moderate to high probability of mortality according to MEDS score. The mean Pro-BNP levels of survivor and non-survivor patients were 489.69 ± 327.47 and 3954.98 ± 2717.85 pg/ml, respectively (p < 0.0001). Sensitivity and specificity of Pro-BNP (in 1000 pg/ml cut off ) and MEDS score (in level 3) in prediction of 1-month mortality were 93.6 (83.7-97.9), 94.8 (84.7-98.6), 65.0 (51.9-76.3), and 98.2 (89.5-99.9), respectively. Area under the ROC curve of the two tests were 97.36 (95% CI: 92.92-94.48) and 92.31 (95% CI: 86.35-96.53), respectively (p = 0.0543). Conclusion: Pro-BNP and MEDS score both have excellent diagnostic accuracy in predicting 1-month mortality of sepsis patients. How- ever, considering the higher sensitivity as well as availability and ease of calculation, it seems that Pro-BNP can be considered an appropriate tool for screening patients with high risk of mortality following sepsis in ED. Keywords: Pro-brain natriuretic peptide; sepsis; mortality; Emergency department; dimensional measurement accuracy © Copyright (2018) Shahid Beheshti University of Medical Sciences Cite this article as: Shojaee M, Safari S, Sabzghabaei A, Alavi-Moghaddam M, Arhami Dolatabadi A, Kariman H, Soltani S. Pro-BNP versus MEDS Score in Determining the Prognosis of Sepsis Patients; a Diagnostic Accuracy Study. Emergency. 2018; 6(1): e4. 1. Introduction S eptic response is in fact a chain of inflammatory and anti-inflammatory processes and hormone-cell reac- tions that manifest clinically via systemic disorders such as: shock, myocardial disorder, activation of coagula- tion system and disseminated endothelial injury (1-4). Based on the report of world health organization (WHO), in United States of America more than 1 million people are affected with sepsis, about half of which lose their lives and sepsis is the third cause of mortality after heart diseases and can- cers (5, 6). Timely identification of patients at risk of mortal- ∗Corresponding Author: Soheil Soltani; Emergency Department, Imam Hos- sein Hospital, Imam Hossein Square, Shahid Madani Avenue, Tehran, Iran. Email: soheil_soltani@yahoo.com Tel: 00989122196632 ity due to sepsis can be helpful in selecting type of interven- tion, treatment protocol, diagnostic method, and probably improve their final outcome (7-9). Mortality in Emergency Department Sepsis (MEDS) score is an acceptable scale for predicting the prognosis of sepsis, which is not commonly used nowadays due to having many variables and its calcula- tion being time consuming (10). Currently, biomarkers such as: Pro-BNP, lactate, CRP, D-Dimer, Pro-Calcitonin and Tro- ponin are considered for determining the prognosis of pa- tients with sepsis (11-13). Natriuretic peptides are used in diagnosis and risk stratification of patients with acute coro- nary syndrome and congestive heart failure but the role of these factors in prognosis and diagnosis of patients with sep- tic shock is still under debate (14). Natriuretic peptides play an important role in maintenance of cardiovascular home- ostasis and circulating blood volume. Pro-BNP is secreted in response to stretching in the atrium or ventricle wall or due to This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com M. Shojaee et al. 2 myocardial ischemia in 2 shapes of N-terminal Pro-BNP (NT- Pro BNP) and C-terminal peptide (BNP) (15, 16). It has been shown that Pro-BNP can act as an independent predictor of mortality in patients with cardiac shock, septic shock, and se- vere sepsis (17-23). However, this marker has not been able to predict short-term mortality in critically ill patients hospital- ized in critical care unit with hypoxic respiratory failure (24). Therefore, in search for finding an accurate as well as easy and available scale for replacing MEDS score in emergency department (ED), the present study was designed with the aim of evaluating the diagnostic value of pro-BNP in compar- ison with MEDS score in determining the prognosis of sepsis patients ED. 2. Methods 2.1. Study design and setting The present study is a prospective cross-sectional one that was designed and performed with the aim of determining the diagnostic accuracy of Pro-BNP in predicting the prog- nosis of patients with sepsis in ED of Imam Hossein Hospi- tal, Tehran, Iran, from September 2014 to March 2016. Pro- tocol of this study was approved by the ethics committee of Shahid Beheshti University of Medical Sciences and the re- searchers adhered to the principles indicated in the declara- tion of Helsinki regarding medical ethics. Before entering the study, the patient or their relative signed an informed written consent for participating in the study. The research team did not interfere in the diagnostic and treatment processes of the patients. 2.2. Participants Patients over the age of 14 years with sepsis presenting to the ED were randomly included in the study. For random sam- pling, cases were selected on random days (these days were randomly selected on the first day of each month). Simul- taneous presence of systemic inflammatory response syn- drome (SIRS) criteria and definitive evidence of infection based on American College of Chest Physicians/Society of Critical Care Medicine consensus was considered as the def- inition of sepsis (25). Patients with diagnosis of sepsis, severe sepsis, or septic shock were included in the study. Cases with history of proven cardiac or kidney failure, electrocardiogram (ECG) indicating new cardiac problems and children were ex- cluded from the study. 2.3. Data gathering Demographic data (age, sex), source of infection (pneumo- nia; urinary tract infection; digestive disease; bedsore and . . . ), variables needed for calculation of MEDS score (10), serum level of Pro-BNP on admission as well as 1-month out- come of the patients regarding mortality were gathered for all the patients via a pre-designed checklist. Source of infec- tion was determined according to the final decision of the in charge physician and considering all the laboratory, imaging, and clinical evidence. Pro-BNP measurement was done us- ing 1cc of the patient’s venous blood sample drawn on ad- mission for measuring other laboratory parameters ordered by the in charge physician. Blood drawing was done by a lab- oratory technician who was not aware of the study. Pro-BNP level measurement was done via chemilumenescence sand- wich immunoassay using Elecsys 2010 kit, Roche diagnostic, Mannheim, Germany. The kit was a high-quality and rapid (taking 18 minutes) kit that finally reports Pro-BNP rate as pg/cc. To calculate MEDS score, a medical calculator named emcalculator was used and according to the results the pa- tients were divided into 5 groups regarding mortality risk: very low, low, moderate, high, and very high. Finally, the mor- tality status of the patients 30 days after admission to ED was followed and recorded via phone calls. It should be noted that treatment of sepsis patients in the mentioned center is done according to the protocol of surviving sepsis campaign by emergency service with consultation of infection service (26). One emergency medicine resident was in charge of data gathering, calculating MEDS score and follow up of patients by phone. 2.4. Statistical Analysis Minimum sample size required for the present study consid- ering 95% confidence interval, 80% power, 7% error, and 0.63 area under the receiver operating characteristic (ROC) curve of Pro-BNP (27), was calculated to be 122 cases. Data were analyzed via SPSS 21 and STATA 11 software. Qualitative data were reported as frequency and percentage, and quantitative data as mean ± standard deviation (SD). Finally, for deter- mining the diagnostic accuracy of Pro-BNP in predicting the 1-month prognosis of sepsis patients, ROC curve was drawn and the best cut-off was calculated. In addition, sensitivity, specificity, positive and negative predictive values, and pos- itive and negative likelihood ratios with 95% confidence in- terval were calculated and reported using VassarStats medi- cal calculator. To compare the diagnostic accuracies of the 2 tests in predicting the risk of 1-month mortality in patients with sepsis presenting to the ED, comparison of area under the ROC curve was done via chi square test. In all analy- ses p<0.05 was considered as level of significance. Accuracy of 0.90-0.100 was considered as excellent, 0.80-0.90 as good, 0.70-0.80 as fair, 0.60-0.70 as poor, and 0.50-.60 as fail. 3. Results 3.1. Baseline characteristics 155 patients were candidates for entering the study, 9 (5.80%) were excluded due to history of kidney failure and 7 (4.52%) This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com 3 Emergency. 2018; 6 (1): e4 Figure 1: Mean and standard deviation of patients’ Pro-BNP level based on their MEDS score level. Figure 2: Area under the receiver operating characteristic (ROC) curve of Pro-BNP and MEDS score in determining the risk of 1- month mortality among sepsis patients in emergency department (p=0.0543). due to cardiac failure. In addition, 18 (11.61%) were excluded due to missing data and problem in follow-up. Finally, 121 patients with the mean age of 75.87±11.82 (20-98) years were evaluated (55.4% male). Table 1 shows the baseline char- acteristics of the patients. 85 (70.25%) patients had moder- ate to high probability of mortality according to MEDS scale. The mean Pro-BNP levels of survivor and non-survivor pa- tients were 489.69 ± 327.47 and 3954.98 ± 2717.85 pg/ml, re- spectively (p < 0.0001). Figure 1 depicts mean Pro-BNP level based on the patients’ MEDS score. 3.2. Screening performance characteristics Of all the studied patients, 63 (52.1%) died after 1 month. Area under the ROC curve of Pro-BNP and MEDS score in prediction of 1-month mortality of sepsis patients admitted to ED were 97.36 (95% CI: 92.92-94.48) and 92.31 (95% CI: 86.35-96.53), respectively (p = 0.0543, figure 2). According to the area under the ROC curve, the best cut off to predict 1- month mortality was 1000 pg/ml for Pro-BNP and level 3 for Table 1: Baseline characteristics of studied patients Variable Number (%) Sex Male 67 (55.4) Female 54 (44.6) Age (year) 20 - 40 3 (2.5) 40 - 60 7 (5.8) ≥ 60 111 (91.7) Source of infection Pneumonia 49 (40.5) Urinary tract infection 45 (37.2) Digestive problem 14 (11.6) Bedsore 6 (4.9) Multiple sources 7 (5.8) MEDS score 1 21 (17.4) 2 15 (12.4) 3 43 (33.5) 4 15 (12.4) 5 27 (22.3) Pro-BNP (pg/ml) < 1000 59 (48.8) ≥ 1000 62 (51.2) Data are reported as frequency and percentage. MEDS scale. Table 2 shows the screening performance char- acteristics of MEDS score and Pro-BNP in the mentioned cut offs. 3.3. Discussion: Based on the findings of the present study, Pro-BNP over 1000 pg/ml and MEDS scale both have excellent diagnostic accuracy in predicting 1-month mortality of sepsis patients. However, considering the higher sensitivity of Pro-BNP (93.6 vs. 65.0) as well as availability and ease of calculation, it seems that Pro-BNP can be considered an appropriate tool for screening patients with high risk of mortality following sepsis in ED. Presently, various tools have received attention for determining the prognosis of critically ill patients (26, 28, 29). In a study by Varpula et al. in 2007, it was shown that Pro-BNP changes in the first 3 days of hospitalization is a good prognostic scale in septic patients (19). In Kimmoun et al. study in 2013 Pro-BNP level directly correlated with the hemodynamic changes of septic patients (30). A comparison regarding increase in Pro-BNP among cardiac patients and sepsis patients showed a significant increase in the level of this marker in sepsis or septic shock patients compared to cardiac patients (31). In Carpenter et al. study, it was shown that 44% of those with sepsis or septic shock have levels of systolic disorders and in line with this finding they deemed higher serum BNP levels in these patients associated with worse outcome (32). Another study has also suggested the concentration of this biomarker on the 5th day after hospi- This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com M. Shojaee et al. 4 Table 2: Screening performance characteristics of Pro-BNP in 1000 pg/ml cut-off and MEDS score in level 3 (moderate to high risk of death) Characteristics Pro-BNP MEDS True positive 59 41 True negative 55 57 False positive 3 1 False negative 4 22 Sensitivity 93.6 (83.7-97.9) 65.0 (51.9-76.3) Specificity 94.8 (84.7-98.6) 98.2 (89.5-99.9) Positive predictive value 95.2 (85.6-98.7) 97.6 (85.9-99.8) Negative predictive value 93.2 (82.7-97.8) 72.2 (60.8-81.4) Positive likelihood ratio 19.66 (6.51-59. 3) 41.0 (5.9-284. 4) Negative likelihood ratio 0.07 (0.02-0.18) 0.3 (0.2-0. 5) Measures are presented with 95% confidence interval. talization as an index for prognosis of critically ill patients with a higher risk of mortality (33). On the other hand, ac- curacy of MEDS score in prediction of 1-month risk of mor- tality for sepsis patients has been evaluated and declared in numerous studies. Kuo et al. studied 431 patients with pyo- genic liver abscesses and affirmed the high accuracy of this scale in this regard (34). Another study in 2003 introduced this scale as a proper tool for triage and making decisions re- garding treatment of sepsis patients (10). A study in Nether- lands in 2012 indicated the higher power of this scale com- pared to biomarkers such as CRP and Lactate in prediction of 28-day mortality of sepsis patients in ED (35). Area under the ROC curve of MEDS score in this regard was estimated to be 0.81 in a study by Macdonald et al. (36). Considering all the mentioned points as well as some other important points including ease of calculation, availability, higher sensitivity, and similar accuracy and specificity it seems that Pro-BNP can be considered a more beneficial tool compared to MEDS score in prediction of mortality risk and screening of patients with sepsis in ED. It is obvious that doing this screening can help in concentrating the treatment system on more critically ill patients and using more rigorous diagnostic and treatment modalities in trying for improving their outcome. However, it should be noted that in developing countries such as Iran Pro-BNP measurement is still expensive and it might appear not cost effective to do this evaluation for all sepsis patients. Yet it should be taken into account that proper allocation of funds leads to overall decrease in treatment costs and im- provement of patients’ outcome. 3.4. Limitations Not differentiating the patients with sepsis, severe sepsis, and septic shock can be one of the limitations of this study. It might have been better to differentiate these patients and evaluate the diagnostic accuracy of the tests in each of the 3 groups to improve the overall accuracy of the study. 4. Conclusion Based on the findings of the present study, Pro-BNP and MEDS score both have excellent diagnostic accuracy in pre- dicting 1-month mortality of sepsis patients. However, con- sidering the higher sensitivity as well as availability and ease of calculation, it seems that Pro-BNP can be considered an appropriate tool for screening patients with high risk of mor- tality following sepsis in ED. 5. Appendix 5.1. Acknowledgements This study is extracted from Dr. Soheil Soltani’s thesis for achieving the specialist degree in emergency medicine. The emergency department staff members of Imam Hossein Hospital are acknowledged for their effort and cooperation throughout the study. 5.2. Author contribution All the authors meet the standard authorship criteria accord- ing to the recommendations of international committee of medical journal editors. 5.3. Funding/Support No funds have been received for carrying out this project. 5.4. Conflict of interest The authors declare that there is no conflict of interest re- garding the present study. References 1. Gullo A, Bianco N, Berlot G. Management of severe sep- sis and septic shock: challenges and recommendations. Critical care clinics. 2006;22(3):489-501. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com 5 Emergency. 2018; 6 (1): e4 2. Hotchkiss RS, Karl IE. The pathophysiology and treat- ment of sepsis. New England Journal of Medicine. 2003;348(2):138-50. 3. Landry DW, Oliver JA. The pathogenesis of va- sodilatory shock. New England Journal of Medicine. 2001;345(8):588-95. 4. Lee WL, Slutsky AS. Sepsis and endothelial permeability. The New England journal of medicine. 2010;363(7):689. 5. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Critical care medicine. 2001;29(7):1303-10. 6. Gaieski DF, Edwards JM, Kallan MJ, Carr BG. Benchmark- ing the incidence and mortality of severe sepsis in the United States. Critical care medicine. 2013;41(5):1167-74. 7. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of hypotension before initia- tion of effective antimicrobial therapy is the critical de- terminant of survival in human septic shock. Critical care medicine. 2006;34(6):1589-96. 8. Rangel-Frausto MS, Pittet D, Costigan M, Hwang T, Davis CS, Wenzel RP. The natural history of the systemic in- flammatory response syndrome (SIRS): a prospective study. Jama. 1995;273(2):117-23. 9. Zambon M, Ceola M, Almeida-de-Castro R, Gullo A, Vin- cent J-L. Implementation of the Surviving Sepsis Cam- paign guidelines for severe sepsis and septic shock: we could go faster. Journal of critical care. 2008;23(4):455-60. 10. Shapiro NI, Wolfe RE, Moore RB, Smith E, Burdick E, Bates DW. Mortality in Emergency Department Sep- sis (MEDS) score: a prospectively derived and vali- dated clinical prediction rule. Critical care medicine. 2003;31(3):670-5. 11. Scheer C, Fuchs C, Rehberg S. Biomarkers in Severe Sep- sis and Septic Shock: Just Listen to the Heart? Critical care medicine. 2016;44(4):849-50. 12. Biteker FS, Caylak SD, Sozen H. Biomarkers in sep- sis. The American journal of emergency medicine. 2016;34(5):924-5. 13. Brueckmann M, Huhle G, Lang S, Haase KK, Bertsch T, WeiB C, et al. Prognostic value of plasma N-terminal pro- brain natriuretic peptide in patients with severe sepsis. Circulation. 2005;112(4):527-34. 14. Maeder M, Fehr T, Rickli H, Ammann P. Sepsis-associated myocardial dysfunction: diagnostic and prognostic im- pact of cardiac troponins and natriuretic peptides. Chest. 2006;129(5):1349-66. 15. Levin ER, Gardner DG, Samson WK. Natriuretic peptides. New England Journal of Medicine. 1998;339(5):321-8. 16. Goetze J, Christoffersen C, Perko M, Arendrup H, Rehfeld J, Kastrup J, et al. Increased cardiac BNP expression as- sociated with myocardial ischemia. The FASEB Journal. 2003;17(9):1105-7. 17. Januzzi JL, Morss A, Tung R, Pino R, Fifer MA, Thompson BT, et al. Natriuretic peptide testing for the evaluation of critically ill patients with shock in the intensive care unit: a prospective cohort study. Critical care. 2006;10(1):R37. 18. Roch A, Allardet-Servent J, Michelet P, Oddoze C, Forel J-M, Barrau K, et al. NH2 terminal pro–brain natriuretic peptide plasma level as an early marker of prognosis and cardiac dysfunction in septic shock patients. Critical care medicine. 2005;33(5):1001-7. 19. Varpula M, Pulkki K, Karlsson S, Ruokonen E, Pettila V, Group FS. Predictive value of N-terminal pro–brain natri- uretic peptide in severe sepsis and septic shock. Critical care medicine. 2007;35(5):1277-83. 20. Shah KB, Nolan MM, Rao K, Wang DJ, Christenson RH, Shanholtz CB, et al. The characteristics and prognos- tic importance of NT-ProBNP concentrations in crit- ically ill patients. The American journal of medicine. 2007;120(12):1071-7. 21. Kotanidou A, Karsaliakos P, Tzanela M, Mavrou I, Kopterides P, Papadomichelakis E, et al. Prognostic im- portance of increased plasma amino-terminal pro-brain natriuretic peptide levels in a large noncardiac, general intensive care unit population. Shock. 2009;31(4):342-7. 22. Almog Y, Novack V, Megralishvili R, Kobal S, Barski L, King D, et al. Plasma level of N terminal pro-brain natriuretic peptide as a prognostic marker in critically ill patients. Anesthesia & Analgesia. 2006;102(6):1809-15. 23. Meyer B, Huelsmann M, Wexberg P, Delle Karth G, Berger R, Moertl D, et al. N-terminal pro-B-type natriuretic peptide is an independent predictor of outcome in an unselected cohort of critically ill patients. Critical care medicine. 2007;35(10):2268-73. 24. Jefic D, Lee JW, Jefic D, Savoy-Moore RT, Rosman HS. Utility of B-Type Natriuretic Peptide and N-terminal Pro B-Type Natriuretic Peptide in Evaluation of Respiratory Failure in Critically Ill Patients. Chest. 2005;128(1):288- 95. 25. American College of Chest Physicians/Society of Crit- ical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med. 1992;20(6):864-74. 26. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med. 2013;41(2):580-637. 27. Varpula M, Pulkki K, Karlsson S, Ruokonen E, Pettila V, Group ftFS. Predictive value of N-terminal pro–brain na- triuretic peptide in severe sepsis and septic shock*. Criti- cal Care Medicine. 2007;35(5):1277-83. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com M. Shojaee et al. 6 28. Alireza Baratloo M, Shokravi M, Saeed Safari M. Predic- tive value of Glasgow Coma Score and Full Outline of Un- responsiveness Score on the outcome of multiple trauma patients. Archives of Iranian medicine. 2016;19(3):215. 29. Hermans MAW, Leffers P, Jansen LM, Keulemans YC, Stassen PM. The value of the Mortality in Emergency De- partment Sepsis (MEDS) score, C reactive protein and lactate in predicting 28-day mortality of sepsis in a Dutch emergency department. Emergency Medicine Journal. 2011. 30. Kimmoun A, Ducrocq N, Mory S, Delfosse R, Muller L, Perez P, et al. Cardiac contractile reserve parameters are related to prognosis in septic shock. BioMed research in- ternational. 2013;2013. 31. Rudiger A, Gasser S, Fischler M, Hornemann T, von Eckardstein A, Maggiorini M. Comparable increase of B- type natriuretic peptide and amino-terminal pro-B-type natriuretic peptide levels in patients with severe sep- sis, septic shock, and acute heart failure*. Critical Care Medicine. 2006;34(8):2140-4. 32. Charpentier J, Luyt C-E, Fulla Y, Vinsonneau C, Cariou A, Grabar S, et al. Brain natriuretic peptide: a marker of my- ocardial dysfunction and prognosis during severe sepsis. Critical care medicine. 2004;32(3):660-5. 33. Post F, Weilemann LS, Messow C-M, Sinning C, MÃijnzel T. B-type natriuretic peptide as a marker for sepsis- induced myocardial depression in intensive care pa- tients. Critical care medicine. 2008;36(11):3030-7. 34. Kuo SH, Lee YT, Li CR, Tseng CJ, Chao WN, Wang PH, et al. Mortality in Emergency Department Sepsis score as a prognostic indicator in patients with pyogenic liver ab- scess. Am J Emerg Med. 2013;31(6):916-21. 35. Hermans M, Leffers P, Jansen L, Keulemans Y, Stassen P. The value of the Mortality in Emergency Department Sepsis (MEDS) score, C reactive protein and lactate in predicting 28-day mortality of sepsis in a Dutch emer- gency department. Emerg Med J. 2012;29(4):295-300. 36. Macdonald SP, Arendts G, Fatovich DM, Brown SG. Com- parison of PIRO, SOFA, and MEDS scores for predicting mortality in emergency department patients with severe sepsis and septic shock. Academic Emergency Medicine. 2014;21(11):1257-63. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com Introduction Methods Results Conclusion Appendix References