Pages Navigation Menu

Meet 'em, greet 'em, treat 'em and street 'em

SGEM#120: One Thing or Two for Community Acquired Pneumonia?

SGEM#120: One Thing or Two for Community Acquired Pneumonia?

Podcast Link: SGEM120

Date: May 7th, 2015

Guest Skeptics: Victor Tsang graduated from University of Waterloo School of Pharmacy and is currently doing a residency at London Health Sciences Centre in London, Ontario.

Cassandra McEwan graduate of McGill University and University of Waterloo. She is also completing a residency program at London Health Sciences Centre.

Case: 62-year-old man presents to your emergency department with a four day history of increasing shortness of breath, purulent cough, fever, and generally feeling quite unwell. You measure her vitals, and he is mildly tachycardic and tachypneic, normotensive, O2 Sats are 96% on room air and his temperature is 38.2C.

You obtain routine blood work including a lactate level and blood cultures. A chest x-ray reveals a left lower lobe infiltrate in his lungs suggestive of pneumonia, and the decision is made to admit her to your general medicine floor You are about to start her on empiric antibiotics and wonder whether to use β-Lactam monotherapy or β-Lactam plus macrolide combination therapy

Background: Community-acquired pneumonia (CAP) is a common cause of emergency department visits and hospital admissions. Over the years, studies have looked at ways of scoring severity of illness to decide where a patient should be treated and how long a patient should receive antibiotics.

I remember starting out and having to calculate the FINE criteria to determine severity and whether or not to admit to hospital. Now we can just go to MD Calc and open up the PORT score or pneumonia severity index (PSI) App as a clinical decision tool to help us risk stratify patients for outpatient or in-patient management. This man scored 92 or moderate risk based on his age, sex, tachycardia and tachypnea.

There is still controversy surrounding what antibiotic(s) should be our empiric go-to. This has been increasingly difficult to pinpoint given the rise of antimicrobial resistance. In a 10-year span (1995 to 2005), Canada saw a dramatic rise in macrolide-resistant S. pneumoniae from 3.7% to 19.0%. Antimicrobial resistance is largely driven by overuse/misuse of antibiotics.

Clinical Question: In adult patients admitted to hospital with moderately severe CAP, which is better at achieving clinical stability: β-Lactam monotherapy or β-Lactam plus macrolide combination therapy?

Reference: Garin et al. β-Lactam Monotherapy vs. β-Lactam–Macrolide Combination Treatment in Moderately Severe Community-Acquired Pneumonia: A Randomized Noninferiority Trial. JAMA Intern Med. 2014

  • Population: Immunocompetent adults presenting to the ED with CAP and who were subsequently hospitalized
    • Inclusion criteria: Age >18 years, >2 clinical findings suggestive of pneumonia and Presence of new infiltrate on chest radiograph
    • Exclusion criteria: Severe immunosuppression, Recent hospitalization (<14 days), Residency in nursing home, Severe pneumonia, as per IDSA 2007 rule or PSI category V, or Administration of any antibiotic for >24 hours before inclusion
  • Intervention: Monotherapy
    • Cefuroxime 1.5 g TID IV followed by cefuroxime 500 mg BID PO,
    • Amoxicillin/clavulanic acid 1.2 g IV QID followed by amoxicillin/clavulanic acid 625 mg TID PO
  • Comparison: Combination Therapy
    • Monotherapy regimen PLUS clarithromycin 500 mg BID IV or PO
  • Outcome:
    • Primary Outcome: Percent of patients not reaching clinical stability at day-7 (All 5 vital signs reached and maintained for a minimum of 24 h: HR <100 bpm, SBP >90 mmHg, tympanic temperature <38.0 degrees C, RR <24 breaths/min, O2 sat by pulse oximetry >90% on room air)
    • Secondary Outcomes: Intensive care unit admission, complicated pleural effusion, length of stay, change in initial antibiotic treatment, in-hospital death, 30-day mortality, 90-day mortality, 30-day readmission, 90-day readmission, new pneumonia within 30 days

Authors’ Conclusions: β-Lactam monotherapy was not found to be non-inferior to β-Lactam plus macrolide combination therapy in hospitalized adult patients with moderately severe CAP. However, “patients infected with atypical pathogens or with PSI category IV pneumonia had delayed clinical stability with monotherapy” based on secondary outcome analyses.

Quality Checklist for Randomized Clinical Trials:

  1. checklist-cartoonThe study population included or focused on those in the ED: YES – all patients presented initially to the emergency department
  2. The patients were adequately randomized: YES“computer generated and stratified by center, with a 1:1 ratio, in randomly alternating blocks of 6, 8 and 10”
  3. The randomization process was concealed. YES – The patients were allocated to treatment arms by means of consecutive, numbered, sealed, and opaque envelopes
  4. The patients were analyzed in the groups to which they were randomized. YES “all analyses were performed in the intention-to-treat populations”
    • Even in patients who ended up receiving a macrolide in the monotherapy group due to a positive Legionella urine test, they still did intention-to-treat analysis.
  5. The study patients were recruited consecutively (i.e. no selection bias). YES
  6. The patients in both groups were similar with respect to prognostic factors. Unsure. 
    • Legionella was more common in the monotherapy group, occurring in 4.1% of patients vs. only 1.4% in the combination group. This could mean the monotherapy group were less likely to achieve clinical stability.
  7. All participants (patients, clinicians, outcome assessors) were unaware of group allocation. NO – patients and clinicians were aware of allocation.
    • However, the two separate outcome assessors were blinded to treatment allocation and the primary outcome was objective, so should not be vulnerable to bias.
  8. All groups were treated equally except for the intervention. YES
  9. Follow-up was complete (i.e. at least 80% for both groups). YES – There was 100% follow-up at day-30. Follow-up was done by phone if patients were discharged before then.
  10. All patient-important outcomes were considered. Unsure
    • Vital signs indicating clinical stability are technically surrogate outcomes. Ideal primary outcomes would have been clinical cure, hospital readmission, mortality, or length of stay.
  11. The treatment effect was large enough and precise enough to be clinically significant. NO
    • The upper limit of the 1-sided 90% confidence interval was 13.0%, exceeding the predefined non-inferiority boundary of 8%.
    • The FDA recommends using a non-inferiority boundary of 10% for assessing clinical success of a new anti-infectious treatment. So setting it at a conservative 8% showed good intentions a priori.

Screen Shot 2015-05-07 at 9.41.22 PMKey Results: A total of 580 patients were included in the analysis with 291 receiving monotherapy and 289 receiving combination treatment.

  • Primary Outcome (Patients not reaching clinical stability at day 7):

41.2% in mono vs. 33.6% in combo (ARR= 7.6% NNT=14)

  • Subgroup Analysis:
    • Combination better for atypical pathogens and pneumonia severity index category IV patients
    • Monotherapy patients were more likely to be readmitted in 30days
    • No difference in mortality, intensive care unit admission, complications, length of stay and recurrence of pneumonia within 90 days

Screen Shot 2015-04-25 at 3.11.12 PMStudy was designed well (good internal validity) with a reasonable primary outcome that was patient-oriented. The timing of patient follow-up makes sense, not many patients were lost to follow-up. In addition the data analysis was a true intention to treat analysis.

Even though the absolute difference was 7.6%, the upper limit of the 90% CI was 13%. Since this is above the a priori non-inferiority boundary of 8%, non-inferiority of monotherapy was not demonstrated.

There were some threats to validity. The study was supposed to target patients with moderately severe CAP, but actually accepted all CAP patients. This was evidenced by 10% PSI I, 50% II/III, and 40% IV; so the study artificially met its power because it accepted patients that were mild (I) as well as severe (IV).

It is important to note that there was more Legionella randomized to the monotherapy group. Legionella predisposes patients to being sicker. This imbalance at baseline could have contributed to the monotherapy being less effective.

A limitation of this study was the antibiotics used are not available in Canada (IV amoxiclav and IV cefuroxime). We also usually use azithromycin rather than clarithromycin as our macrolide in CAP.

While the study was conducted across multiple sites, all were in Switzerland. We need to consider the differences in antimicrobial resistance patterns compared to our own country. Ideally, we’d like to see a similar study done in Canada.

We contacted the authors and had them send us their antibiogram, which we compared to ours at own and surprisingly, the patterns was quite similar!

One final thing to comment upon was there were no safety differences between the two arms, which is reassuring for us.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: The author’s conclusions are similar to our conclusion. The evidence shows β-Lactam monotherapy is not  non-inferior to β-Lactam + macrolide combo therapy in admitted community acquired pneumonia patients. As such, this data re-affirms our current practice, which is to use β-Lactam + macrolide as empiric therapy in admitted community acquired pneumonia patients.

We are hesitate to make the same conclusions that the authors made regarding secondary outcomes, due to issues with proper powering.

SGEM Bottom Line: In patients presenting to the emergency department with community acquired pneumonia severe enough to warrant admission, we should continue to use β-Lactam + macrolide empiric therapy.

Case Resolution: You decide to start the patient empirically on ceftriaxone IV + azithromycin PO. Two days later his blood cultures come back showing strep pneumo sensitive to cefuroxime.

At this point, he has improved clinically. You make the decision to write him a script for cefuroxime to complete his course of antibiotics, and discharge his home.

Victor Tsang

Victor Tsang

Clinical Application: According to the Ontario “Anti-infective Guidelines for Community-acquired Infections” – an evidence-based, peer-reviewed publication – community acquired pneumonia is divided into three categories: mild to moderate, severe requiring hospitalization and severe requiring intensive care unit.

Notice how none of these categories align 100% with the article’s “moderately severe community acquired pneumonia”. Recommended empiric therapies also differ. β-Lactam monotherapy is only recommended for mild to moderate community acquired pneumonia patients safely treated as out-patients.

Cassandra McEwan

Cassandra McEwan

Once admitted, the only monotherapy recommended is a fluoroquinolone. However, at our hospital, we are trying to decrease our fluoroquinolone use, as our antibiogram shows increasing resistance to it and also because it is associated with higher risk of C. difficile infection compared to other options.

For our community acquired pneumonia patients sick enough to be admitted, we empirically use β-Lactam  + macrolide, which is supported by the results of this study.

What Do I Tell My Patient? You have community-acquired pneumonia and we need to admit you to hospital for treatment. You will be started on two different antibiotics to treat the infection (one oral and one intravenous). We will keep a close eye on you and hope to see some improvement in the next 24-48 hours. Some of the blood test we took may be able to tell us how to better treat your pneumonia but they will take a few days to get final results.


Postma et al just published a study April 2nd in the NEJM. It was a non-inferiority of the β-Lactam monotherapy vs.β-Lactam–macrolide combination therapy vs. fluoroquinolone monotherapy. Their primary outcome was 90-day mortality, in an ITT analysis, using a non-inferiority margin of 3% and a two-sided 90% confidence interval.

  • Authors Conclusions: “Among patients with clinically suspected CAP admitted to non-ICU wards, a strategy of preferred empirical treatment with beta-lactam monotherapy was noninferior to strategies with a beta-lactam–macrolide combination or fluoroquinolone monotherapy with regard to 90-day mortality”.

So there primary outcome of non-inferiority for mortality agrees with the secondary outcome of the study we just reviewed that showed no mortality difference.

Keener Kontest: Last weeks winner was Brenda Palsa from South Huron Hospital. Brenda knew Sir Peter Kerley was the Knighted Radiologist who described the A, B and C lines in the lungs on chest X-ray.

Listen to the podcast for this weeks keener question. If you know the answer send an email to with “keener” in the subject line. The first correct response will win a cool skeptical prize.

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine.


Screen Shot 2014-11-15 at 7.07.05 PM