Show Summary Details
Page of

Diagnosis and management of atypical pneumonia 

Diagnosis and management of atypical pneumonia
Diagnosis and management of atypical pneumonia

Martin Langer

and Edoardo Carretto

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2020. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 30 November 2020

Key points

  • ‘Atypical pneumonia’ is not rare in the critically-ill patient and ‘atypical micro-organisms’ are intrinsically resistant to β‎-lactams. Specific empiric antimicrobial treatment (macrolides, newer fluorquinolones) is mandatory.

  • ‘Atypical’ pneumonia belongs mainly to the community-acquired pneumonia (CAP) and consequently CAP-guidelines include diagnosis and treatment of the ‘atypical’ pneumoniae.

  • Hospital-acquired ‘atypical’ pneumonia are almost exclusively caused by L. pneumophila and may occur in immunocompromised patients with cancer, organ transplant, or undergoing immunosuppressive treatment, who are exposed to contaminated water in the hospital.

  • For L. pneumophila infections, the urinary antigen test is a rapid assay, easy to perform, and widely available. Its major limitation is the specificity only for serogroup 1 (>70%, but <100%). Since the isolation of L. pneumophila from respiratory tract is still considered the diagnostic gold standard in microbiology, an attempt to culture the micro-organism should be made whenever is possible.

  • Evaluation of serological response (IgM and IgG) in both, acute, and convalescent (3–4 weeks) sera, permit the diagnosis of M. pneumoniae and C. pneumoniae. In the near future, it is expected that Nucleic Acid Amplification tests (NAAT) techniques will represent the future key points for the diagnosis.


The term ‘atypical pneumonia’ is probably obsolete and also questionable [1]‌, but still widely used, and this chapter refers to pneumonia caused by Legionella pneumophila, Mycoplasma pneumoniae, and Chlamydophila pneumoniae [2]. ‘Atypical’ pneumonia belongs mainly to the type community-acquired pneumonia (CAP) occurring both in immunocompromised and in immunocompetent patients, and consequently all CAP-guidelines extensively discuss clinical features, diagnosis, and treatment of the so-called atypical pneumoniae. Hospital-acquired (mainly not intensive care unit (ICU)-acquired) ‘atypical’ pneumonia are almost exclusively caused by L. pneumophila and may occur in immunocompromised patients with cancer, organ transplant, or undergoing immunosuppressive treatment, who are exposed to contaminated water in the hospital [3].

As shown in Table 118.1 and Fig. 118.1 the definition of ‘atypical pneumonia’ is not inclusive of viral pneumonia, and attempts to identify viruses that are of particular importance for epidemiological purposes [4,5,6].

Table 118.1 The most frequently isolated ‘typical’ and ‘atypical’ micro-organisms in CAP patients with increasing severity


‘Typical’ pathogens

‘Atypical’ pathogens

Other pathogens


Streptococcus pneumoniae, Haemophilus influenzae, Moraxella spp.

Mycoplasma pneumoniae, Chlamydophila pneumoniae, Legionella pneumophila

Respiratory viruses: influenza A-B, adenovirus, respiratory syncytial, para-influenza, H1N1

Inpatient: non-ICU

Streptococcus pneumoniae, Haemophilus influenzae, Moraxella spp., Gram-negative bacilli

Legionella pneumophila, Mycoplasma pneumoniae, Chlamydophila pneumoniae

  • Respiratory viruses: influenza A-B, adenovirus, respiratory syncytial, para-influenza, H1N1

  • Aspiration

Inpatient: ICU

Streptococcus pneumoniae, Staphylococcus aureus, Haemophilus influenzae, Gram-negative bacilli

Legionella pneumophila (Mycoplasma pneumoniae)

  • Respiratory viruses (H1N1): Aspergillus spp., Pneumocistis jiroveci

  • Aspiration

Adapted from Mandell LA et al., ‘Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults’, Clinical Infectious Diseases, 2007, 44(Suppl. 2), pp. S27–72, by permission of Infectious Diseases Society of America. Data from File TM, ‘Community-acquired pneumonia’, Lancet, 2003, 362, pp. 1991–2001.

Fig. 118.1 Flow chart—‘atypical pneumonia’—data from [5] and [6]. B. A. Cunha extensively investigated ‘atypical pneumonia’, and his work is helpful to understand similarities and differences with respect to ‘typical’ CAP.

Fig. 118.1 Flow chart—‘atypical pneumonia’—data from [5] and [6]. B. A. Cunha extensively investigated ‘atypical pneumonia’, and his work is helpful to understand similarities and differences with respect to ‘typical’ CAP.

* For psittacosis, Q fever, and tularaemia see specific literature.

Data from Cunha BA, 'Clinical features of legionnaires' disease', Seminars in Respiratory Infections, 1998, 13, pp. 116–27; and Cunha BA, 'The atypical pneumonias: clinical diagnosis and importance', Clinical Microbiology and Infection, 2006, 12(Suppl. 3), pp. 12–24.

The British Thoracic Society reports prevalence data for patients admitted to the ICUs with CAP: 17.8% of infections are due to Legionella pneumophila, and 2.7% and 2.2% to Mycoplasma and Chlamydophila, respectively [7]‌. In the Italian GiViTI—ICU surveillance project from 2009, Legionella pneumophila was identified in 4.6% of patients admitted for severe CAP. Legionellosis must therefore be considered in all patients admitted to the ICU with severe CAP. Mycoplasma pneumoniae and Chlamydophila pneumoniae causes infrequently severe illness in immunocompetent adults, but are more often diagnosed in children, the elderly, or immunocompromised hosts.

Clinical diagnosis

The clinical diagnosis of ‘atypical pneumonia’ in patients admitted to the ICU because of severe sepsis or respiratory failure is difficult and there are no specific clinical or radiological findings. ‘Atypical pneumonia,’ more than just a pneumonia, is a syndrome with pulmonary and extrapulmonary manifestations. Pneumonia with concomitant non-specific central nervous system (CNS) abnormalities, such as headache, mental confusion, lethargy, renal, and hepatic involvement (microscopic haematuria, increased creatinine, transaminases, and increased lactate dehydrogenase), cardiac abnormalities (relative bradycardia), diarrhoea, electrolyte abnormalities (hyponatraemia, hypophosphataemia), but also absence of sputum production and ‘no growth’ from standard cultures may suggest ‘atypical’ micro-organisms, and request adequate and urgent empirical treatment and diagnostic work-up. In an observational study, legionellosis, probably because not suspected/treated in a timely manner, was independently associated with early treatment failure and this increases the risk of death [8]‌.

Radiology, albeit not specific, is a cornerstone of the diagnosis of ‘atypical pneumonia’ in patients admitted to the ICU. Severe CAP due to Legionella spp. usually cause patchy, localized infiltrates in the lower lobes, being frequently multilobar and even bilateral. In rare instances, legionellosis is associated with cavitations, more frequently in patients treated with steroids. Clinical history and evidence of contact with animals are important elements that support the suspicion of ‘atypical’ pneumonia due to infrequent diseases such as Q fever, psittacosis, and tularaemia, which are hardly ever aetiologies seen outside a specific context.

Knowledge of the diseases and clinical suspicion are important, but not sufficient for diagnosis. Only specific microbiological investigations allow a reliable diagnosis, which is useful for the obligatory reporting and of great value for epidemiological purposes. A reliable microbiological diagnosis permit the tailoring of the type and duration of the treatment, even in cases where an adequate empiric, guideline-orientated treatment had been prescribed.

Microbiological diagnosis

Legionella infections

From the microbiological point of view, to date more than 50 Legionella species have been classified. Legionella pneumophila was the first species described and it causes 80–90% of reported cases of legionellosis. L. pneumophila is divided into 16 different serogroups, but serogroup 1 accounts for more than 70% of all of the legionellosis [9,10].

The diagnostic tools of Legionella infections have been focused on culture and serological investigation. A significant improvement was achieved using methods able to detect the urinary antigens of the micro-organism. Recently, nucleic acid amplification tests (NAATs) revealing the DNA of Legionella strains from environmental and clinical samples have been made available.

The urinary antigen test is based on the qualitative detection of a soluble antigen specific for Legionella pneumophila serogroup 1 [11]‌, revealed through a rapid immunochromatographic assay. This antigen has been detected in urine as early as 3 days after the onset of symptoms and it may persist for months [12]. These tests provide quick results (15 minutes) using a common urine sample, and detect the antigen in early, as well as later, stages of disease, and antimicrobial treatment does not influence the result of the test. In infections due to L. pneumophila serogroup 1 the sensitivity of the method varies from 60 to 100%, with a specificity of more than 99% [13]. The major disadvantage of this test is its inability to diagnose infections caused by serogroups 2–16. Furthermore, the prolonged persistence of the antigen in the urinary samples may mislead the diagnosis if the patient had an asymptomatic Legionella infection, or a Pontiac fever, in the months before the episode of pneumonia that is to be investigated.

Isolation of L. pneumophila from respiratory specimens (e.g. sputum, tracheal aspirate, broncho-alveolar lavage, etc.) using selective media is considered to be the diagnostic gold standard. However, this kind of culture has to be specifically requested by the laboratory, since specific media are necessary, which are not routinely used. The culture is also time-consuming, because Legionella spp. take 3–7 days to grow. However, although the sensitivity of this method varies from laboratory to laboratory, an attempt to culture the micro-organism should be made whenever is possible. Through cultural methods it is possible to diagnose infections caused by Legionella pneumophila (serogroups different from 1) or by Legionella spp. different from Legionella pneumophila (most frequently involving Legionella micdadei, now reclassified as Tatlockia micdadei). Moreover, the isolation of the strain will allow epidemiological studies or could be required for comparison, with molecular techniques, of the isolate to environmental strains if necessary. If a patient is positive to the urinary antigen, allowing epidemiological studies, the clinician should immediately collect a respiratory sample, trying to isolate and identify the micro-organism. Even if there are no data about the decreased sensitivity of the culture in patients treated with effective antibiotics, it can be postulated that every day of effective therapy decreases the likelihood of isolating the micro-organism.

As far as serological analysis is concerned, different techniques are available and should be performed on paired sera (acute and convalescent). The most commonly used techniques are enzyme immunoassays (EIAs), even if detection of antibodies with indirect fluorescence has better performances, but they are time-consuming and require skilled personnel. The main disadvantage of serology is that the host response is typically slow, and sometimes specific antibodies can be detected only months after the onset of symptoms. Thus, serological testing is useful for epidemiological purposes, but has little or no impact for the diagnosis and treatment of the early stage of the disease [14].

Recently, nucleic acid-based methods have been applied to the detection of Legionella species, both in clinical and in environmental samples. The polymerase chain reaction (PCR) techniques, if used in combination with the urinary-antigen test, increase the diagnostic chances in the early stage of the disease compared with the use of each test alone [15]. These tests have been successfully used in reference and research laboratories, although most of these experiences were with ‘home-made’ techniques. At the time of writing, only one out of the three commercial available assays is FDA-cleared. Although these methods have high capability for the future, they should be used only in reference settings and in selected cases, such as suspected infections in patients exposed to L. pneumophila strains different from serogroup 1 (e.g. for environmental exposure).

Mycoplasma pneumoniae and Chlamydophila pneumoniae

The aetiological diagnosis of these two entities implies common problems. The diagnosis of both pathogens is based on serological tests and on NAATs. Culture is very difficult and time-consuming, and can be performed only in reference laboratories.

For Mycoplasma infections, serological tests should be performed on paired sera (acute and convalescent), using complement fixation test and EIAs. However, this approach is useful only for epidemiological purposes. The diagnosis of acute infections can be achieved by evaluating the IgM antibodies even if they are lacking in children under 6 months of age, in some cases of primary infections and during re-infections. For M. pneumoniae, IgMs are produced 3–4 days or later after the onset of symptoms, and sometimes persist for several weeks to months.

Regarding serological tests for C. pneumoniae, determined in paired sera, the micro-immunofluorescence (MIF) test is able to document infections and is useful for epidemiological purposes and is currently considered the gold standard.

Both for Mycoplasma pneumoniae and Chlamydophila pneumoniae, PCR systems have been developed and used for diagnosis. For these micro-organisms, comparison of NAATs with culture and/or serology is difficult, since there are problems in defining a reliable gold standard. To date, there are no commercial FDA-cleared assays, although a wide variety of PCR-based protocols, using different target genes, have been developed in research laboratories. These techniques, and in particular real-time PCR, will represent the future key points for the diagnosis [16].

What should clinicians do for the diagnosis?

If an ‘atypical pneumonia’ is suspected, clinicians should:

  • Collect a urinary sample for detection of Legionella pneumophila serogroup 1 antigen.

  • If legionellosis is strongly suspected, or if the urinary antigen is positive, appropriate respiratory samples should be collected as soon as possible and sent to the laboratory, with a previous communication to the microbiologist.

  • Collect acute and convalescent (3–4 weeks) sera for IgM and IgG serology for L. pneumophila, M. pneumoniae, and C. pneumoniae.

  • Contact the laboratory to discover whether NAATs are available for the diagnosis and also to find the correct way to collect samples to be used with these techniques.

Transmission within the ICU and isolation practice

While hospital acquisition of legionellosis from contaminated water in hospital is of great concern, and may present in clusters or outbreaks, a person-to-person transmission has never been reported [17]. Isolation of Legionella-infected patients is therefore unnecessary.


Starting from the well-known unresponsiveness of ‘atypical’ bacteria to β‎-lactam antibiotics and the impossibility to perform ‘standard’ sensitivity testing, the cornerstone of treatment is empirical therapy with macrolides or newer fluorquinolones. In the absence of adequate clinical studies, all available guidelines are based on in vitro investigations, observational studies and expert opinions.

Every CAP-guideline includes ‘anti-atypical’ drugs as their first line CAP treatment, mainly a combination of a respiratory fluorquinolone or a macrolide (with a β‎-lactam). In many severe and poorly responsive cases diagnosed as legionellosis, combination therapy, adding rifampicin to the fluorquinolone has been tried, but the efficacy is still controversial [3,18]. From 2 to 3 weeks of treatment are usually considered adequate, but this information too is not ‘evidence based.’

Once the microbiological diagnosis becomes available, the treatment of ‘atypical’ pneumonia may also be targeted according to the pathogen [19]:

  • Mycoplasma pneumoniae: doxycycline, macrolide, newer fluoroquinolones (most data available for levofloxacin).

  • Chlamydophila pneumoniae: doxycycline, macrolide, fluoroquinolones.

  • Legionella spp.: newer fluoroquinolones (most data available for levofloxacin), macrolide (azitromycine preferred) +/– rifampicin.


1. Murdoch DR and Chambers ST. (2009). Atypical pneumonia—time to breathe new life into a useful term? Lancet—Infectious Diseases, 9, 512–19.Find this resource:

2. Plouffe JF. (2000). Importance of atypical pathogens of community-acquired pneumonia. Clinical Infectious Diseases, 31(Suppl. 2), S35–9.Find this resource:

3. Carratala J and Garcia-Vidal C. (2010). An update on Legionella. Current Opinions in Infectious Diseases, 23, 152–7.Find this resource:

4. Mandell LA, Wunderink RG, Anzueto A, et al. (2007). Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clinical Infectious Diseases, 44(Suppl. 2), S27–72.Find this resource:

5. Cunha BA. (1998). Clinical features of legionnaires’ disease. Seminars in Respiratory Infections, 13, 116–27.Find this resource:

6. Cunha BA. (2006). The atypical pneumonias: clinical diagnosis and importance. Clinical Microbiology and Infection, 12(Suppl. 3), 12–24.Find this resource:

7. Lim WS, Baudouin SV, George RC, et al. (2009). British Thoracic Society guidelines for the management of community acquired pneumonia in adults: update 2009. Thorax, 64(Suppl III), iii1–iii55.Find this resource:

8. Roson B, Carratala J, Fernandez-Sabe N, Tubau F, Manresa F, and Gudiol F. (2004). Causes and factors associated with early failure in hospitalized patients with community-acquired pneumonia. Archives of Internal Medicine, 164, 502–8.Find this resource:

9. Yu VL, Plouffe JF, Pastoris MC, et al. (2002). Distribution of Legionella species and serogroups isolated by culture in patients with sporadic community-acquired legionellosis: an international collaborative survey. Journal of Infectious Diseases, 186, 127–8.Find this resource:

10. Stout JE and Yu VL. (1997). Legionellosis. New England Journal of Medicine, 337, 682–7.Find this resource:

11. Berdal BP, Farshy CE, and Feeley JC. (1979). Detection of Legionella pneumophila antigen in urine by enzyme-linked immunospecific assay. Journal of Clinical Microbiology, 9, 575–8.Find this resource:

12. Kohler RB, Winn WC, Jr, and Wheat LJ. (1984). Onset and duration of urinary antigen excretion in Legionnaires disease. Journal of Clinical Microbiology, 20, 605–7.Find this resource:

13. Tronel H and Hartemann P. (2009). Overview of diagnostic and detection methods for legionellosis and Legionella spp. Letters in Applied Microbiology, 48, 653–6.Find this resource:

14. Den Boer JW and Yzerman EP. (2004). Diagnosis of Legionella infection in Legionnaires’ disease. European Journal of Clinical Microbiology & Infectious Diseases, 23, 871–8.Find this resource:

15. Diederen BM. (2008). Legionella spp. and Legionnaires’ disease. Journal of Infection, 56, 1–12.Find this resource:

16. Diaz MH and Winchell JM. (2012). Detection of Mycoplasma pneumoniae and Chlamydophila pneumoniae directly from respiratory clinical specimens using a rapid real-time polymerase chain reaction assay. Diagnostic Microbiology and Infectious Disease, 73, 278–80.Find this resource:

17. Newton HJ, Ang DK, van Driel IR, and Hartland EL. (2010). Molecular pathogenesis of infections caused by Legionella pneumophila. Clinical Microbiology Reviews, 23, 274–98.Find this resource:

18. Pedro-Botet ML, Garcia-Cruz A, Tural C, et al. (2006). Severe Legionnaires’ disease successfully treated with levofloxacin and azithromycin. Journal of Chemotherapy, 18, 559–61.Find this resource:

19. Woodhead M, Blasi F, Ewig S, et al. (2011). Guidelines for the management of adult lower respiratory tract infections—full version. Clinical Microbiology and Infection, 17(Suppl. 6), E1–59.Find this resource: