Family Practice

Family Practice Advance Access originally published online on December 12, 2008
Family Practice 2009 26(1):10-21; doi:10.1093/fampra/cmn095

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 26/1/10    most recent cmn095v1 E-letters: Submit a response Alert me when this article is cited Alert me when E-letters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Google Scholar Articles by Falk, G. Articles by Fahey, T. Search for Related Content PubMed PubMed Citation Articles by Falk, G. Articles by Fahey, T. Social Bookmarking          What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.

C-reactive protein and community-acquired pneumonia in ambulatory care: systematic review of diagnostic accuracy studies

Gavin Falk and Tom Fahey

Division of Population Health Sciences, Department of General Practice, RCSI Medical School, Beaux Lane House, Mercer Street, Dublin 2, Ireland

Correspondence to Tom Fahey, Division of Population Health Sciences, Department of General Practice, RCSI Medical School, Beaux Lane House, Mercer Street, Dublin 2, Ireland; Email: tomfahey{at}rcsi.ie

Received 8 May 2008; Revised 10 October 2008; Accepted 11 November 2008.

	Abstract

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
Background. There is uncertainty regarding the diagnostic value of C-reactive protein (CRP) in patients presenting with symptoms suggestive of community-acquired pneumonia (CAP) in community or ambulatory settings.

Objective. We assessed the diagnostic value of CRP in primary care and accident and emergency departments in terms of ruling in or ruling out CAP.

Methods. Diagnostic accuracy systematic review, we searched PubMed from January 1966 to September 2008 and EMBASE from January 1980 to September 2008 using a diagnostic accuracy search filter. We included cross-sectional or cohort studies that assess the diagnostic utility of CRP at different cut-points against a reference standard of chest X-ray. We calculated pooled positive and negative likelihood ratios (LRs) and assessed heterogeneity using the I² index.

Results. Eight studies incorporating 2194 patients were included. The median prevalence of CAP was 14.6% (range 5%–89%). At a CRP cut-point of 20 mg/l, the pooled positive LR+ was 2.1 [95% confidence interval (CI) 1.8–2.4] and the pooled negative LR– was 0.33 (95% CI 0.25–0.43). At the two other CRP cut-points (50, >100 mg/l), the results were heterogeneous, so the pooled results should be interpreted with caution.

Conclusions. CRP may be of value in ruling out a diagnosis of CAP in situations where the probability of CAP >10%, typically accident and emergency departments. In primary care, additional diagnostic testing with CRP is unlikely to alter the probability of CAP sufficiently to change subsequent management decisions such as antibiotic prescribing or referral to hospital.

Keywords. Ambulatory care, C-reactive protein, diagnosis, pneumonia.

	Introduction

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
GPs regularly assess patients presenting with symptoms suggestive of lower respiratory tract infection (LRTI), which can include acute bronchitis and pneumonia.¹ The dilemma of differentiating between these diagnoses, using history and physical examination, is challenging.² Recent studies have shown that relying on clinical findings alone do not predict persistence of cough duration or functional impairment in terms of the subsequent clinical course of illness in individual patients.³

C-reactive protein (CRP) is an acute-phase protein, which is synthesized by the liver in response to infection or tissue inflammation. A range of cytokines, including interleukin-6, interleukin-1β and tumour necrosis factor, stimulate its production.⁴ CRP was first identified in 1930 from the observation of patients with pneumonia. Its name was derived from the fact that it reacted with the pneumococcal C-polysaccharide in patients’ plasma during the acute phase of pneumococcal pneumonia. Subsequently, it has been recognized as a marker of inflammation, a so-called ‘acute-phase protein’. An acute-phase protein has been described as a protein whose plasma concentration increases or decreases by at least 25% during an inflammatory process.⁵ Serum concentration of CRP is normally <3 mg/l; but following severe infection or inflammation, it can rise above 500 mg/l.

CRP near-patient tests are now widely available and can accurately detect CRP levels >8 mg/l.⁶,⁷ A normal CRP in a patient with symptoms of respiratory infection most likely indicates a self-limiting infection that does not require referral to hospital or antibiotic treatment. However, it is important that physicians are aware that a normal value may be observed early in an illness, before the ‘acute-phase’ response has been mounted.⁸

The ability to differentiate pneumonia from other LRTIs, such as acute bronchitis, is important for several reasons. First, missing a diagnosis of pneumonia may have fatal consequences for a patient.⁹,¹⁰ Second, inappropriate prescribing of antibiotics to patients with acute bronchitis increases the risk of side effects—more commonly rash and gastrointestinal upset and less commonly anaphylactic reactions. This practice can thereby ‘medicalize’ a self-limiting illness and contribute to the public health problem of antibiotic-resistant organisms in the community.¹¹,¹²

Uptake of CRP testing in primary care has been increasing throughout Europe, for instance a CRP assay is now widely used in the Scandinavian countries and in Switzerland.¹³ In Norway, a survey carried out in 2002 revealed that 41% of all patients consulting with an airway infection had a CRP test. The purpose of this systematic review is to evaluate CRP as a near-patient test in the diagnosis of community-acquired pneumonia (CAP) in the community.

	Methods

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
Search strategy
An electronic search was performed following the guidelines of Haynes et al.¹⁴ PubMed was searched from January 1966 to September 2008 and EMBASE from January 1980 to July 2008. Combinations of the phrases ‘C-reactive protein’, ‘pneumonia’, ‘community-acquired pneumonia’, ‘General Practice’, ‘Family Practice’, ‘Emergency Department’ and ‘Primary care’ were used. This search was supplemented by hand-checking references of filtered papers, searching Google Scholar and searching the Cochrane database. The MEDION database at the University of Maastricht, which contains over 300 reviews published on diagnostic and screening studies, was also searched.¹⁵ Because of funding constraints allied to evidence concerning the more limited contribution of non-English language papers to systematic reviews, we limited our search to English language papers only.¹⁶

Study selection
The inclusion and exclusion criteria were as follows:

• Population—participants in each study were to be recruited from a community, primary care setting or ambulatory setting, for example emergency departments, and have symptoms suggestive of acute respiratory infection suggestive of LRTI.
  
• Study design and reference standard—the study should assess the diagnostic accuracy by means of a cross-sectional study and use chest X-ray as the reference (gold) standard for determining presence of CAP.
  
• Index test—the study must assess the use of CRP in diagnosing CAP.
  
• Outcome measures—studies must present data that will allow the construction of 2 x 2 tables for the assessment of diagnostic accuracy of CRP in diagnosing CAP.
  

Assessment of study quality
In order to assess the quality of each included study, a nine-point modified Quality Assessment of Studies of Diagnostic Accuracy (QUADAS) quality analysis score was applied.¹⁷

Data extraction and analysis
Data were extracted from individual studies and 2 x 2 tables constructed in relation to each diagnostic accuracy study of CRP. We pooled individual diagnostic accuracy studies using the statistical package Meta-DiSc¹⁸ and calculated pooled positive and negative likelihood ratios (LRs) using a random-effects model. LRs indicate by how much a given diagnostic test result will raise or lower the pretest probability of a target disorder, in this instance the presence of CAP, to a post-test probability. LRs of >1 increase the probability of CAP and <1 decrease the probability of CAP. In general terms, LRs 5 or 0.2 generate ‘moderate’ shifts, while LRs of 2 or 0.5 generate ‘small’ shifts in pretest to post-test probability.¹⁹ We visually examined the presence or absence of heterogeneity and quantified heterogeneity by means of the chi-square statistic called the ‘inconsistency index’ or ‘I² index’. This index describes the percentage of total variation across studies due to heterogeneity rather than chance.²⁰ Higgins et al. suggest ‘assigning the adjectives low, moderate and high (heterogeneity) to I² of 25%, 50% and 75%’, respectively.²⁰ In our analysis, the following approach was adopted. When I² 50%, data were taken to be sufficiently homogeneous to allow pooling, and summary estimates [and their 95% confidence interval (CI)] of diagnostic test accuracy in the form of positive or negative LRs were calculated. An I² >50% was taken to indicate significant between-study heterogeneity. In these instances, the ranges of positive and negative LRs from each of the included studies are presented.

We analysed and pooled results utilizing three different cut-points of CRP (20, 50, >100 mg/l) in order to assess the diagnostic value of CRP in ‘ruling in’ or ‘ruling out’ CAP. We applied these different cut-points across the spectrum of prior probability for CAP in order to assess the diagnostic value of a ‘positive’ or ‘negative’ (depending of the cut-point) CRP result.

	Results

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
Our search yielded eight papers that satisfied our inclusion criteria for the study.²¹–²⁷ Figure 1 illustrates the flow of studies through the review process.

View larger version (14K): [in this window] [in a new window] [Download PowerPoint slide]   FIGURE 1 Flow of studies through review process

 
Characteristics and quality of included studies
The webtable 1 (see Appendix) outlines the important characteristics of each study. Among the 2194 patients included in the eight studies, there was substantial variation in the prevalence of CAP (median 16%, range 5%–89%). The use of CRP in diagnosing CAP was assessed using cut-offs ranging from 11 to 200 mg/l. Overall quality of included studies was moderate (webtable 1) with clear selection criteria, all recruited patients receiving index and reference standard tests incorporating blind outcome assessment in six studies. However, the possibility of spectrum bias was high in some of the included studies with a wide range of prevalence of CAP being reported and variation in the inclusion criteria.

Definition of the reference standard test for CAP
All studies used chest X-ray as their reference standard but the criteria for defining CAP varied between studies. Four studies were non-specific, using terms like ‘new infiltrate’ or ‘radiologically diagnosed CAP’. Two of these four studies explicitly stated that radiologists read the radiographs. Two other studies were more specific, specifying a lateral and posterior-anterior chest X-ray and interpretation by two or more radiologists; one of whom was a senior member of staff. The remaining two studies combined an infiltrate on X-ray with symptoms suggestive of a LRTI.

Diagnostic value of CRP
The pooled LRs for different CRP cut-points are shown in Figure 2. Only at the cut-point of a CRP of 20 mg/l were the individual LRs reported from six studies sufficiently homogeneous to allow pooling, with the pooled positive LR+ being 2.1 (95% CI 1.8–2.4) and the pooled negative LR– being 0.33 (95% CI 0.25–0.43). At the two other CRP cut-points (50, >100 mg/l), the results were heterogeneous, so the pooled results should be interpreted with caution (Fig. 2).

View larger version (27K): [in this window] [in a new window] [Download PowerPoint slide]   FIGURE 2 Positive and negative LRs for CRP cut-points (1) cut-point 20 mg/l (2) cut-point 50 mg/l (3) cut-point > 100 mg/l

 

	Discussion

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
Clinical interpretation of results
Our findings suggest that CRP testing when the cut-point is set at 20 mg/l has limited use in ruling out CAP in patients presenting with symptoms of LRTI in community or ambulatory settings. The critical element to this interpretation relates to the prior probability of CAP, which we found to be highly variable in the included studies in this review, with the prevalence of CAP varying from 5% to 89%. The nomogram (Fig. 3) shows the combinations of prior and corresponding posterior probabilities calculated using the lower and upper bounds of the 95% CI around the pooled negative LR estimate when the CRP cut-point is 20 mg/l. The largest absolute difference between prior and posterior probabilities occurs when the prior for CAP is in the intermediate range (40%–80%). However, in community settings, the prevalence of CAP in a recent community-based cohort study appears to be around 6%,¹ so in situations of low prior prevalence such as this a negative test result will not substantially reduce the probability of CAP (Fig. 3). In situations where the prior probability of CAP is higher, for example in hospital accident and emergency settings, then a negative CRP test result when the cut-point is set at 20 mg/l may provide important diagnostic information in terms of ruling out CAP. These findings highlight the different diagnostic value of CRP, depending on the setting of care.

View larger version (29K): [in this window] [in a new window] [Download PowerPoint slide]   FIGURE 3 The probability nomogram for the post-test probability of CAP: higher and lower 95% CI estimates of pooled negative LR for a CRP cut-point of 20 mg/l

 
In the same way, we estimated the diagnostic value of a positive test result when the CRP cut-point is at >100 mg/l based on the range of positive LRs reported in the included studies (Fig. 2). In this instance, the diagnostic range of a positive test result is wide, with the highest reported LR+ (51.8) having a substantial impact on ruling in CAP,²¹ while the lowest reported LR+ (2.3),²⁸ having a negligible impact and not convincingly raising the posterior probability of CAP (Fig. 4). Further research is needed to assess and refine the diagnostic utility of CRP at a cut-point of >100 mg/l, in terms of ruling in a diagnosis of CAP, irrespective of the setting of care.

View larger version (31K): [in this window] [in a new window] [Download PowerPoint slide]   FIGURE 4 The probability nomogram for the post-test probability of CAP: highest and lowest estimates of reported positive LR for a CRP cut-point of >100 mg/l

 
Context of previous studies
Our pooled estimates are consistent with a previous review which concluded that CRP was neither sufficiently sensitive to ‘rule out’ nor sufficiently specific to ‘rule in’ an infiltrate on chest X-ray.²⁹ However, we qualify our interpretation of the diagnostic value of CRP in two ways. First, in relation to the prior prevalence of CAP which in turn relates to the setting of care each individual patient presents to. A negative test result at cut-point 20 mg/l is not likely to be informative in community/ambulatory setting but could be useful in accident and emergency settings where the initial prior probability of CAP is in the intermediate range. Second, a negative or positive CRP result may produce additional diagnostic information if the patient presenting has additional symptoms or signs that revises the initial prior probability of CAP into the intermediate range. For example, Hopstaken et al.²⁶ show that in situations where the prior of CAP was 13%, patients presenting with a dry cough, diarrhoea or recorded temperature 38°C should have the probability of CAP revised upwards, as these clinical features are independently associated with CAP. A positive or negative CRP result (cut-point 20 mg/l) in these situations is likely to be more diagnostically informative.

Shortcomings of the review
Our analysis clearly contains considerable population heterogeneity. Some studies take place in a general practice setting and others in the accident and emergency departments of different-sized hospitals. The studies are geographically diverse and the prior probabilities range from 5% to 89%. It is essential that GPs recognize and understand the implications of using a diagnostic test derived in a clinical setting different from their own. Similarly, it is critical that any interpretation of a near-patient test such as CRP is made having taken into account the clinical features in an individual patient and then revising the probability of CAP before proceeding to CRP testing. If a patient has few or no features of CAP, then further diagnostic testing in the community is likely to be uninformative.

Future studies
Further studies are required that assess the diagnostic value of CRP at different cut-points, so that the precision and accuracy of CRP can be established. In addition, the predictive value of clinical signs, particularly focal lung signs, is less substantial than GPs assume while at the same time being strongly associated with the probability of antibiotic prescribing.³,³⁰ Future studies should examine the incremental value of symptoms and signs alongside CRP testing. Ideally, clinical prediction rules for the presence/absence of CAP should be developed and validated, and their impact on antibiotic prescribing, referral to hospital and on patient outcome, be assessed.³¹ Combinations of symptoms and signs in accident and emergency settings provide additional diagnostic value, prior to radiological or haematological investigation.² Future studies in community settings should adopt the same approach.

	Conclusions

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
This systematic review of diagnostic accuracy studies suggest that the value of CRP (cut-point 20 mg/l) may be of value in ruling out a diagnosis of CAP in situations where the probability of CAP is higher than 10%, typically accident and emergency departments. In primary care settings where the probability of CAP appears to be between 5% and 10%, additional diagnostic testing with CRP is unlikely to alter the probability of CAP sufficiently to change subsequent management decisions such as antibiotic prescribing or referral to hospital. The diagnostic value of CRP to rule in a diagnosis of CAP requires further study.

	Declaration

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
Funding: Irish College of General Practitioners and Health Research Board (HRC/2007/1) Centre for Primary Care Research to GF.

Ethical approval: None.

Conflicts of interest: None.

	Appendix-Webtable 1

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 

Summary of diagnostic accuracy studies of CRP in relation to the diagnosis of CAP

Author year; no. of patients Prior (%) Patient population Age; sex: male Prevalence of symptoms/signs Diagnostic tests assessed; cut-points Reference standard; definition Diagnostic utility positive LR+ or odds ratio for CRP cut-points Average levels Notes Almirall et al.28 2004; N = 226 88.9 Patients >14 years with CAP reported by physicians working in public primary health care centres, private clinics and emergency department of hospitals in Barcelona, Spain Patients >14 years; mean with CAP: 57 years; male: 57.7% Not reported CRP mg/l Chest X-ray and symptoms of acute LRTI CRP LR+ CRP mg/l Although the reference standard includes symptoms, data were reported on the sensitivity and specificity of chest X-ray alone. The LR+ reported was calculated from these data CAP versus healthy 11.0 >33.2 1.48 Confirmed 110.7 >50 2.11 Unconfirmed 31.9 CAP versus 0unconfirmed pneumonia 33.2 >100 2.30 Healthy 1.9 Flanders et al.21 2004; N = 168 11.9 Patients 18 year of age presenting to emergency department or acute care ambulatory clinic of University of California, San Francisco with acute cough of duration 3 weeks Patients 18 years; median with CAP: 59 years; median no. of CAP: 37 years; range: 20–88 years; mean: 40 years; male: 45% Symptom or finding CAP CRP Chest X-ray CRP LR+ Mean CRP mg/l Fever 65 11 11 1.9 Confirmed 60 Muscle pain 70 40 40 6.9 CAP Fatigue 70 100 100 52.0 Healthy 9 Runny nose 45 Throat pain 45 Cough 100 Exclusion criteria: existence of any co-morbid conditions that affect CRP levels: pregnancy, inflammatory disorders, other infections, severe tissue damage in previous 7 days, myocardial infarct or unstable angina, cancer, HIV Dry or scant phlegm 50 White phlegm 30 Yellow phlegm 35 Green phlegm 20 Blood in sputum 30 Wheezing 55 Shortness of breath 70 Painful breathing 45 Heart rate 100 bpm 55 Systolic blood pressure 90 mmHg 0 Respiratory rate 24 7 Temperature 37.8°C 53 O2 sat 93% 32 Any vital sign abnormality 70 Toxic appearance 21 Tonsillar exudates 5 Lymphadenopathy 5 Prolonged expiration 5 Decreased breath sounds 60 Rales 50 Rhonchi 40 Wheezes 25 Decreased breath sounds and rales 30 Muller et al.22 2007; N = 545 72.7 Patients with suspected LRTIs, presenting to emergency department in Basel, Switzerland Patients >18 years; mean: 67.1 years; male: 62.6% Symptoms Prevalence CRP (mg/l) Chest X-ray CRP (mg/l) LR+ CRP (mg/l) Cough 91.2 Procalcitonin (µg/l) >40 1.85 Mean 127.9 Sputum 71.6 WBCC (x109/l) >50 2.47 Median 103.4 Dyspnoea 71.9 >100 4.89 Range 0.5–512 Signs CRP (mg/l) Procalcitonin (µg/l) >200 8.88 Rales 72.3 >40 >0.1 Procalcitonin (µg/l) >50 >0.25 Mean 3.1 >100 >0.5 Median 0.32 >200 >1.0 Range 0.02–234.7 WBC (x109/l) Median 12.9 Lagerstrom et al.23 2006; N = 177 46.3 Patients attending a GP in regular working hours Monday to Friday in Sweden Mean with CAP: 51 years; median with CAP: 53; male: 53% Symptoms and signs % CRP (mg/l) Chest X-ray CRP (mg/l) LR+ CRP (mg/l) Inclusion criteria were fever (>38°C) and cough for <1 week, or a long-standing (1–4 weeks) dry cough ± fever Dry cough 21 ESR (mm/hour) >20 2.28 Median 65.0 Additionally 1 of the following signs or symptoms: (i) lateral chest pain; (ii) crackles or wheezes on auscultation; (iii) an ‘apparently’ sick patient Productive cough 79 WBCC (x109/l) >50 3.48 Range 5–150 Lateral chest pain 40 Nasopharyngeal swabs Auscultation signs 83 Sputum for culture ESR (mm/hour) Duration of symptoms (days) CRP (mg/l) Median 53 Median 6 >20 Range 5–100 Range 1–8 >50 WBC (x109/l) IQ 4, 8 Median 9.8 <1 day: no. of patients 3 Range 4.4–21.4 Melbye et al.24 1988; N = 71 15.5 Adults treated with antibiotics by a GP for a suspected pneumonia, Norway Patients >15 years; mean: 48 years; mean with CAP: 44 years; male: 60.6% Not given Test Cut-points Chest X-ray lateral and PA CRP (mg/l) LR+ Not given N = 71; however, to calculate sensitivities and specificities N = 68 only CRP 50 mg/l >11 2.05 ESR >35 mm/hour >50 21.45 WBCC >10 400/mm3 Temperature 37.5°C Melbye et al.25 1992; N = 402 5.0 Adults with symptoms suggestive of respiratory or throat infection in general practice, Norway Patients >18 years; mean age 33.2 years; male: 70% Not given Test Cut-points Chest X-ray CRP (mg/l) LR+ Test Mean A chest X-ray was only ordered only when (i) pneumonia considered to be a possibility along with ESR 20 mm/hour or CRP 20 mg/l; (ii) when a patient had an ESR 50 mm/hour or CRP 60 mg/l, or 3) they were one of 97 patients randomly chosen for X-ray; =>124 patients did not have a chest X-ray CRP 50 mg/l >20 2.83 CRP 23.2 ESR >35 mm/hour >50 5.03 ESR 11.0 Patients with dyspnoea severe enough to need treatment and pregnant women were excluded WBCC >10 400/mm3 >100 4.78 WBCC 9.5 Temperature 37.5°C The average age is just 33.2 years, may help explain the very low prior of just 5.0%. Hopstaken et al.26 2003;N = 246 13.2 Patients presenting to their GP with signs and symptoms of an lower respiratory tract infection in southern Holland Mean: 52 years; range: 18–89 years; male: Unknown Symptoms % History Chest X-ray lateral and PA CRP LR+ Not given N = 246; however, to calculate sensitivities and specificities N = 243 only 65 years 28.8 Dry coughs >10 1.84 Recent cough < 2 days 5.1 Chills >20 3.55 Dry cough 23.9 Nausea >50 6.04 Inclusion criteria: (i) 18 years old AND; (ii) new (28 days) or increasing cough AND; (iii) 1 of the following four: SOB, wheezing, chest pain, auscultation abnormalities AND; (iv) 1 of the following four: reported fever, perspiring, headache, myalgia AND; (v) Dx of LRTI by GP Sputum purulence 54.7 Diarrhoea Dyspnoea 77.4 Physical exam CRP OR Thoracic pain 59.7 General impression Moderate/severe illness >10 14.1 Fever 35.0 >20 9.9 Chills 50.2 Temperature 38°C >50 21.4 Confusion 3.3 ESR OR Nausea 16.0 Blood test Cut-points >10 12.6 Diarrhoea 7.8 CRP 10/20/50 >20 4.7 Smoking 33.3 ESR 10/20/40 >40 8.2 Smoking in the past 61.7 Co-morbidity Asthma 19.3 COPD 13.2 Physical signs General impress: moderate/severe illness 26.7 Respiratory rate > 20/minutes 3.7 Percussion dullness 4.5 Auscultation abnorm 84.0 Bronchial breathing 26.3 Crackles 20.6 Temperature > 38°C 23.9 Clin dx of pneumonia 8.6 Holm et al.27 2007; N = 364 13.2 Adults with a clinical diagnosis of LRTI by their GP in Denmark Patients > 18 years % CRP (mg/l) Chest X-ray CRP LR+ Not given Current smoker 45 PCT (µg/l) >20 2.08 Exclusion criteria: pregnancy, hospitalization in preceding 7 days, severity of illness equiring hospitalization, and former participation in the study Former smoker 26 WBCC (x109/l) Influenza immunization 19 CRP >20 mg/l Pneumococcal immunization 6 WCC >10 x 109/l COPD 9 Any underlying disease 36 Cough 98 Dyspnoea 72 Sputum production 81 Chest pain 64 Abnormal auscultation 37 Temperature > 37.5°C 43 Abnormal auscultation 37

	Notes

 
Falk G and Fahey T. C-reactive protein and community-acquired pneumonia in ambulatory care: systematic review of diagnostic accuracy studies. Family Practice 2009; 26: 10–21.

	References

Top Abstract Introduction Methods Results Discussion Conclusions Declaration Appendix-Webtable 1 References

 
¹ Macfarlane J, Holmes W, Gard P, et al. Prospective study of the incidence, aetiology and outcome of adult lower respiratory tract illness in the community. Thorax (2001) 56:109–114.[Abstract/Free Full Text]

² Metlay JP, Fine MJ. Testing strategies in the initial management of patients with community-acquired pneumonia. Ann Intern Med (2003) 138:109–118.[Abstract/Free Full Text]

³ Holmes WF, Macfarlane JT, Macfarlane RM, Hubbard R. Symptoms, signs, and prescribing for acute lower respiratory tract illness. Br J Gen Pract (2001) 51:177–181.[Web of Science][Medline]

⁴ Clyne B, Olshaker JS. The C-reactive protein. J Emerg Med (1999) 17:1019–1025.[CrossRef][Web of Science][Medline]

⁵ Morley JJ, Kushner I. Serum C-reactive protein levels in disease. Ann N Y Acad Sci (1982) 389:406–418.[Web of Science][Medline]

⁶ Dahler-Eriksen BS, Lassen JF, Petersen PH, Lund ED, Lauritzen T, Brandslund I. Evaluation of a near-patient test for C-reactive protein used in daily routine in primary healthcare by use of difference plots. Clin Chem (1997) 43:2064–2075.[Abstract/Free Full Text]

⁷ Esposito S, Tremolati E, Begliatti E, Bosis S, Gualtieri L, Principi N. Evaluation of a rapid bedside test for the quantitative determination of C-reactive protein. Clin Chem Lab Med (2005) 43:438–440.[Medline]

⁸ Putto A, Ruuskanen O, Meurman O, et al. C reactive protein in the evaluation of febrile illness. Arch Dis Child (1986) 61:24–29.[Abstract/Free Full Text]

⁹ Meehan TP, Fine MJ, Krumholz HM, et al. Quality of care, process, and outcomes in elderly patients with pneumonia. J Am Med Assoc (1997) 278:2080–2084.[Abstract/Free Full Text]

¹⁰ Gleason PP, Meehan TP, Fine JM, Galusha DH, Fine MJ. Associations between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia. Arch Intern Med (1999) 159:2562–2572.[Abstract/Free Full Text]

¹¹ Dorca J, Torres A. Lower respiratory tract infections in the community: towards a more rational approach. Eur Respir J (1996) 9:1588–1589.[Medline]

¹² Simpson JC, Macfarlane JT, Watson J, Woodhead MA. A national confidential enquiry into community acquired pneumonia deaths in young adults in England and Wales. British Thoracic Society Research Committee and Public Health Laboratory Service. Thorax (2000) 55:1040–1045.[Abstract/Free Full Text]

¹³ Melbye H, Stocks N. Point of care testing for C-reactive protein—a new path for Australian GPs? Aust Fam Physician (2006) 35:513–517.[Medline]

¹⁴ Haynes RB, Wilczynski N, McKibbon KA, Walker CJ, Sinclair JC. Developing optimal search strategies for detecting clinically sound studies in MEDLINE. J Am Med Inform Assoc (1994) 1:447–458.[Abstract/Free Full Text]

¹⁵ Medion Database. University of Maastricht. http://www.mediondatabase.nl/. (last accessed 26 November 2008).

¹⁶ Moher D, Pham B, Klassen TP, et al. What contributions do languages other than English make on the results of meta-analyses? J Clin Epidemiol (2000) 53:964–972.[CrossRef][Web of Science][Medline]

¹⁷ Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol (2003) 3:25.[CrossRef][Medline]

¹⁸ Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol (2006) 6:31.[CrossRef][Medline]

¹⁹ Grimes DA, Schulz KF. Refining clinical diagnosis with likelihood ratios. Lancet (2005) 365:1500–1505.[CrossRef][Web of Science][Medline]

²⁰ Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J (2003) 327:557–560.[Free Full Text]

²¹ Flanders SA, Stein J, Shochat G, et al. Performance of a bedside c-reactive protein test in the diagnosis of community-acquired pneumonia in adults with acute cough. Am J Med (2004) 116:529–535.[CrossRef][Web of Science][Medline]

²² Müller B, Harbarth S, Stolz D, et al. Diagnostic and prognostic accuracy of clinical and laboratory parameters in community-acquired pneumonia. BMC Infect Dis (2007) 7:10.[CrossRef][Medline]

²³ Lagerstrom F, Engfeldt P, Holmberg H. C-reactive protein in diagnosis of community-acquired pneumonia in adult patients in primary care. Scand J Infect Dis (2006) 38:964–969.[Medline]

²⁴ Melbye H, Straume B, Aasebo U, Brox J. The diagnosis of adult pneumonia in general practice. The diagnostic value of history, physical examination and some blood tests. Scand J Prim Health Care (1988) 6:111–117.[Medline]

²⁵ Melbye H, Straume B, Brox J. Laboratory tests for pneumonia in general practice: the diagnostic values depend on the duration of illness. Scand J Prim Health Care (1992) 10:234–240.[Medline]

²⁶ Hopstaken RM, Muris JW, Knottnerus JA, Kester AD, Rinkens PE, Dinant GJ. Contributions of symptoms, signs, erythrocyte sedimentation rate, and C-reactive protein to a diagnosis of pneumonia in acute lower respiratory tract infection. Br J Gen Pract (2003) 53:358–364.[Web of Science][Medline]

²⁷ Holm A, Nexoe J, Bistrup LA, et al. Aetiology and prediction of pneumonia in lower respiratory tract infection in primary care. Br J Gen Pract (2007) 57:547–554.[Medline]

²⁸ Almirall J, Bolibar I, Toran P, et al. Contribution of C-reactive protein to the diagnosis and assessment of severity of community-acquired pneumonia. Chest (2004) 125:1335–1342.[Abstract/Free Full Text]

²⁹ van der Meer V, Neven AK, van den Broek PJ, Assendelft WJ. Diagnostic value of C reactive protein in infections of the lower respiratory tract: systematic review. Br Med J (2005) 331:26.[Abstract/Free Full Text]

³⁰ Hopstaken RM, Butler CC, Muris JW, et al. Do clinical findings in lower respiratory tract infection help general practitioners prescribe antibiotics appropriately? An observational cohort study in general practice. Fam Pract (2006) 23:180–187.[Abstract/Free Full Text]

³¹ Reilly BM, Evans AT. Translating clinical research into clinical practice: impact of using prediction rules to make decisions. Ann Intern Med (2006) 144:201–209.[Abstract/Free Full Text]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				J. W L Cals, F. H F Chappin, R. M Hopstaken, M. E van Leeuwen, K. Hood, C. C Butler, and G.-J. Dinant C-reactive protein point-of-care testing for lower respiratory tract infections: a qualitative evaluation of experiences by GPs Fam. Pract., December 18, 2009; (2009) cmp088v1. [Abstract] [Full Text] [PDF]

				A. D Hay and K. V Juttner Antibiotics for acute cough in primary care BMJ, May 5, 2009; 338(may05_1): b834 - b834. [Full Text]

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 26/1/10    most recent cmn095v1 E-letters: Submit a response Alert me when this article is cited Alert me when E-letters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Google Scholar Articles by Falk, G. Articles by Fahey, T. Search for Related Content PubMed PubMed Citation Articles by Falk, G. Articles by Fahey, T. Social Bookmarking          What's this?

Online ISSN 1460-2229 - Print ISSN 0263-2136

Copyright © 2010 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics