Kaohsiung J Med Sci May 2008 • Vol 24 • No 5 248
Fever is one of the more common chief complaints of patients who visit the emergency department (ED). As the cause of fever may be the common cold or fatal sepsis, it is a great challenge to emergency room physi-cians to make the correct diagnosis. In addition to the patient’s symptoms and signs, many tests are used to Received: Jun 21, 2007 Accepted: Dec 24, 2007
Address correspondence and reprint requests to: Dr Hon-Man Chan, Department of Emergency Medicine, Kaohsiung Medical University Hos-pital, 100 Shih-Chuan 1stRoad, Kaohsiung 807, Taiwan.
E-mail: [email protected]
C
HARACTERISTICS OF
F
EBRILE
P
ATIENTS WITH
N
ORMAL
W
HITE
B
LOOD
C
ELL
C
OUNTS AND
H
IGH
C-
REACTIVE
P
ROTEIN
L
EVELS IN
AN
E
MERGENCY
D
EPARTMENT
Kuan-Ting Liu,1Tzeng-Jih Lin,1,2and Hon-Man Chan1,3
1Department of Emergency Medicine, Kaohsiung Medical University Hospital, Kaohsiung Medical University, and Departments of 2Emergency Medicine and
3Surgery, Faculty of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan.
Fever is one of the more common chief complaints of patients who visit emergency departments (ED). Many febrile patients have markedly elevated C-reactive protein (CRP) levels and normal white blood cell (WBC) counts. Most of these patients have bacterial infection and no previous underlying disease of impaired WBC functioning. We reviewed patients who visited our ED between November 2003 and July 2004. The WBC count and CRP level of patients over 18 years of age who visited the ED because of or with fever were recorded. Patients who had normal WBC count (4,000–10,000/μL) and high CRP level (> 100 mg/L) were included. The data, including gender, age and length of hospital stay, were reviewed. Underlying diseases, diagnosis of the febrile disease and final condition were recorded according to the chart. Within the study period, 54,078 patients visited our ED. Of 5,628 febrile adults, 214 (3.8%) had elevated CRP level and nor-mal WBC count. The major cause of febrility was infection (82.24%). Most of these patients were admitted (92.99%). There were 32 patients with malignant neoplasm, nine with liver cirrhosis, 66 with diabetes mellitus and 11 with uremia. There were no significant differences in age and gen-der between patients with and those without neoplasm. However, a higher inhospital mortality rate and other causes of febrility were noted in patients with neoplasm. It was not rare in febrile patients who visited the ED to have a high CRP level but normal WBC count. These patients did not necessarily have an underlying malignant neoplasm or hematologic illness. Factors other than malignant neoplasm or hematologic illness may be associated with the WBC response, and CRP may be a better indicator of infection under such conditions.
Key Words:C-reactive protein, emergency department, fever
examine febrile patients. Sometimes, they have signif-icant symptoms and signs, and the diagnosis can be made by suitable examination. On other occasions, however, patients do not have obvious symptoms and signs, although physicians need an indicator to avoid a failure to diagnose severe disease.
In febrile patients, white blood cell (WBC) count is a common examination. In addition, procalcitonin, C-reactive protein (CRP) and interleukin-6 levels will be elevated in cases of severe infection [1–7]. Some studies suggest that these examinations could help to differentiate between less threatening fever and septic patients. CRP is a common available examination item in Taiwan’s hospitals. In practice, however, WBC count and CRP are not always elevated at the same time [8–10]. Some obviously septic patients do not have ele-vated WBC count but their CRP is markedly eleele-vated. Such a condition can also be found in some patients with hematologic disease and neoplasm [11,12].
We found that many patients without hematologic disease and neoplasm have normal WBC count and markedly elevated CRP. These patients usually have obvious infection or inflammation. Therefore, we ana-lyzed the characteristics of these patients, and then compared them with those of patients with malignancy.
M
ETHODSWe retrospectively reviewed patients who visited the ED of Kaohsiung Medical University Hospital between November 2003 and July 2004 because of fever or high body temperature (tympanic temperature > 38.3°C). Because the period of study was within 1 year of the severe acute respiratory syndrome (SARS) outbreak, all febrile patients received blood examinations including WBC count and CRP level. Adult patients (> 18 years) who had normal WBC counts (4,000–10,000/μL) and high CRP levels (>100mg/L) were included for further analysis. Patient characteristics including gender and age were recorded. Underlying diseases including dia-betes mellitus, end-stage renal disease, liver cirrhosis and malignant neoplasm were recorded by history taking and examination in hospital. The diagnosis and final condition on discharge from hospital were deter-mined according to the chart records filled in by the doctor in charge of the ward or ED. Student’s t test was used to compare age and days of hospitalization between patients with and those without underlying
malignant neoplasm and/or hematologic disease. χ2 and Fisher’s exact tests were used to examine the correlation between gender, cause of fever, hospital-ization, type of infection and mortality with underly-ing malignant neoplasm and/or hematologic disease.
R
ESULTSWithin the study period, 54,078 patients visited our ED. Of 5,628 febrile adults, 214 (3.8%) had an elevated CRP level and normal WBC count. The age of these patients ranged from 20 to 97 years. The characteris-tics of these patients are shown in Table 1. There were 32 patients with malignant neoplasm, nine with liver cirrhosis, 66 with diabetes mellitus and 11 with uremia. The major cause of febrility was infection (82.24%). Pneumonia and urinary tract infection were the leading diagnoses of infection (Table 2). Most of the patients
Table 1.Characteristics of patients with high C-reactive protein levels and normal white blood cell counts* Age (yr) 61.9± 18.2 Gender Female 96 (44.86) Male 118 (55.14) Underlying disease Malignancy 32 (14.95) Liver cirrhosis 9 (4.21) Diabetes mellitus 66 (30.84) Uremia 11 (5.14)
*Data presented as mean ± standard deviation or n (%).
Table 2.Causes of febrility and patient outcome* Infection 176 (82.24) Infection focus
Urinary tract 44 (25) Lung & bronchus 70 (39.77) Gastrointestinal tract 10 (5.68) Soft tissue 9 (5.11) Liver 12 (6.82) Multisite 11 (6.25) Others 20 (11.36) Number of hospitalizations 199 (92.99) Days of hospitalization 14.06± 12.91 Mortality 9 (4.52)
were admitted (92.99%). There were no significant dif-ferences in age and gender between patients with and those without neoplasm (Table 3). However, a higher inhospital mortality rate and other causes of febrility were noted in patients with neoplasm.
D
ISCUSSIONFebrile patients who have normal WBC counts and elevated CRP levels usually have infection (82.24%). CRP is an acute-phase reactant produced by the liver that can increase markedly in response to infection or inflammation. In a previous study [13], markedly increased CRP level (>100mg/L) was highly associated with severe sepsis. This makes it possible to distinguish pyelonephritis from cystitis, bacterial pneumonia from acute bronchitis, acute bronchitis from uncomplicated acute or chronic obstructive pulmonary disease, and bacterial meningitis from aseptic meningitis. However, the range of elevation is large: the higher the CRP level, the more sensitivity there is to an association with sepsis. For this reason, a cutoff point of 100 mg/L was selected in this study for its higher ability to detect the factor results in normal WBC counts in those patients. In Putto et al’s study [14], CRP of > 40 mg/L could detect 79% of bacterial infection with 90% specificity. However, CRP of 20–40 mg/L has been recorded in both viral and bacterial infec-tions. Many studies found that CRP was more sensi-tive than WBC counts in distinguishing bacterial infection [1,10].
Many studies have focused on the use of CRP in patients with malignancy, hematologic disease or neu-tropenia, because these patients do not have normal WBC response to infection [3,4,12,15–22]. Such studies have shown that CRP could help to diagnose sepsis in such patients. In a study of children with cancer, Santolaya et al [18] showed that patients with CRP level > 40 mg/L had bacterial infection (sensitivity of 100%, specificity of 76.6%). Arber et al [12] found that levels of CRP in sepsis were higher than in graft-versus-host disease. Although CRP is elevated in cancer itself, fever with elevated CRP could still reveal infection. Most of the patients in our study did not have malig-nancy or hematologic disease, but WBC count did not increase in those with infections. Although some of these patients had chronic disease, further study is needed to determine the cause of impaired WBC re-sponse in these patients. Other biomarkers like CRP may be more suitable to detect infection in such patients. In our study, there were no differences in age and gender between patients with and those without malignancy. Although most causes of febrility were infection in both types of patients, patients with malig-nancy still had higher incidences of causes of febrility other than infection. The inhospital mortality rate was higher in patients with malignancy. Our data did not attribute the mortality to the difference in severity of infection or underlying malignancy. The patients with malignancy had a greater possibility of having a rare infection or multiple site infection.
As fever is one of the most common complaints of patients who visit the ED, it is very important to Table 3.Comparison of characteristics and outcomes between patients with and those without malignancy*
With malignancy (n= 32) Without malignancy (n= 182) p Age (yr) 61.9± 19.3 61.8± 10.6 0.969†
Male gender 17 (53.1) 101 (55.5) 0.804‡
Infection 19 (59.4) 157 (86.3) < 0.005‡
Infection focus < 0.005‡
Urinary tract 3 (15.8) 41 (26.1) Lung & bronchus 10 (52.6) 60 (38.2) Multisite 5 (26.3) 6 (3.8) Others 1 (5.3) 50 (31.8)
Hospitalization 30 (93.8) 169 (92.9) 0.855‡
Days of hospitalization 17.4± 12.2 13.5± 13.4 0.131†
Mortality 4 (13.3) 5 (3) 0.031‡ *Data presented as mean ± standard deviation or n (%); †t test; ‡χ2test or Fisher’s exact test.
determine whether the cause is severe infection or inflammation. Many standard medical tests have aided clinical diagnosis, such as WBC counts, interleukin-6, interleukin-8, CRP, procalcitonin, soluble Fcγ recep-tor type III and mannose-binding protein [1,4,5,7,23]. Although studies have shown the value of these exam-inations, the majority, except for CRP and WBC count, are unavailable in the ED of most hospitals. Most infections can be diagnosed by clinical symptoms and signs, but diagnosis may be difficult in patients who cannot express their symptoms well, such as children. Accordingly, CRP could be used in febrile children [1,7,16,18,23–26] to distinguish bacterial infection. Furthermore, the causes of febrility may be difficult to distinguish in some situations including trauma [27] and bone marrow transplantation [12]. The CRP level test has value in such cases. Furthermore, many infectious or inflammatory diseases have no specific symptoms; marked elevation of CRP has significant diagnostic value in such cases as well.
There are several limitations to this study. First, the major goal of the study was to analyze the charac-teristics of febrile patients with normal WBC count and high CRP level. We lacked the data to confirm the roles in differential sepsis in this study. Secondly, this study did not determine whether CRP itself affects the disposition of the doctors. Further study is necessary to determine if doctors tend to suggest that patients with high CRP level be hospitalized. Finally, this study analyzed the data in an ED, so the results can be applied to patients in an ED, but it did not deter-mine if these patients had normal WBC counts through-out the course of disease or whether some patients developed high CRP levels during the course of the disease.
It was not rare for febrile patients who visited the ED to have high CRP level but normal WBC count. These patients usually had significant infection or inflammation and needed hospitalization and fur-ther treatment, but they did not necessarily have an underlying malignant neoplasm or hematologic illness. This suggests that some factors other than malignant neoplasm or hematologic illness may be associated with the WBC response, and that CRP may be a better indicator of infection under such conditions. Further studies are needed to elucidate what these factors may be. We believe that it is reasonable to check CRP level in addition to WBC count for patients who visit the ED due to fever.
R
EFERENCES1. Galetto-Lacour A, Zamora S, Gervaix A. Bedside procal-citonin and C-reactive protein tests in children with fever without localizing signs of infection seen in a referral center. Pediatrics 2003;112:1054–60.
2. Kawasaki Y, Hosoya M, Katayose M, et al. Correlation between serum interleukin 6 and C-reactive protein concentrations in patients with adenoviral respiratory infection. Pediatr Infect Dis J 2002;21:370–4.
3. Kallio R, Bloigu A, Surcel H, et al. C-reactive protein and erythrocyte sedimentation rate in differential diag-nosis between infections and neoplastic fever in patients with solid tumours and lymphomas. Support Care Cancer 2001;9:124–8.
4. Engel A, Mack E, Kern P, et al. An analysis of interleukin-8, interleukin-6 and C-reactive protein serum concen-trations to predict fever, gram-negative bacteremia and complicated infection in neutropenic cancer patients. Infection 1998;26:213–21.
5. Herrmann J, Blanchard H, Brunengo P, et al. TNF alpha, IL-1 beta and IL-6 plasma levels in neutropenic patients after onset of fever and correlation with the C-reactive protein (CRP) kinetic values. Infection 1994;22:309–15. 6. Fassbender K, Pargger H, Muller W, et al. Interleukin-6
and acute-phase protein concentrations in surgical intensive care unit patients: diagnostic signs in nosoco-mial infection. Crit Care Med 1993;21:1175–80.
7. Lehrnbecher T, Venzon D, de Haas M, et al. Assess-ment of measuring circulating levels of interleukin-6, interleukin-8, C-reactive protein, soluble Fc gamma receptor type III, and mannose-binding protein in febrile children with cancer and neutropenia. Clin Infect Dis 1999;29:414–9.
8. Verkkala K, Valtonen V, Jarvinen A, et al. Fever, leuco-cytosis and C-reactive protein after open-heart surgery and their value in the diagnosis of postoperative infec-tions. Thorac Cardiovasc Surg 1987;35:78–82.
9. Gozzard D, Yin J, Delamore I. The clinical usefulness of C-reactive protein measurements. Br J Haematol 1986; 63:411–4.
10. Povoa P, Almeida E, Moreira P, et al. C-reactive protein as an indicator of sepsis. Intensive Care Med 1998;24: 1052–6.
11. Grutzmeier S, von Schenck H. C-reactive protein dur-ing chemotherapy for acute leukemia with special refer-ence to non-infective causes of fever. Med Oncol Tumor Pharmacother 1986;3:71–5.
12. Arber C, Passweg J, Fluckiger U, et al. C-reactive protein and fever in neutropenic patients. Scand J Infect Dis 2000;32:515–20.
13. Morley J, Kushner I. Serum C-reactive protein levels in disease. Ann NY Acad Sci 1982;389:406–18.
14. Putto A, Ruuskanen O, Meurman O, et al. C reactive protein in the evaluation of febrile illness. Arch Dis Child 1986;61:24–9.
15. Timonen T, Koistinen P. C-reactive protein for detec-tion and follow-up of bacterial and fungal infecdetec-tions in
severely neutropenic patients with acute leukaemia. Eur J Cancer Clin Oncol 1985;21:557–62.
16. Katz J, Mustafa M, Bash R, et al. Value of C-reactive protein determination in the initial diagnostic evaluation of the febrile, neutropenic child with cancer. Pediatr Infect Dis J 1992;11:708–12.
17. Rintala E, Irjala K, Nikoskelainen J. Value of measure-ment of C-reactive protein in febrile patients with hema-tological malignancies. Eur J Clin Microbiol Infect Dis 1992;11:973–8.
18. Santolaya M, Cofre J, Beresi V. C-reactive protein: a valu-able aid for the management of febrile children with cancer and neutropenia. Clin Infect Dis 1994;18:589–95. 19. Gunther G, Gardlund B, Hast R, et al. Endotoxaemia
and inflammatory mediators in febrile patients with haematological disease. J Intern Med 1995;237:27–33. 20. Manian F. A prospective study of daily measurement of
C-reactive protein in serum of adults with neutropenia. Clin Infect Dis 1995;21:114–21.
21. Kostiala A, Kostiala I, Valtonen V, et al. Levels of C-reactive protein in patients with hematologic malig-nancies. Scand J Infect Dis 1985;17:407–10.
22. Harris R, Stone P, Hudson A, et al. C reactive protein rapid assay techniques for monitoring resolution of infection in immunosuppressed patients. J Clin Pathol 1984;37:821–5.
23. Heney D, Lewis I, Evans S, et al. Interleukin-6 and its relationship to C-reactive protein and fever in chil-dren with febrile neutropenia. J Infect Dis 1992;165: 886–90.
24. Lembo R, Marchant C. Acute phase reactants and risk of bacterial meningitis among febrile infants and chil-dren. Ann Emerg Med 1991;20:36–40.
25. Gervaix A, Galetto-Lacour A, Gueron T, et al. Usefulness of procalcitonin and C-reactive protein rapid tests for the management of children with urinary tract infection. Pediatr Infect Dis J 2001;20:507–11.
26. Pulliam P, Attia M, Cronan K. C-reactive protein in febrile children 1 to 36 months of age with clinically undetectable serious bacterial infection. Pediatrics 2001; 108:1275–9.
27. Miller P, Munn D, Meredith J, et al. Systemic inflamma-tory response syndrome in the trauma intensive care unit: who is infected? J Trauma 1999;47:1004–8.
!"VS==S==ON= !"VS==NO==OQ= !"#$%&' !"!#$ %&'( UMT !"#$NMM
