Take a Deep Breath: Treatment of Pneumonia
Editor’s Note: At the time of writing, Maria Wilkins was a fourth-year pharmacy student at Shenandoah University’s Bernard J. Dunn School of Pharmacy. She will be starting her PGY1/PGY2 in Health System Pharmacy Administration and Leadership at Allegheny General Hospital in Pittsburgh, Pennsylvania. She will also attain a Masters in Public Health from the University of Pittsburgh during her two-year residency. In the little spare time she has now (and will have these next couple years), she loves to explore places to eat, run, and chill with her cat, Steela (named after the Pittsburgh Steelers).
A huge thanks to Maria for taking on this post while finishing school and gearing up for residency. A+ on time management already!!
Having previously shared his pharmacy wisdom with tl;dr here, Dev Chatterji returns to tl;dr after working with Maria on rotation. He currently practices as a Clinical Specialist at Inova Fairfax Hospital in the DC Metro area doing ID and Internal Medicine.
What is Pneumonia?
Pneumonia is an infection of the lower respiratory tract that can cause varying degrees of illness in patients of all ages. At some point in our lives, we’ve probably had (or have known someone with) pneumonia. I was nine years old when I had pneumonia, and I remember because I was coughing non-stop throughout The Lion King on Broadway when my parents decided to take me to the hospital (true story!).
Pneumonia causes over 500,000 emergency department visits each year and leads to nearly 50,000 deaths yearly in the United States according to the CDC.
Classification of Pneumonia
The first step in managing patients with pneumonia is determining what type of pneumonia the patient has. As will be discussed below, the initial pharmacotherapeutic management of pneumonia differs based on classification as well as patient-specific risk factors. Pneumonia can be classified as follows:
The definitions above may seem self-explanatory, but there can be confusion between VAP and HAP (which came first, the chicken or the egg). Luckily the treatment is essentially the same with HAP/VAP (yay, fewer things to memorize!), but just don’t be that pharmacist using the term VAP when the patient has never been intubated.
You may have also heard the term HCAP (Healthcare-Associated Pneumonia) floating around. This classification was introduced in 2005 by two ‘no-name’ organizations known as the American Thoracic Society (ATS) and the Infectious Disease Society of America (IDSA).
HCAP included patients presenting from the community with pneumonia that had previous contact with the healthcare system, thus putting them at risk for drug-resistant pathogens. This would include patients living in a nursing home/long-term care facility, receiving dialysis or IV antibiotics/chemotherapy, or recently hospitalized in an acute-care hospital. However, the concept of HCAP has been somewhat controversial, and there hasn’t really been a solid consensus on the value of HCAP as a predictor for drug-resistant pathogens. For some light reading on this subject I refer to you to a couple good articles summarizing this topic.
Furthermore, in the most current 2016 IDSA guidelines on HAP/VAP, they actually failed to address the management of HCAP because they too now feel that the evidence has not shown HCAP to be a good predictor for the risk of drug-resistant pathogens. We’ll see if this gets re-addressed in the update to the CAP guidelines that should be coming out shortly, but for now, say goodbye to HCAP.
Diagnosis of Pneumonia
Pneumonia generally presents with hallmark symptoms of infection (fever/chills, leukocytosis) along with respiratory symptoms such as cough, shortness of breath, and chest pain. However, from a diagnostic standpoint, there is no ‘gold standard’ test to diagnose pneumonia, and it can be challenging to determine the etiology. That’s why we leave most of this to our physicians and advanced practice clinicians.
Proper diagnosis is very important and has many implications from an antibiotic stewardship standpoint (i.e., not treating viral pneumonia with antibiotics) as well as for the purpose of optimizing patient outcomes.
To be complete, we’ll touch some key aspects of diagnosis to note when you are working up a patient with bacterial pneumonia:
Respiratory (sputum) and blood cultures are routinely drawn for hospitalized patients—but, they are often negative so not as helpful for determining etiology
There’s also the possibility of a low quality sputum culture. The clue for this is lots of epithelial cells. (When your patient hacks up some sputum for the sample, it has to pass through the oral cavity - and if this isn’t collected correctly, you just might pick up some normal oropharyngeal flora - along with those epithelial cells.)
Nasal PCR for detecting MRSA
Urine antigen tests
Legionella urine antigen - rapid turnaround; this assay can help detect infection caused by Legionella, but not other atypical bacteria
Streptococcus pneumoniae urine antigen - rapid turnaround; but most empiric antibiotic therapy will cover S. pneumoniae anyhow…
New and consistent pulmonary infiltrates on chest imaging (typically one-sided) are generally more consistent with bacterial pneumonia
Of note, upon repeat imaging these lung consolidations should not typically resolve quickly in a patient with true bacterial pneumonia
Can help distinguish viral from bacterial disease.
It’s generally easier to think of this as a “double check” for physicians.
A procalcitonin less than 0.25 mcg/mL is typically not associated with bacterial infection; however, it’s imperfect and a lot of things can falsely increase (trauma, CKD, malaria) or decrease (severe immunosuppression, fungal infections, parapneumonic effusions) procalcitonin values. So you have take those values with a grain of salt.
The most important thing to remember about procalcitonin is to use clinical judgement over everything.
Procalcitonin may be ordered serially to see if it’s trending downwards which reassures the team that we are on the right track with therapy.
Management of Community-Acquired Pneumonia (CAP)
Now to get into the management of pneumonia… and we’ll start with CAP.
The first key issue you first encounter when managing patients presenting with CAP is determining whether the patient needs to be admitted to the hospital or not. (If you’ve worked in the hospital at all, you know just how precious bed space is!)
So how does the team make this decision? We do this by using severity-of-illness indices, and there are a few different options for determining this.
The first is the CURB-65 score. CURB-65 is a fast, easy to remember and commonly used index that tells us whether we should treat inpatient or outpatient based on mortality risk. It takes into consideration mental status, BUN, respiratory rate, blood pressure, and age over 65 years old (see below).
Another common tool to determine either inpatient or outpatient management is the Pneumonia Severity Index (PSI). This tool uses many of the same factors from CURB-65…but it is much longer (aka harder to memorize). It also was pretty comparable to the CURB-65 in its usefulness, despite being more comprehensive. So why complicate things?
If, from these severity indices AND clinical judgment, it is determined that a patient needs to admitted to the hospital for management, the next step is to determine if the patient is at risk for multi-drug resistant organisms (MDRO). The main reason to consider all of this is to help you choose optimal initial antibiotic therapy.
Risk Scoring of Pneumonia
If the patient’s severity of illness qualifies them for an inpatient admission, the next major decision point is to determine the risk of MDRO. As mentioned earlier, the sub-classification of HCAP is no longer endorsed in the most recent IDSA HAP/VAP guidelines. Their recommendation is to use validated scoring tools to determine if a patient is at risk of MDROs rather than just based on exposure to the healthcare setting alone.
A number of risk prediction models have surfaced to help risk stratify patients. The Drug Resistance In Pneumonia (DRIP) score is one example of a prediction tool. It seems to be the most accurate and validated prediction model to date, and it has performed better overall than other models that have been published.
Bacterial Pathogens in CAP
This is important because the empiric therapy we use in CAP is based on coverage of these organisms. We’ll focus on treating for both typical and atypical organisms empirically.
Let’s start with the first section for previously healthy patients… Although this is what the most recent (*cough* 2007) IDSA CAP guidelines suggest, this monotherapy regimen has pretty much disappeared now for several reasons.
One reason may be because Streptococcus pneumoniae resistance to macrolides has exceeded 25% in most parts of the United States. As mentioned earlier, S. pneumoniae is one of our most common CAP pathogens, so we wouldn’t want to take that heavy of a gamble with just a single drug.
The other reason might be that while macrolides and doxycycline work really well for atypicals, there is evidence that shows atypicals are not as common as we originally thought in adult CAP patients. The etiologies of CAP were elucidated in a 2015 multi-center study, and most of the time it was a viral cause. They also found that even if it was bacterial, most of the time it wasn’t an atypical organism for adults.
So the bottom line is that you do not want to use a monotherapy regimen empirically.
Now we move onto our next option which will either be monotherapy with a fluoroquinolone (yay for adherence) OR a combination of a beta-lactam (think penicillins or cephalosporins) WITH a macrolide or a fluoroquinolone. We’ll talk about the decision making for which agent to choose later, but the great thing is that you can mix-and-match these! Beta-lactams cover the typicals, and macrolides/fluoroquinolones cover the atypicals. So as long as you have those two agents - you’re golden!
CAP Therapy – Inpatient (High Risk)
Given that the guidelines are over 10 years old, inpatient management of CAP can get a little tricky…
For inpatients deemed non-severe, with low risk of MDR pathogens (i.e., DRIP negative), empiric treatment is geared to cover common bacterial pathogens as outlined above:
Intravenous non-Pseudomonal beta-lactam (usually ceftriaxone) plus a macrolide (usually azithromycin) or doxycycline as an alternative
Fluoroquinolone (either levofloxacin or moxifloxacin) as monotherapy
With the ‘bad press’ that the fluoroquinolones have gotten recently, this option should probably only be used when other options are not possible.
For inpatients deemed to be severe (ICU admission) or having a high risk of MDR pathogens (i.e., DRIP positive), empiric treatment is geared to cover common bacterial pathogens as outlined above with additional consideration for Pseudomonas and MRSA, if applicable:
Intravenous non-Pseudomonal beta-lactam (usually ceftriaxone) plus a macrolide (usually azithromycin) or doxycycline as an alternative
If Pseudomonal coverage is needed
Substitute non-Pseudomonal beta-lactam with cefepime, piperacillin/tazobactam, or meropenem PLUS
Ciprofloxacin or levofloxacin
Aminoglycoside plus azithromycin
Aminoglycoside plus moxifloxacin
If MRSA coverage is needed
Vancomycin or linezolid
Duration of Therapy for CAP
Evidence shows there is no additional benefit or difference in outcomes to using a 10 day course of therapy versus 5 days for CAP. So for most CAP patients, we’ll stick with a 5 day course. But that may be lengthened depending on the infection and if the patient is clinically stable.
Management of Hospital-Acquired Pneumonia (HAP & VAP)
Beyond CAP, the other main type of pneumonia is pneumonia that occurs while in the hospital—hospital acquired pneumonia (HAP) and ventilator associated pneumonia (VAP). As a reminder, VAP is a type of HAP that occurs more than 48 hours after endotracheal intubation. This type of pneumonia generally carries a higher mortality risk due to more resistant and virulent pathogens.
Bacterial Pathogens in HAP/VAP
Once again, these pathogens differ from the most common isolates in CAP, so we’ll want to empirically treat differently. In HAP/VAP, we’ll focus more on covering Pseudomonas and Methicillin-resistant Staphylococcus aureus (MRSA).
Empiric Therapy of HAP/VAP
Due to the higher rate of mortality and the virulence of associated pathogens in HAP/VAP, it is of the utmost importance to select appropriate empiric therapy.
Antibiotics should primarily be selected on the basis of risk factors for MDR pathogens and local susceptibility patterns (i.e., your institution’s antibiogram). All patients with HAP/VAP should receive empiric antibiotic treatment targeting Staphylococcus aureus and Pseudomonas aeruginosa. Patients at a high risk of having an MDRO should receive two anti-Pseudomonal agents from different classes as outlined above.
High risk would include patients with IV antibiotic use in the past 90 days, septic shock at time of VAP, acute respiratory distress syndrome (ARDS) before VAP, and acute renal replacement therapy before VAP. It is reasonable to omit double-Pseudomonal coverage if a patient has low risk of having a MDRO.
Duration of Therapy for HAP/VAP
The guidelines recommend a treatment of 7 days for most HAP/VAP patients rather than longer courses of 10-15 days. This is especially important because it increases the number of antibiotic-free days (less costly and fewer side effects) and reduces the risk of recurrence with MDROs.
Similar to CAP days of therapy, the course may be prolonged for some patients if they are clinically unstable.
An Overview of Medications Used to Treat Pneumonia
Now we’ll briefly review each class of medication we mentioned earlier.
(Disclaimer: The following descriptions are just an overview of the important things to remember, but not all inclusive of EVERYTHING you need to know about these drugs. You still have to do some of your own research :) )
You’ll notice how this particular class gives us “the most bang for our buck.” As you can see, it covers many organisms and could potentially be used as monotherapy. So why don’t we just put everyone with pneumonia on a fluoroquinolone?
I hope that statement alone strikes a cord in your brain. Like any other drug, there are contraindications, warnings, and other considerations to keep in mind.
Fluoroquinolones in particular have been getting a lot of bad press lately and may not be a considerable agent in some patients due to the risks of tendonitis/tendon rupture, aortic dissections/tears, CNS effects (anxiety, confusion, etc), and peripheral neuropathy just to name a few.
I hope you noticed how ciprofloxacin isn’t recommended for empiric CAP treatment.
You might tempted to say, “Cipro isn’t considered a respiratory fluoroquinolone because it doesn’t penetrate the lungs.”
And you’d be so very wrong.
WE NEED TO STOP that misconception. The reason it is not considered a “respiratory fluoroquinolone” in the setting of CAP is because it has poor activity against Streptococcus pneumoniae (which, as we just learned, is one of the common pathogens in CAP). If there is one thing you get out of this article, it is the real reason ciprofloxacin is not used in CAP.
Macrolides are great agents because they cover atypicals. Azithromycin is favored because of its ease in dosing (no renal adjustments, comes in a Z-Pak, 1:1 IV to PO ratio). Macrolides are very common in any setting and are generally well tolerated, but one thing to remember is that they can cause QT prolongation.
Now we aren’t going to check ECGs in every patient that has pneumonia, but it is something you should consider in patients presenting with heart conduction issues (afib, arrhythmias, etc.) or in those already taking several other QT-prolonging agents.
FYI, one of my patients actually presented with sepsis secondary to CAP and due to all the electrolyte abnormalities, he had a QTc around 540 milliseconds! So it IS real! We immediately had to think of an alternative agent (hint: we’re talking about it next).
Doxycycline is our alternative to using a macrolide in CAP. Not extremely common, but possible depending on the patient. Avoid these agents in children <8 years old (due to teeth discoloration), pregnancy, and breastfeeding.
It’s important to counsel patients on separating doses from antacids, iron, and dairy products. Also, remind them to take with a full glass of water and stay upright for at least 30 minutes to prevent esophagitis/ulceration.
Penicillins are known for their gram-positive coverage, but for pneumonia, we favor using a penicillin combined with a beta-lactamase inhibitor like Zosyn (piperacillin-tazobactam) or Unasyn (ampicillin-tazobactam) to extend our gram-negative coverage. (However, that doesn’t mean there isn’t a place for high-dose, plain amoxicillin in select suspected S. pneumoniae CAP cases (high dose = 1 g TID)).
Since penicillin allergies are common, it’s important to verify whether the patient actually experienced anaphylaxis or not. A lot of times, the reaction may be an intolerance or maybe the patient has had the medication recently without any problems. This is especially important because we don’t want to avoid penicillins (or any of the other beta-lactam antibiotics) completely since they are one of our safer antibiotic options.
One thing to note is that the IDSA guidelines only list a few cephalosporin agents, but theoretically you could use any of the agents in the 3rd and 4th generation because of their gram-negative coverage and high distribution to the lung tissues.
FYI, the 2nd generation cephalosporin cefuroxime is technically indicated for treatment of CAP as well. Although you may need to reach for it due to cost reasons in some patients, there is a higher chance of resistance with S. pneumoniae, so it shouldn’t really be your go-to.
There are lots of agents to choose from in these classes, so your recommendation will rely on patient specific factors. Your patient needs an IV agent but you want to avoid renal dose adjustments...use ceftriaxone. You may want Pseudomonas coverage in a CAP patient, so maybe you’ll choose cefepime. The possibilities are almost endless!
Once again, don’t forget to verify the type of penicillin allergy a patient has since cephalosporins can exhibit cross reactivity (1-2%) in those patients. On this note, let’s have a moment for antibiotic side chains. Cross-reactivity between penicillins and cephalosporins and amongst different types of cephalosporins appears to be at least partially determined by the chemical structure of 2 of the side chains (R3 and R7).
These drugs are commonly referred to as the “big guns” because they are some of broadest spectrum agents available on the market. Generally speaking, you wouldn’t necessarily want to use this broad of an antibiotic empirically, but it is possible if there were contraindications or resistance to the other agents. Carbapenems may have minimal cross reactivity in patients with a penicillin allergy (very low, ~1%; again, verify the severity) and also have a risk of seizures.
But really, all beta lactams carry some risk of CNS toxicity, so monitor your patients and make sure you’re adjusting doses appropriately!
Ertapenem is the only carbapenem that DOES NOT cover Pseudomonas, so you won’t see it being used in HAP/VAP because that’s one of the major pathogens we’re trying to cover. You may still see it in CAP (way down the line if you’re really desperate) because Pseudomonas isn’t a common CAP pathogen.
Aztreonam is reserved for patients that have true anaphylaxis to penicillins, including the agents that have potential for cross reactivity (cephalosporins and carbapenems). Aztreonam has gram negative activity only, including Pseudomonas. Generally we’d want to use an agent like a fluoroquinolone before aztreonam, but if the patient has VAP or HAP, we need 2 different anti-Pseudomonal drugs on board. If you do use aztreonam, it requires renal dose adjustments when CrCl < 30 mL/min.
Technically, aminoglycosides should be preferred over fluoroquinolones as part of our double-Pseudomonal coverage based on evidence from a few studies. Let’s say the pathogen is resistant to your beta-lactam (aka, your first anti-Pseudomonal agent). Fluoroquinolones are less likely to have activity compared to the aminoglycosides.
This means we are more likely to cover the bug with the addition of an aminoglycoside versus a fluoroquinolone.
The reason you may not see this much is because of the risks (ototoxicity and nephrotoxicity) and the seemingly cumbersome drug monitoring. The peak confirms efficacy whereas the trough ensures safety. You can read more about aminoglycoside dosing here.
Vancomycin is usually our drug of choice in pneumonia patients when we need to cover for MRSA. Similar to aminoglycosides, vancomycin also requires drug monitoring to ensure the trough is within a therapeutic goal, usually between 15-20 mg/L for a severe infection like pneumonia. Read here for more information about hitting the sweet spot with vanc.
Vancomycin also has risks of nephrotoxicity and ototoxicity. Another important thing to remember with vancomycin is to slowly infuse it to avoid Red Man syndrome (severe pruritus and erythematous rash to the face, neck, and upper torso).
Linezolid is also an agent that covers MRSA, but it’s not as common due to high cost compared to vancomycin. The adverse events associated with linezolid are serotonin syndrome (watch those drug interactions!) and hematologic effects such as leukopenia and thrombocytopenia. There are studies that suggest using linezolid may achieve clinical cure faster than vancomycin, but generally both agents will provide the same clinical outcomes.
Prevention of Pneumonia
Remember that the IDSA guidelines recommend empiric therapy, so you’ll need to consider your institution’s antibiogram and the individual patient to determine the best therapy.
Once you get sensitivities back, narrow your therapy accordingly. (And if you don’t have sensitivities because those sputum cultures can be wily, use your best judgment based on the patient’s clinical progress to determine a step-down plan.)
Pay attention to duration of therapy—remember we’re in sort of a “less is more” era.
Don’t forget the real reason why ciprofloxacin isn’t considered a “respiratory quinolone.”