|Year : 2020 | Volume
| Issue : 3 | Page : 121-127
A study of clinical and laboratory evaluation and outcome of patients with acute febrile illness with thrombocytopenia
S Sumangala, Sangram Biradar, Mohd Zeeshan Ali, Mushtaq Saudagar
Department of General Medicine, MR Medical College, Gulbarga, Karnataka, India
|Date of Submission||14-Oct-2019|
|Date of Decision||03-Jan-2020|
|Date of Acceptance||01-Feb-2020|
|Date of Web Publication||15-Jul-2020|
Dr. S Sumangala
Soogareddy, No. 12 Sajjalashree, Gangamma Garden, Malagala Main Road, Near SBI, Second Stage, Nagarbhavi, Bengaluru, Karnataka
Source of Support: None, Conflict of Interest: None
Introduction: Fever with thrombocytopenia has become a common presenting problem. Infection is the main cause of thrombocytopenia. Fever with thrombocytopenia is frequently associated with an increased risk of morbidity and mortality. Infections such as malaria, dengue, enteric fever, and septicemia are some of the major causes of fever with thrombocytopenia in India. Aims and Objectives: To evaluate clinical profile of febrile thrombocytopenia, to identify different causes and the most common cause of febrile thrombocytopenia, and to assess the outcome and complications associated with febrile thrombocytopenia. Materials and Methods: A cross-sectional study of 160 patients was carried out at Basaveshwar Teaching and General Hospital. Patients with thrombocytopenia who were ≥18 years of age at admission between November 2016 and May 2018 were observed and followed up during their stay in hospital, diagnoses were made, and bleeding manifestations and requirement of platelet transfusion were also recorded. Results: Febrile thrombocytopenia affected all age groups ranging from 18 to 75 years of age but was common in 21–30 years of age group (40.60%), with a male-to-female ratio 57:43. Infection (88.12%) was the most common cause of thrombocytopenia, while dengue (53.13%) was the most common of the infections followed by malaria (15.63%) and septicemia (8.75%). Bleeding manifestations were seen in 33.13% of patients. 81% of the patients with bleeding tendencies had petechiae/purpura as the most common bleeding manifestation, followed by spontaneous bleeding in 54.70%. Melena (34.48%) was the most common among spontaneous bleeding. Bleeding manifestations were more common when the platelet counts were <20,000 cells/cumm. Good recovery was noted in 92.50%, while 5% had mortality and 2.50% cases were referred. Septicemia was the major cause of mortality. Conclusion: In our setup, infection such as dengue fever was the common cause of fever with thrombocytopenia followed by malaria and septicemia. In majority of patients, thrombocytopenia was transient and asymptomatic, but in significant number of cases, there were bleeding manifestations. On treating the specific cause, drastic improvement in the platelet count was noted. Mortality in febrile thrombocytopenia is not directly associated with degree of thrombocytopenia but with concomitant involvement of other organs, leading to multiorgan dysfunction.
Keywords: Dengue, fever, malaria, spontaneous bleeding, thrombocytopenia
|How to cite this article:|
Sumangala S, Biradar S, Ali MZ, Saudagar M. A study of clinical and laboratory evaluation and outcome of patients with acute febrile illness with thrombocytopenia. APIK J Int Med 2020;8:121-7
|How to cite this URL:|
Sumangala S, Biradar S, Ali MZ, Saudagar M. A study of clinical and laboratory evaluation and outcome of patients with acute febrile illness with thrombocytopenia. APIK J Int Med [serial online] 2020 [cited 2020 Aug 14];8:121-7. Available from: http://www.ajim.in/text.asp?2020/8/3/121/289795
| Introduction|| |
Fever with thrombocytopenia has become a common presenting problem. Patients with a history of fever in a tropical country like India usually have an infectious etiology, and many of them have associated thrombocytopenia. Many undifferentiated febrile illnesses are common in tropical countries like India which may mimic infections such as dengue, enteric fever, malaria, leptospirosis, and influenza. Sometimes, noninfectious causes such as primary hematological disorders may also present with febrile thrombocytopenia. Thrombocytopenia may be defined as a subnormal number of platelets in the circulating blood. A normal human platelet count ranges from 150,000 to 450,000/cumm of the blood. Often, patients with thrombocytopenia are asymptomatic and are diagnosed by routine complete blood count. Occasionally, there may be bruising, purpura, petechiae, nose bleeding, and gum bleeding. Rarely, platelet count may be as low as 5000/cumm predisposing the patients to life-threatening bleeding in the central nervous system (CNS) or from the gastrointestinal and genitourinary tracts.
Thrombocytopenia occurs due to decreased platelet production, which occurs in conditions such as Vitamin B12 and folate deficiency, leukemia, sepsis (bacterial or viral infection), and hereditary disease. Thrombocytopenia may also occur due to increased destructions, such as idiopathic thrombocytopenic purpura (ITP), thrombotic thrombocytopenic purpura (TTP), hemolytic–uremic syndrome (HUS), disseminated intravascular coagulation (DIC), paroxysmal nocturnal hemoglobinuria, systemic lupus erythematosus (SLE), antiphospholipid syndrome, posttransfusion purpura, and hypersplenism. Drugs, which can cause thrombocytopenia, are quinine, valproic acid, methotrexate, carboplatin, interferon, isotretinoin, and heparin.,, Diseases which commonly present with fever and thrombocytopenia are malaria, leptospirosis, rickettsial infections, septicemia, typhoid, borreliosis, and arboviruses such as dengue or yellow fever and rodent-borne viruses such as Hanta and Lassa fever, human immunodeficiency virus, and TTP–HUS.,, Thrombocytopenia in fever, being a prognostic factor, can predict the cause and thus helps in early diagnosis and treatment of the same, preventing further fatal outcome associated with it such as intracerebral bleed, hemorrhage into vital organs, and shock. In this study, special emphasis is on clinical profile of febrile thrombocytopenia and to determine different etiologies of these febrile illnesses. This study also aims to analyse different bleeding manifestations to find their incidence in fever with thrombocytopenia. This study also aims to study the relation of bleeding manifestations with platelet count and to document various complications in these clinical settings. Hence, the study of correlation between platelet counts and hemorrhagic manifestations helps us to know the correct time for infusion of platelets, thus avoiding unnecessary platelet transfusion.
Aims and objectives
To evaluate clinical profile of febrile thrombocytopenia, to identify different causes and the most common cause of febrile thrombocytopenia, and to assess the outcome and complications associated with febrile thrombocytopenia.
| Materials and Methods|| |
It is an observational type of descriptive study (cross-sectional study) conducted in the Department of General Medicine at Basaveshwar Teaching and General Hospital, Kalaburgi, Karnataka, India, over 1½ years, i.e., from November 2016 to May 2018. The number of patients was 160.
All inpatients presenting with <1 week of fever, i.e., AM temperature of >98.9°F and PM temperature of >99.9°F, platelet count of <150,000/cumm, and age ≥18 years.
Patients with platelet count of >150,000/cumm, patients aged <18 years, patients on long-term medications which cause thrombocytopenia (heparin, rifampicin, carbamazepine, acetaminophen, etc.), patients with known chronic illnesses such as SLE, ITP, chronic liver diseases, malignancies, and autoimmune diseases, previously diagnosed cases of leukemia and myelodysplastic syndromes.
Once the patients with fever and thrombocytopenia were admitted, a careful history was recorded, and general physical examination was done. Detailed examination of various systems, including vital, respiratory system, gastrointestinal system, CNS, and cardiovascular system examination, was done. Routine investigation (chest X-ray, complete blood count, random blood sugar, liver function test, renal function test, electrocardiograph, and serum electrolytes) were also done. Specific/special investigations were done as and when indicated.
Details of history, general physical examination, and laboratory and imaging studies were noted down from time to time. Once the specific diagnoses such as enteric fever, septicemia, and dengue were reached, the patients were treated for it specifically and symptomatically. Mixed infections were also included in the study. They were followed up from admission till recovery, discharge, or death whichever was earlier. In 3.1% of cases, the cause was unidentified. Such cases were also included in the study. In our institution, tests done to find out infectious agents were limited to tests for dengue virus (dengue NS 1 Ag, dengue immunoglobulin M [IgM], and dengue IgG enzyme-linked immunosorbent assay [ELISA]), leptospira (leptospira IgM ELISA), malaria (peripheral smear examination and malaria rapid optimal test), Salmonella typhi (blood culture and Salmonella typhi IgM), and infectious hepatitis (hepatitis B virus [HBV] surface antigen ELISA, HCV-anti-HCV IgM ELISA, HEV-anti-HEV IgM ELISA, and HAV-anti-HAV IgM ELISA). Other common infectious causes for febrile thrombocytopenia in our community such as chikungunya, Scrub typhus, and Kyasanur forest disease were not actively looked up on, owing to financial and technical constraints.
Data entry and analysis
Descriptive and inferential statistical analyses have been carried out in the present study. P <0.005 was considered statistically significant.
| Results|| |
From November 2016 to May 2018, a total of 160 patients (n = 160) with febrile thrombocytopenia who met the inclusion criteria were studied. 91 (56.88%) were male and 69 (43.12%) were female. The maximum cases were found to be in the age group of 21–30 years (40.62%).
Infection was the established diagnosis in 88.125% (n = 141) of the cases. Among them, hematological conditions were 12 (7.5%) cases [Table 1]. Megaloblastic anemia was also associated with fever.
The most common symptom of fever with thrombocytopenia was myalgia in 82 cases followed by headache in 80 patients [Table 2]. On examination, 59 (36.87%) patients had pallor. In this present study, at the time of admission, 64 patients (40%) had platelet count in the range of 50,001–100,000. In the present study, maximum number of patients (64%) had platelet count in the range of 50,001–100,000 cells/cumm. Maximum numbers of cases were in the range of 50,001–100,000.
|Table 2: Distribution of various clinical symptoms in febrile thrombocytopenia|
Click here to view
There were 12 patients (7.5%) in the range of 10,001–20,001 and eight cases (5%) below 10,000 platelet count [Table 3]. In our study, of 53 patients with bleeding manifestation, petechiae/purpura was seen in 43 patients (81%) and spontaneous bleeding in 29 (54.7%) patients.
|Table 3: Distribution of lowest platelet count during course of hospitalization|
Click here to view
In this present study, of 160 patients, a total of 53 cases had bleeding manifestations accounting to 33.12% [Table 4]. Bleeding tendencies correlated with the lowest platelet count during admission, and maximum tendency was noted with platelet count <20,000 cells/cumm. Bleeding tendencies were more commonly seen in patients with platelet count <20,000 (91%) [Figure 1].
|Figure 1: Distribution of bleeding tendencies based on platelet count (lowest platelet count during admission)|
Click here to view
Of the 85 cases of dengue, 28 (32.94%) had bleeding manifestations. Of the 25 cases of malaria, 4 (16%) had bleeding manifestations. Of the 14 cases of septicemia, 9 (64.28%) had bleeding manifestations, and of nine cases of enteric fever, no patient had bleeding manifestation [Table 5]. Bleeding manifestations were seen in 1 (25%) case out of four mixed infection.
|Table 5: Correlation of petechiae/purpura and spontaneous bleeding with thrombocytopenia|
Click here to view
In the present study, among 85 cases of dengue, five cases were found in the range of ≤10,000 platelet count (lowest during admission). 54% of dengue cases had platelet count >50,000 (mild-to-moderate thrombocytopenia) [Table 6].
In this study, anemia (hemoglobin <11 g/dl) was seen in 72 cases (45%), leucopenia in 50 cases (31.25%), leukocytosis (total count >11,000 cells/cumm) in 18 (11.25%) cases, and abnormal renal functions in 21 cases (13.12%). No cases of polycythemia due to hemoconcentration were seen in dengue fever patient. Abnormal serum bilirubin and serum glutamate pyruvate transaminase (SGPT) were seen in 37 (23.12%) and 29 (18.12%) cases, respectively.
The diagnoses of all cases of leukemia and eight cases of megaloblastic anemia were confirmed by bone marrow aspiration study [Table 7]. There was maximum increment in the mean platelet count was on day 5 of admission [Figure 2]. Persistently, decreased platelet count was seen in 13 patients (8.13%) [Figure 3]. In our study, platelet transfusion was given in 42 (26.25%) patients.
|Table 7: Laboratory and etiological profile of febrile thrombocytopenia casess|
Click here to view
|Figure 2: Mean platelet counts on follow-up days, overall comparison of platelet count (Friedman's test)|
Click here to view
In our study, 108 (67.5%) patients did not show any major complications; 52 (32.5%) patients had notable complications [Table 8]. Acalculous cholecystitis was also seen as a common complication in dengue fever.
Of 160 patients, 148 patients (91.88%) of them had good recovery, 8 (5.63%) patients expired, and 4 (2.5%) were referred to higher centers. In the present study, in the cases of septicemia, we had 5 (29.41%) mortalities out of 14 cases. In dengue cases, we had 2 (2.35%) out of 85 cases. One patient (11.1%) out of nine falciparum malaria cases expired. The range of platelet count in the mortality cases was <10,000/cumm in two cases, followed by 21–50,000/cumm in four cases, >50,000/cumm in one case, and 10,001–20,000 cells/cumm in one case. Of eight cases of mortality, 50% were due to septicemia with multiorgan dysfunction. All the expired patients had hypotension (100%). Abnormal renal function and acute respiratory distress syndrome (ARDS) was seen in 75% and 62.50% of the expired patients respectively. 2 (25%) cases had myocarditis and two expired patients were cases of dengue shock syndrome (DSS).
| Discussion|| |
In the present study, a total of 160 patients admitted over 1½ years in our hospital were studied. Of 160 patients studied, 91 (56.80%) were male and 69 (43.12%) were female. The affected populations were in the age group of 18–75 years, with the maximum prevalence of fever with thrombocytopenia in the age group of 21–30 years (40.62%) and more common in males (56.80%). Febrile thrombocytopenia is a common clinical condition and is caused by infectious and noninfectious etiology. In this study, infection was the established diagnosis in 88.12% of the cases. Dengue (53.12%) was the most common cause of fever with thrombocytopenia followed by malaria (15.62%), septicemia (8.75%), hematological conditions (7.50%), enteric fever (5.63%), unexplained cause (3.13%), mixed infections (dengue and malaria) (2.50%), leptospirosis (1.88%), DIC 2 (1.25%), and viral hepatitis (hepatitis B) (0.62%). Among hematological conditions, megaloblastic anemia (5%) was most common followed by acute myeloid leukemia (2.5%). In the study by Nair et al., septicemia with 29 cases was the leading cause of fever associated with thrombocytopenia. In a study by Saini, infections (83.50%) were also the most common cause. Dengue (47%) was the most common cause followed by malaria (20%). In a study by Lohitashwa et al., infections were also the most common cause, but here, malaria (41%) was the most common cause of fever with thrombocytopenia. In the clinical presentation of 160 cases in the present study, all had fever, 51.25% had myalgia followed by headache in 50% of patients, 41.25% had arthralgia, and 37.50% had nausea and vomiting. In a study conducted by Gondhali et al., of 100 cases, the most common symptoms were headache (90%) and body ache (92%). 22% cases had pallor, 28% had icterus, 12 cases had hepatomegaly, and 19 cases had splenomegaly. In our study, the most common sign was pallor (36%) followed by hypotension (18%), hepatosplenomegaly (16%), icterus (14%), hepatomegaly (12%), pedal edema (7%), splenomegaly (5%), and clubbing (0.6%).
In our study, the distribution of platelet counts at admission in the range of more than 50,000/cumm was seen in 111 (69%) cases as compared to 62 (56.80%) cases in the study by Nair et al. and 24% cases in the study by Saini. Platelet counts in the range of 21,000–50,000/cumm were seen in 39 (24.37%) cases in our study as compared to 28 (25.70%) in the study by Nair et al. Platelet counts <10,000/cumm were seen in 4 (2.50%) cases in our study as compared to 19 (17.50%) in the study by Nair et al. In our study, about the distribution of lowest platelet count during hospitalization, we found that 63 patients (39.37%) had platelet count in the range of 50,001–100,000, 49 patients (30.62%) in the range of 20,001–50,000, 28 patients (17.50%) in the range of 100,001–150,000, 12 patients (7.50%) in the range of 10,001–20,001, and 8 cases (5%) below 10,000 platelet count. This range correlated with bleeding tendencies and manifestations (P < 0.001). The clinical manifestations of thrombocytopenia were there in 53 (33.12%) patients, and petechiae/purpura (81.10%) was the most common bleeding manifestation followed by spontaneous bleeding (54.72%).
In septicemia, there may be a striking propensity toward intravascular fibrin deposition and thrombosis (consumption coagulopathy), which causes increased spontaneous bleeding [Table 9]. Infectious disease causes thrombocytopenia by impaired platelet production and increased destruction, and petechiae/purpura is the earliest manifestation of thrombocytopenia. A platelet count of approximately 5000–10,000 is required to maintain vascular integrity in the microcirculation. When the count is markedly decreased, petechiae first appear in the areas of increased venous pressure, the ankles and feet in an ambulatory patient. These findings were noted in our study also. Wet purpura and blood blisters are thought to denote an increased risk of life-threatening hemorrhage in the thrombocytopenic patient. Excessive bruising is seen in disorders of both platelet number and function.
In our study, bleeding tendency in dengue was about 32.94%, in septicemia was 64.28%, and malaria was 16%. In a study conducted by Saini, mixed infections with both dengue and malaria had the highest bleeding tendency (57.14%) followed by 55.30% of cases of dengue, 23.80% of cases of septicemia, and 27.12% of cases of malaria.
Anemia was seen in 45%, leucopenia was seen in 31.25%, leukocytosis was present in 11.25%, and abnormal renal functions were seen in 13.12% of cases. Abnormal serum bilirubin and SGPT were seen in 23.12% and 18.12% of cases, respectively. Dengue was the most common cause of anemia (29.16%) and leucopenia (66%). Majority of deranged renal function and leukocytosis were seen in septicemia cases. This correlated with the study by Gondhali et al., where 45% had leucopenia. Of those, 86.67% were dengue cases. Leukocytosis and abnormal renal functions were seen in 29% and 24%, of which majority were found in septicemia.
Saini showed increasing trend of platelet counts in 95% of cases. Lohitashwa et al. and Patil et al. also found the results consistent with our study, which showed increasing trend of platelet counts in 63.30% and 61% and decreasing trend of platelet counts in 36.70% and 39%, respectively. Good outcome was seen in 92.50% of patients with increasing trends in platelet count at the time of discharge. Mortality was noticed in 5% of patients, and 2.50% of cases of leukemia were referred to higher center. Major cause for mortality was septicemia in 62.50% patients followed by dengue (25%) and falciparum malaria (12.50%). Of eight cases of mortality, 50% were due to septicemia with multiorgan dysfunction. The expired patients had hypotension (100%), abnormal renal function (75%) and 62.50% had ARDS, two (25%) cases had myocarditis, and two cases of DSS. This indicates that deaths were related to underlying etiology and concomitant involvement of other organs, leading to multiorgan dysfunction. Renal failure was the most common organ dysfunction. In conclusion, septicemia was the major cause of mortality. The findings were consistent with existing literature. A study done on mortality in sepsis by Finfer et al. shows 37.50% mortality. Another study by Kaukonen et al. showed 18.4% mortality by sepsis. Variations in the definition of severe sepsis can explain differences in mortality rates among septic patients.
Common range of platelet count at the time of admission was 50,001–100,000 in 64 cases (40%), followed by ≥100,000 in 29.37% and 20,001–50,000 in 24.37%. There were 3.75% and 2.50% in the range of 10,001–20,001 and below 10,000 platelet count, respectively. Clinical manifestations of thrombocytopenia were present in 33.13% and absent in 66.87%. Among 53 cases with bleeding manifestations, petechiae/purpura was most common and was seen in 43 patients (81%) and spontaneous bleeding in 29 (54.70%) patients. Among spontaneous bleeding, melena (34.48%) was most common. Bleeding tendencies were more commonly seen in patients with platelet count <20,000 (91%). Among major causes, bleeding manifestations was seen in 32.94% of dengue cases, 16% of malaria cases, and 64.28% in septicemia cases. Majority of dengue cases (54%) had platelet count >50000 (mild–to-moderate thrombocytopenia). Among 85 dengue cases, six patients (7.06%) had dengue hemorrhagic fever and two cases (2.35%) were DSS. In our study, 148 patients (91.88%) had good recovery and 8 (5.63%) patients had mortality. There was no significant correlation between platelet count and mortality. In our study, septicemia was the major cause of mortality [Table 10].
Mortality depends on severity of disease, diagnosis made, available treatment and care, time of initiation of treatment, and associated other medical illness. Mortality can be reduced by early and right diagnosis and timely and effective treatment and care. In future, various pathological and microbiological imaging modalities should be needed for research and diagnosis of many viral hemorrhagic fevers.
| Conclusion|| |
Thrombocytopenia, a common observation in hemograms, needs a systematic evaluation to find out the underlying cause which can be of infective or noninfective etiology. All cases of thrombocytopenia may not have a bleeding manifestation and may be asymptomatic at initial presentation. However, in few cases, it may lead to severe bleeding which may be life-threatening for the patient.
Based on the results of our study, it can be inferred that platelet transfusion should be considered in all patients with platelet count below 20,000/cumm. Hence, the prompt diagnosis and immediate specific treatment of underlying etiology of febrile thrombocytopenia with maintenance of platelet count and hemostatic function give good recovery. In our setup, infections such as dengue fever, malaria, enteric fever, septicemia, leptospirosis, and noninfectious causes such as megaloblastic anemia are usually associated with febrile thrombocytopenia.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Suneetha DK, Inbanathan J, Sahna E, Shashank MS. Common etiology of acute fever with thrombocytopenia in a tertiary care hospital Mysuru. Int J Sci Stud 2016;4:61-4.
Nair PS, Jam A, Khanduri U, Kumar V. A study of fever associated thrombocytopenia. JAPI 2003;51:1173.
Barbara A. Konkle. Disorders of platelets and vessel wall. In: Fauci AS, Braunwald E, Kasper DL, Longo DL, Jameson JL, Loscalzo J, et al
., editors. Harrison's Principles of Internal Medicine. 19th
ed., Vol. 2, Ch. 140. New York, NY: McGrawHill; 2015. p. 725-32.
Dinarello CA, Porat R. Fever and hyperthermia. In: Kasper DL, Fauci AS, Hauser SL, Longo D, Jameson JL, Loscalzo J, editors. Harrison's Principles of Internal Medicine 19/E. Vol. 1. New York: McGraw Hill Professional; 2015. p. 123.
Craig JI, McClelland DB, Watson HG. Thrombocytopenia. In: College NR, Walker BR, Ralston SH, editors. Davidson's Principles and Practice of Medicine. 21st
ed. Edinburgh: Churchill Livingstone Elsevier; 2010. p. 1003-4.
Bichile SK. Platelet disorder. In: Munjal YP, editor. API Textbook of Medicine. 9th
ed., Vol. 1. New Delhi: Jaypee Brothers; 2012. p. 987-8.
Lee GR, Foerster J, Lukens J, Paraskevas F, Greer JP, Rodgers GM. Shirley parker levine-Miscellaneous causes of thrombocytopenia. In: Wintrobe's Clinical Haematology. 10th
ed., Vol. 2. Philadelphia: Lippincott Williams; 1999. p. 1623-9.
Kumar P, Chandra K. A clinical study of febrile thrombocytopenia: A hospital-based retrospective study. Indian J Clin Pract 2014;24:952-7.
Modi TN, Mehta AD, Sriram AS. Clinical profile of febrile thrombocytopenia. Journal of research in medical and dental science. 2016;4:115-20.
Saini KC. Clinical and etiological profile of fever with thrombocytopenia – A tertiary care hospital based study. J Assoc Physicians India 2018:66;33-6.
Lohitashwa SB, Gutthi LP, Vegesna S, Pundarikaksha V, Kolla S, Gundapaneni M. Clinical and lab profile of fever with thrombocytopenia. Int J Contemp Med Res 2017;4:1057-1061.
Gondhali MP, Vethekar M, Bhangale D, Choudhry K, Chaudhry M, Patrike G, et al
. Clinical assessment of fever with thrombocytopenia - A prospective study. Int J Med Res Health Sci 2016;5:258-77.
Patil P, Solanke P, Harshe G. To study clinical evaluation and outcome of patients with febrile thrombocytopenia. Int J Sci Res Publ 2014;4:1-3.
Finfer S, Bellomo R, Lipman J, French C, Dobb G, Myburgh J. Adult population incidence of severs sepsis in Australian and New Zealand ICU. Intensive Care Med 2004;30:589-96.
Kaukonen KM, Bailey M, Suzuki S, Pilcher D, Bellomo R. Mortality related to severe sepsis and septic shock among critically ill patients in Australia and New Zealand, 2000-2012. JAMA 2014;311:1308-16.
Vincent JL, Opal SM, Marshall JC, Tracey KJ. Sepsis definitions: Time for change. Lancet 2013;381:774-5.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7], [Table 8], [Table 9], [Table 10]