• 
• 

Three Part Question

In [a febrile child with acute otitis media] is [measurement of C - reactive protein] a good screening test [to differentiate between bacterial and viral infection ]?

Clinical Scenario

You are a senior house officer in a paediatric assessment unit and commonly see children with acute otitis media who are febrile but otherwise well. You are not sure about the role of antibiotics in this condition and note that various existing guidelines have different recommendations about use of antibiotics. You wonder if measurement of C-reactive protein can be used as a screening test to differentiate between viral and bacterial otitis media and to decide which group of children require antibiotic therapy. You decide to find out more.

Search Strategy

Primary Sources:
Medline (1951 to date) using interface Dialog DataStar
Secondary Sources:
Bestbets website and Cochrane database:
Keywords: "C-reactive protein" or "Acute phase reactants" or "Acute phase protein" or "CRP" AND "otitis media" or "upper respiratory tract infection" or "urti".

Search Outcome

A total of 148 articles were found. This was limited to 82 articles by selecting those in English language and human studies related to children (up to 18 years of age). Each abstract was read and 6 relevant studies were found. Two of those studies were relevant but were designed to look for acute phase reactants in acute bacterial otitis media only. Both of them were subsequently excluded from analysis.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Tejani N R et al. 1995 USA	185 children with acute otitis media aged 3 months to 7 years of age.Tympanocentasis, bacterial and viral studies of the middle ear fluid, virologic studies of nasal wash specimens, measurement of serum antibody titres to respiratory viruses, full blood counts and quantitation of serum C-RP concentration were performed on all of them.	Case- Series (level 4)	Serum C-RP concentrations were compared among subjects with AOM who were divided into four groups. I-bacterial infection, II- Bacterial and viral infection, III- viral infection, IV- no identifiable pathogen.	The sensitivity of C-RP value higher than 2 mg/dl in the diagnosis of bacterial AOM was 22%, whereas the specificity was 94%; The positive predictive value was 94.2%; and the negative predictive value was 21%.	Small number of children with viral acute otitis media.
Karma P et al., 1987 Finland	79 children with acute otitis media aged 4 months to 5 years	Case series (level 4)	C-RP values in children with S. pneumoniae or H. influanzae growth in MEF and in children with negative MEF culture or growth of minor otitis pathogen (S. aureus, CONS, diptheroids or S. pyogens	27 % children with S. pneumoniae or H. influenzae growth in MEF had C-RP < 10 mg/dl. 25 of the 41 children with major otits pathogen and 9 of the 38 children without major otitis pathogen had C-RP value of more than 20 mg/dl. (P<0.01)	Viral cultures were not done in children with negative MEF culture Small sample size.Children who were receiving antibiotics for otitis media andChildren with other simultaneous infection e.g. pneumonia. tonsillitis were also included in analysis.
Komoroski E M et al., 1987 USA	199 children (aged 1-24 months) with acute otiis media diagnosed by pneumatic otoscopy and confirmed by tympanocentesis. Rectal temperature, WBC with differential, blood culture and CRP were performed on all of them.	Case- control study (level 3b)	C-RP, WBC and absolute neutrophil count in children with pure growth of a bacterial pathogen and children with sterile middle ear effusion	C-RP was elevated (>1.0 mg/dl) in 34% children, 49% children with S. pneumoniae or H. influenzae growth from MEF had C-RP less than 1.0 mg/dl. C-RP values were lower in patients with sterile effusion compared with S. pneumoniae (P< 0.001, two tailed) or H. influenzae (P=0.04, two tailed). There was no statistically significant difference in C-RP values between patients with B. catarrhalis and those with sterile effusions.	Viral cultures were not done in children with sterile middle ear effusion. Only children between 1-24 months of age were included in the study.
Principi N et al., 1986 USA	Study group: 67 children with acute otitis mediaControl group: 67 matched controls affected by non-infectious neurological disorder	Case- control study (level 3 b)	C-RP values in children with bacterial growth from middle ear fluid and with sterile middle ear fluid. C-RP value in children affected by noninfectious neurological disorders.	The upper limit of C-RP in controls was 15. 71% of children with bacterial AOM and 67% of AOM children with sterile middle ear fluid had C-RP value greater than 15. C-RP greater than 15 mg/lit. showed sensitivity of 72%, specificity of 33%, predictive value of a positive test of 66% and predictive value of negative test of 40% in detecting Bacterial AOM.	Viral cultures were not done in children with sterile middle ear fluid.

Comment(s)

Acute otitis media is a common paediatric condition, which is caused by both bacterias and viruses. However bacterial and viral otitis media can not be differentiated on clinical grounds alone. The necessity of antibiotic therapy in AOM has been a subject of debate and existing guidelines (like SIGN or AAP guidelines) differ in their recommendations about use of antibiotics. It is desirable to determine the cause of AOM (bacterial or viral) early in the course of the illness to decide which group of children requires antibiotic therapy. C-reactive protein (C-RP), a prototypic acute phase reactant is one of the common parameter used to differentiate between invasive bacterial and viral infection. I found four studies looking at the use C-reactive protein in differentiating between bacterial and viral AOM (see table). Study by Tejani and colleagues (1) was methodologically sound and virologic studies of middle ear fluid and nasal wash specimens along with measurement of serum antibody titre to common respiratory viruses were done. However, in this study number of children with viral AOM was small. Study by Karma and colleagues (2), Komoroski and colleagues (3) and Principi N and colleagues (3) had weaknesses in their methodology and in none of them viral studies of the middle ear fluid to look for viral pathogens, were done. One of them (2) included children who were receiving antibiotics for otitis media and children with other simultaneous infection e.g. pneumonia, tonsillitis in the study as well, a selection bias which can alter the results. All the studies used different upper limit of C-RP value as normal (Tejani NR <2 mg/dl., Karma P <10 mg/dl., Principi N <15 mg/dl., Komoroski E M <1 mg/dl.). Two of these studies (1, 2) found sensitivity of 22% and 72% respectively, of high CRP value in detecting bacterial otitis media. Third study (3) found that 27% of children with S. pnumoniae or H. influenzae had C-RP less than10 mg/dl whereas fourth study found no statistical difference in C-RP values in children with B. catarrahalis growth from middle ear fluid and sterile middle ear effusion. Despite their weaknesses all studies concluded that single C-RP value should not be used to differentiate between acute bacterial and viral otitis media. Two other studies (5,6)(not appraised in this article as these were designed to look at the C-RP values in bacterial otitis media only) found that significant proportion of children with bacterial otitis media can have normal C-RP values. An observation, which indirectly supports the finding of four studies, appraised here. At this stage, on the basis of published evidence, it can be concluded that a single C-RP value in children with AOM should not be used to differentiate between viral and bacterial infection.

Clinical Bottom Line

In a child with acute otitis media, measurement of single C-reactive protein value is not useful in differentiating between bacterial and viral infection.

References

1. Tejani NR, Chonmaitree T, Rassin DK, Howie VM, Owen MJ, Golgman AS, Use of C-reactive protein in differentiation between acute bacterial and viral otitis media. Pediatrics 1995 May; 95: 664-669
2. Karma P, Sipila Mkoskela M Peltola H. C-reactive protein in acute otitis media Acta otolaryngol (Stockh.). 1987; 103:395-399
3. Komoroski EM, VanhareG, Shurin PA,et al. Quantitavi measurement of C-reactive protein in. acute otitis media J Pediatr 1987; 111:81-84
4. Principi N, Marchisio P, Bigalli L, Massironi E, C-reactive protein in acute otitis media, , Pediatr Infect Dis J J 1986; 525-527