Add Legionella to Your Differential Diagnosis for Pneumonia

Paul G. Auwaerter, MD


August 14, 2017

Hello. This is Paul Auwaerter with Medscape Infectious Diseases, and professor of medicine at the Johns Hopkins University School of Medicine. It has been more than 40 years since the outbreak in Philadelphia that has since been described as Legionnaires disease. As an infectious diseases physician, I feel like I'm still learning about Legionella as a cause of pneumonia.

Some have said that Legionella is perhaps the most typical of atypical pneumonias, meaning that you often see purulent sputum, lobar consolidation is not unusual, and patients may have pleural effusions. There is significant mortality much like we see in pneumococcal disease, with rates of about 9%.[1] Tip-offs for Legionella pneumonia include purulent sputum with no organisms, neurologic complaints (eg, headache or ataxia), gastrointestinal difficulties (eg, nausea, vomiting, diarrhea), and hyponatremia.

I think most ID physicians are quite familiar with whom to be worried about: older patients over 60, smokers, those with chronic obstructive pulmonary disease, and the immunocompromised. Of course, people can be exposed to this environmental waterborne pathogen—which can be aerosolized—when traveling and staying in a new hotel or residence, attending fairs, and so on.

The Centers for Disease Control and Prevention (CDC)[2] reported on over 27 outbreaks from 2000 to 2014. It was interesting to me that one third of these outbreaks were associated with healthcare facilities. In these hospital systems, every one of them had some deficit in water-management control and prevention, and 48% had two or more [issues]. Eighty-five percent of deaths from these outbreaks occurred in the hospital setting. A recent CDC report[3] found a case fatality rate of 25% for definite healthcare-associated cases, contrasting to the 9% overall. Of course, this is probably because patients in hospitals are more ill. Certainly, some vigilance with water systems could lead to prevention of these cases.

Many community-acquired and even hospital-acquired pneumonias are without a clear-cut diagnosis. I wonder if, indeed, we are not looking as carefully as we should for Legionella because it does not always jump up on the list [of pathogens] for hospital-acquired pneumonia. It is something to think about, especially in the more ill patients with purulent sputum and no other obvious diagnosis.

For diagnosis of Legionella pneumophila serotype 1, the urinary antigen works quite well, with 70%-80% sensitivity and high specificity.[4] For other Legionella species, such as L micdadei or other L pneumophila serogroups, culture with buffered charcoal yeast extract remains the gold standard, and this has to be ordered specifically. Of course, many laboratories have difficulty growing this pathogen, so it is something that we may not always be detecting quite as carefully. Polymerase chain reaction (PCR)[5] and molecular techniques[6] have been helpful in understanding the panoply of microorganisms capable of causing respiratory infections, but they have yet to be validated or US Food and Drug Administration–approved for testing, and so we use home-brewed tests in our institutions or tests offered by commercial labs. How well PCR performs, we really do not know. One recent systematic review[5] suggests that it might have about an 80% sensitivity—again, only for L pneumophila.

Overall, I have begun to think about Legionella a bit more carefully when I'm evaluating patients in the hospital for pneumonia, and I may consider ordering urine antigen and/or cultures. Of course, complex that with the decision of whether to cover for an atypical pneumonia such as Legionella with a macrolide fluoroquinolone.

Prevention in hospital water systems is no doubt one of the key attributes and continues to be something emphasized by the CDC. Because this is typically a spring/summer and early fall pathogen, I thought this might be a timely report.

Thank you so much for listening. This is Paul Auwaerter speaking.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.