Infection of the lung (NEJM Review 2015).
Besides blood and sputum cultures, there are a variety of diagnostic tests depending on which pathogen is a concern. Gram stains have always been questionable for making a diagnosis, but I can't give them up. For CAP:
"Gram stain of sputum samples was useful in guiding microbiological diagnosis of CAP in 23% of patients. The Gram stain and culture of sputum samples obtained from patients who have received antibiotic treatment was unreliable. The presence of gram-positive diplococci and gram-negative coccobacilli was highly specific for the culture of S. pneumoniae and H. influenzae, respectively (PubMed)."
"In a meta-analysis examining respiratory specimen Gram stain for diagnosis of ventilator-associated pneumonia, absence of bacteria on Gram stain had a high negative predictive value, but a positive Gram stain correlated poorly with organisms recovered in culture (PubMed)."
My take? A good sputum gram stain is one of many pieces of information needed to determine what the cause of the pneumonia is/might be.
I tend not to think in terms of VAP, HAP, CAP which are CRAP (except you should follow the treatment guidelines). I think of pneumonias as differentiated by risks, CXR pattern, gram stain and culture.
A sputum culture is no worse than a BAL for a microbiologic diagnosis (PubMed).
Legionella and Pneumococcal urine antigens are often routinely ordered:
"A total of 474 episodes of CAP were included. Streptococcus pneumoniae was the causative pathogen in 171 cases (36.1%). It was detected exclusively by urinary antigen test in 75 cases (43.8%). Sixty-nine patients had CAP caused by a pathogen other than S pneumoniae. Specificity was 96%, positive predictive value ranged from 88.8% to 96.5%, and the positive likelihood ratio ranged from 14.6 to 19.9 (PubMed)."
So not a bad test. And it may allow you to narrow antibiotics. HAHAHAHAHAHAHAHAHA. Like anyone does that when the cultures come back. I kill me.
The Pneumococcal antigen can be positive for a prolonged time after disease: "18 (52.9%) of the 34 patients in the first month after pneumonia diagnosis. In 12 of these positive cases, the test was still positive in the second month, in six patients after 4 months, and in two cases 6 months after the diagnosis of pneumonia (PubMed)."
So many variations, depending on the organism, the host and the exposures.
PPI's are a risk in the first month after starting them, he says with a tortured sentence structure (PubMed).
In the elderly, the use of anti-psychotics increases the risk of CAP in a dose dependent manner (PubMed). I am glad MY voices tell me good things.
Post-obstructive pneumonia may be more common; tends to be slow in onset and have a cavity (Pubmed).
Drowning. 90% aspirate and when cultured grow Aeromonas species > Staphylococcus aureus > Pseudomonas species = Haemophilus influenzae > Streptococcus pneumoniae, Klebsiella pneumoniae, Enterobacter cloacae, Acinetobacter species, and Stenotrophomonas maltophilia (PubMed).
And environmental disasters spew soil pathogens into the air for people to inhale. For example, after tornados, Alcaligenes xylosoxidans, Burkholderia species, Aspergillus, Pseudomonas putida, and Enterobacter aerogenes are found in patients with pneumonia (PubMed).
Curiously, obesity is protective for survival (PubMed).
Over 40% of patients with CAP will have undiagnosed diabetes or pre-diabetes (PubMed).
The lung does have a low density microbiome (Pubmed) that can be altered in disease. If stool transplant is barf-o-genic, wait until they try sputum transplant.
50% of the time we get no diagnosis and the cause has changed over the decades. For example, pneumococcus has become uncommon for a variety of reasons, going from 90% of cases to 10% (PubMed). As the French say, the more things change, the more things change.
25% of patients will have a mixed viral/bacterial cause and may have an increased death rate (PubMed). And about 40% of those being admitted with a viral cause will get a bacterial superinfection, usually, but not always, with S. pneumonia (PubMed).
In one adult cohort, the viral cause of pneumonia as Respiratory Syncytial virus 6.1%, Human metapneumovirus 4.5% patients, and influenza in 6.5% of patients (PubMed).
There are always new pathogens cropping up, be it SARS in Asia or coronavirus a Saudi Arabia (PubMed). One day a new pathogen will arise and sweep the world killing us all.
- Bacterial: S. pneumonia (still number one. At least 13.5% of patients with invasive disease will have some sort of hypogammaglobulinemia (PubMed)), Actinomycosis, C. pneumonia, C. psittaci, Coxiella burnetti, K. pneumonia, Legionella spp (probably under-diagnosed; PCR if used routinely can markedly increase the diagnosis, in one study it was from 1 in 9 to 1 in 20 cases of pneumonia (PubMed)), Mycobacterium spp, Mycoplasma pneumoniae, Nocardia spp, P. aeruginosa, S. aureus (for CAP, MRSA accounts for only 2.5% of causative organisms (PubMed)).
- most of the time ESBL E. coli and K pneumonia in ventilated patients are colonization but when found in infection-related ventilator-associated events it doesn't stop the use of unneeded carbapenems (PubMed).
When should you worry about MRSA as a cause of CAP? Probably never. S. aureus causes less that 2% of community acquired pneumonia (except as a complication of influenza), yet 30% of patients get put on anti-Staphylococcal antibiotics. Go figure.
"MRSA pneumonia was observed more frequently in patients with a previous history of MRSA infection (OR = 6.05; P < 0.001), a PSI score ≥120 (OR = 2.40; P = 0.015), intravenous antibiotic treatment within 30 days of pneumonia (OR = 2.23; P = 0.018) (PubMed)."
A negative nares for MRSA had a high specificity and negative predictive value (almost 100%) for ruling out MRSA pneumonia, particularly for CAP/HCAP (PubMed). So MRSA nasal negative? Don't give vancomycin. If the cultures and nasal swab are negative for MRSA it is reasonable to stop antibiotics (PubMed). This is especially true for HAP/ VAP (PubMed).
Intubation, tracheotomy, and 2 bronchoscopy increase the risk of pneumococcal pneumonia (PubMed).
- Parasite: usually as part of a Loeffler's syndrome as worms pass through the lung often with eosinophilia. Round worms Ascaris lumbricoides, Strongyloides stercoralis and hookworms Ancylostoma duodenale and Necator americanus, Trichinella spiralis and Schistosomiasis.
Candida almost NEVER causes pneumonia and can almost always be ignored, even in a BAL. "We concede, by the present data, that it is convincingly proven that Candida species are at most a very rare cause of pneumonia (PubMed)".; 0.07% of positive BAL's will actually be the real deal.
Like most fungal infections, the diagnosis is slow to be made as people do not consider it, even in endemic areas (Pubmed). Take a damn exposure history and know what is in your community.
- Community Acquired Lobar ('typical') Pneumonia: S. pneumonia (still number one, patients with this disease have a marked increased risk for acute MI, CHF and arrhythmia (PubMed)), K. pneumonia (especially ETOH abusers), Legionella spp., S. aureus (after influenza or other viral URI; always suspect R sided endocarditis when there is S. aureus in the sputum).
In one study (PubMed) they found an etiology in 90% of patients, with S. pneumonia in combination with a virus about 25% of the time.
- Community Acquired Interstitial ('atypical') Pneumonia: C. pneumonia, C. psittaci, Coccidioides immitis (29% of people presenting with pneumonia in Az had cocci) Coxiella burnetti, Mycoplasma pneumoniae, Legionella spp (is increased shortly after hot, humid, thundershower weather (PubMed)), Influenza, PJP.
- Aspiration Pneumonia: if rasty dentition (i.e. no WAY would you kiss that person) and chronic loss of consciousness from drugs, ETOH, seizures or trauma) then anaerobes and Streptococci. They usually present as lung abscess and empyema. People with good dentition (i.e. kissable) do not have a big load of anaerobes in their mouth. If the patient has no teeth, they have no anaerobes. Repeat that last sentence until you understand it. Most other aspirations are chemical and do not require antibiotics, but try to stop anyone from giving them. Can't do it. Seriously. An aspiration event is not not not not an aspiration pneumonia and whether community, hospital or ventilator no study or guideline that I can find suggests that anaerobes are important in acute aspiration events. Treating anaerobes because of an acute aspiration event is so stupid it makes my teeth hurt.
- Cavitary Pneumonia: mixed anaerobes PLUS Streptococci (lung abscess), S. aureus, Mycobacterium (Tb and atypical), Actinomycosis, Nocardia, Aspergillus, Mucormycosis, Coccidiomycosis, Cryptococcus, Histoplasma, PJP (on inhaled pentamidine).
- Hospital Acquired or Ventilator Acquired Pneumonia: mostly gram negative rods of the Pseudomonas variety or S. aureus (especially if a tracheostomy). Occasionally Legionella spp. NOT anaerobes. Ever. So quit using metronidazole.
From the IDSA Guidelines. Follow the guidelines. When compared to patients who do not receive antibiotics as recommended by the guidelines, those who get therapy by the IDSA have decreased 48 hour (PubMed) and long term mortality (PubMed).
Home or admit? To decide, use the pneumonia severity index.You really should use the calculator, it will improve care and survival (PubMed). Here is an online severity index calculator. And here is the local version.
Do not forget the vaccines; everyone does but it works. By the way, if the patient is getting better, they only need one CXR (PubMed).
As you read the recommendations remember in systemic reviews, for all CAP inpatient/outpatient disease, given as monotherapy or in combination with a beta-lactam, a quinolone leads to more clinical cure than macrolide for all kinds of pneumonia (PubMed). BUT. Azithromycin and other macrolides with a beta-lactam consistently demonstrate decreased mortality when the cause is Pneumococcus, and Pneumococcus is the most common cause of CAP.
Specific examples follow. From the IDSA guidelines:
1. Previously healthy and no use of antimicrobials within the previous 3 months: A macrolide >> doxycycline. But amoxicillin use is associated with decreased mortality if they have pneumococcus (PubMed). Except if you are really worried about Mycoplasma pneumoniae: Resistance to macrolides can be as high at 69% (but may make no difference in outcome) in some series and is associated with a prolonged fever if you use a macrolide. Go figure. Treat a patient with an antibiotic that does not work and the patient doesn't do as well.
2. Presence of comorbidities such as chronic heart, lung, liver or renal disease; diabetes mellitus; alcoholism; malignancies; asplenia; immunpsuppressing conditions or use of immunpsuppressing drugs; or use of antimicrobials within the previous 3 months (in which case an alternative from a different class should be selected) .
A beta-lactam plus a macrolide.
This is the best option. In meta-analysis, "macrolide-based regimens were associated with a significant 22% reduction in mortality compared with non macrolides; however, this benefit did not extend to patients studied in RCTs or patients that received guideline-concordant antibiotics. Our findings suggest guideline concordance is more important than choice of antibiotic when treating CAP (PubMed)." Of course there is Newtons second law of medical literature. For every study there is an equal and opposite study: "Among patients with clinically suspected CAP admitted to non-ICU wards, a strategy of preferred empirical treatment with beta-lactam monotherapy was non-inferior to strategies with a beta-lactam–macrolide combination or fluoroquinolone monotherapy with regard to 90-day mortality (PubMed) (PubMed)."
Damned if you do, damned if you do.
A respiratory quinolone.
And you can give the quinolone po with no change in outcome (Pubmed).
3. In regions with a high rate of infection with high-level (MIC > 16 mg/mL) macrolide-resistant Streptococcus pneumoniae, consider use of alternative agents listed above in (2) for patients without comorbidities.
Inpatients, non-ICU treatment
A beta-lactam plus a macrolide. I think this is definitely the best option and the data all points to a decrease in mortality with the combination, espcially if bacteremia, and who can predict up front (PubMed) (Pubmed)(PubMed). Doxycycline may be equal to the quinolone (PubMed), but not in the guidelines.
There is a suggestion, not statistically significant (which like as not means it is not real as p was 0.06 and a real significant result should be 0.005, not 0.05 (5 out of 4 Americans Do Not Understand Statistics)), that giving the macrolide first decreases mortality (PubMed). I really doubt it but the literature is what it is.
Or go with
A respiratory quinolone.
Of interest, using a quinolone may be more likely to select for subsequent resistant organisms (PubMed). I increasingly think of quinolones as evil, since most evil is done by those that think they are doing good.
Inpatients, ICU treatment
A beta-lactam (cefotaxime, ceftriaxone, or ampicillin-sulbactam) plus either azithromycin or a respiratory quinolone (for penicillin-allergic patients, a respiratory fluoroquinolone and aztreonam are recommended)
The above beta-lactam plus an aminoglycosides and ciprofloxacin (for penicillin-allergic patients, substitute aztreonam for above beta-lactam).
If CA-MRSA is a consideration, add vancomycin or linezolid.
No comorbidities No recent antibiotics: A macrolide OR doxycycline. Recent antibiotics: moxifloxacin, levofloxacin, or gemifloxacin alone, erythromycin, azithromycin, or clarithromycin PLUS high-dose amoxicillin. OR erythromycin, azithromycin, or clarithromycin PLUS high-dose amoxicillin./clavulanate.
Recent antibiotics: moxifloxacin, levofloxacin, gemifloxacin alone OR erythromycin, azithromycin, clarithromycin PLUS a beta-lactam Suspected aspiration with infection: amoxicillin-clavulanate OR clarithromycin (I prefer amoxicillin. OR a third generation cephalosporins PLUS metronidazole).
How long to treat? Shorter is better than longer, recent studies with a long acting azithromycin (PubMed) suggest a single dose will do. Usually it is 5,7,10 or 14, mostly based on the number of fingers we have or the number of days in the week. I think if we had 8 fingers and 12 day weeks, then patients would be getting some multiple or 8 and 12 days of therapy (PubMed). For uncomplicated CAP, 5 days suffices, and maybe even 3 days total (PubMed).
And, as should come to no surprise to anyone, patients are more adherent with their compliance (or is it compliant with their adherence?) if given a once a day antibiotic (PubMed).
Depending on the population, up to 66% of Mycoplasma pneumoniae, can be macrolide resistant.
I REALLY prefer the macrolide/beta lactam combination. It probably decreases mortality (PubMed). A meta analysis (and I never meta-analysis I liked)(PubMed) suggests it is a waste of time to cover atypicals, but several studies have demonstrated combination therapy with a macrolide (PubMed) increases survival of patients with pneumococcal bacteremia. What kills people is the pneumococcus, not the 'atypicals'. Usually. Some epidemiologic studies suggest that CAP mortality doubles if treatment for atypical pathogens is not given (PubMed).
Suspected aspiration with infection: amoxicillin-clavulanate OR clarithromycin or penicillin/beta-lactamase inhibitors or (I prefer amoxicillin. OR a third generation cephalosporins PLUS metronidazole).
Is linezolid superior to vancomycin? Data keeps going back and forth with a 2010 meta analysis suggesting they are equally a piece of crap (PubMed). As of 2012, the preponderance of data and the best clinical trial (PubMed) point to the slight superiority of linezolid for MRSA pneumonia with 10% higher cure rate but a similar 60 day mortality. But then a 2014 review suggests they are equal (PubMed). Sigh. The trend suggests linezolid for pneumonia (PubMed) and the better outcomes may be due to suppression of toxins, at least in the mouse model (PubMed). Although not supported by data, I tend towards linezolid when the MIC to vancomycin is >= 1.0, but I am hesitant for bacteremic MRSA infections off all sorts. For patients that are STS, I have been giving ceftaroline and either clindamycin or linezolid based on a hope and a prayer, which is something for an atheist.
How long to treat CAP? Shorter is better than longer, recent studies with a long acting azithromycin suggest a single dose will do. Usually it is 5,7,10 or 14, mostly based on the number of fingers we have or the number of days in the week. I think if we had 8 fingers and 12 day weeks, that would determine most of the duration of antibiotic therapy. That being said, for CAP that is doing well clinically (afebrile, WBC normal, etc) 3 or 4 days of antibiotics is all that is needed, so saith the guidelines. Even severe CAP probably only needs 7 days of therapy (Ref).
Despite the guidelines that suggest 5 days is all most people need, at least 70% will get an extended duration they do not need. How long. 10 days of course. People just can't resist the allure of counting fingers and basing the antibiotic duration on the result (PubMed).
In a meta analysis, (PubMed) mild to moderate disease, 5 days total therapy is probably enough.
Even if bacteremic with S. pneumoniae, it clinically stable ("cough and shortness of breath are improving, patient is afebrile for at least 8 hours, white blood cell count is normalizing, and oral intake and gastrointestinal tract absorption are adequate"), it is safe to change to oral (PubMed). 5 days of levofloxacin is as good as 10 days.
The only disease that has probably not been treated with steroids is Cushings, and CAP is no exception. In a Lancet study, 5 mg a day for 4 days decreased the length of hospitalization in nonimmunocompromised patients by a day, but did not alter short term morbidity or mortality. Would I recommend it? Naw. The steroid group had more hyperglycemia and readmission (PubMed). I would be more interested if dexamethasone decreased long term mortality. Then again, 7 days of prednisone also shortens illness without increasing mortality (PubMed). A 2015 meta analysis suggests " For hospitalized adults with CAP, systemic corticosteroid therapy may reduce mortality by approximately 3%, need for mechanical ventilation by approximately 5%, and hospital stay by approximately 1 day" (PubMed). Another suggests "Corticosteroids treatment was associated with an decreased risk of adult respiratory distress syndrome (RR, 0.21; 95% CI, 0.080.59), and may reduce the lengths of hospital and intensive care unit stay, the duration of intravenous antibiotic treatment, and the time to clinical stability. Corticosteroid were not associated with increased rates of adverse events (PubMed)." For me giving steroids for CAP is an unnatural act and it is harder and harder to change as I age.
Fine Risk Class III/IV CAP.
If the pneumonia severity score is less than 110, they only need 3 days of IV and need NOT go home on oral (PubMed). But no one will do this. If the pneumonia severity score is class IV or V (severe) and patient stable (defined as respiratory rate <25/min, oxygen saturation >90% or arterial oxygen pressure >55 mm Hg, hemodynamically stable, >1°C decrease in temperature in case of fever, absence of mental confusion, and the ability to take oral drugs) can change to oral therapy after 3 days (PubMed).
Regardless, when patient is afebrile and vital signs are stable, time for discharge. It is way stupid to watch them on po for 24 hours as pharmacokinetically they are on iv four half lives, usually a day. And it is safe to do (PubMed).
If the cultures are negative I always recommend they be sent home on doxycycline or amoxicillin. as the vast majority of patients are cured at the time of discharge and the last thing they need is a prescription for an expensive quinolone or macrolide.
CAP requiring Intensive Care Unit:
Besides antibiotics, pneumonia should maybe perhaps receive hydrocortisone 200 mg bolus followed by 10mg/hr for 4 days. It decreases mortality (PubMed).
The patient should receive a macrolide for its immunomodulatory effects whether or not the organism is susceptible to a macrolide; it decreases mortality (PubMed).
Pseudomonas infection is not an issue: cefotaxime or ceftriaxone or ampicillin-sulbactam (hah, a drug I still this is a Yugo: underpowered and overpriced) PLUS either erythromycin, azithromycin, clarithromycin or moxifloxacin, levofloxacin, or gemifloxacin.
Pseudomonas infection is an issue
Pseudomonas infection is an issue but the patient has a beta-lactam allergy: Either aztreonam PLUS levofloxacin (I prefer ciprofloxacin) OR aztreonam PLUS moxifloxacin, or ciprofloxacin, with or without an aminoglycoside. How long to treat? Most of the time 8 days of IV (PubMed).
Hospital Acquired or Ventilator Acquired Pneumonia: (cefepime or ceftazidime OR imipenem or meropenem OR piperacillin /tazobactam) PLUS (antipseudomonal fluoroquinolone (ciprofloxacin or levofloxacin) OR aminoglycoside (amikacin, gentamicin, or tobramycin))PLUS linezolid or vancomycin (if worried about MRSA, and who isn't) PLUS (azithromycin or quinolone (if worried about Legionella). Clarithromycin, perhaps due to immunomodulatory effects at a dose of 1 gm a day for three days markedly decreases days on the vent and may decrease mortality (PubMed).
There is an oh so stupid 'rule' in medicine: once antibiotics are started they can't be stopped and once a class of antibiotic is started, it can't be changed. I call shenanigans. De-escalation of antibiotics is reasonable and safe. You can change from multiple antibiotics to a single antibiotic if Pseudomonas aeruginosa is not present (and most of the time you do not need to two antibiotics for Pseudomonas), shorten the therapy to < 5 days if the cultures are negative and there have been > 48 hours of defervescence, and change from a broad to a narrow antibiotic based on culture data
How long to treat? Most of the time 8 days of IV (PubMed, PubMed). However, and pay attention here, if the CPIS score on day three is < 6, antibiotics can be discontinued as the patient probably doesn't have pneumonia. Click here to calculate a CPIS score.
1-3 days antibiotics is long enough in suspected VAP if "daily minimum PEEP of ≤5 cm H2O and daily minimum FiO2 less than 40% for at least 3 days from the first day of antibiotics (and) may be suitable candidates for early antibiotic discontinuation" (PubMed).
And a curious literature, like Alice's curiouser and curiouser, that adjunctive (Pubmed) clarithromycin for three days decreases mortality. This is not the only study to show such a benefit and perhaps has more to do with the immunomodulatory effects of clarithromycin than its antibacterial effects (Pubmed).
Hospitalized Nursing Home Patient: Same as for medical ward and ICU.
Why is the patient still febrile after antibiotics? From the 2014 NEJM Review:
- Correct organism but inappropriate antibiotic choice or dose
- Resistance of organism to selected antibiotic
- Wrong dose (e.g., in a patient who is morbidly obese or has fluid overload)
- Antibiotics not administered
- Correct organism and correct antibiotic but infection is loculated (e.g., most commonly empyema)
- Obstruction (e.g., lung cancer, foreign body)
- Incorrect identification of causative organism
- No identification of causative organism and empirical therapy directed toward wrong organism
- Noninfectious cause
- Drug-induced fever
- Presence of an unrecognized, concurrent infection
Older patients will often not recover completely, having accelerated functional decline (PubMed). To quote another article (PubMed): "A year after hospitalization for community-acquired pneumonia, moderate-to-severe impairment in multiple cognitive domains affected one-third of patients ≥ 65 years old and 20% of younger patients, and another third of survivors had mild cognitive impairment." Infections are short term and long term bad for people.
Cardiac events are common with CAP and more than doubles mortality (PubMed). The risk of MI goes up with acute viral pneumonias as well. Influenza > RSV > other viri (PubMed), especially during the first week of infection. Tere is a "small but strongly significant associations in the timing of respiratory infection (with HMPV, RSV, influenza, rhinovirus, and adenovirus) and MI or ischemic stroke hospitalizations in the elderly (PubMed)."
Even if they survive their CAP, and most should, the risk if death is increased for a year after CAP and that death risk is increased for a decade after Pneumococcal pneumonia. Infections are bad for you (PubMed).
Pleural effusions at admission are a risk for death and have a longer hospital stay (if they live)(Pubmed).
As for so many infections, prior use of NSAIDS increases like likelihood of badness compared to those who did not take NSAIDS: "pleural empyema and lung cavitation (37.5% vs. 7%; ) and had a trend to more invasive disease, with a higher frequency of pleural empyema (25% vs. 5%, P=0.014) and bacteremia (PubMed)."
Could try selective oropharyngeal decontamination with topical application of tobramycin, colistin, and amphotericin B in the oropharynx and stomach for four days. Decreased mortality at 30 days by 2.9% (PubMed). Overall, I am not a fan.
Room air O2 saturations in CAP of < 92% have increased morbidity and mortality (PubMed).
There is always the urge to add inhaled antibiotics. A 2016 meta analysis suggests "There is insufficient evidence for the use of inhaled antibiotic therapy as primary or adjuvant treatment of VAP or VAT" (PubMed). Not that insufficient evidence ever stops anyone. Hell sufficient evidence shows acupuncture is nothing but placebo and people keep on using it.
For severe pneumonia, one meta-analysis suggests zinc supplementation will decrease mortality (PubMed).
I tend to go on and on that infections are inflammatory, that inflammation is prothrombotic, and that every infection that has been evaluated has an increased in vascular events (stroke, MI or PE) in the months following the infection. It may be why people with hospitalized infections have higher death rates. But it is like the weather. Everyone complains about it but no one tries to change it.
Well, now you can. 300 mg of aspirin a day for a month decreased cardiac events and deaths in patients with pneumonia (PubMed). I suspect that in a decade, once the clinical trials are complete, that everyone with an acute infection will go home on ASA.
But. NSAIDS during an acute respiratory infections increase MI risk 3.4x (PubMed).
What continues to flabber my gaster is how people order tests then ignore them For pneumonia a negative procalcitonin combined with a postive viral PCR does not result in antibiotics being stopped (PubMed).
Relevant links to my Medscape blog
Last Update: 07/06/18.