Usefulness of aetiological tests for guiding antibiotic therapy in community-acquired pneumonia

Transcription

1 International Journal of Antimicrobial Agents 31 (2008) 3 11 Review Usefulness of aetiological tests for guiding antibiotic therapy in community-acquired pneumonia K. Strålin Department of Infectious Diseases, Örebro University Hospital, SE Örebro, Sweden Abstract The goal with antibiotic therapy in community-acquired pneumonia (CAP) is to cure the patient, ideally without causing side effects and without contributing to the further development of antibiotic resistance. Although patients with severe CAP should be treated with broad-spectrum antibiotics, patients with non-severe CAP should preferably receive pathogen-directed therapy. Rapid aetiological tests, such as sputum Gram stain and urinary antigen tests, are useful for targeting initial pathogen-directed therapy. Non-rapid tests, such as cultures, can subsequently support a switch from initial broad-spectrum therapy to narrow-spectrum therapy and direct therapy changes in case of treatment failure. As conventional diagnostic methods often fail to identify the aetiology of CAP, PCR (polymerase chain reaction) tests for respiratory pathogens have become useful and should be further developed. Based on the test specificities, aetiological tests may provide diagnoses with varying reliability, i.e. definite aetiologies (e.g., blood culture and Legionella urinary antigen test), probable aetiologies (e.g., sputum culture and PCR for Mycoplasma pneumoniae), or possible aetiologies (e.g., culture of nasopharyngeal secretions and PCR for Streptococcus pneumoniae). A definite or probable aetiology can often be used to target antibiotic therapy Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Community-acquired pneumonia; Aetiology; Diagnostic tests; PCR; Urinary antigen test; Antibiotics 1. Introduction Community-acquired pneumonia (CAP) is a common and sometimes severe disease with an annual incidence of about 1% [1] and a mortality rate of 0 30% [2]. Thus, selection of antibiotic therapy in CAP is important for the emergence of antibiotic resistance in society and for the outcome in the patients [3]. Ideally, antibiotic therapy should be directed against the pathogen that is causing the pneumonia. However, as the aetiology is often not known at presentation, patients must initially receive empirical antibiotic treatment. This review will focus on how aetiological tests can be used to guide antibiotic therapy in adult patients with CAP. Present address: Department of Infectious Diseases, Örebro University Hospital, SE Örebro, Sweden. Tel.: ; fax: address: kristoffer.stralin@orebroll.se. 2. Empirical antibiotic treatment in community-acquired pneumonia In patients with severe CAP, under-use of antibiotics and use of narrow-spectrum antibiotics can be associated with a risk of death [3]. Thus, international guidelines [4 9] recommend treatment with broad-spectrum antibiotics for severe CAP, as defined by the CURB-65 (confusion, urea, respiratory rate and blood pressure aged 65 or more) score or Pneumonia Severity Index rules. As CAP due to multiple pathogens can occur [10,11], patients with severe CAP should be treated with a broad antibacterial agent(s) for the first 2 days, even if an aetiological agent has been identified. In patients with non-severe CAP, overuse of broadspectrum antibiotics can increase the risk of adverse events [12,13] and the development of antibiotic resistance among respiratory pathogens [14] and other bacterial species [15]. Thus, for patients with non-severe CAP, the guidelines from the British Thoracic Society (BTS) [7] and the Swedish Society of Infectious Diseases [8] recommend monotherapy with penicillin or an aminopenicillin, unless atypical pathogens /$ see front matter 2007 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi: /j.ijantimicag

2 4 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) 3 11 are suspected. Penicillin is a reliable therapy for CAP caused by Streptococcus pneumoniae [16], even in cases when the bacterium has reduced in vitro susceptibility to penicillin [17,18]. As S. pneumoniae is the commonest microbiological cause of CAP [2,5,11] and the most commonly identified cause of CAP death [2], it should always be covered by the empirical CAP treatment [8]. The recommendation to treat patients with non-severe CAP with -lactam monotherapy is supported by two meta-analyses that showed that such treatment was not inferior to treatment with agents with activity against atypical pathogens, unless Legionella was the cause [19,20]. However, the North American guidelines [4,5], recommend routine antibiotic coverage of atypical pathogens for hospitalised patients with non-severe CAP. 3. Prediction of aetiology based on clinical, laboratory, radiological and epidemiological factors Although single clinical, laboratory and radiological features cannot reliably predict CAP aetiology [11,21], by combining a number of features the possibility of making a correct prediction increases. Acute onset of illness, pleuritic chest pain, high leukocyte count, and lobar chest X-ray infiltrates are associated with pneumococcal aetiology, while young age, gradual onset of illness, and non-productive cough have been associated with mycoplasmal aetiology [22 24]. With a prediction model based on these features, Farr et al. [23] correctly predicted pneumococcal aetiology in 45%, and mycoplasmal aetiology in 77%, of patients who were positive by microbiological testing. Thus, the presentation of pneumococcal pneumonia is variable. This fact further supports the recommendation [8] that S. pneumoniae should always be covered in empirical CAP treatment. Legionella aetiology has been associated with diarrhoea, headache or confusion, liver dysfunction and hyponatraemia [24,25]. Certain risk factors or epidemiological conditions can also indicate that the pneumonia is caused by a certain pathogen. For instance, associations between different conditions and pathogens have been identified as follows: smoking and/or chronic obstructive pulmonary disease Haemophilus influenzae; alcoholism S. pneumoniae; exposure to birds Chlamydophila psittaci; recent hotel stay Legionella; and influenza virus infection S. pneumoniae, Staphylococcus aureus, and H. influenzae [4]. In a recent study by van der Eerden et al. [12] initial antibiotic treatment, guided by the clinical and epidemiological presentation, was at least as effective as treatment guided by the results of rapid microbiological investigations. Based on clinical criteria, patients were treated with penicillin, erythromycin, amoxicillin/clavulanate, or flucloxacillin ± gentamicin [12]. This study supports the usefulness of the clinical and epidemiological presentation for the selection of antibiotic therapy in CAP, at least when all treatment alternatives are effective against S. pneumoniae. 4. The role of aetiological testing for antimicrobial therapy in CAP A major role of aetiological testing in CAP is to enable the use of pathogen-directed therapy, and thus reduce the use of broad-spectrum antibiotics. As it is recommended that the antibiotic therapy should be started within 4 h of hospital admission [4,8], rapid tests with a shorter analysis time can be used to influence the choice of first-line antibiotic therapy. Sputum Gram strain, urinary antigen tests, and real-time PCR for respiratory pathogens are examples of such rapid tests. Less rapid tests, such as cultures, conventional PCR for respiratory pathogens, and serology, can provide useful information that may support ongoing antibiotic therapy, support narrowing of broad-spectrum therapy, and support therapy changes in case of treatment failure [3]. Culture of blood and respiratory specimens may be essential for the identification of unexpected or uncommon CAP aetiologies that the empirical treatment does not cover for, e.g., Pseudomonas spp., methicillin-resistant S. aureus, and other highly resistant pathogens. Culture remains a cornerstone of the diagnostic techniques, as it can provide information about antibiotic susceptibility. In patients who receive narrow-spectrum antibiotics based on clinical prediction of aetiology, identification of the pathogen by an aetiological test is helpful for determining the need for treatment alterations [26]. Another major role of aetiological testing is to survey the spectrum of CAP aetiologies in order to guide future empirical antimicrobial policies. 5. Definite, probable and possible aetiological diagnoses When an aetiological test is positive, it is important to question the reliability of the result. Depending on the test specificities, positive results of different tests can be categorised to represent definite, probable or possible CAP aetiologies [27]. Based on a classification by Marston et al. [27], a new categorisation of bacterial aetiologies, which includes additional and newer tests, is presented in Table 1. Tests with high specificities can provide definitive bacterial aetiologies and can target safely antibiotic therapy, while tests with low specificities can provide possible aetiologies. When a possible aetiology is identified by a diagnostic test, the result should be related to the clinical presentation, and to the response to initial antibiotic therapy, before treatment alteration should be directed from it. Probable aetiological diagnoses can often be used to target antibiotic therapy. If microbiological testing provides two (or more) bacterial aetiologies with different reliability in a patient, the aetiology with the strongest reliability is most likely to represent the true aetiology. The role of mixed bacterial aetiologies in CAP is not yet fully understood.

3 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) Table 1 Categorisation of positive diagnostic test results into definite, probable, and possible bacterial aetiologies of community-acquired pneumonia Definite aetiology Isolation of bacterial pathogens in culture from blood or pleural fluid Isolation of Legionella spp. from respiratory secretions Identification of Legionella antigen in urine Identification of Legionella DNA in respiratory secretions Four-fold rise in antibody titre to Legionella spp., Mycoplasma pneumoniae, Chlamydophila pneumoniae or Chlamydophila psittaci Probable aetiology Identification of pneumococcal antigen in urine a Isolation of bacterial pathogens in culture from purulent sputum or specimens obtained by bronchoscopy Moderate to large amounts of Gram-positive diplococci seen in sputum Gram stain of purulent sputum Identification of M. pneumoniae DNA or C. pneumoniae DNA in respiratory secretions b A high IgM antibody titre to M. pneumoniae or an elevated IgM antibody titre to C. pneumoniae or C. psittaci Possible aetiology Predominance of a bacterial pathogen other than diplococci, as seen in sputum Gram stain Isolation of Streptococcus pneumoniae or Haemophilus influenzae in culture from nasopharyngeal secretions or non-purulent sputum Identification of pneumococcal DNA or H. influenzae DNA in respiratory secretions c a A positive result of the Binax NOW test, with a colour intensity of the result line at least as intense as that of the control line provides a definite aetiology. b Well-validated quantitative PCR assays with high detection limits may provide definite aetiologies. c Well-validated quantitative PCR assays with high detection limits may provide probable aetiologies. Tests that provide possible aetiological diagnoses, such as PCR for S. pneumoniae, or culture of nasopharyngeal secretions, have suboptimal specificities, although they may have high sensitivities [28]. Negative results of highly sensitive tests may be used to rule out CAP aetiologies and direct the physicians to question the CAP diagnosis and/or to search for less common aetiological agents. 6. Reliability and usefulness of different diagnostic tests 6.1. Blood culture Although blood culture has a low sensitivity for the confirmation of CAP aetiology [29], it has a high specificity, and a positive result provides a definite diagnosis. As bacteraemia in CAP is associated with significant mortality [2], blood cultures should always be obtained from hospitalised CAP patients, in order to identify bacteraemia and thus enable optimal pathogen-directed therapy. Meehan et al. [30] found collection of blood culture within 24 h from hospital arrival to be associated with improved survival in patients with CAP. However, if a patient has bacteraemia with a bacterium that is non-specific for the respiratory tract, such as Gram-negative enteric bacilli, the physician should question if the patient has a primary infection at another site. Infiltrates on chest X- ray may be caused by non-infectious conditions [31]. Double infection with CAP and another community-acquired infection should also be considered Sputum culture Sputum culture is often used for the identification of probable/presumed aetiologies in CAP [11 13,21,27], either singly [13,21] or with the simultaneous identification of the compatible organism in sputum Gram stain [11,12,27]. It is usually recommended that the quality of the sputum samples should be determined by microscopy [4,8]. Samples with a high ratio between the leukocyte concentration and squamous epithelial cells are more likely to originate from the lower respiratory tract than samples with a low ratio [32,33], and should thus be considered more reliable (Table 1). In addition, in order to minimise the impact of contamination from oropharyngeal flora, a bacterial concentration of 10 5 colony-forming units (CFU)/mL of sputum is often used as the detection limit for a positive sputum culture [34]. The concentration of colonising bacteria in the respiratory tract is probably usually below that limit [35]. It is particularly important to perform sputum culture in cases of necrotising pneumonia, since S. aureus aetiology in such cases has been reported to have a high mortality [36]. A negative culture result of good-quality sputum has been suggested to be strong evidence against S. aureus aetiology in CAP [4]. Gram-negative enteric bacilli often colonise the respiratory tract of elderly persons and patients who are treated with antibiotics [37]. Thus, sputum cultures positive for such bacteria should be interpreted with caution. As Legionella spp. do not normally colonise the respiratory tract, the isolation of Legionella in lower respiratory tract culture provides a definite diagnosis [38]. In a patient who cannot produce an adequate sputum sample, sputum production can be induced by inhalation of 3% NaCl [39] Culture of bronchoscopic-derived samples In cases of therapy failure, when conventional methods have failed to identify the CAP aetiology, diagnostic bronchoscopy should be considered [6]. Cultures from bronchoalveolar lavage (BAL) and protected specimen brush (PSB) are both sensitive and specific [40]. However, since bacteria from the upper respiratory tract may be introduced into the lower respiratory tract by the bronchoscope [41], false-positive results may occur. Thus, BAL and PBS specimens should preferably be cultured quantitatively [4]. Commonly used cut-offs for positive results are 10 4 CFU/mL for BAL fluid [11] and 10 3 CFU/mL for PSB [42].

4 6 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) Culture of nasopharyngeal secretions In cases when no sputum sample can be obtained, secretions from the nasopharynx can be obtained for culture [8,28]. Such a culture may identify the pathogen that is responsible for the infection, and may yield useful information about antibiotic susceptibility. For S. pneumoniae, nasopharyngeal swab culture has shown sensitivities of 27 58%, and specificities of 82 97% [43,44].ForH. influenzae,we[44] found nasopharyngeal swab culture to have a sensitivity of 75% and a specificity of 95%. Among 621 asymptomatic Swedish adults in four studies [28,45 47], culture positivity rates of nasopharyngeal swabs were as low as 2.1% for S. pneumoniae and 2.1% for H. influenzae. These results support the clinical usefulness of nasopharyngeal culture for identification of these two bacteria in adults Culture of pleural fluid Diagnostic thoracocentesis should be performed when a significant pleural effusion is present [5]. A positive culture of pleural fluid provides a definite CAP aetiology Sputum Gram stain The interpretation of Gram-stained sputum samples has been associated with inter- and intra-reader variability [48,49], and the sensitivities and specificities for the detection of pneumococcal pneumonia have varied between different studies [49 52]. However, applied to sputum samples of good quality, Gram stain has been found useful and reliable for targeting pathogen-directed first-line antibiotic therapy in CAP patients [53,54]. In a study by Boerner and Zwadyk [53],42 of 43 pneumonia patients with a sputum Gram stain showing Gram-positive diplococci responded to penicillin or ampicillin monotherapy. In a prospective study by Roson et al. [54], 93 CAP patients with a sputum Gram stain showing Gram-positive diplococci were treated with a single antibiotic agent with activity against S. pneumoniae. Therapy was changed in only three of them. Sputum Gram stain has rarely been studied for bacteria other than S. pneumoniae. Fine et al. [49] found the sensitivity of sputum Gram stain to be less for identification of H. influenzae than for the identification of S. pneumoniae.however, the study by Roson et al. [54] indicates that when sputum Gram stain identifies Gram-negative coccobacilli, which is usually interpreted as H. influenzae, it may be clinically useful information. Thirty of 36 patients with such results were given -lactam monotherapy and none of them was subjected to therapy change [54] Pneumococcal urinary antigen Among adult CAP patients, the Binax NOW S. pneumoniae urinary antigen test has shown sensitivities of 79 87% compared with blood culture [55 57], but only 54%, when cultures of respiratory tract secretions were included in the reference standard [57]. However, the specificity is high. Among adults without respiratory symptoms in three summarised studies [57 59], the false positivity rate was 0.6% (2/336) and it was 8% (1/13) among subjects colonised with S. pneumoniae in the nasopharynx. As false positivity may be caused by a recent pneumococcal infection [58,60], a positive result provides a probable, but not a definite, pneumococcal aetiology (Table 1). However, a positive result with a colour intensity of the result line at least as intense as that of the control line is probably more specific [55,57], and may provide a definite pneumococcal aetiology. Concentration of the urine prior to testing will enhance the sensitivity but lower the specificity of the test. Among 219 young patients with mild pneumonia, Guchev et al. [61] prospectively targeted antibiotic therapy according to the results of the NOW S. pneumoniae urinary antigen test. The patients with positive test results (22% of the cases) were treated with amoxicillin, while those with negative results (78%) were treated with claritromycin. The clinical success rates were 90% in the amoxicillin group and 94% in the claritromycin group (non-significant). However, since the NOW S. pneumoniae test has a suboptimal sensitivity [57], it cannot be used to rule out pneumococcal pneumonia. Thus, we [26] studied the clinical course of 152 hospitalised patients (median age 74 years) who were empirically treated with -lactam monotherapy and divided the patients according to the result of the urinary antigen test, which was positive in 38 patients (25%) and negative in 114 patients. As the analyses were performed at the end of the study, the physicians did not know the results. A successful outcome was noted in 92% of those with a positive test result, and in 76% of those with a negative test result (P = 0.034). Among the 30 patients who did not have a successful outcome, only one patient died, while 29 patients were cured after treatment alteration [26]. These two studies [26,61] indicate that the urinary antigen test can safely target narrow-spectrum -lactam treatment in CAP. Our study [26] supports the use of -lactam monotherapy as empirical therapy also in test-negative patients, although physicians should be aware that therapy changes will be needed more frequently than in test-positive patients. Thus, a negative test result motivates additional aetiological testing. Contrary to cultures, the pneumococcal urinary antigen test often remains positive during antibiotic therapy [56]. This makes the test useful in patients who have received antibiotic treatment prior to sampling Legionella urinary antigen The Legionella urinary antigen test has a high specificity for Legionella pneumophila serogroup 1 [38]. A Dutch study [62] indicates that antibiotic management in Legionnaires disease can be guided by the results of the Legionella urinary antigen test. Early adequate therapy did not influence the

5 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) outcome among patients with Legionnaires disease who had a negative Legionella urinary antigen test, although it reduced the risk of intensive care unit admission and death by 38% in those with a positive antigen test [62] Polymerase chain reaction (PCR) methods Since conventional diagnostic methods often fail to identify a CAP aetiology, the sensitive PCR technique appears promising for the identification of aetiological agents. However, PCR methods have not yet been standardised for respiratory pathogens other than Mycobacterium tuberculosis and Legionella spp. [63]. While PCR for these two pathogens provides definite diagnosis, PCR for Mycoplasma pneumoniae and Chlamydophila pneumoniae can provide probable aetiologies, as they may colonise the respiratory tract of healthy persons [64,65] and persist in the respiratory tract for a long period after an acute infection [66,67]. PCR methods for S. pneumoniae and H. influenzae applied to respiratory tract specimens have rarely been studied. The usefulness of such PCR methods is dependent on the specificity of the DNA target of the method, and on the detection limit of the assay, since bacteria closely related to these pathogens are present in the normal flora of the upper respiratory tract, and since the respiratory tract of healthy adults may be colonised with small amounts of these organisms [44]. Before the clinical use of such methods, they should be evaluated on samples from patients with lower respiratory tract infection and from control subjects. Fig. 1 shows the yield of a multiplex PCR for S. pneumoniae, H. influenzae, M. pneumoniae and C. pneumoniae applied to nasopharyngeal aspirates from CAP patients and controls [28]. Compared with conventional reference standards this assay showed the following sensitivities and specificities: for S. pneumoniae 91% and 85%; for H. influenzae 75% and 86%; for M. pneumoniae 90% and 97%; and for C. pneumoniae 100% and 100%, respectively [28,44]. We correlated the results of PCR for S. pneumoniae applied to nasopharyngeal aspirates to the clinical course in 136 hospitalised CAP patients, who were treated with lactam monotherapy [26]. The PCR results were not known to the physicians who treated the patients. Clinical improvement and discharge without treatment change, and survival at day 30, was seen in 93% (n = 53) of 57 patients with positive PCR and in 70% (n = 55) of 79 patients with PCR negative for S. pneumoniae (P = 0.001) (previously unpublished data). These results indicate that PCR for S. pneumoniae may be clinically useful. We have also studied PCR for respiratory pathogens on BAL fluid from patients with lower respiratory tract infection [68]. The major advantages compared with culture were seen in patients who were treated with antibiotics prior to bronchoscopy. Among 103 such patients, S. pneumoniae was identified by BAL culture in three cases and by BAL PCR in 32 cases (P < 0.001) [68]. As pneumococcal aetiology with a slow response to adequate antibiotic therapy and mycoplasmal aetiology have been reported to be two major causes of therapy failure in CAP [69], PCR for S. pneumoniae and M. pneumoniae may be useful complements to culture of BAL in patients who are subjected to bronchoscopy. Multiplex PCR assays provide several results in one analysis. If such an assay has a high sensitivity, a negative result may indicate that none of the tested pathogens has caused the pneumonia. Accordingly, negative results for several pathogens support the likelihood that a positive result is truly positive. Real-time PCR provides a rapid result, which may influence the choice of the first-line antibiotic therapy in CAP. It may also provide quantification of the DNA content of the sample. This enables the selection of clinically useful detection limits that could differentiate between infection and colonisation. Thus, the specificity and reliability of a positive result may be improved (Table 1, footnotes b and c). Wellvalidated commercial quantitative real-time multiplex PCR assays will probably be valuable aetiological tests. PCR methods for the identification of bacterial pathogens in blood samples are under development [70] and may become useful tools in the management of CAP [42] Rapid tests for viral pathogens Fig. 1. Positivity rates of a multiplex PCR for Streptococcus pneumoniae, Haemophilus influenzae, Mycoplasma pneumoniae and Chlamydophila pneumoniae applied to nasopharyngeal aspirate in patients with communityacquired pneumonia (CAP) and controls [28]. Commercial antigen tests for the influenza virus, based on immunofluorescence or immunochromatographic techniques, have high specificities for the detection of influenza [71,72]. The sensitivities of the tests are also high, although viral culture and PCR have even higher sensitivities [71]. It has been well known for several decades that influenza often precedes severe S. aureus pneumonia [73]. Thus, patients with severe CAP who are positive for the influenza virus by a rapid test should reasonably be treated with antibiotic therapy that includes coverage of S. aureus. Falsey et al. [74] found rapid influenza virus testing to be of importance for antimicrobial therapy in 166 patients with documented influenza. Patients with positive (n = 86) and negative (n = 80) rapid influenza antigen tests had

6 8 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) 3 11 similar frequencies of chest X-ray infiltrates (20% and 24%), although antibiotic treatment was given less often to those with a positive antigen test (86% versus 99%; P = 0.002). In addition, a positive influenza antigen test resulted in higher rates of discontinuation of antibiotic therapy (14% versus 2%; P = 0.004) and the use of antiviral agents (73% versus 8%; P < 0.001). Real-time PCR for viral pathogens was recently reported to double the positivity rate of viral pathogens in patients with lower respiratory tract infection [75]. However, the results caused partial or total cessation of antibiotic treatment in only 11% of the tested patients. The total antibiotic use and cost was not reduced Serological tests Serological testing is useful mainly for systematic surveys for the establishment of frequencies of different CAP aetiologies. In such surveys paired sera should be used. However, a high immunoglobulin antibody titre against M. pneumoniae, detected by enzyme-linked immunosorbent assay (IgM) or complement fixation test, may be used for the establishment of probable mycoplasmal aetiology [76]. Accordingly, IgM titres of 1/16 for C. pneumoniae or C. psittaci, detected by the microimmunofluorescence test, may be used for the establishment of probable Chlamydophila infection [77] (Table 1). A single elevated antibody titre is too non-specific for the establishment of Legionella infection [38]. 7. Experience of first-line antibiotic therapy guided by rapid tests The usefulness of sputum Gram stain and S. pneumoniae urinary antigen test for selecting first-line antibiotic treatment has been discussed in Sections 6.6 and 6.7. In a prospective randomised study, van der Eerden et al. [12] compared pathogen-directed therapy with empiric broadspectrum antibiotic therapy in 303 patients with CAP of different severities. Pathogen-directed therapy was based on the clinical presentation (n = 72), or on the results of rapid microbiological investigations (n = 62), i.e. Gram stain of sputum or pleural fluid, pneumococcal antigen detection in sputum or pleural fluid, and Legionella urinary antigen. The rates of therapy failure were similar ( 20%) in all sub-groups of the study. However, side effects of antibiotic therapy occurred more frequently in patients treated empirically with broad-spectrum antibiotics than in those treated with pathogen-directed therapy (60% versus 17%; P < 0.001) [12]. 8. Experience of antibiotic therapy guided by non-rapid aetiological testing Among CAP patients with an identified aetiological agent, therapy changes were carried out according to these findings in 12% in a study by Lidman et al. [13] and in 32% in a study by Ewig et al. [78]. In a study of patients with severe CAP [79], the results of microbiological investigations led to a change in therapy in 42% of cases. The commonest change was simplification of the treatment. 9. Economic considerations of diagnostic testing The main argument against aetiological testing is cost. However, among hospitalised CAP patients, traditional aetiological testing contributes only to a small proportion of the total cost [80]. The largest cost in the management of CAP is the in-patient cost [81]. If the CAP aetiology is known in the early course of disease, optimal therapy with a low risk of adverse events can be given, and the hospital stay can probably be short. In addition, if the aetiology is known, it is possible that the physician will feel more confident to treat the patient as an outpatient or discharge the patient earlier. Table 2 Diagnostic tests for bacterial pathogens in patients with severe community-acquired pneumonia, as recommended by the international guidelines Guidelines a Diagnostic test b IDSA/ATS [4] Canadian [5] ERS [6] BTS [7] Swedish [8] Blood culture Gram stain and culture of lower respiratory tract specimens c d d Pleural fluid culture if pleural effusion Legionella culture e Legionella UAT Pneumococcal UAT Nasopharyngeal culture PCR for M. pneumoniae, C. pneumoniae, and Legionella spp. f a IDSA/ATS, Infectious Diseases Society of America/American Thoracic Society; Canadian, Canadian Infectious Diseases Society and Canadian Thoracic Society; ERS, European Respiratory Society; BTS, British Thoracic Society; Swedish, Swedish Society of Infectious Diseases. b UAT, urinary antigen test. c Sputum, bronchoalveolar lavage, or protected specimen brush. d Sputum Gram stain only for examination of the amount of leukocytes and squamous epithelial cells. e Legionella culture should be performed if the Legionella UAT is positive. f If well-validated test available.

7 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) This could have positive economic consequences. Studies are needed to evaluate if aetiological testing is cost-effective. 10. Recommended aetiological testing Recommendations by international guidelines International guidelines recommend increased efforts to identify the causative organism in severe CAP [4 8] (Table 2), although therapy should not be delayed if there is difficulty in obtaining adequate samples [5]. As the patient is improving, the initial broad-spectrum therapy can be de-escalated or narrowed, according to the results of the diagnostic tests [4,5,8]. However, it is recommended that the therapy should not be narrowed until concerns regarding mixed infection have been appropriately addressed [5,6]. For non-severe CAP, diagnostic testing to identify an aetiology is optional, according to the guidelines of the Infectious Diseases Society of America/American Thoracic Society [4], and should be based on clinical factors, epidemiological factors and prior antibiotic therapy, according to the BTS and Swedish guidelines [7,8] Suggested strategy for diagnostic testing In Table 3, a strategy for diagnostic testing in hospitalised CAP patients is presented. It is based on disease severity and Table 3 Recommended strategy for aetiological testing in hospitalised patients with community-acquired pneumonia (CAP) Groups of CAP patients Recommended microbiological investigations a No antibiotics taken prior to Blood culture sampling Culture and Gram stain of sputum, culture of NPS if sputum is unobtainable Suspected atypical pathogen Blood culture Culture and PCR for respiratory pathogens (M. pneumoniae, C. pneumoniae, Legionella spp., S. pneumoniae and H. influenzae) applied to sputum (or NPS), Legionella urinary antigen test Antibiotic treatment prior to Blood culture sampling S. pneumoniae urinary antigen test PCR for respiratory pathogens applied to sputum (or NPS) Severe CAP, e.g., CURB-65 Blood culture 3 5, or Pneumonia Severity Index score > 110 or Culture, Gram stain, and PCR for respiratory pathogens applied to sputum/bal b (or NPS) Failure to the initial antibiotic therapy S. pneumoniae urinary antigen test, Legionella urinary antigen test a NPS, nasopharyngeal secretions; BAL, bronchoalveolar lavage. b If PCR for Legionella spp. cannot be performed, Legionella culture should be performed instead. presence/absence of prior antibiotic therapy. Ongoing antibiotic treatment often causes false-negative cultures, although PCR tests [82] and urinary antigen tests [56] often remain positive in spite of the treatment. Thus, the preferred diagnostic techniques could be culture in non-treated patients, and PCR and urinary antigen tests in patients with ongoing treatment. However, in order to identify unexpected pathogens that are not covered by the empirical therapy, blood culture should be performed in all hospitalised CAP patients and culture from respiratory secretions should be performed in all patients with severe CAP, and in cases of therapy failure. This strategy recommends that at least one respiratory tract sample should be collected from all hospitalised CAP patients for culture and/or PCR, in order to enable frequent identification of aetiologies. 11. Conclusion In order to cure CAP patients without causing unnecessary side effects and without contributing to the development of antibiotic resistance, antibiotic therapy should be carefully selected. Although patients with severe CAP should be treated with broad-spectrum antibiotics, patients with non-severe CAP should preferably receive pathogen-directed therapy. While rapid aetiological tests may be useful for targeting initial pathogen-directed therapy, non-rapid tests may support switch from broad- to narrow-spectrum antibiotic therapy and support therapy changes in the case of treatment failure. Positive results of the different tests can be categorised to represent definite, probable or possible CAP aetiologies. While a definite or probable aetiology can often be used to target antibiotic therapy, the clinical presentation and the response to the initial antibiotic therapy should be considered prior to therapeutic decisions based on a possible aetiology. As conventional diagnostic methods often fail to identify the CAP aetiology, PCR tests for respiratory pathogens have become useful and should be further developed. Funding: The Research Committee of Örebro County Council (Sweden). Competing interest: None declared. Ethical approval: Not required. References [1] Jokinen C, Heiskanen L, Juvonen H, et al. Incidence of communityacquired pneumonia in the population of four municipalities in eastern Finland. Am J Epidemiol 1993;137: [2] Fine MJ, Smith MA, Carson CA, et al. Prognosis and outcomes of patients with community-acquired pneumonia: a meta-analysis. JAMA 1996;275: [3] Niederman MS. Use of broad-spectrum antimicrobials for the treatment of pneumonia in seriously ill patients: maximizing clinical outcomes and minimizing selection of resistant organisms. Clin Infect Dis 2006;42(Suppl. 2):S72 81.

8 10 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) 3 11 [4] Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007;44(Suppl. 2):S [5] Mandell LA, Marrie TJ, Grossman RF, Chow AW, Hyland RH; The Canadian Community-Acquired Pneumonia Working Group. Canadian guidelines for the initial management of community-acquired pneumonia: an evidence-based update by the Canadian Infectious Diseases Society and the Canadian Thoracic Society. Clin Infect Dis 2000;31: [6] Woodhead M, Blasi F, Ewig S, et al. Guidelines for the management of adult lower respiratory tract infections. Eur Respir J 2005;26: [7] Macfarlane J, Boswell T, Douglas G, et al. BTS guidelines for the management of community-acquired pneumonia in adults 2004 update. URL: MACAPrevisedApr04.pdf, [8] Hedlund J, Strålin K, Örtqvist Å, Holmberg H; Community-Acquired Pneumonia Working Group of the Swedish Society of Infectious Diseases. Swedish guidelines for the management of communityacquired pneumonia in immunocompetent adults. Scand J Infect Dis 2005;37: [9] Schouten JA, Prins JM, Bonten MJ, et al. Revised SWAB guidelines for antimicrobial therapy of community-acquired pneumonia. Neth J Med 2005;63: [10] Lieberman D, Schlaeffer F, Boldur I, et al. Multiple pathogens in adult patients admitted with community-acquired pneumonia: a one year prospective study of 346 consecutive patients. Thorax 1996;51: [11] Ruiz M, Ewig S, Marcos MA, et al. Etiology of community-acquired pneumonia: impact of age, comorbidity, and severity. Am J Respir Crit Care Med 1999;160: [12] van der Eerden MM, Vlaspolder F, de Graaff CS, et al. Comparison between pathogen directed antibiotic treatment and empirical broad spectrum antibiotic treatment in patients with community acquired pneumonia: a prospective randomised study. Thorax 2005;60: [13] Lidman C, Burman LG, Lagergren A, Ortqvist A. Limited value of routine microbiological diagnostics in patients hospitalized for community-acquired pneumonia. Scand J Infect Dis 2002;34: [14] Bronzwaer SL, Cars O, Buchholz U, et al. A European study on the relationship between antimicrobial use and antimicrobial resistance. Emerg Infect Dis 2002;8: [15] Paterson DL. Collateral damage from cephalosporin or quinolone antibiotic therapy. Clin Infect Dis 2004;38(Suppl. 4):S [16] Peterson LR. Penicillins for treatment of pneumococcal pneumonia: does in vitro resistance really matter? Clin Infect Dis 2006;42: [17] Falagas ME, Siempos II, Bliziotis IA, Panos GZ. Impact of initial discordant treatment with beta-lactam antibiotics on clinical outcomes in adults with pneumococcal pneumonia: a systematic review. Mayo Clin Proc 2006;81: [18] Yu VL, Chiou CC, Feldman C, et al. An international prospective study of pneumococcal bacteremia: correlation with in vitro resistance, antibiotics administered, and clinical outcome. Clin Infect Dis 2003;37: [19] Mills GD, Oehley MR, Arrol B. Effectiveness of beta lactam antibiotics compared with antibiotics active against atypical pathogens in non-severe community acquired pneumonia: meta-analysis. BMJ 2005;330:456. [20] Shefet D, Robenshtok E, Paul M, Leibovici L. Empirical atypical coverage for inpatients with community-acquired pneumonia: systematic review of randomized controlled trials. Arch Intern Med 2005;165: [21] Fang GD, Fine M, Orloff J, et al. New and emerging etiologies for community-acquired pneumonia with implications for therapy: a prospective multicenter study of 359 cases. Medicine (Baltimore) 1990;69: [22] Holmberg H, Bodin L, Jonsson I, Krook A. Rapid aetiological diagnosis of pneumonia based on routine laboratory features. Scand J Infect Dis 1990;22: [23] Farr BM, Kaiser DL, Harrison BD, Connolly CK. Prediction of microbial aetiology at admission to hospital for pneumonia from the presenting clinical features: British Thoracic Society Pneumonia Research Subcommittee. Thorax 1989;44: [24] Woodhead MA, Macfarlane JT. Comparative clinical and laboratory features of Legionella with pneumococcal and mycoplasma pneumonias. Br J Dis Chest 1987;81: [25] Mulazimoglu L, Yu VL. Can Legionnaires disease be diagnosed by clinical criteria? A critical review. Chest 2001;120: [26] Strålin K, Holmberg H. Usefulness of the Streptococcus pneumoniae urinary antigen test in the treatment of community-acquired pneumonia. Clin Infect Dis 2005;41: [27] Marston BJ, Plouffe JF, File Jr TM, et al.; The Community-Based Pneumonia Incidence Study Group. Incidence of community-acquired pneumonia requiring hospitalisation. Results of a population-based active surveillance study in Ohio. Arch Intern Med 1997;157: [28] Strålin K, Törnqvist E, Kaltoft MS, Olcén P, Holmberg H. Etiologic diagnosis of adult bacterial pneumonia by culture and PCR applied to respiratory tract samples. J Clin Microbiol 2006;44: [29] Scott JA, Hall AJ. The value and complications of percutaneous transthoracic lung aspiration for the etiologic diagnosis of communityacquired pneumonia. Chest 1999;116: [30] Meehan TP, Fine MJ, Krumholz HM, et al. Quality of care, process, and outcomes in elderly patients with pneumonia. JAMA 1997;278: [31] Rome L, Murali G, Lippmann M. Nonresolving pneumonia and mimics of pneumonia. Med Clin North Am 2001;85, , xi. [32] Murray PR, Washington JA. Microscopic and baceriologic analysis of expectorated sputum. Mayo Clin Proc 1975;50: [33] Kalin M, Lindberg AA, Tunevall G. Etiological diagnosis of bacterial pneumonia by gram stain and quantitative culture of expectorates: leukocytes or alveolar macrophages as indicators of sample representativity. Scand J Infect Dis 1983;15: [34] Kalin M, Lindberg AA. Diagnosis of pneumococcal pneumonia: a comparison between microscopic examination of expectorate, antigen detection and cultural procedures. Scand J Infect Dis 1983;15: [35] Pirtle JK, Monroe PW, Smalley TK, Mohr JA, Rhoades ER. Diagnostic and therapeutic advantages of serial quantitative cultures of fresh sputum in acute bacterial pneumonia. Am Rev Respir Dis 1969;100: [36] Gillet Y, Issartel B, Vanhems P, et al. Association between Staphylococcus aureus strains carrying gene for Panton-Valentine leukocidin and highly lethal necrotising pneumonia in young immunocompetent patients. Lancet 2002;359: [37] Örtqvist Å. Commentary to Community-acquired pneumonia due to Escherichia coli by Marrie TJ et al. Clin Microbiol Infect 1998;4: [38] Murdoch DR. Diagnosis of Legionella infection. Clin Infect Dis 2003;36:64 9. [39] Fishman JA, Roth RS, Zanzot E, Enos EJ, Ferraro MJ. Use of induced sputum specimens for microbiologic diagnosis of infections due to organisms other than Pneumocystis carinii. J Clin Microbiol 1994;32: [40] Cook DJ, Fitzgerald JM, Guyatt GH, Walter S. Evaluation of the protected brush catheter and bronchoalveolar lavage in the diagnosis of nosocomial pneumonia. J Intensive Care Med 1991;6: [41] Kirkpatrick MB, Bass Jr JB. Quantitative bacterial cultures of bronchoalveolar lavage fluids and protected brush catheter specimens from normal subjects. Am Rev Respir Dis 1989;139: [42] Menendez R, Cordoba J, de La Cuadra P, et al. Value of the polymerase chain reaction assay in noninvasive respiratory samples for diagnosis of community-acquired pneumonia. Am J Respir Crit Care Med 1999;159: [43] Hedlund J, Örtqvist Å, Kalin M. Nasopharyngeal culture in the pneumonia diagnosis. Infection 1990;18:283 5.

9 K. Strålin / International Journal of Antimicrobial Agents 31 (2008) [44] Strålin K. Diagnostic methods for bacterial etiology in adult community-acquired pneumonia (PhD thesis). Linköping: Linköping University, [45] Fransen H, Wolontis S. Infections with viruses, Mycoplasma pneumoniae and bacteria in acute respiratory illness. A study of hospitalized patients, patients treated at home, and healthy subjects. Scand J Infect Dis 1969;1:31 7. [46] Holmberg H. Aetiology of community-acquired pneumonia in hospital treated patients. Scand J Infect Dis 1987;19: [47] Gunnarsson RK, Holm SE, Söderström M. The prevalence of potential pathogenic bacteria in nasopharyngeal samples from healthy children and adults. Scand J Prim Health Care 1998;16:13 7. [48] Cooper GM, Jones JJ, Arbique JC, Flowerdew GJ, Forward KR. Intra and inter technologist variability in the quality assessment of respiratory tract specimens. Diagn Microbiol Infect Dis 2000;37: [49] Fine MJ, Orloff JJ, Rihs JD, et al. Evaluation of housestaff physicians preparation and interpretation of sputum Gram stains for communityacquired pneumonia. J Gen Intern Med 1991;6: [50] Reed WW, Byrd GS, Gates Jr RH, Howard RS, Weaver MJ. Sputum gram s stain in community-acquired pneumococcal pneumonia A metaanalysis. West J Med 1996;165: [51] Musher DM, Montoya R, Wanahita A. Diagnostic value of microscopic examination of gram-stained sputum and sputum cultures in patients with bacteremic pneumococcal pneumonia. Clin Infect Dis 2004;39: [52] Gleckman R, DeVita J, Hibert D, Pelletier C, Martin R. Sputum gram stain assessment in community-acquired bacteremic pneumonia. J Clin Microbiol 1988;26: [53] Boerner DF, Zwadyk P. The value of the sputum gram s stain in community-acquired pneumonia. JAMA 1982;247: [54] Roson B, Carratala J, Verdaguer R, Dorca J, Manresa F, Gudiol F. Prospective study of the usefulness of sputum Gram stain in the initial approach to community-acquired pneumonia requiring hospitalization. Clin Infect Dis 2000;31: [55] Dominguez J, Gali N, Blanco S, et al. Detection of Streptococcus pneumoniae antigen by a rapid immunochromatographic assay in urine samples. Chest 2001;119: [56] Smith MD, Derrington P, Evans R, et al. Rapid diagnosis of bacteremic pneumococcal infections in adults by using the Binax NOW Streptococcus pneumoniae urinary antigen test: a prospective, controlled clinical evaluation. J Clin Microbiol 2003;41: [57] Strålin K, Kaltoft MS, Konradsen HB, Olcén P, Holmberg H. Comparison of two urinary antigen tests for establishment of pneumococcal etiology of adult community-acquired pneumonia. J Clin Microbiol 2004;42: [58] Marcos MA, Jimenez de Anta MT, de la Bellacasa JP, et al. Rapid urinary antigen test for diagnosis of pneumococcal community-acquired pneumonia in adults. Eur Respir J 2003;21: [59] Palmu A, Kaijalainen T, Leinonen M, Mäkelä PH, Kilpi T. Nasopharyngeal carriage and pneumococcal urine antigen test in healthy elderly subjects. In: 4th International Symposium on Pneumococci and Pneumococcal Disease [60] Murdoch DR, Laing RT, Cook JM. The NOW S. pneumoniae urinary antigen test positivity rate 6 weeks after pneumonia onset and among patients with COPD. Clin Infect Dis 2003;37: [61] Guchev IA, Yu VL, Sinopalnikov A, Klochkov OI, Kozlov RS, Stratchounski LS. Management of nonsevere pneumonia in military trainees with the urinary antigen test for Streptococcus pneumoniae: an innovative approach to targeted therapy. Clin Infect Dis 2005;40: [62] Lettinga KD, Verbon A, Weverling GJ, et al. Legionnaires disease at a Dutch flower show: prognostic factors and impact of therapy. Emerg Infect Dis 2002;8: [63] Murdoch DR. Molecular genetic methods in the diagnosis of lower respiratory tract infections. APMIS 2004;112: [64] Gnarpe J, Lundback A, Sundelof B, Gnarpe H. Prevalence of Mycoplasma pneumoniae in subjectively healthy individuals. Scand J Infect Dis 1992;24: [65] Hyman CL, Roblin PM, Gaydos CA, Quinn TC, Schachter J, Hammerschlag MR. Prevalence of asymptomatic nasopharyngeal carriage of Chlamydia pneumoniae in subjectively healthy adults: assessment by polymerase chain reaction-enzyme immunoassay and culture. Clin Infect Dis 1995;20: [66] Foy HM, Grayston JT, Kenny GE, Alexander ER, McMahan R. Epidemiology of Mycoplasma pneumoniae infection in families. JAMA 1966;197: [67] Hammerschlag MR, Chirgwin K, Roblin PM, et al. Persistent infection with Chlamydia pneumoniae following acute respiratory illness. Clin Infect Dis 1992;14: [68] Strålin K, Korsgaard J, Olcén P. Evaluation of a multiplex PCR for bacterial pathogens applied to bronchoalveolar lavage. Eur Respir J 2006;28: [69] Örtqvist Å, Kalin M, Lejdeborn L, Lundberg B. Diagnostic fiberoptic bronchoscopy and protected brush culture in patients with communityacquired pneumonia. Chest 1990;97: [70] Peters RP, van Agtmael MA, Danner SA, Savelkoul PH, Vandenbroucke-Grauls CM. New developments in the diagnosis of bloodstream infections. Lancet Infect Dis 2004;4: [71] Petric M, Comanor L, Petti CA. Role of the laboratory in diagnosis of influenza during seasonal epidemics and potential pandemics. J Infect Dis 2006;194(Suppl. 2):S [72] Hurt AC, Alexander R, Hibbert J, Deed N, Barr IG. Performance of six influenza rapid tests in detecting human influenza in clinical specimens. J Clin Virol 2007;39: [73] Cunha BA. Influenza: historical aspects of epidemics and pandemics. Infect Dis Clin North Am 2004;18: [74] Falsey AR, Murata Y, Walsh EE. Impact of rapid diagnosis on management of adults hospitalized with influenza. Arch Intern Med 2007;167: [75] Oosterheert JJ, van Loon AM, Schuurman R, et al. Impact of rapid detection of viral and atypical bacterial pathogens by real-time polymerase chain reaction for patients with lower respiratory tract infection. Clin Infect Dis 2005;41: [76] Daxboeck F, Krause R, Wenisch C. Laboratory diagnosis of Mycoplasma pneumoniae infection. Clin Microbiol Infect 2003;9: [77] Dowell SF, Peeling RW, Boman J, et al. Standardizing Chlamydia pneumoniae assays: recommendations from the Centers for Disease Control and Prevention (USA) and the Laboratory Centre for Disease Control (Canada). Clin Infect Dis 2001;33: [78] Ewig S, Bauer T, Hasper E, Marklein G, Kubini R, Luderitz B. Value of routine microbial investigation in community-acquired pneumonia treated in a tertiary care center. Respiration 1996;63: [79] Rello J, Bodi M, Mariscal D, et al. Microbiological testing and outcome of patients with severe community-acquired pneumonia. Chest 2003;123: [80] Bartlett JG. Community-acquired pneumonia. N Engl J Med 1996;334:863. [81] Bauer TT, Welte T, Ernen C, et al. Cost analyses of communityacquired pneumonia from the hospital perspective. Chest 2005;128: [82] Wheeler J, Murphy OM, Freeman R, Kearns AM, Steward M, Lee MJ. PCR can add to detection of pneumococcal disease in pneumonic patients receiving antibiotics at admission. J Clin Microbiol 2000;38:3907.