- Email: [email protected]
The C-reactive protein1
The Journal of Emergency Medicine, Vol. 17, No. 6, pp. 1019 –1025, 1999 Copyright © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0736-4679/99 $–see front matter
PII S0736-4679(99)00135-3
Clinical Laboratory in Emergency Medicine
THE C-REACTIVE PROTEIN Brian Clyne,
MD*
and Jonathan S. Olshaker,
MD*†
*Division of Emergency Medicine, Department of Surgery, University of Maryland Medical System and †Department of Emergency Medicine, Veterans Affairs Medical Center, Baltimore, Maryland Reprint Address: Brian Clyne, MD, Division of Emergency Medicine, University of Maryland Medical System, 22 South Greene Street, Baltimore, MD 21201
e Abstract—C-reactive protein (CRP) was identified in 1930 and was subsequently considered to be an “acute phase protein,” an early indicator of infectious or inflammatory conditions. Since its discovery, CRP has been studied as a screening device for inflammation, a marker for disease activity, and as a diagnostic adjunct. Improved methods of quantifying CRP have led to increased application to clinical medicine. In the emergency department (ED), CRP must be interpreted in the clinical context; no single value can be used to rule in or rule out a specific diagnosis. We conclude that CRP has limited utility in the ED. It may be a useful adjunct to serial examinations in equivocal presentations of appendicitis in those centers without ready access to computed tomography (CT) scan. It may be elevated with complications or treatment failures in patients with pneumonia, pancreatitis, pelvic inflammatory disease (PID), and urinary tract infections. In patients with meningitis, neonatal sepsis, and occult bacteremia, CRP is usually elevated. However, CRP has no role in diagnosing these clinical entities, and a normal CRP level should never delay antibiotic coverage. © 1999 Elsevier Science Inc.
of patients acutely infected with pneumococcal pneumonia that formed a precipitate when combined with polysaccharide C of Streptococcus pneumoniae (1). Subsequently, it was found that this reaction was not unique to pneumococcal pneumonia but could be found with a variety of other acute infections. This was early evidence of the body’s chemical response to inflammatory states and led to the characterization of other so-called “acute phase proteins” (2,3). Like many acute phase proteins, CRP is normally present in trace levels in serum but increases rapidly and dramatically in response to a variety of infectious or inflammatory conditions (4 – 6). Since its discovery, CRP has been studied as a screening device for occult inflammation, as a marker of disease activity, and as a diagnostic tool (7). Recently, more rapid and precise methods of quantifying CRP have led to a renewed interest in its value in clinical medicine (8,9). This article will focus on the potential use of CRP as an adjunct diagnostic test in the emergency department (ED).
e Keywords—C-reactive protein; acute phase proteins; clinical laboratory tests; emergency department
BIOCHEMISTRY/PHYSIOLOGY C-reactive protein is synthesized by hepatocytes. It is a pentameric protein consisting of five noncovalently bonded identical subunits with an overall molecular weight of approximately 118,00 daltons. In response to
INTRODUCTION C-reactive protein (CRP) was first detected in 1930 by Tillet and Frances, who identified a substance in the sera
Clinical Laboratory in Emergency Medicine is coordinated by Jonathan S. Olshaker, MD, of the University of Maryland Medical Center, Baltimore, Maryland
RECEIVED: 26 June 1998; FINAL ACCEPTED: 21 January 1999
SUBMISSION RECEIVED:
6 January 1999; 1019
1020
B. Clyne and J. S. Olshaker
infection or tissue inflammation, CRP production is stimulated by cytokines, particularly IL-6, IL-1, and tumor necrosis factor (4,10). Levels in healthy individuals are normally less than 10 mg/L; however, in disease states, this level increases in the first 6 to 8 h and can reach peak levels approaching 350 – 400 mg/L after approximately 48 h (4,5,7,8,11). Once in the tissue, CRP undergoes calcium-dependent binding to choline phosphatides such as lecithin, lysolecithin, and sphyngomyelin, as well as to polysaccharides and peptopolysaccharides present on bacteria, parasites, and fungi (6). The functional properties of CRP include the ability to activate complement through the classic complement pathway, and the ability to modulate the function of phagocytic cells. Although the exact function of CRP in vivo is not known, these properties suggest that it has a role in opsonization of infectious agents and damaged cells (4,6). Upon resolution of inflammation or tissue destruction, CRP levels rapidly decline with an elimination half-life estimated at 4 –9 h (5,11,12). This rapid postinflammatory decline makes it useful as a marker for disease activity. Thus, CRP has been found useful in monitoring exacerbations in chronic inflammatory conditions like rheumatoid arthritis, inflammatory bowel disease, and a number of vasculitic syndromes (11). In contrast to the erythrocyte sedimentation rate (ESR), CRP rises more rapidly and peaks earlier in response to inflammatory stimuli. It also returns to normal levels more quickly upon resolution of stimuli whereas ESR may not return to normal for several weeks despite clinical improvement. CRP levels are not affected by anemia, polycythemia, protein levels, red blood cell shape, patient age, or gender. All of these factors have been found to affect the ESR, and lead some to conclude that CRP is a better test to monitor acute inflammation and tissue injury (12–14).
results can be obtained in 30 min with an analytical sensitivity of ⬃0.04 mg/L (12). Normal human serum will contain CRP concentrations less than 10 mg/L with the mean level increasing slightly with age. There is no difference in mean concentrations between men and women although higher levels are found late in pregnancy (11). It is generally accepted that mild inflammation and viral infections cause elevation of CRP in the 10 – 40 mg/L range while active inflammation and bacterial infection produce levels of 40 –200 mg/L. Levels over 200 mg/L are found in severe bacterial infections and burns (10,16). A value above 100 mg/L is more likely to be associated with a bacterial infection although viral infections have caused CRP levels this high (6,8,17).
APPLICATION OF CRP IN THE ED Numerous studies conclude that CRP aids in the diagnosis of invasive bacterial disease, implying a role in the ED; however, there are a great deal of conflicting data regarding the use of CRP for diagnostic purposes (6,17–21). Some argue that as a nonspecific indicator of inflammation, CRP, by definition, cannot accurately differentiate among the many sources of potential tissue destruction (16). In addition, conclusive data have not been reproduced consistently. Young cites several design flaws inherent in many of the studies that support the use of CRP for diagnosis. These include small study groups, inappropriate reference ranges, and the reporting of peak CRP levels rather than initial levels at presentation (11). The following are examples of disease entities commonly seen in the ED to which CRP testing has been applied.
APPENDICITIS MEASUREMENT Until the late 1970s, CRP was measured using qualitative or semi-quantitative laboratory techniques, most commonly latex agglutination, which precluded its use as a differential diagnostic test because any degree of inflammation produced positive results (12,15). Presently, accurate and rapid quantitative measures of CRP are obtained using laser nephelometry, rate immunonephelometry or turbidimetry, and enzyme immunoassay. Nephelometric assays, which measure the formation of antigen-antibody complexes when anti-CRP antiserum reacts with patient serum, are the most widely used technique for measuring CRP today (14,15). Accurate
C-reactive protein has been studied in a variety of surgical conditions, mostly as an indicator of postoperative complications (22–24). Preoperatively, elevated CRP levels have been reported to aid in the diagnosis of acute appendicitis. The overall sensitivity of CRP in the studies we reviewed ranges from 40 – 87%, with a specificity of 53– 82% (25–29). Early in the course of appendicitis, the white blood cell count has shown the best diagnostic sensitivity among laboratory tests (30,31). After 12–24 h of symptoms, CRP is deemed by some to be useful, especially when serial levels show an increase (27,32–34). Consistent data are lacking, and many studies report elevated CRP levels in patients with normal appendixes, as well
C-Reactive Protein
as normal CRP levels in patients with gangrenous appendixes (33–37). In a study of 56 patients, Albu found that 26 of 26 patients (100%) with a CRP value greater than 25 mg/L after 12 h of symptoms had histologically diagnosed appendicitis (38). In another study, the combination of WBC ⬎ 10 (⫻1012/L), or CRP ⬎ 12 mg/L, or neutrophil count ⬎ 70% was 98% sensitive in acute appendicitis. The relative contribution of each of these three tests to the overall diagnostic accuracy is not addressed (39). Using the combination of elevated WBC or band neutrophils or CRP, Marchand reports a 100% sensitivity for acute appendicitis. In this study, CRP ⬎ 12 mg/L alone was 70% sensitive with a neutrophil count ⬎ 7.88 (⫻109/L) showing the highest individual sensitivity, 84% (31). Albu’s study found a negative predictive value of 100% and concluded that all patients with CRP less than 25 mg/L after 12 h of symptoms will have spontaneous resolution (38). In three studies, a normal CRP along with a normal WBC count and normal neutrophil count excluded acute appendicitis with a negative predictive value approaching 100% (31,36,39). It should be noted that these studies had small sample sizes, and employed different reference ranges and cutoff levels for an elevated CRP value. It is also unclear from these studies if patients without appendicitis were clinically improved, making CRP testing redundant. In a meta-analysis of the accuracy of CRP in diagnosing appendicitis, Hallan found a wide range of data for sensitivity (40 –99%) and specificity (27–90%). He cites differing patient populations and variations in cutoff values as sources of this disparate data. He concludes that “CRP is a test of medium diagnostic accuracy and is a little inferior to the total leukocyte count.” We agree with Hallan that based on the available literature, it is impossible to draw firm conclusions regarding the clinical usefulness of CRP in acute appendicitis (40). Among the myriad causes of acute abdominal pain, CRP cannot accurately identify acute appendicitis, and should not be used to make management decisions in equivocal cases. There is no substitute for serial clinical examinations during an observation period for abdominal pain. In most EDs with access to it, spiral computed tomography (CT) scan with or without rectal contrast will provide a quicker and more definitive diagnosis of appendicitis (41). However, in centers that have access to CRP testing and not to a spiral CT scan, there might be a role for CRP.
CHOLECYSTITIS Juvonen found that a CRP level ⬎ 30 mg/L had a sensitivity of 78% in 108 patients with histologically
1021
diagnosed cholecystitis. In the same study, ultrasound had a sensitivity of 79%. When positive ultrasound findings were combined with a CRP ⬎ 30 mg/L, sensitivity was reported as 97% (42). C-reactive protein was found to be a more sensitive marker than ESR, white blood cell count, or fever in identifying acute cholecystitis. The study included 18 patients with elevated CRPs or ultrasound evidence suggestive of cholecystitis and no histologic evidence of disease. It also used peak CRP levels to calculate sensitivities, with peak levels typically occurring 2–5 days after hospital admission. The study is the only one in the literature to address CRP in cholecystitis and provides no compelling evidence that CRP would be of use in the ED. Given the lack of data, routine CRP testing for cholecystitis cannot be advocated, especially since serial examinations and sonography provide sufficient diagnostic discrimination for this entity.
PANCREATITIS C-reactive protein has been used to diagnose and predict the severity of pancreatitis (43,44). Wilson found that peak CRP levels ⱖ 210 mg/L discriminated between patients with clinically mild and severe pancreatitis with a sensitivity of 83% and a specificity of 85% (45). Heath obtained similar results, with peak concentrations ⬎ 150 mg/L predicting severe attacks (46). In these studies, pancreatitis was diagnosed by serum amylase and was graded on severity based on clinical impression. While the cutoff point of serum CRP to differentiate mild from severe forms of acute pancreatitis is debated, it is generally agreed upon that the initial value on admission is an insensitive marker of disease severity, implying limited use in the ED (45,47).
PELVIC INFLAMMATORY DISEASE C-reactive protein has been studied along with ESR in the diagnosis and prediction of severity of pelvic inflammatory disease (PID). Lehtinen concludes that both CRP and ESR should be routinely used to augment the clinical diagnosis of PID. However, in this study, no statistical significance was found between CRP levels in women presenting with endometritis only and no PID at all. The overall sensitivity and specificity of CRP in the diagnosis of PID was 74% and 67%, respectively, using a cutoff level of 20 mg/L. No significant difference was found when compared with the ESR in the diagnosis of PID (48). Soper advocates the routine use of both the ESR and CRP to increase the overall specificity of the diagnosis of PID but offers no cutoff CRP value to indicate likelihood
1022
B. Clyne and J. S. Olshaker
of infection (49). Miettinen found that an ESR ⱖ 40 mm/h combined with a CRP ⱖ 60 mg/L had a sensitivity and specificity of 97% and 61%, respectively, in detecting severe PID diagnosed by laparoscopy (50). In the studies reviewed, there are significant numbers of patients with other gynecologic problems causing raised CRP. Causes of false-positives include ectopic pregnancy, spontaneous abortions, and uterine fibroids. Based on the data, we conclude that higher CRP levels correspond to more severe cases of PID and that persistently elevated levels imply treatment failure or complications. CRP is potentially useful for monitoring response to antibiotic therapy in patients who are not clinically improving. In the ED, there is no convincing data that CRP should influence patient disposition or antibiotic regimen in suspected PID.
PNEUMONIA In the diagnosis of pneumonia, CRP is not useful for discriminating between bacterial and viral infections (51,52). However, it has been found useful as a marker for antibiotic treatment failure or the development of infectious complications (53,54). This has been shown in both the non-HIV and HIV populations (55). Although higher CRP levels usually correspond to bacterial pneumonia, especially due to S. pneumoniae, the ED management should be based on traditional parameters and on the overall clinical assessment (56). There are no data to suggest that CRP is a better prognostic indicator than fever, hypoxia, respiratory effort, or radiography in patients with pneumonia. In outpatients treated with antibiotics and those diagnosed with viral pneumonia who present to the ED with persistent symptoms, an elevated CRP level greater than 100 mg/L likely indicates treatment failure.
URINARY TRACT INFECTION Most data on CRP in urinary tract infection come from the pediatric literature. In all patients with clinically obvious pyelonephritis (14 of 14), Jodal found CRP levels ⬎ 25 mg/L, while all patients with uncomplicated infection (10 of 10) had CRP levels ⬍ 25 mg/L (21). C-reactive protein is not a useful marker to help distinguish between simple and complicated urinary tract infection in patients without clinical signs of acute pyelonephritis (21,57). False positives may occur in acute cystitis with extensive bladder wall irritation. Higher CRP levels, in general, are associated with antibiotic failures and with pyelonephritis, especially in infants (58,59).
MENINGITIS Serum CRP has been reported to accurately diagnose bacterial meningitis and to correspond with resolution of symptoms after successful antibiotic treatment (60 – 62). Recent studies have focused on cerebrospinal fluid (CSF) CRP to distinguish bacterial from viral meningitis (60,63–72). Sensitivities of CSF CRP in diagnosing bacterial meningitis have been reported as high as 97–100% on initial lumbar puncture (63,64,66). In some cases, CSF CRP was a more sensitive indicator than more traditional parameters like CSF neutrophil count and Gram’s stain (63– 66). Corall reported a sensitivity of 100% and a specificity of 94% for detecting culturepositive bacterial meningitis by measuring CSF CRP on initial lumbar puncture. This study, however, was limited to patients with CSF pleocytosis ⬎ 10 WBC/mm3 (65). Benjamin found that CSF CRP had a sensitivity of only 66% and concluded that it is too insensitive to be useful, while the serum CRP is too nonspecific for routine application (71). Donald found considerable overlap in the CSF CRP levels among patients with bacterial and viral meningitis, limiting clinical usefulness (73). In a study of neonates, Philip found that CSF CRP did not distinguish between infants with meningitis and those with no infection (74). He argues against the routine use of CSF CRP measurements in suspected neonatal sepsis and meningitis. Using a semiquantitative method, Komorowski found that even in the setting of pleocytosis with CSF leukocyte count ⬎ 10/L, a CSF CRP ⬎ 8 mg/L was still only 35% sensitive in identifying bacterial meningitis (70). In studies of serum and CSF CRP, there are significant numbers of false positives and false negatives (67,68). Reasons for false positives include viral meningitis, brain abscess, other occult sources of inflammation, and the use of semiquantitative methods of measuring CRP. False negatives are most likely due to early bacterial infection, poor diffusion of CRP into the CSF, group B streptococcal infections, or a diminished inflammatory response in neonates (72). We conclude that the discrimination between viral and bacterial meningitis cannot be made with any certainty by serum or CSF CRP, nor should these tests influence patient management in the ED. Any clinical suspicion of meningitis, even in patients with equivocal CSF findings, should prompt early broad spectrum antibiotic coverage.
NEONATAL SEPSIS AND PEDIATRIC INFECTIONS In the neonatal period, CRP has been studied extensively for its ability to detect early infection (75– 80). Berger
C-Reactive Protein
1023
found that after the first 3 days of life, CRP is the single best test in early detection of neonatal septicemia, and that serial measurements parallel the course of infection and the efficacy of antibiotic treatment (75). Forest found that serial CRP values lead to increased accuracy of detecting bacterial infection in neonates. In this study, all neonates with proven bacterial infection had a CRP level above 80 mg/L on at least one occasion during serial measurements 12 h apart (76). Similarly, Mathers states that the main role of CRP lies in the exclusion or confirmation of infection 24 h after the initial clinical concern. He concludes, like others, that CRP has no role in the early diagnosis of neonatal sepsis and does not support the withholding of antibiotics based on a normal value (77). In a thorough review, DaSilva concludes that CRP is the best single diagnostic test in the evaluation of neonates with suspected sepsis. However, the report does not address the length of time required before CRP reaches optimal sensitivity. He emphasizes that clinicians cannot base management decisions on the available literature because of wide variability in measurement techniques and study designs (78). Bacteremia occurs in 3– 4% of febrile children evaluated in an ED setting, frequently without an obvious source of infection (20). In general, higher serum CRP levels are found on admission in patients who are bacteremic (81– 83). However, one study reported 8 bacteremic patients with initially normal CRP measurements and another found no significant difference in CRP values among patients with or without bacteremia (79,84). The frequency of false negatives in neonatal sepsis and bacteremia render CRP unhelpful in the ED setting.
SUMMARY We regularly encounter equivocal presentations of disease in the ED. In these situations, we often turn to the clinical laboratory to help confirm or disprove our suspicions. The measurement of C-reactive protein has enjoyed periodic emphasis over the years as a measure of general illness and as an adjunct to physical examination and other laboratory data. Unfortunately, the nonspecific nature of the acute phase response prevents CRP from being a useful discriminatory diagnostic test. Based on the literature, we conclude that CRP has a limited role, if any, in the diagnosis and management of commonly seen ED problems. Faster and more interpretable tests are available to help diagnose equivocal cases of appendicitis, cholecystitis, and PID. C-reactive protein adds little to the diagnosis of pneumonia, urinary tract infections, and pancreatitis. False negative CRP results in the setting of meningitis, neonatal sepsis, and bacteremia are com-
mon and therefore unhelpful. A single CRP value should not factor into the decision to treat these patients.
REFERENCES 1. Tillett WS, Francis T. Serological reactions in pneumonia with a non-protein somatic fraction of the pneumococcus. J Exp Med 1930;52:561–71. 2. Abernathy TJ, Avery OT. The occurrence during acute infection of a protein not normally present in the blood. I. Distribution of the reactive protein in patients’ sera and the effects of calcium on the flocculation reaction with C polysaccharide of pneumococcus. J Exp Med 1941;73:173. 3. MacLeod CM, Avery OT. The occurrence during acute infections of a protein not normally present in blood. II. Isolation and properties of the C-reactive protein. J Exp Med 1941;73:183. 4. Kolb-Bachofen V. A review on the biological properties of C-reactive protein. Immunobiology 1991;183:133– 45. 5. Gewurz H, Mold C, Siegal J, Fiedel B. C-reactive protein and the acute phase response. Adv Int Med 1982;27:345–72. 6. Ballou SP, Kushner I. C-reactive protein and the acute phase response. Adv Intern Med 1992;37:313–36. 7. Pepys MB. C-reactive protein fifty years on. Lancet 1981;1:653–7. 8. Palosuo T, Husman T, Koistinen J, Aho K. C-reactive protein in population samples. Acta Med Scand 1986;220:175–9. 9. Harmoinen A, Hallstrom O, Gronroos P. Rapid quantitative determination of C-reactive protein using laser nephelometer. Scand J Clin Lab Invest 1980;40:293. 10. Jupe D. The acute phase response and laboratory testing. Aust Fam Phys 1996;25:324 –9. 11. Young B, Gleeson M, Cripps, AW. C-reactive protein: a critical review. Pathology 1991;23:118 –24 12. Jaye DL, Waites KB. Clinical applications of C-reactive protein in pediatrics. Pediatr Infect Dis J 1997;16:735– 47. 13. International Committee for Standardization in Haematology. Guidelines on selection of laboratory tests for monitoring the acute phase response. J Clin Pathol 1988;41:1203–12. 14. Deodar S. C-reactive protein: the best laboratory indicator available for monitoring disease activity. Cleve Clin J Med 1989;56: 126 –30. 15. Powell LJ. C-reactive protein-a review. Am J Med Tech 1979;45: 138 – 48. 16. Stuart J, Whicher JT. Tests for detecting and monitoring the acute phase response. Arch Dis Child 1988;63:115–7. 17. Morley JJ, Kushner I. Serum C-reactive protein levels in disease. Ann NY Acad Sci 1982;389:406 –18. 18. Nudelman R, Kagan BM. C-reactive protein in pediatrics. Adv Pediatr 1983;30:517– 47. 19. Okamura JM, Miyagi JM, Terada K, Hokama Y. Potential clinical applications of C-reactive protein. J Clin Lab Anal 1990;4:231–5. 20. Singer JI, Buchino JJ, Chabali R. Selected laboratory in pediatric emergency care. Emerg Med Clin North Am 1986;4:377–96. 21. Jodal U, Lindberg U, Lincoln K. Level diagnosis of symptomatic urinary tract infections in childhood. Acta Paediatr Scand 1975; 64:201– 8. 22. Mustard RA, Bohnen JMA, Haseeb S, Kasina R. C-reactive protein levels predict postoperative septic complications. Arch Surg 1987; 122:69 –73. 23. Orriss DE. Serial serum C-reactive protein levels as an indicator of infection in cardiac transplant patients. Med Lab Sci 1988;45:116 – 20. 24. Boralessa H, DeBeer FC, Manchie A, Whitman JG, Pepys MB. C-reactive protein in patients undergoing cardiac surgery. Anaesthesia 1986;41:11–5. 25. Chi CH, Shiesh SC, Chen KW, Wu MH, Lin XZ. C-reactive protein for the evaluation of acute abdominal pain. Am J Emerg Med 1996;14:254 – 6. 26. Thompson MM, Underwood MJ, Dookeran KA, Lloyd DM, Bell
1024
27.
28.
29.
30.
31.
32. 33.
34.
35.
36. 37.
38. 39.
40.
41.
42.
43. 44.
45.
46.
47.
48.
49. 50.
PRF. Role of sequential leucocyte counts and C-reactive protein measurements in acute appendicitis. Br J Surg 1992;79:822– 4. Oosterhuis WP, Zwinderman AH, Teeuwen M, et al. C-reactive protein in the diagnosis of acute appendicitis. Eur J Surg 1993; 159:115–9. Van Lente F, Marchand A, Galen RS. Evaluation of the nephelometric assay of CRP using the Beckman ICS. Clin Chem 1981; 27:1075. Al-Saigh AH. C-reactive protein in the differential diagnosis of the acute abdomen, especially acute appendicitis. J R Coll Surg Edinb 1992;37:238 – 40. Gronroos JM, Forsstrom JJ, Irjala K, et al. Phospholipase A2, C-reactive protein, and white blood cell count in the diagnosis of acute appendicitis. Clin Chem 1994;40:1757– 60. Marchand A, Van Lente F, Galen R. The assessment of laboratory tests in the diagnosis of acute appendicitis. Am J Clin Pathol 1983;80:369 –74. Chen SC, Wang SM. C-reactive protein in the diagnosis of acute appendicitis. Am J Emerg Med 1996;14:101–3. Thimsen DA, Tong GK, Gruenberg, JC. Prospective evaluation of C-reactive protein in patients suspected to have acute appendicitis. Am Surg 1989;55:466 – 8. Eriksson S, Granstrom L, Olander B, et al. Sensitivity of interleukin-6 (IL-6) and C-reactive protein concentrations in the diagnosis of acute appendicitis. Eur J Surg 1995;161:41–5. Peltola H, Ahlqvist J, Rapola J, et al. C-reactive protein compared with white blood cell count and erythrocyte sedimentation rate in the diagnosis of acute appendicitis in children. Acta Chir Scand 1986;152:55– 8. Dueholm S, Bagi P, Bud M. Laboratory aid in the diagnosis of acute appendicitis. Dis Colon Rectum 1989;32:855–9. Mikaelsson C, Arnbjornsson E. The value of C-reactive protein determinations in patients with suspected acute appendicitis. Ann Chir Gyn 1984;73:281– 4. Albu E, Miller BM, Choi Y, et al. Diagnostic value of C-reactive protein in acute appendicitis. Dis Colon Rectum 1994;37:49 –51. Van Dieijen-Visser MP, Go PMNYH, Brombacher PJ. The value of laboratory tests in patients suspected of acute appendicitis. Eur J Clin Chem Clin Biochem 1991;29:749 –52. Hallan S, Asberg A. The accuracy of C-reactive protein in diagnosing acute appendicitis—a meta-analysis. Scand J Clin Lab Invest 1997;57:373– 80. Rao PM, Rhea JT, Novelline RA, et al. Helical CT technique for the diagnosis of appendicitis: prospective evaluation of a focused appendix CT examination. Radiology 1997;202:139 – 44. Juvonen T, Kiviniemi H, Niemela O, et al. Diagnostic accuracy of ultrasonography and C-reactive protein concentration in acute cholecystitis: a prospective clinical study. Eur J Surg 1992;158:365–9. Mazlam MZ, Hodgson HJF. Why measure C-reactive protein? Gut 1994;35:5–7. Sternby B, O’Brien JF, Zinsmeister AR, et al. What is the best biochemical test to diagnose acute pancreatitis? A prospective clinical study. Mayo Clin Proc 1996;71:1138 – 44. Wilson C, Heads A, Shenkin A, et al. C-reactive protein, antiproteases and complement factors as objective markers of severity in acute pancreatitis. Br J Surg 1989;76:177– 81. Heath DI, Cruickshank A, Gudgeon M, et al. Role of IL-6 in mediating the acute phase protein response and potential as early means of severity assessment in acute pancreatitis. Gut 1993;34: 41–5. Pezilli R, Billi P, Cappelletti O, Barakat B. Serum C-reactive protein in acute biliary pancreatitis. Is it a reliable marker for the early assessment of severity of the disease? Ital J Gastro Hepatol 1997;29:554 –7. Lehtinen M, Laine S, Heinonen PK, et al. Serum C-reactive protein determination in acute pelvic inflammatory disease. Am J Obstet Gynecol 1986;154:158 –9. Soper DE. Diagnosis and laparoscopic grading of acute salpingitis. Am J Obstet Gynecol 1991;164:1370 – 6. Miettinen AK, Heinonen PK, Laippala P. Test performance of erythrocyte sedimentation rate and C-reactive protein in assessing
B. Clyne and J. S. Olshaker
51.
52.
53. 54.
55.
56.
57.
58. 59.
60. 61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
the severity of acute pelvic inflammatory disease. Am J Obstet Gynecol 1993;169:1143–9. Korppi M, Heiskanen-Kosma T, Leinonen M. White blood cells, C-reactive protein and erythrocyte sedimentation rate in pneumococcal pneumonia in children. Eur Respir J 1997;10:1125–9. McCarthy PL, Frank AL, Ablow RC, et al. Value of the C-reactive protein in the differentiation of bacterial and viral pneumonia. J Paediatr 1978;92:454 – 6. Smith RP, Lipworth BJ. C-reactive protein in simple communityacquired pneumonia. Chest 1995;107:1028 –31. Smith RP, Lipworth BJ, Cree IA, et al. C-reactive protein, a clinical marker in community-acquired pneumonia. Chest 1995; 108:1288 –91. Storgaard M, Laursen AL, Andersen PL. The C-reactive protein responses in HIV infected patients with pneumonia. Scand J Infect Dis 1993;25:305–9. Ortqvist A, Hedlund J, Wretlind B, Carlstrom A, Kalin M. Diagnostic and prognostic value of IL-6 and C-reactive protein in community-acquired pneumonia. Scand J Infect Dis 1995;27:457– 62. Hellerstein S, Duggan E, Welchert E, Mansour F. Serum C-reactive protein and the site of urinary tract infections. J Pediatr 1982;100:21–5. Hanson LA, Jodal U, Sabel K, Wadsworth C. The diagnostic value of C-reactive protein. Ped Infect Dis 1983;2:87–90. Saxstad J, Nilsson LA, Hanson LA. C-reactive protein in serum from infants as determined with immunodiffusion techniques. Infants with various infections. Acta Paediatr Scand 1970;59:676 – 80. Peltola HO. C-reactive protein for rapid monitoring of infections of the central nervous system. Lancet 1982;1:980 –3. Lembo RM, Marchant CD. Acute phase reactants and risk of bacterial meningitis among febrile infants and children. Ann Emerg Med 1991;20:36 – 44. Clarke D, Cost K. Use of serum C-reactive protein in differentiating septic from aseptic meningitis in children. J Pediatr 1983; 102:718 –20. Gershom EB, Briggeman-Mol GJJ, de Zegher F. Cerebrospinal fluid C-reactive protein in meningitis: diagnostic value and pathophysiology. Eur J Pediatr 1986;145:246 –9. MacFarlane DE, Narla VR. Cerebrospinal fluid C-reactive protein in the laboratory diagnosis of bacterial meningitis. Acta Paediatr Scand 1985;74:560 –3. Corrall CJ, Pepple JM, Moxon ER, Hughes WT. C-reactive protein in spinal fluid of children with meningitis. J Pediatr 1981;1981: 365–9. Abramson JS, Hampton KD, Babu S, et al. The use of C-reactive protein from cerebrospinal fluid for differentiating meningitis from other central nervous system diseases. J Infect Dis 1985;151: 854 – 8. Hansson LO, Axelsson G, Linne T, Aurelius E, Lindquist L. Serum C-reactive protein in the differential diagnosis of acute meningitis. Scand J Infect Dis 1993;25:625–30. Sabal KG, Hanson LA. The clinical usefulness of C-reactive protein determination in bacterial meningitis and septicemia in infancy. Acta Paediatr Scand 1974;63:381– 8. Stearman M, Southgate HJ. The use of cytokine and C-reactive protein measurements in cerebrospinal fluid during acute infective meningitis. Ann Clin Biochem 1994;31:255– 61. Komorowski RA, Farmer SG, Knox KK. Comparison of cerebrospinal fluid C-reactive protein and lactate for diagnosis of meningitis. J Clin Micro 1986;24:982–5. Benjamin DR, Opheim KE, Brewer L. Is C-reactive protein useful in the management of children with suspected bacterial meningitis? Am J Clin Pathol 1984;81:779 – 82. Gray BM, Simmons DR, Mason H, Barnum S, Volanakis JE. Quantitative levels of C-reactive protein in cerebrospinal fluid in patients with bacterial meningitis and other conditions. J Pediatr 1986;108:665–70. Donald PR, Strachan AF, Schoeman JF, DeBeer FC. Cerebrospinal
C-Reactive Protein
74. 75. 76. 77. 78.
fluid C-reactive protein in infective meningitis in childhood. J Lab Clin Med 1985;106:424 –7. Philip AGS, Baker CJ. Cerebrospinal fluid C-reactive protein in neonatal meningitis. J Pediatr 1983;102:715–7. Berger C, Uehlinger J, Ghelfi D, Blau N, Fanconi S. Comparison of C-reactive protein and white blood cell count with differential in neonates at risk for septicaemia. Eur J Pediatr 1995;154:138 – 44. Forest J, Lariviere F, Dolce P, Masson M, Nadeau L. C-reactive protein as biochemical indicator of bacterial infection in neonates. Clin Biochem 1986;19:192– 4. Mathers NJ, Pohlandt F. Diagnostic audit of C-reactive protein in neonatal infection. Eur J Paediatrics 1987;146:147–51. DaSilva O, Ohlsson A, Kenyon C. Accuracy of leukocyte indices and C-reactive protein for diagnosis of neonatal sepsis: a critical review. Pediatr Infect Dis J 1995;14:362– 6.
1025 79. Peltola H, Jaakkola M. C-reactive protein in early detection of bacteremic versus viral infections in immunocompetent and compromised children. J Pediatr 1988;113:641– 6. 80. Pourcyrous M, Bada HS, Korones SB, Baselski V, Wong SP. Significance of serial C-reactive protein responses in neonatal infection and other disorders. Pediatrics 1993;92:431–5. 81. Bennish M, Vardiman J, Beem MO. The zeta sedimentation ratio in children. J Pediatr 1984:104:249. 82. McCarthy PL, Jekel JK, Dolan TF. Comparison of acute phase reactants in pediatric patients with fever. Pediatrics 1978;62:716. 83. Kohli V, Singhi S, Sharma P, Ganguly NK. Value of serum C-reactive protein concentrations in febrile children without apparent focus. Ann Trop Pediatr 1993;13:373– 8. 84. McCabe RE, Remington JA. C-reactive protein in patients with bacteremia. J Clin Microbiol 1984;20:317–91.
PII S0736-4679(99)00135-3
Clinical Laboratory in Emergency Medicine
THE C-REACTIVE PROTEIN Brian Clyne,
MD*
and Jonathan S. Olshaker,
MD*†
*Division of Emergency Medicine, Department of Surgery, University of Maryland Medical System and †Department of Emergency Medicine, Veterans Affairs Medical Center, Baltimore, Maryland Reprint Address: Brian Clyne, MD, Division of Emergency Medicine, University of Maryland Medical System, 22 South Greene Street, Baltimore, MD 21201
e Abstract—C-reactive protein (CRP) was identified in 1930 and was subsequently considered to be an “acute phase protein,” an early indicator of infectious or inflammatory conditions. Since its discovery, CRP has been studied as a screening device for inflammation, a marker for disease activity, and as a diagnostic adjunct. Improved methods of quantifying CRP have led to increased application to clinical medicine. In the emergency department (ED), CRP must be interpreted in the clinical context; no single value can be used to rule in or rule out a specific diagnosis. We conclude that CRP has limited utility in the ED. It may be a useful adjunct to serial examinations in equivocal presentations of appendicitis in those centers without ready access to computed tomography (CT) scan. It may be elevated with complications or treatment failures in patients with pneumonia, pancreatitis, pelvic inflammatory disease (PID), and urinary tract infections. In patients with meningitis, neonatal sepsis, and occult bacteremia, CRP is usually elevated. However, CRP has no role in diagnosing these clinical entities, and a normal CRP level should never delay antibiotic coverage. © 1999 Elsevier Science Inc.
of patients acutely infected with pneumococcal pneumonia that formed a precipitate when combined with polysaccharide C of Streptococcus pneumoniae (1). Subsequently, it was found that this reaction was not unique to pneumococcal pneumonia but could be found with a variety of other acute infections. This was early evidence of the body’s chemical response to inflammatory states and led to the characterization of other so-called “acute phase proteins” (2,3). Like many acute phase proteins, CRP is normally present in trace levels in serum but increases rapidly and dramatically in response to a variety of infectious or inflammatory conditions (4 – 6). Since its discovery, CRP has been studied as a screening device for occult inflammation, as a marker of disease activity, and as a diagnostic tool (7). Recently, more rapid and precise methods of quantifying CRP have led to a renewed interest in its value in clinical medicine (8,9). This article will focus on the potential use of CRP as an adjunct diagnostic test in the emergency department (ED).
e Keywords—C-reactive protein; acute phase proteins; clinical laboratory tests; emergency department
BIOCHEMISTRY/PHYSIOLOGY C-reactive protein is synthesized by hepatocytes. It is a pentameric protein consisting of five noncovalently bonded identical subunits with an overall molecular weight of approximately 118,00 daltons. In response to
INTRODUCTION C-reactive protein (CRP) was first detected in 1930 by Tillet and Frances, who identified a substance in the sera
Clinical Laboratory in Emergency Medicine is coordinated by Jonathan S. Olshaker, MD, of the University of Maryland Medical Center, Baltimore, Maryland
RECEIVED: 26 June 1998; FINAL ACCEPTED: 21 January 1999
SUBMISSION RECEIVED:
6 January 1999; 1019
1020
B. Clyne and J. S. Olshaker
infection or tissue inflammation, CRP production is stimulated by cytokines, particularly IL-6, IL-1, and tumor necrosis factor (4,10). Levels in healthy individuals are normally less than 10 mg/L; however, in disease states, this level increases in the first 6 to 8 h and can reach peak levels approaching 350 – 400 mg/L after approximately 48 h (4,5,7,8,11). Once in the tissue, CRP undergoes calcium-dependent binding to choline phosphatides such as lecithin, lysolecithin, and sphyngomyelin, as well as to polysaccharides and peptopolysaccharides present on bacteria, parasites, and fungi (6). The functional properties of CRP include the ability to activate complement through the classic complement pathway, and the ability to modulate the function of phagocytic cells. Although the exact function of CRP in vivo is not known, these properties suggest that it has a role in opsonization of infectious agents and damaged cells (4,6). Upon resolution of inflammation or tissue destruction, CRP levels rapidly decline with an elimination half-life estimated at 4 –9 h (5,11,12). This rapid postinflammatory decline makes it useful as a marker for disease activity. Thus, CRP has been found useful in monitoring exacerbations in chronic inflammatory conditions like rheumatoid arthritis, inflammatory bowel disease, and a number of vasculitic syndromes (11). In contrast to the erythrocyte sedimentation rate (ESR), CRP rises more rapidly and peaks earlier in response to inflammatory stimuli. It also returns to normal levels more quickly upon resolution of stimuli whereas ESR may not return to normal for several weeks despite clinical improvement. CRP levels are not affected by anemia, polycythemia, protein levels, red blood cell shape, patient age, or gender. All of these factors have been found to affect the ESR, and lead some to conclude that CRP is a better test to monitor acute inflammation and tissue injury (12–14).
results can be obtained in 30 min with an analytical sensitivity of ⬃0.04 mg/L (12). Normal human serum will contain CRP concentrations less than 10 mg/L with the mean level increasing slightly with age. There is no difference in mean concentrations between men and women although higher levels are found late in pregnancy (11). It is generally accepted that mild inflammation and viral infections cause elevation of CRP in the 10 – 40 mg/L range while active inflammation and bacterial infection produce levels of 40 –200 mg/L. Levels over 200 mg/L are found in severe bacterial infections and burns (10,16). A value above 100 mg/L is more likely to be associated with a bacterial infection although viral infections have caused CRP levels this high (6,8,17).
APPLICATION OF CRP IN THE ED Numerous studies conclude that CRP aids in the diagnosis of invasive bacterial disease, implying a role in the ED; however, there are a great deal of conflicting data regarding the use of CRP for diagnostic purposes (6,17–21). Some argue that as a nonspecific indicator of inflammation, CRP, by definition, cannot accurately differentiate among the many sources of potential tissue destruction (16). In addition, conclusive data have not been reproduced consistently. Young cites several design flaws inherent in many of the studies that support the use of CRP for diagnosis. These include small study groups, inappropriate reference ranges, and the reporting of peak CRP levels rather than initial levels at presentation (11). The following are examples of disease entities commonly seen in the ED to which CRP testing has been applied.
APPENDICITIS MEASUREMENT Until the late 1970s, CRP was measured using qualitative or semi-quantitative laboratory techniques, most commonly latex agglutination, which precluded its use as a differential diagnostic test because any degree of inflammation produced positive results (12,15). Presently, accurate and rapid quantitative measures of CRP are obtained using laser nephelometry, rate immunonephelometry or turbidimetry, and enzyme immunoassay. Nephelometric assays, which measure the formation of antigen-antibody complexes when anti-CRP antiserum reacts with patient serum, are the most widely used technique for measuring CRP today (14,15). Accurate
C-reactive protein has been studied in a variety of surgical conditions, mostly as an indicator of postoperative complications (22–24). Preoperatively, elevated CRP levels have been reported to aid in the diagnosis of acute appendicitis. The overall sensitivity of CRP in the studies we reviewed ranges from 40 – 87%, with a specificity of 53– 82% (25–29). Early in the course of appendicitis, the white blood cell count has shown the best diagnostic sensitivity among laboratory tests (30,31). After 12–24 h of symptoms, CRP is deemed by some to be useful, especially when serial levels show an increase (27,32–34). Consistent data are lacking, and many studies report elevated CRP levels in patients with normal appendixes, as well
C-Reactive Protein
as normal CRP levels in patients with gangrenous appendixes (33–37). In a study of 56 patients, Albu found that 26 of 26 patients (100%) with a CRP value greater than 25 mg/L after 12 h of symptoms had histologically diagnosed appendicitis (38). In another study, the combination of WBC ⬎ 10 (⫻1012/L), or CRP ⬎ 12 mg/L, or neutrophil count ⬎ 70% was 98% sensitive in acute appendicitis. The relative contribution of each of these three tests to the overall diagnostic accuracy is not addressed (39). Using the combination of elevated WBC or band neutrophils or CRP, Marchand reports a 100% sensitivity for acute appendicitis. In this study, CRP ⬎ 12 mg/L alone was 70% sensitive with a neutrophil count ⬎ 7.88 (⫻109/L) showing the highest individual sensitivity, 84% (31). Albu’s study found a negative predictive value of 100% and concluded that all patients with CRP less than 25 mg/L after 12 h of symptoms will have spontaneous resolution (38). In three studies, a normal CRP along with a normal WBC count and normal neutrophil count excluded acute appendicitis with a negative predictive value approaching 100% (31,36,39). It should be noted that these studies had small sample sizes, and employed different reference ranges and cutoff levels for an elevated CRP value. It is also unclear from these studies if patients without appendicitis were clinically improved, making CRP testing redundant. In a meta-analysis of the accuracy of CRP in diagnosing appendicitis, Hallan found a wide range of data for sensitivity (40 –99%) and specificity (27–90%). He cites differing patient populations and variations in cutoff values as sources of this disparate data. He concludes that “CRP is a test of medium diagnostic accuracy and is a little inferior to the total leukocyte count.” We agree with Hallan that based on the available literature, it is impossible to draw firm conclusions regarding the clinical usefulness of CRP in acute appendicitis (40). Among the myriad causes of acute abdominal pain, CRP cannot accurately identify acute appendicitis, and should not be used to make management decisions in equivocal cases. There is no substitute for serial clinical examinations during an observation period for abdominal pain. In most EDs with access to it, spiral computed tomography (CT) scan with or without rectal contrast will provide a quicker and more definitive diagnosis of appendicitis (41). However, in centers that have access to CRP testing and not to a spiral CT scan, there might be a role for CRP.
CHOLECYSTITIS Juvonen found that a CRP level ⬎ 30 mg/L had a sensitivity of 78% in 108 patients with histologically
1021
diagnosed cholecystitis. In the same study, ultrasound had a sensitivity of 79%. When positive ultrasound findings were combined with a CRP ⬎ 30 mg/L, sensitivity was reported as 97% (42). C-reactive protein was found to be a more sensitive marker than ESR, white blood cell count, or fever in identifying acute cholecystitis. The study included 18 patients with elevated CRPs or ultrasound evidence suggestive of cholecystitis and no histologic evidence of disease. It also used peak CRP levels to calculate sensitivities, with peak levels typically occurring 2–5 days after hospital admission. The study is the only one in the literature to address CRP in cholecystitis and provides no compelling evidence that CRP would be of use in the ED. Given the lack of data, routine CRP testing for cholecystitis cannot be advocated, especially since serial examinations and sonography provide sufficient diagnostic discrimination for this entity.
PANCREATITIS C-reactive protein has been used to diagnose and predict the severity of pancreatitis (43,44). Wilson found that peak CRP levels ⱖ 210 mg/L discriminated between patients with clinically mild and severe pancreatitis with a sensitivity of 83% and a specificity of 85% (45). Heath obtained similar results, with peak concentrations ⬎ 150 mg/L predicting severe attacks (46). In these studies, pancreatitis was diagnosed by serum amylase and was graded on severity based on clinical impression. While the cutoff point of serum CRP to differentiate mild from severe forms of acute pancreatitis is debated, it is generally agreed upon that the initial value on admission is an insensitive marker of disease severity, implying limited use in the ED (45,47).
PELVIC INFLAMMATORY DISEASE C-reactive protein has been studied along with ESR in the diagnosis and prediction of severity of pelvic inflammatory disease (PID). Lehtinen concludes that both CRP and ESR should be routinely used to augment the clinical diagnosis of PID. However, in this study, no statistical significance was found between CRP levels in women presenting with endometritis only and no PID at all. The overall sensitivity and specificity of CRP in the diagnosis of PID was 74% and 67%, respectively, using a cutoff level of 20 mg/L. No significant difference was found when compared with the ESR in the diagnosis of PID (48). Soper advocates the routine use of both the ESR and CRP to increase the overall specificity of the diagnosis of PID but offers no cutoff CRP value to indicate likelihood
1022
B. Clyne and J. S. Olshaker
of infection (49). Miettinen found that an ESR ⱖ 40 mm/h combined with a CRP ⱖ 60 mg/L had a sensitivity and specificity of 97% and 61%, respectively, in detecting severe PID diagnosed by laparoscopy (50). In the studies reviewed, there are significant numbers of patients with other gynecologic problems causing raised CRP. Causes of false-positives include ectopic pregnancy, spontaneous abortions, and uterine fibroids. Based on the data, we conclude that higher CRP levels correspond to more severe cases of PID and that persistently elevated levels imply treatment failure or complications. CRP is potentially useful for monitoring response to antibiotic therapy in patients who are not clinically improving. In the ED, there is no convincing data that CRP should influence patient disposition or antibiotic regimen in suspected PID.
PNEUMONIA In the diagnosis of pneumonia, CRP is not useful for discriminating between bacterial and viral infections (51,52). However, it has been found useful as a marker for antibiotic treatment failure or the development of infectious complications (53,54). This has been shown in both the non-HIV and HIV populations (55). Although higher CRP levels usually correspond to bacterial pneumonia, especially due to S. pneumoniae, the ED management should be based on traditional parameters and on the overall clinical assessment (56). There are no data to suggest that CRP is a better prognostic indicator than fever, hypoxia, respiratory effort, or radiography in patients with pneumonia. In outpatients treated with antibiotics and those diagnosed with viral pneumonia who present to the ED with persistent symptoms, an elevated CRP level greater than 100 mg/L likely indicates treatment failure.
URINARY TRACT INFECTION Most data on CRP in urinary tract infection come from the pediatric literature. In all patients with clinically obvious pyelonephritis (14 of 14), Jodal found CRP levels ⬎ 25 mg/L, while all patients with uncomplicated infection (10 of 10) had CRP levels ⬍ 25 mg/L (21). C-reactive protein is not a useful marker to help distinguish between simple and complicated urinary tract infection in patients without clinical signs of acute pyelonephritis (21,57). False positives may occur in acute cystitis with extensive bladder wall irritation. Higher CRP levels, in general, are associated with antibiotic failures and with pyelonephritis, especially in infants (58,59).
MENINGITIS Serum CRP has been reported to accurately diagnose bacterial meningitis and to correspond with resolution of symptoms after successful antibiotic treatment (60 – 62). Recent studies have focused on cerebrospinal fluid (CSF) CRP to distinguish bacterial from viral meningitis (60,63–72). Sensitivities of CSF CRP in diagnosing bacterial meningitis have been reported as high as 97–100% on initial lumbar puncture (63,64,66). In some cases, CSF CRP was a more sensitive indicator than more traditional parameters like CSF neutrophil count and Gram’s stain (63– 66). Corall reported a sensitivity of 100% and a specificity of 94% for detecting culturepositive bacterial meningitis by measuring CSF CRP on initial lumbar puncture. This study, however, was limited to patients with CSF pleocytosis ⬎ 10 WBC/mm3 (65). Benjamin found that CSF CRP had a sensitivity of only 66% and concluded that it is too insensitive to be useful, while the serum CRP is too nonspecific for routine application (71). Donald found considerable overlap in the CSF CRP levels among patients with bacterial and viral meningitis, limiting clinical usefulness (73). In a study of neonates, Philip found that CSF CRP did not distinguish between infants with meningitis and those with no infection (74). He argues against the routine use of CSF CRP measurements in suspected neonatal sepsis and meningitis. Using a semiquantitative method, Komorowski found that even in the setting of pleocytosis with CSF leukocyte count ⬎ 10/L, a CSF CRP ⬎ 8 mg/L was still only 35% sensitive in identifying bacterial meningitis (70). In studies of serum and CSF CRP, there are significant numbers of false positives and false negatives (67,68). Reasons for false positives include viral meningitis, brain abscess, other occult sources of inflammation, and the use of semiquantitative methods of measuring CRP. False negatives are most likely due to early bacterial infection, poor diffusion of CRP into the CSF, group B streptococcal infections, or a diminished inflammatory response in neonates (72). We conclude that the discrimination between viral and bacterial meningitis cannot be made with any certainty by serum or CSF CRP, nor should these tests influence patient management in the ED. Any clinical suspicion of meningitis, even in patients with equivocal CSF findings, should prompt early broad spectrum antibiotic coverage.
NEONATAL SEPSIS AND PEDIATRIC INFECTIONS In the neonatal period, CRP has been studied extensively for its ability to detect early infection (75– 80). Berger
C-Reactive Protein
1023
found that after the first 3 days of life, CRP is the single best test in early detection of neonatal septicemia, and that serial measurements parallel the course of infection and the efficacy of antibiotic treatment (75). Forest found that serial CRP values lead to increased accuracy of detecting bacterial infection in neonates. In this study, all neonates with proven bacterial infection had a CRP level above 80 mg/L on at least one occasion during serial measurements 12 h apart (76). Similarly, Mathers states that the main role of CRP lies in the exclusion or confirmation of infection 24 h after the initial clinical concern. He concludes, like others, that CRP has no role in the early diagnosis of neonatal sepsis and does not support the withholding of antibiotics based on a normal value (77). In a thorough review, DaSilva concludes that CRP is the best single diagnostic test in the evaluation of neonates with suspected sepsis. However, the report does not address the length of time required before CRP reaches optimal sensitivity. He emphasizes that clinicians cannot base management decisions on the available literature because of wide variability in measurement techniques and study designs (78). Bacteremia occurs in 3– 4% of febrile children evaluated in an ED setting, frequently without an obvious source of infection (20). In general, higher serum CRP levels are found on admission in patients who are bacteremic (81– 83). However, one study reported 8 bacteremic patients with initially normal CRP measurements and another found no significant difference in CRP values among patients with or without bacteremia (79,84). The frequency of false negatives in neonatal sepsis and bacteremia render CRP unhelpful in the ED setting.
SUMMARY We regularly encounter equivocal presentations of disease in the ED. In these situations, we often turn to the clinical laboratory to help confirm or disprove our suspicions. The measurement of C-reactive protein has enjoyed periodic emphasis over the years as a measure of general illness and as an adjunct to physical examination and other laboratory data. Unfortunately, the nonspecific nature of the acute phase response prevents CRP from being a useful discriminatory diagnostic test. Based on the literature, we conclude that CRP has a limited role, if any, in the diagnosis and management of commonly seen ED problems. Faster and more interpretable tests are available to help diagnose equivocal cases of appendicitis, cholecystitis, and PID. C-reactive protein adds little to the diagnosis of pneumonia, urinary tract infections, and pancreatitis. False negative CRP results in the setting of meningitis, neonatal sepsis, and bacteremia are com-
mon and therefore unhelpful. A single CRP value should not factor into the decision to treat these patients.
REFERENCES 1. Tillett WS, Francis T. Serological reactions in pneumonia with a non-protein somatic fraction of the pneumococcus. J Exp Med 1930;52:561–71. 2. Abernathy TJ, Avery OT. The occurrence during acute infection of a protein not normally present in the blood. I. Distribution of the reactive protein in patients’ sera and the effects of calcium on the flocculation reaction with C polysaccharide of pneumococcus. J Exp Med 1941;73:173. 3. MacLeod CM, Avery OT. The occurrence during acute infections of a protein not normally present in blood. II. Isolation and properties of the C-reactive protein. J Exp Med 1941;73:183. 4. Kolb-Bachofen V. A review on the biological properties of C-reactive protein. Immunobiology 1991;183:133– 45. 5. Gewurz H, Mold C, Siegal J, Fiedel B. C-reactive protein and the acute phase response. Adv Int Med 1982;27:345–72. 6. Ballou SP, Kushner I. C-reactive protein and the acute phase response. Adv Intern Med 1992;37:313–36. 7. Pepys MB. C-reactive protein fifty years on. Lancet 1981;1:653–7. 8. Palosuo T, Husman T, Koistinen J, Aho K. C-reactive protein in population samples. Acta Med Scand 1986;220:175–9. 9. Harmoinen A, Hallstrom O, Gronroos P. Rapid quantitative determination of C-reactive protein using laser nephelometer. Scand J Clin Lab Invest 1980;40:293. 10. Jupe D. The acute phase response and laboratory testing. Aust Fam Phys 1996;25:324 –9. 11. Young B, Gleeson M, Cripps, AW. C-reactive protein: a critical review. Pathology 1991;23:118 –24 12. Jaye DL, Waites KB. Clinical applications of C-reactive protein in pediatrics. Pediatr Infect Dis J 1997;16:735– 47. 13. International Committee for Standardization in Haematology. Guidelines on selection of laboratory tests for monitoring the acute phase response. J Clin Pathol 1988;41:1203–12. 14. Deodar S. C-reactive protein: the best laboratory indicator available for monitoring disease activity. Cleve Clin J Med 1989;56: 126 –30. 15. Powell LJ. C-reactive protein-a review. Am J Med Tech 1979;45: 138 – 48. 16. Stuart J, Whicher JT. Tests for detecting and monitoring the acute phase response. Arch Dis Child 1988;63:115–7. 17. Morley JJ, Kushner I. Serum C-reactive protein levels in disease. Ann NY Acad Sci 1982;389:406 –18. 18. Nudelman R, Kagan BM. C-reactive protein in pediatrics. Adv Pediatr 1983;30:517– 47. 19. Okamura JM, Miyagi JM, Terada K, Hokama Y. Potential clinical applications of C-reactive protein. J Clin Lab Anal 1990;4:231–5. 20. Singer JI, Buchino JJ, Chabali R. Selected laboratory in pediatric emergency care. Emerg Med Clin North Am 1986;4:377–96. 21. Jodal U, Lindberg U, Lincoln K. Level diagnosis of symptomatic urinary tract infections in childhood. Acta Paediatr Scand 1975; 64:201– 8. 22. Mustard RA, Bohnen JMA, Haseeb S, Kasina R. C-reactive protein levels predict postoperative septic complications. Arch Surg 1987; 122:69 –73. 23. Orriss DE. Serial serum C-reactive protein levels as an indicator of infection in cardiac transplant patients. Med Lab Sci 1988;45:116 – 20. 24. Boralessa H, DeBeer FC, Manchie A, Whitman JG, Pepys MB. C-reactive protein in patients undergoing cardiac surgery. Anaesthesia 1986;41:11–5. 25. Chi CH, Shiesh SC, Chen KW, Wu MH, Lin XZ. C-reactive protein for the evaluation of acute abdominal pain. Am J Emerg Med 1996;14:254 – 6. 26. Thompson MM, Underwood MJ, Dookeran KA, Lloyd DM, Bell
1024
27.
28.
29.
30.
31.
32. 33.
34.
35.
36. 37.
38. 39.
40.
41.
42.
43. 44.
45.
46.
47.
48.
49. 50.
PRF. Role of sequential leucocyte counts and C-reactive protein measurements in acute appendicitis. Br J Surg 1992;79:822– 4. Oosterhuis WP, Zwinderman AH, Teeuwen M, et al. C-reactive protein in the diagnosis of acute appendicitis. Eur J Surg 1993; 159:115–9. Van Lente F, Marchand A, Galen RS. Evaluation of the nephelometric assay of CRP using the Beckman ICS. Clin Chem 1981; 27:1075. Al-Saigh AH. C-reactive protein in the differential diagnosis of the acute abdomen, especially acute appendicitis. J R Coll Surg Edinb 1992;37:238 – 40. Gronroos JM, Forsstrom JJ, Irjala K, et al. Phospholipase A2, C-reactive protein, and white blood cell count in the diagnosis of acute appendicitis. Clin Chem 1994;40:1757– 60. Marchand A, Van Lente F, Galen R. The assessment of laboratory tests in the diagnosis of acute appendicitis. Am J Clin Pathol 1983;80:369 –74. Chen SC, Wang SM. C-reactive protein in the diagnosis of acute appendicitis. Am J Emerg Med 1996;14:101–3. Thimsen DA, Tong GK, Gruenberg, JC. Prospective evaluation of C-reactive protein in patients suspected to have acute appendicitis. Am Surg 1989;55:466 – 8. Eriksson S, Granstrom L, Olander B, et al. Sensitivity of interleukin-6 (IL-6) and C-reactive protein concentrations in the diagnosis of acute appendicitis. Eur J Surg 1995;161:41–5. Peltola H, Ahlqvist J, Rapola J, et al. C-reactive protein compared with white blood cell count and erythrocyte sedimentation rate in the diagnosis of acute appendicitis in children. Acta Chir Scand 1986;152:55– 8. Dueholm S, Bagi P, Bud M. Laboratory aid in the diagnosis of acute appendicitis. Dis Colon Rectum 1989;32:855–9. Mikaelsson C, Arnbjornsson E. The value of C-reactive protein determinations in patients with suspected acute appendicitis. Ann Chir Gyn 1984;73:281– 4. Albu E, Miller BM, Choi Y, et al. Diagnostic value of C-reactive protein in acute appendicitis. Dis Colon Rectum 1994;37:49 –51. Van Dieijen-Visser MP, Go PMNYH, Brombacher PJ. The value of laboratory tests in patients suspected of acute appendicitis. Eur J Clin Chem Clin Biochem 1991;29:749 –52. Hallan S, Asberg A. The accuracy of C-reactive protein in diagnosing acute appendicitis—a meta-analysis. Scand J Clin Lab Invest 1997;57:373– 80. Rao PM, Rhea JT, Novelline RA, et al. Helical CT technique for the diagnosis of appendicitis: prospective evaluation of a focused appendix CT examination. Radiology 1997;202:139 – 44. Juvonen T, Kiviniemi H, Niemela O, et al. Diagnostic accuracy of ultrasonography and C-reactive protein concentration in acute cholecystitis: a prospective clinical study. Eur J Surg 1992;158:365–9. Mazlam MZ, Hodgson HJF. Why measure C-reactive protein? Gut 1994;35:5–7. Sternby B, O’Brien JF, Zinsmeister AR, et al. What is the best biochemical test to diagnose acute pancreatitis? A prospective clinical study. Mayo Clin Proc 1996;71:1138 – 44. Wilson C, Heads A, Shenkin A, et al. C-reactive protein, antiproteases and complement factors as objective markers of severity in acute pancreatitis. Br J Surg 1989;76:177– 81. Heath DI, Cruickshank A, Gudgeon M, et al. Role of IL-6 in mediating the acute phase protein response and potential as early means of severity assessment in acute pancreatitis. Gut 1993;34: 41–5. Pezilli R, Billi P, Cappelletti O, Barakat B. Serum C-reactive protein in acute biliary pancreatitis. Is it a reliable marker for the early assessment of severity of the disease? Ital J Gastro Hepatol 1997;29:554 –7. Lehtinen M, Laine S, Heinonen PK, et al. Serum C-reactive protein determination in acute pelvic inflammatory disease. Am J Obstet Gynecol 1986;154:158 –9. Soper DE. Diagnosis and laparoscopic grading of acute salpingitis. Am J Obstet Gynecol 1991;164:1370 – 6. Miettinen AK, Heinonen PK, Laippala P. Test performance of erythrocyte sedimentation rate and C-reactive protein in assessing
B. Clyne and J. S. Olshaker
51.
52.
53. 54.
55.
56.
57.
58. 59.
60. 61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
the severity of acute pelvic inflammatory disease. Am J Obstet Gynecol 1993;169:1143–9. Korppi M, Heiskanen-Kosma T, Leinonen M. White blood cells, C-reactive protein and erythrocyte sedimentation rate in pneumococcal pneumonia in children. Eur Respir J 1997;10:1125–9. McCarthy PL, Frank AL, Ablow RC, et al. Value of the C-reactive protein in the differentiation of bacterial and viral pneumonia. J Paediatr 1978;92:454 – 6. Smith RP, Lipworth BJ. C-reactive protein in simple communityacquired pneumonia. Chest 1995;107:1028 –31. Smith RP, Lipworth BJ, Cree IA, et al. C-reactive protein, a clinical marker in community-acquired pneumonia. Chest 1995; 108:1288 –91. Storgaard M, Laursen AL, Andersen PL. The C-reactive protein responses in HIV infected patients with pneumonia. Scand J Infect Dis 1993;25:305–9. Ortqvist A, Hedlund J, Wretlind B, Carlstrom A, Kalin M. Diagnostic and prognostic value of IL-6 and C-reactive protein in community-acquired pneumonia. Scand J Infect Dis 1995;27:457– 62. Hellerstein S, Duggan E, Welchert E, Mansour F. Serum C-reactive protein and the site of urinary tract infections. J Pediatr 1982;100:21–5. Hanson LA, Jodal U, Sabel K, Wadsworth C. The diagnostic value of C-reactive protein. Ped Infect Dis 1983;2:87–90. Saxstad J, Nilsson LA, Hanson LA. C-reactive protein in serum from infants as determined with immunodiffusion techniques. Infants with various infections. Acta Paediatr Scand 1970;59:676 – 80. Peltola HO. C-reactive protein for rapid monitoring of infections of the central nervous system. Lancet 1982;1:980 –3. Lembo RM, Marchant CD. Acute phase reactants and risk of bacterial meningitis among febrile infants and children. Ann Emerg Med 1991;20:36 – 44. Clarke D, Cost K. Use of serum C-reactive protein in differentiating septic from aseptic meningitis in children. J Pediatr 1983; 102:718 –20. Gershom EB, Briggeman-Mol GJJ, de Zegher F. Cerebrospinal fluid C-reactive protein in meningitis: diagnostic value and pathophysiology. Eur J Pediatr 1986;145:246 –9. MacFarlane DE, Narla VR. Cerebrospinal fluid C-reactive protein in the laboratory diagnosis of bacterial meningitis. Acta Paediatr Scand 1985;74:560 –3. Corrall CJ, Pepple JM, Moxon ER, Hughes WT. C-reactive protein in spinal fluid of children with meningitis. J Pediatr 1981;1981: 365–9. Abramson JS, Hampton KD, Babu S, et al. The use of C-reactive protein from cerebrospinal fluid for differentiating meningitis from other central nervous system diseases. J Infect Dis 1985;151: 854 – 8. Hansson LO, Axelsson G, Linne T, Aurelius E, Lindquist L. Serum C-reactive protein in the differential diagnosis of acute meningitis. Scand J Infect Dis 1993;25:625–30. Sabal KG, Hanson LA. The clinical usefulness of C-reactive protein determination in bacterial meningitis and septicemia in infancy. Acta Paediatr Scand 1974;63:381– 8. Stearman M, Southgate HJ. The use of cytokine and C-reactive protein measurements in cerebrospinal fluid during acute infective meningitis. Ann Clin Biochem 1994;31:255– 61. Komorowski RA, Farmer SG, Knox KK. Comparison of cerebrospinal fluid C-reactive protein and lactate for diagnosis of meningitis. J Clin Micro 1986;24:982–5. Benjamin DR, Opheim KE, Brewer L. Is C-reactive protein useful in the management of children with suspected bacterial meningitis? Am J Clin Pathol 1984;81:779 – 82. Gray BM, Simmons DR, Mason H, Barnum S, Volanakis JE. Quantitative levels of C-reactive protein in cerebrospinal fluid in patients with bacterial meningitis and other conditions. J Pediatr 1986;108:665–70. Donald PR, Strachan AF, Schoeman JF, DeBeer FC. Cerebrospinal
C-Reactive Protein
74. 75. 76. 77. 78.
fluid C-reactive protein in infective meningitis in childhood. J Lab Clin Med 1985;106:424 –7. Philip AGS, Baker CJ. Cerebrospinal fluid C-reactive protein in neonatal meningitis. J Pediatr 1983;102:715–7. Berger C, Uehlinger J, Ghelfi D, Blau N, Fanconi S. Comparison of C-reactive protein and white blood cell count with differential in neonates at risk for septicaemia. Eur J Pediatr 1995;154:138 – 44. Forest J, Lariviere F, Dolce P, Masson M, Nadeau L. C-reactive protein as biochemical indicator of bacterial infection in neonates. Clin Biochem 1986;19:192– 4. Mathers NJ, Pohlandt F. Diagnostic audit of C-reactive protein in neonatal infection. Eur J Paediatrics 1987;146:147–51. DaSilva O, Ohlsson A, Kenyon C. Accuracy of leukocyte indices and C-reactive protein for diagnosis of neonatal sepsis: a critical review. Pediatr Infect Dis J 1995;14:362– 6.
1025 79. Peltola H, Jaakkola M. C-reactive protein in early detection of bacteremic versus viral infections in immunocompetent and compromised children. J Pediatr 1988;113:641– 6. 80. Pourcyrous M, Bada HS, Korones SB, Baselski V, Wong SP. Significance of serial C-reactive protein responses in neonatal infection and other disorders. Pediatrics 1993;92:431–5. 81. Bennish M, Vardiman J, Beem MO. The zeta sedimentation ratio in children. J Pediatr 1984:104:249. 82. McCarthy PL, Jekel JK, Dolan TF. Comparison of acute phase reactants in pediatric patients with fever. Pediatrics 1978;62:716. 83. Kohli V, Singhi S, Sharma P, Ganguly NK. Value of serum C-reactive protein concentrations in febrile children without apparent focus. Ann Trop Pediatr 1993;13:373– 8. 84. McCabe RE, Remington JA. C-reactive protein in patients with bacteremia. J Clin Microbiol 1984;20:317–91.