Severe CAP: The 'Roid Debate Rages On


A 72 year old man presents to the ED with Shortness of breath, productive cough and fever of 102.1 F.

It's Phil.

  • Physical Exam: Decreased breath sounds on the left with positive egophony in the left lower lung field. No wheezing or evidence of bronchospasm.

  • Vitals: Tachypnic and febrile, otherwise stable.

  • Significant Labs: WBC 20, PCT 2.1 and mild AKI.

You order a lactate, a set of blood cultures and start ceftriaxone/azithro for CAP and....

You decide to order 40mg of PO prednisone daily.

Maybe you feel like Dwight??

Maybe you feel like Dwight??

Steroids in CAP - maybe you agree, or maybe you don't. In the literature, steroids as an adjunct therapy in pneumonia dates back as far as 1956 - and they've been controversial ever since. 



Cortisol is an endogenous steroid hormone- a hydrophobic molecule derived from cholesterol. These don't need to bind to a receptor outside of the cell and are able to go past membrane to bind to receptor inside cell. Due to this they have widespread systemic effects

  • It is released in response to activty by HPA axis in acute stress (infection or bear)

  • Cortisol secreting hormone released by hypothalamus, which goes to pituitary gland; ACTH is produced which goes to adrenal cortex and secretes cortisol

  • Three important roles:

    • Primes body for improved performace (incr HR, bronchodilation)

    • Mobilizes nutrient resources to sustain that performance

    • Calms down inflammation and supresses immune system

So what is rationale for exogenous steroids in CAP?

Local bacterial infections produce local inflammatory effects which become systemic (SIRS)

  • High burdens of inflammatory cytokines are BAD - high PNA severity index scores, higher risk of antimicrobial treatment failure, ARDS, ARF; higher mortality

CIRCI: Critical illness related corticosteroid insufficiency

  • A phenomenon described by SCCM/ESICM. When someone is critically ill there is dysregulation of HPA axis, decreased cortisol production and tissue resistance to cortisol

So by supressing the immune system, reducing the expression and secretion of pro-inflammatory cytokines and reducing leukocyte trafficking we SHOULD be able to calm inflammatory storm and benefit mortality

the literature

There are thousands of articles to sift through, but we are going to briefly touch on three studies relevant to the discussion.

1) 2015 - Corticosteroid Therapy for Patients Hospitalized With Community-Acquired Pneumonia: A Systematic Review and Meta-analysis

Link to study

  • Six studies and just over 400 patients with pneumonia, both ICU & acute care patients

    • Primary outcome: mortality

    • Secondary outcome:

      • Time to resolution of symptoms/clinical stability, pneumonia recurrence, mechanical ventilation/pressosr support, for ICU patients ICU discharge time

  • Findings: corticosteroids didn't reduce mortality among patients with PNA but it did find improvement in almost all secondary outcomes (but not in pediatric patients with RSV)

  • Evidence quality: LOW (small, single studies, poor blinding, small number of participants)

No formal recommendations made.

2) 2015 - Corticosteroids for Pneumonia

Link to study

  • A SRMA study of 13 trials and ~2000 patients that specifically looked at steroids for hospitalized patients with CAP

    • Primary outcome: all cause mortality

    • Secondary Outcome:

      • Need for mechanical ventilation, development of ARDS, hospital LOS

  • When it was published, started to frame the conversation about WHICH patients would benefit from steroids

  • Findings: corticosteroid therapy was associated with a ~5% absolute reduction in need for mechanical ventilation and development of ARDS; reduced hospital LOS ~1 day

    • Benefit in mortality varied on degree of CAP - severe (defined by CURB65, APACHE II and ATS definitions at time) had ~3% absolute reduction

    • Most common adverse effect: hyperglycemia (but authors of included RCTs often excluded patient sat risk for adverse events related to steroids such as GIB, immunosuppression)

3) 2017 Cochrane Review - Is treatment with corticosteroids beneficial and safe for people with pneumonia?

Link to study

  • 17 RCTs and just over 2200 patients

    • Primary outcome: all cause mortality

  • Findings:

    • Among all patients all cause mortality was decreased in corticosteroid therapy - driven primarily by cohort of patients with severe CAP (NNT 18, relative risk of death 0.58)

    • Improvement in secondary morbidity outcomes in severe and non severe CAP

  • Evidence quality: moderate for mortality reduction and high for reduction in early clinical failure

    • Most common adverse effect: hyperglycemia

  • Limitations

    • Mean age of patients in this study was 69.8

    • Lack of data prevented authors from performing subgroup analysis

    • Concern that patients at high risk of adverse reactions may have not been enrolled

    • In all but 3 studies, classification of severe CAP was done at trial level, not patient level - may have reduced external validity

    • High risk of publication/reporting bias - possibly exacerbated in setting of SRMA

    • Dose and schedule of steroids varied

so what's our takeaway?

  • Should we give steroids?

  • What dose to use?

  • How do you treat steroid induced leukocytosis?

  • Final nail in the coffin?

Looking Ahead

  • ESCAPe trial (extended steroid in CAP) - longer duration

  • Trial by Hassan et. al looking at IV steroids given according to a variety of pneumonia severity scoring systems - trying to evaluate which is best/which patients benefit most

  • Updated ATS/IDSA Guidelines this summer


“Curiosity”, "Let Go", "Insight", "Waiting for the Moment that Never Comes", "You're enough", "Keep trying", "Vancouver City Centre" and "Oakridge 41st" by Lee Rosevere is licensed under CC BY-NC 4.0 / Song has been cropped in length from original form

“храпи и переворачивайся” by Kosta T is licensed under CC BY-NC-SA 3.0 / Song has been cropped in length from original form

Primal scream therapy video: