Return to Frequently Asked Questions

FAQ FOR PHYSICIANS AND HEALTHCARE WORKERS

What are your recommendations concerning thermal springs and hot baths for immunocompromised patients or transplant receipts.

       The risk is low, but not zero. The patient should avoid breathing in vapors. Also, we recommend that transplant patients and highly immunosuppressed patients not drink tap water. Instead, boil water and cool it for drinking.  When traveling or not at home, bottled water might be preferred.

 

Can Legionella cause endocarditis?  Are there documented cases that legionella bacteria have migrated to the heart causing vegetations on the heart valves and requiring valve replacement?

        Yes, numerous cases of hospital-acquired legionella endocarditis have been described – the largest series is from Stanford University. The source was the drinking water of the hospital.  Legionella is transmitted to the heart in 2 ways: bacteremia from hospital-acquired pneumonia or possibly contiguous spread from the use of contaminated water post-cardiac surgery. If you have such a case, it may be reportable as a case report in the peer-review literature. The water supply of the hospital must be cultured for legionella, and ideally, the results of Legionella serology and urinary antigen should be available for the patient.  Pericarditis has also been reported

 

We had a patient test positive by urine antigen in January who is now testing positive again 4 months later- if recurrence is rare, is there a specific time period that the test will convert to negative following the disease.  Or, does a persistently positive test imply that Legionella infection still persists? Or that it wasn't completely eradicated from her system?

        Prolonged urinary antigen excretion has been reported. Most of the patients that we have evaluated converted to urinary antigen negative between 30-60 days. Factors related to prolonged excretion have included severity of illness and immunosuppression.  This was recently reviewed by - Sopena N, Sabria M, et al.  Factors related to

persistence of Legionella urinary antigen excretion in patients with Legionnaires' disease.  Eur. J. Clin. Microbiol. Infect. Dis. 2002. 21:845-848.

  

 

I am the patient/nurse who had Legionnaires' disease and was waiting for results from a hot tub and shower.  I had a positive urine antigen.  The hot tub was culture positive but the showerhead was culture negative. We are still waiting to see if the culture yields L. pneumophila type 1.  The person who stayed after me had Pontiac Fever according to the epidemiologist. Question: if he had symptoms several weeks ago, are there any laboratory tests that may confirm this diagnosis.  The epidemiologist thought it was too late to test him.         

        The antibody serology (blood test) can remain positive for 3 – 6 months (or longer in some cases) after the infection.

 

Legionella is in the water supply of my long-term care facility

... A series of questions from a long-term care facility...

Recently a resident of our long-term care facility was found to have Legionella pneumonia. Tests performed by the health department revealed that Legionella is in our water supply. Is Legionella often found in long-term care facilities?

        Yes. It is present in the drinking water of many long term care facilities.  No formal survey has yet been performed – but 20-70% of hospitals harbor Legionella.  In addition, numerous cases of Legionnaires’ disease have been reported from long-term care facilities and nursing homes (Fang GD, Med 90; Marrie TJ 86, Loeb M 99, Maesaki 92, Brennen C 87).  We are now recommending that all such facilities culture their water supply (Seenivasan M, Journal of American Geriatrics Society, 53:875-880, 2005).

 

We have had 4 other cases of pneumonia in which urine antigen tests have been negative- however, two of these patients did not really improve clinically until they were put on Zithromax for their antibiotic therapy.  I also raised the question of aspiration and was told this is not a problem, however you indicate that it is.  Now, I am flushing the nasogastric tube of my patient with sterile water until we have treated our water supply

         We agree with your approach.  Aspiration has been shown to be a major mode of transmission for Legionnaires’ disease and studies in long-term care facilities show that aspiration occurring from nasogastric tubes, indeed, is a risk factor for aspiration (Loeb MJ 02, Am Geriatrics 99; Seenivasan M, in press). 

 

He has already had documented aspiration pneumonia 2 times and is one of the patients that did not improve until Zithromax was given.  If his urine antigen test was negative can we feel confident that he did not have Legionella?

        No.  Although the sensitivity is about 85%-90% for Legionella pneumophila, serogroup 1, it will not detect other Legionella spp or serogroups.  Azithromycin (Zithromax) is effective therapy.  Levofloxacin (or another quinolone) might also be used empirically for nursing home pneumonia of uncertain etiology if the water supply harbors Legionella.  Quinolones also cover other common pathogens in nursing home pneumonia including gram-negative rods.

 

We have documented colonization with Legionella in the hospital water supply of our small hospital.  We have not observed cases of Legionnaires’ disease at this time (but underdiagnosis is probable), and after a successful trial of superheat and flush, colonization recurred.  It is technically impossible to resume that disinfection method (old distribution system).  Thus, we cannot disinfect the water supply and we cannot shut down the hospital.  Is it reasonable to withhold drinking water from our patients and to institute systematic laboratory testing for all patients with hospital acquired pneumonia in an attempt to treat with anti-Legionella antibiotics with minimal delay?

       Yes, your approach is very reasonable.  We would not recommend disinfection of your water supply since your patients are not high risk.  We would recommend that your clinical microbiology lab adopt the urinary antigen test if the Legionella in your water supply is serogroup 1. Physicians should order it for patients who have contracted hospital-acquired pneumonia.

       If your patients contract hospital-acquired pneumonia of uncertain etiology, we recommend adding a quinolone (e.g., levofloxacin) as part of empiric therapy.
        Please download the Allegheny County Guidelines on the Home Page of www.legionella.org . These Guidelines, which are simple and cost-effective, will protect your patients from Legionnaires’ disease. Your suggested approach is in concordance with the Guidelines. Knowledge of colonization can eliminate mortality without tremendous expense by adopting laboratory testing and using effective antibiotics.

 

Legionella anisa was found in a surprising number of the distal water sites in our hospital.  No cases of Legionnaires’ disease have ever been detected, although we do not have Legionella culture available as a diagnostic test.  Since the urinary antigen does not identify L. anisa, should we disinfect our water supply?

        The majority (>90%) of cases of Legionnaires' disease reported in the U.S. are caused by Legionella pneumophila.  L. anisa is included in the ”Other species" group which is less than 10% for all other Legionella species.  There are 48 named species of Legionella, with approximately half having been implicated in human disease.

         L. anisa is frequently isolated from environmental specimens but very rarely causes disease.  Disease caused by other Legionella species, like L. anisa, occurs almost exclusively in immunocompromised individuals.  Only a handful of cases attributed to L. anisa have been reported.  We consider this species as nonpathogenic (Yu, J Infect Chemother 2004, Stout, in press).  And, we would NOT disinfect your water supply.

 

Legionella has been found in our hospital and the water supply was successfully disinfected with copper-silver ionization.  Nevertheless, showers for our transplant patients have been banned.  This is a major inconvenience for patients and nursing staff.  If, as you claim, aerosolization is not the mode of transmission, why should showers be banned?  Our engineering staff also monitors showerheads by removing them and cleaning them with disinfectants although this is done on an erratic basis.        

       Showers are not important disseminators for Legionella.  Our view has credibility since Dr. Victor L. Yu was a co-author of the article published in the Annals of Internal Medicine 1981 that suggested Legionella might be transmitted via showers. Subsequent case-control studies showed our original conclusion was erroneous, although no retraction has ever been published.  Subsequent studies from Belgium, Netherlands, University of Virginia, Wadsworth VA Medical Center, University of Iowa, Lackland Air Force Base, University of Pittsburgh, and University of Arizona also showed this conclusion was erroneous.  The article by Sabria in Lancet Infectious Disease 2002 on the Home Page gives an overview of the studies.  Most (but not all) U.S. transplant centers have quietly rescinded their ban on showering and cases of Legionnaires’ disease attributed to showering have not occurred.  We agree with this policy. 

          Studies also show that disinfection of showerheads by chemicals or cleaning is ineffective long term given the fact that Legionella recolonizes the showerheads from existing biofilms in the pipes of the plumbing system.

 

Antimicrobial therapy for Legionnaires' disease

Is there any data that supports that one quinolone is more effective than another?  A physician in our hospital thinks that gatifloxacin is superior, yet I don't see it recommended as a primary regimen in The Sanford Guide to Antimicrobial Therapy 2003 edition.

          In the textbook Antimicrobial Therapy and Vaccines and the companion website, www.antimicrobe.org, we explicitly caution against the use of gatifloxacin for Legionnaires' disease. To our knowledge, there is not a single culture-confirmed case ever cured with gatifloxacin. We believe it should not have been FDA-approved without that clinical experience.

        The most potent quinolone in the intracellular model is levofloxacin. The largest clinical experience by far is with levofloxacin with an extraordinarily high rate of cure (Yu Chest 2004).  Ciprofloxacin has also been successfully used.  Moxifloxacin and gemifloxacin may be equally effective, but clinical data is minimal. 

 

In terms of efficacy – ignoring side effects -, is there any advantage in using levofloxacin ( or the newest quinolones) instead of erythromycin, once I have the etiologic diagnosis of Legionella pneumonia?

        Erythromycin is minimally active against Legionella in in vitro models. The quinolones are the most active agents and levofloxacin is the most active quinolone.  We no longer need erythromycin/rifampin now that levofloxacin is available.  Levofloxacin, moxifloxacin, gemifloxacin OR azithromycin are the current drugs of choice.  Please go to www.antimicrobe.org for more details. The textbook Antimicrobial Therapy and Vaccines, Vol I. Microbes gives a detailed assessment of antibiotic therapy for Legionella including discussion of macrolides, tetracyclines, trimethoprim-sulfamethoxazole, and imipenem.

 

Discussion on Laboratory Diagnosis for Legionnaires' disease

Is the best test the urine antigen test?

       The urinary antigen is not the most sensitive test, but it is the most widely available test.  It can detect only Legionella pneumophila, serogroup 1.  The most sensitive test is culture on selective media.  We believe every hospital which has Legionella in their drinking water should perform this test in-house. 

 

Our laboratory  currently sends specimens to a reference lab for the urinary Legionella antigen test. Some doctors in our hospitals would like to have this test done in-house to get faster results.  Is there any justification for stat testing when we can get results in 1-3 days?  Are there specific indications for doing the test or should all patients with pneumonia be tested?

        The urinary antigen should be performed on every patient with pneumonia, especially if the patient has COPD (emphysema) or is immunosuppressed. In numerous large scale surveys, Legionella is the 3rd or 4th most common cause of community acquired pneumonia that requires hospital admission and is a common cause of hospital-acquired pneumonia if Legionella is in hospital water (possibly 70% of US hospitals). It is a relatively-inexpensive rapid test and the result is available within minutes of performance of the test. It makes no sense to send it outside to a reference lab when it can be performed in-house without the necessity for special training. The mortality for Legionella is notably higher than for most causes of pneumonia. An immediate result will be life-saving for that patient.   Bring the test in-hospital - absolutely!

 

What does the below Legionella pneumophila AB (G&M) antibody test results mean?

       Legionella AB, IgG = 128

Legionella AB, IgM = <256  

       A single positive antibody (AB) test for Legionella is not considered diagnostic for Legionnaires’ disease.  The IgM and IgG titers can both be elevated in a case of Legionnaires’ disease and will return to normal (below 64-128) over a period of months.  The presence of antibodies may indicate exposure to the bacteria some time in the past.  In some communities, up to 15-20% of the general population can have an antibody level of 128, although this is uncommon.  It is for this reason that 2 blood samples are required for serologic diagnosis of Legionnaires’ disease – one at the time of onset of pneumonia (acute) and the second at a later time (12 weeks is optimal) (convalescent).  A four-fold increase in antibody titer over that time period is considered diagnostic for Legionnaires’ disease (e.g., from a titer of 1:64 to 1:256).  This shows that the antibody titer has risen markedly after onset of pneumonia.

 

What type of media does the Legionella bacterium grow on, and what conditions does it grow best in, eg: temp, atmosphere, etc

        Legionella are grown in the laboratory on media that was specially formulated to support the growth of Legionella pneumophila. This medium is buffered charcoal yeast extract (BCYE) agar, with or without the addition of antimicrobial agents. The plates are inoculated and incubated at 35-37oC in air. Plates can also be incubated with CO2, however it is not required for their growth.  We have formulated dye-containing media that is commercially available from many commercial suppliers for Legionella media (including Remel, Lenexa, KA). 

 

I have a patient in my ER as we speak/type who I suspect of having a Legionella infection. She is a 62 yr old lady who returned two weeks ago from a month stay in Sri Lanka. She has a fever which started 2 days ago.  She also developed hallucinations, diarrhea and a cough. Her CRP is 350 (N range<10). Her X-ray showed an upper right infiltrate. Because of her travel history and clinical signs I suspect a legionella infection. However her urine antigen test (Binax) is negative. What kind of serotype can be found on Sri Lanka?  PubMed shows only 1 hit (Ceylon Med J 2000) where only 1 out of 16 Legionella isolates was tested and yielded serogroup 1.

       We agree the history is suggestive of Legionnaires' disease. The urinary antigen is sensitive only for serogroup 1 which is the predominant serogroup in Hong Kong and Taiwan, so I suspect that it is also the predominant serogroup in Thailand. (But I could be wrong.)  In Australia, Legionella longbeachae is quite common which would not be detectable with the urinary antigen test.  There are species and serogroups of Legionella that are missed by the urinary antigen test as you indicate. The test is highly specific so that if it is positive, the patient has Legionnaires' disease. Your microbiology lab should attempt to culture the sputum with selective Legionella media as described on www.legionella.org.  Serology should be obtained today and later to confirm the diagnosis.  A gram stain of sputum should be done.  Legionella would present with numerous leukocytes, but no visible bacteria.

       I would certainly cover this patient with an anti-legionella antibiotic such as azithromycin or levofloxacin. And then switch when a pathogen is definitively identified.  Keep in mind Meliodosis if she has been in rural areas.

 

I would like to give you the outcome of the patient I presented.  Unfortunately a sputum sample could not be obtained and bronchooscopy was declined. We therefore preformed the Binax urinary antigen test for S. pneumoniae. To my surprise the test was positive. The patient is now doing quite well (no more fever) on penicillin.  Could you tell me please, if you have ever isolated Legionella from the pharynx of patients without legionellosis?

       Dr. Victor L. Yu gargled L. pneumophila, serogroup 1 taken from a patient who died.  L. pneumophila persisted in his oropharynx for 30 minutes, before it gradually disappeared (Yu VL, ICAAC, 1981).

       See the following articles (two can be downloaded from www.legionella.org under Publications dealing with "Mode of Transmission of Legionnaires’ disease": Muder RR. Arch Int Med 1986, Yu VL.  Amer J Med 1993).

       Muder RR.  Isolation of L. pneumophila serogroup 5 from empyema following esophageal perforation.  Source of the organism and mode of transmission. Chest 1992;102:1601-1603,

       Marrie TJ Colonization of the respiratory tract with L. pneumophila for 63 days before onset of pneumonia.  J Infect 1992;24:81-86.

 

We routinely test bronchial aspirates of patients prior to a kidney transplantation using DFA method. In some aspirates I have found Legionella pneumophila. In most cases we use a monoclonal antibody made by Sanofi-Pasteur and it seems to work fine but I am very curious if these findings really reflect Legionella's colonisation of proximal airways.

        Since DFA is highly specific, your findings are important and should be published.  You should also culture your environmental water supply and do culture and monoclonal DFA on the water specimens taken from distal sites of your transplant patients for completeness sake.

       In a study similar to yours, Saravolatz found that 2.9% of renal transplant recipients and 4.8% of hemodialysis patients were colonized in a 2 year prospective study.  Colonization was defined by DFA and endured for 2 - 24 weeks. All colonizing organisms were L. micdadei or L. dumoffi. Saravolatz L, et al. Legionella infections in transplant recipients.  In: Legionella - 2nd International Symposium. ASM 1984:231-233.