Severe community-acquired pneumonia (CAP) · ICU Topics

Cinematic ICU scene of a breathless patient with severe community-acquired pneumonia, a CXR showing dense multilobar consolidation, IV antibiotics and fluids running, a cardiac monitor showing septic physiology, clinical-blue lighting, no faces, no text — FigureSevere CAP — antibiotics within one hour, oxygenation and haemodynamic support, and source control. IDSA/ATS major criteria (ventilation or septic shock) or three or more minor criteria mandate ICU admission; corticosteroids reduce mortality in the high-inflammatory phenotype.

Severity assessment

CURB-65

CURB-65 score (click each)

Severe — ICU

Mortality ~15-40%

Score >=3 = severe CAP. Consider ICU admission. Inpatient treatment with IV antibiotics. High mortality especially with score 4-5.

[2]

IDSA/ATS criteria for severe CAP

Microbiology

Educational schematic of alveolar consolidation with inflammatory exudate, V/Q mismatch and shunt, and cytokine cascade linking severe CAP to septic shock — FigurePathophysiology — alveolar exudate creates shunt and hypoxaemia; a high-inflammatory phenotype drives sepsis, ARDS risk, and the steroid-responsive biology of severe CAP.

Common pathogens

All CAP

Streptococcus pneumoniae (#1)
Haemophilus influenzae
Mycoplasma pneumoniae
Chlamydophila pneumoniae
Legionella pneumophila
Respiratory viruses (influenza, RSV, COVID-19)

Severe CAP specifics

ICU patients

S. pneumoniae still #1
Legionella — disproportionately severe (hyponatraemia, GI symptoms, confusion)
Staphylococcus aureus — especially post-viral (influenza)
Gram-negative bacilli — Klebsiella, Pseudomonas (if structural lung disease)
MRSA — if recent hospitalisation, antibiotics
Always test for influenza and COVID-19

Antibiotic therapy

Educational ICU pathway schematic for severe CAP: severity triage, empiric dual antibiotics, adjunctive corticosteroids, and respiratory support ladder — FigureManagement pathway — antibiotics within one hour (beta-lactam + macrolide or respiratory fluoroquinolone; escalate for MRSA/Pseudomonas risk), consider hydrocortisone in high-inflammatory severe CAP, and escalate oxygen to IMV as needed.

Severe CAP antibiotic protocol

Assess risk factors for resistant organisms

MRSA risk: recent hospitalisation (90 days), IV antibiotics (90 days), nursing home resident, known MRSA colonisation, structural lung disease (bronchiectasis). Pseudomonas risk: bronchiectasis, recent broad-spectrum antibiotics, repeated CAP admissions, structural lung disease.

Standard severe CAP cover (no MRSA/Pseudomonas risk)

Beta-lactam + macrolide: ceftriaxone 2g IV daily + azithromycin 500mg IV daily. OR ampicillin 2g IV Q6H + azithromycin. Alternative: respiratory fluoroquinolone (moxifloxacin 400mg IV daily) — but CAUTION: QT prolongation, tendon rupture, hypoglycaemia.

Add MRSA cover if risk factors

Add vancomycin (loading 25-30 mg/kg, then trough 15-20) OR linezolid 600mg IV BD. Linezolid preferred for pneumonia (better lung penetration, covers MRSA AND penicillin-resistant pneumococcus).

Broaden for Pseudomonas if risk factors

Use anti-pseudomonal beta-lactam: piperacillin-tazobactam 4.5g IV TDS, OR ceftazidime 2g IV TDS, OR meropenem 1g IV TDS. Plus macrolide (azithromycin) for atypical cover. If MRSA risk too: add vancomycin/linezolid.

Test for specific pathogens

Sputum culture + blood cultures (BEFORE antibiotics if possible, but do NOT delay >1 hour). Urine: pneumococcal antigen, Legionella antigen (especially severe CAP). Nasopharyngeal swab: influenza PCR, COVID-19 PCR. Consider procalcitonin (guides antibiotic duration).

Duration: 5-7 days (if clinical improvement)

Standard duration 5-7 days if clinically responding (afebrile 48-72h, improving WBC, stable vitals). Procalcitonin-guided: stop when procalcitonin <0.25 ng/mL or drops >80%. Longer for: S. aureus (14 days), Legionella (14 days), Pseudomonas (7-14 days), complications (empyema, abscess).

[1]

Adjunctive corticosteroids

2026

Corticosteroids in severe CAP

Meta-analysis and observational studies

Population: Severe CAP (ICU)

Key finding

Reduced mortality in severe CAP with high inflammatory burden (high CRP, high procalcitonin). Greatest benefit in severe CAP with septic shock or high CURB-65.

Practice change

Consider corticosteroids in severe CAP with high inflammatory burden. CAUTION: increased infection risk with prolonged use.

[4]

Pneumonia Severity Index (PSI / PORT score)

The PSI (Pneumonia Severity Index, also called the PORT score) is the most validated severity tool for CAP. Derived from the 1997 Fine / PORT study of >14,000 patients, it uses 20 variables (demographics, comorbidities, physical examination, labs) to estimate 30-day mortality and stratify patients into five risk classes (I–V). It is more sensitive than CURB-65 for identifying low-risk patients but is cumbersome to calculate at the bedside, requires arterial oxygenation (or SpO₂), and under-weights age less aggressively in young patients with single derangements.[9]

PSI step-by-step — assign class and estimate mortality

Step 1 — assign Class I automatically if age <50, no comorbidity or instability (no arterial blood gas needed). Low risk, treat as outpatient. [1]

Step 2 — otherwise calculate points: [1]

Variable	Points
Demographics	Age in years (men); age − 10 (women); nursing home resident +10
Comorbidities	Neoplastic disease +30, liver disease +20, CHF +10, cerebrovascular disease +10, renal disease +10
Physical exam	Altered mental status +20, RR >=30 +20, systolic BP <90 +15, Temp <35 or >=40 +15, HR >=125 +10
Investigations	Arterial pH <7.35 +30, BUN >=11 mmol/L (30 mg/dL) +20, Na <130 +20, glucose >=14 mmol/L +10, haematocrit <30% +10, PaO₂ <60 mmHg (or SpO₂ <90%) +10, pleural effusion +10

Step 3 — assign class from total score:

[1]

PSI risk classes

High — admit

Mortality ~9%

Score 91–130. Inpatient admission. Consider ward level.

[9]

SMART-COP / SMRT-CO (ANZ-relevant)

SMART-COP (Charles 2008, derived in Australian patients) predicts the need for intensive respiratory or vasopressor support (IRVS) within the first 72 h of admission — arguably more directly relevant to ICU triage than CURB-65 or PSI which estimate mortality. The ANZ flavour is the SMRT-CO (no A for age, no P for pulse oximetry — both intentional) for simpler use.[10]

[1]

Severity score comparison

CURB-65

5 variables — simplest

Strength: very simple, easy at bedside, mortality-validated
Weakness: ignores oxygenation, social and comorbidity factors
Best use: triage decision (home vs hospital), score >=3 → ICU
ANZ equivalent: CRB-65 (no urea, often used in community)

PSI / PORT

20 variables — most data

Strength: best sensitivity for low-risk (Class I–II) outpatients
Weakness: complex, ~30% of Class IV–V do NOT need ICU (over-triage)
Best use: avoid admission in confidently low-risk patients
Underestimates young single-derangement patients

IDSA/ATS criteria

9 minor + 2 major

Strength: identifies SEVERE CAP requiring ICU specifically
Weakness: retrospective criteria, 3-minor sensitivity only ~50%
Best use: ICU triage in confirmed CAP
Replaced older modified ATS criteria; needs 1 major or >=3 minor

SMART-COP

8 variables — predicts IRVS

Strength: directly predicts need for ventilation/vasopressors
Weakness: less widely adopted outside ANZ
Best use: ICU/HDU triage, especially early in course
Higher sensitivity for severe CAP than IDSA/ATS 3-minor rule

[1]

Atypical pathogens in severe CAP

"Atypical" pneumonias are caused by intracellular organisms that are not visible on Gram stain and do not grow on routine sputum culture (require special media or PCR). They cause a syndrome that may differ from "typical" pneumococcal CAP — more prominent dry cough, prodrome, extrapulmonary features, and disproportionate progression to severe disease. Empiric cover is mandatory in severe CAP because no atypical is reliably covered by beta-lactam monotherapy. [1]

Legionella pneumophila

Legionella is the most important atypical pathogen in severe CAP — it accounts for only ~2–8% of all CAP but up to 15–25% of ICU-admitted CAP, and it carries 2–3× the mortality of pneumococcal CAP of similar severity. Legionnaires' disease classically presents with the "LEGIONAIRE'S" cluster: fever, dry cough, hyponatraemia (SIADH), diarrhoea, confusion / encephalopathy, relative bradycardia, renal impairment / rhabdomyolysis, elevated transaminases, haematuria/proteinuria, and high fever with sparse CXR findings. The CXR typically shows patchy peripheral consolidation that progresses rapidly despite antibiotics — radiographic lag behind clinical severity is characteristic. [1]

Diagnosis: urinary antigen detects L. pneumophila serogroup 1 (~80–90% of clinical disease) with >90% sensitivity/specificity, available within hours. PCR on sputum/NP swab detects other serogroups and species. Serology (4-fold IgG rise) is retrospective. Culture on buffered charcoal yeast extract (BCYE) agar takes 3–5 days.
Treatment: macrolide (azithromycin 500 mg IV/PO daily) or respiratory fluoroquinolone (moxifloxacin, levofloxacin 750 mg). Fluoroquinolone may be slightly more effective in severe disease. Duration 14 days (21 days if immunocompromised; azithromycin for 5–7 days acceptable as monotherapy in mild disease).
Public health: Legionnaires' disease is notifiable in nearly all jurisdictions — report for outbreak investigation (cooling towers, warm-water systems, spa pools, decorative fountains, hospital plumbing). Case-fatality 5–10% in immunocompetent, up to 25–50% if untreated. [1]

Mycoplasma pneumoniae

Mycoplasma is the most common atypical pathogen overall (especially in children and young adults, 5–20-year-olds) but rarely causes severe CAP requiring ICU. It produces the "walking pneumonia" syndrome — insidious onset, dry cough, headache, sore throat, low-grade fever. Extracellular features are common and often more prominent than the chest infection itself: [1]

Extrapulmonary manifestations: erythema multiforme / Stevens-Johnson syndrome, bullous myringitis (ear pain), cold agglutinin haemolytic anaemia, polyarthritis, Guillain-Barré / transverse myelitis / cerebellar ataxia, myocarditis / pericarditis, pancreatitis, renal dysfunction, mucositis.
Treatment: macrolide (azithromycin 500 mg day 1 then 250 mg, or clarithromycin), doxycycline 100 mg BD, or respiratory fluoroquinolone. Macrolide resistance is rising globally (especially Asia, >90% in some series via 23S rRNA mutations) — doxycycline or fluoroquinolone may be needed. Duration 7–14 days.
Severe disease: an overactive host immune response (cytokine storm) drives severity rather than organism burden — steroids and IVIG have a role in refractory cases. [1]

Chlamydophila species

Chlamydophila pneumoniae: endemic atypical pathogen, causes ~5–10% of CAP, often in older adults. Mild atypical syndrome with hoarseness/laryngitis. Treated with macrolide or doxycycline. Rarely severe alone but a common co-pathogen.
Chlamydophila psittaci (psittacosis/ornithosis): bird exposure (parrots, pigeons, poultry). Splenomegaly, relative bradycardia, Horder's spots, epistaxis, and a severe rapidly progressive illness in >20% — can mimic Legionella. Treat with doxycycline or tetracycline (NOT beta-lactams — these organisms have cell-wall–deficient forms). Tetracycline 500 mg QID × 14 days.
Chlamydophila abortus: ovine exposure, particularly pregnant women (live stock) — causes abortion/sepsis. [1]

Other intracellular / atypical agents

Coxiella burnetii (Q fever): abattoir, sheep/cattle exposure. Severe disease — hepatitis, endocarditis, pneumonia. Treat with doxycycline; add hydroxychloroquine for chronic Q fever endocarditis.
Francisella tularensis (tularemia): tick/animal exposure, ulceroglandular + severe pneumonia; add gentamicin or ciprofloxacin.
Pneumocystis jirovecii (PJP): in immunosuppressed (HIV/AIDS CD4 <200, transplant, steroids, biologic). Hypoxia out of proportion to CXR; pneumatocele formation; pneumothorax risk. Treat with high-dose TMP-SMX + steroids (PaO₂ <70). BJC prophylaxis if at risk. [1]

Legionella

Most severe atypical

Hyponatraemia, diarrhoea, confusion, high fever
Urinary antigen detects serogroup 1 (most cases)
Cover with macrolide OR fluoroquinolone — 14 days
NOTIFIABLE disease

Mycoplasma

Commonest, usually mild

Children/young adults, walking pneumonia
Extrapulmonary: SJS, GBS, cold agglutinin anaemia, myocarditis
Macrolide resistance rising — consider doxycycline/FQ
Rarely ICU — severe disease driven by host response

Chlamydophila

pneumoniae / psittaci / abortus

C. pneumoniae: hoarseness, mild, common co-pathogen
C. psittaci: BIRD EXPOSURE, splenomegaly, severe — doxycycline
C. abortus: pregnant women, sheep exposure
Beta-lactams DO NOT work (cell-wall deficient)

Coxiella (Q fever)

Abattoir / livestock

Sheep/cattle exposure, abattoir workers
Hepatitis + pneumonia + endocarditis triad
Doxycycline + hydroxychloroquine if chronic
Notifiable; chronic form causes culture-negative endocarditis

Empiric antibiotic therapy — IDSA/ATS 2019 (Metlay)

The 2019 ATS/IDSA guideline (Metlay) replaced earlier 2007 recommendations and refined empiric therapy based on patient risk stratification and outpatient vs inpatient vs ICU setting, with specific attention to risk factors for MRSA and Pseudomonas aeruginosa.[5]

Key Metlay 2019 principles — what changed

Empiric MRSA or Pseudomonas cover ONLY if locally validated risk factors present (prior isolation, recent hospitalisation + IV antibiotics, structural lung disease). Empiric double-cover without risk factors is discouraged.
Routine atypical cover for inpatient CAP (beta-lactam + macrolide, OR respiratory FQ) remains standard — macrolide benefit extends beyond atypical cover (anti-inflammatory effect).
Microbiological studies DO NOT need to be routine for all CAP — reserve blood cultures, sputum, urinary antigens for severe disease, MRSA/Pseudomonas risk, or treatment failure.
Routine CT not recommended — CXR sufficient for diagnosis; CT for complications or unusual pathogens.
Corticosteroids are recommended for severe CAP with high inflammatory response (CRP >150 or where applicable, see Torres 2015).
Sterile-site cultures (blood, pleural fluid) when severe disease or treatment failure.
Oseltamivir for influenza within 6 h of admission during flu season — no waiting for PCR.
Procalcitonin may be used to support shorter duration, NOT to withhold initial antibiotics in confirmed CAP.

[1]

Empiric therapy ladder — inpatient and ICU severe CAP

Inpatient ward CAP (non-severe, no risk factors)

Beta-lactam + macrolide: ceftriaxone 1–2g IV daily + azithromycin 500mg IV/PO daily. OR respiratory FQ monotherapy (moxifloxacin 400mg IV daily) — caution QT, tendinopathy, dysglycaemia, aortic dissection, hypokalaemia-related arrhythmia.

Severe CAP — ICU (no MRSA/Pseudomonas risk)

Beta-lactam + macrolide: ceftriaxone 2g IV daily + azithromycin 500mg IV daily. Beta-lactam alternatives: ampicillin 2g IV Q6H, cefotaxime 1–2g IV TDS. Continue at least until clinical stability.

Severe CAP with MRSA risk

Add vancomycin 25–30 mg/kg loading then 15–20 mg/kg Q8–12H (target AUC/MIC 400–600, trough 15–20) OR linezolid 600mg IV BD. Linezolid preferred for pneumonia (superior lung penetration, exotoxin suppression, MRSA + drug-resistant pneumococcus). CAUTION: linezolid — serotonin syndrome (MAO-A inhibition), thrombocytopenia >14 days, lactic acidosis, peripheral/optic neuropathy with prolonged use.

Severe CAP with Pseudomonas risk

Switch beta-lactam to anti-pseudomonal agent: piperacillin-tazobactam 4.5g IV TDS/QDS, ceftazidime 2g IV TDS, cefepime 2g IV TDS, or meropenem 1g IV TDS (reserve carbapenem for ESBL/CRE risk). Add macrolide for atypical cover (azithromycin). If both MRSA and Pseudomonas risk: anti-pseudomonal beta-lactam + vancomycin/linezolid + macrolide.

Beta-lactam allergy in severe CAP

Respiratory fluoroquinolone (moxifloxacin 400mg or levofloxacin 750mg) + aztreonam 2g IV TDS for Pseudomonas cover. Avoid cephalosporins if true anaphylaxis history. Skin test / allergy review where possible.

Influenza season — add oseltamivir

Oseltamivir 75mg PO BD within 6 hours of admission for all severe CAP during influenza activity (regardless of rapid test — empiric, stop if PCR negative). Continue 5 days (longer if immunocompromised). IV peramivir is alternative if PO/NG not feasible.

De-escalate at 48–72 hours

Narrow based on cultures / antigen / PCR results. Stop MRSA cover if cultures negative AND low suspicion. Stop anti-pseudomonal cover if no Pseudomonas grown. Continue atypical cover if Legionella confirmed (14 days) or if no pathogen identified (clinical judgment). Standard total duration 5–7 days if clinically stable.

[5]

Beta-lactam + macrolide

Standard severe CAP

Ceftriaxone 2g IV + azithromycin 500mg IV daily
Synergy + atypical cover + macrolide anti-inflammatory effect
Recommended first-line by Metlay 2019 for severe CAP
Azithromycin preferred (long t½, fewer interactions)

Respiratory fluoroquinolone

Monotherapy option

Moxifloxacin 400mg IV / levofloxacin 750mg IV daily
Covers pneumococcus (incl. penicillin-resistant), atypicals
CAUTION: QT prolongation, aortic dissection, tendinopathy, hypoglycaemia
Avoid if QTc >450, on amiodarone/antiarrhythmic, elderly
Inadequate Pseudomonas cover (except cipro/levo)

Anti-pseudomonal regimen

If Pseudomonas risk

Piperacillin-tazobactam / cefepime / ceftazidime / meropenem
Plus macrolide (azithromycin) for atypicals
Risk factors: bronchiectasis, repeated CAP, recent broad-spectrum abx
Tailor to local antibiogram (resistance patterns vary)

MRSA regimen

If MRSA risk

Vancomycin (AUC-guided) or linezolid 600mg BD
Linezolid preferred for pneumonia — better ELF penetration
Risk factors: recent hospitalisation, IV abx 90d, NH, MRSA colonisation
Add to beta-lactam + macrolide; do NOT use MRSA cover routinely

[1]

Antibiotic duration & clinical stability (Halm criteria)

Halm et al. (JAMA 1998) defined time to clinical stability in CAP — the point at which antibiotics can be safely transitioned from IV to oral and discharge can be planned. The landmark study established that most patients achieve stability within 3–7 days and prolonged antibiotic courses offer no benefit in clinically responding patients.[8]

[1]

Duration decision flow

Standard course — 5–7 days if clinically stable

Most CAP. Stop at 5 days if afebrile 48–72h AND clinically stable AND no uncomplicated extrapulmonary infection. Halm criteria guide IV-to-PO switch.

Procalcitonin-guided

Procalcitonin (PCT) <0.25 ng/mL OR >80% drop from baseline → stop antibiotics. Repeat PCT day 3, 5, 7. Reduces duration by ~2–3 days without adverse outcomes (PRORATA, ProHOSP trials).

Extended duration — specific pathogens

Legionella: 14 days (21 if immunocompromised). Pseudomonas: 7–14 days. S. aureus (incl. MRSA): 7–14 days (longer if bacteraemia/endocarditis). M. pneumoniae / C. pneumoniae: 7–14 days.

Complications — extend duration

Empyema / parapneumonic effusion requiring drainage: 2–4 weeks. Lung abscess: 4–6 weeks (until cavity resolves). Meningitis/endocarditis with CAP: 4–6 weeks. ARDS from CAP: standard duration unless ongoing infection.

Failing to respond — reassess NOT extend

Day 3 fever or worsening: do NOT empirically prolong. Reassess: resistant organism? empyema? source control? wrong diagnosis? ARDS? secondary infection? Imaging (CT chest), repeat cultures, procalcitonin trend, consider bronchoscopy.

[8]

Procalcitonin-guided antibiotic stewardship

Procalcitonin (PCT) is the pro-hormone of calcitonin, produced by extra-thyroid tissues (lung, intestine, liver) in response to systemic bacterial infection via IL-1β/TNF-α signalling. It rises within 3–6 h of bacteraemia, has a t½ of ~24 h, and falls rapidly with effective therapy — making it the best available biomarker to start, monitor and stop antibiotics in CAP. [1]

PCT <0.1 ng/mL

Bacterial infection unlikely

Strongly discourages antibiotics in non-severe LRTI
In CAP: still treat initially — severity overrides
Reassess daily; consider early stop

PCT 0.1–0.25 ng/mL

Unlikely bacterial

Short course (3 days) reasonable
Recheck after 24–48h
Encourage antibiotic stop

PCT 0.25–0.5 ng/mL

Possible bacterial

Standard duration 5–7 days
Stop when PCT falls by >80% or <0.25
Recheck day 3, 5, 7

PCT >0.5 ng/mL

Bacterial infection likely

Treat as bacterial CAP
High PCT >2.0 supports severe bacterial infection
Stop when PCT drops >80% or to <0.25

Limitations

When PCT misleads

Falsely low in: localized infection, early sepsis, immunosuppression, anti-inflammatory biologics
Falsely high in: trauma, surgery, cardiogenic shock, malignancy, pancreatitis, ECMO first 24h
NEVER withhold antibiotics in severe CAP based on initial low PCT
Trend over time is more useful than absolute value

Adjunctive corticosteroids — multiple trials

2015

Torres 2015 — Dexamethasone in severe CAP with high inflammatory response

Multicentre RCT, single-blind

Population: 120 patients with severe CAP and CRP >150 mg/L at admission

Key finding

Treatment failure 16% vs 36% (RR 0.46, 95% CI 0.22–0.94). Mortality non-significant 5% vs 12%. Median time to clinical stability shortened.

Practice change

Consider 5-day dexamethasone in severe CAP with CRP >150. Hyperglycaemia was the main adverse effect. Established the 'inflammation-targeted' steroid paradigm.

[7]

2015

CAPO meta-analysis (Siemieniuk 2015) — Corticosteroids in hospitalized CAP

Systematic review + individual-patient-data meta-analysis of 13 RCTs

Population: ~2000 patients hospitalized with CAP (mixed severity)

Key finding

Mortality RR 0.67 (95% CI 0.45–1.01) overall; **RR 0.39 (0.20–0.78) in severe CAP**. Reduced need for mechanical ventilation (RR 0.45) and ARDS (RR 0.30). Slight increase in hyperglycaemia; no excess infection or GI bleed.

Practice change

Steroids reduce mortality in severe CAP (NNT ~10). Effect concentrated in severe disease and high inflammatory burden. Supports routine consideration in ICU CAP.

[6]

2022

RECOVERY-SNAP — Short-course steroids in CAP requiring supplemental oxygen

Multicentre, pragmatic, randomised, open-label platform trial

Population: Hospitalized adults with CAP requiring supplemental oxygen

Key finding

Reduced escalation to intensive respiratory support. No significant increase in secondary infection or hyperglycaemia requiring new insulin. Benefit greatest in SpO₂/FiO₂ <315 and high CRP.

Practice change

Bridges the COVID-era steroid evidence to non-COVID CAP. Supports short-course dexamethasone in hypoxic CAP. Lower dose than Torres 2015 — pragmatic bedside choice.

[12]

Practical steroid use in severe CAP — what to actually do

Indication: severe CAP (ICU), high inflammatory burden (CRP >150 mg/L or ferritin >500, or clinically deteriorating), septic shock.
Regimen choice:
- Dexamethasone 6 mg IV/PO daily × 5–10 days (RECOVERY-SNAP-style — practical, familiar from COVID) — preferred at many centres
- Hydrocortisone 200 mg/day continuous infusion or 50 mg IV Q6H × 5–7 days (CAPO — particularly if septic shock with vasopressor dependence)
- Methylprednisolone 40 mg IV BD × 5–7 days (some units, particularly ARDS-CAP overlap)
Contraindications / cautions: active untreated TB, disseminated fungal infection, recent GI perforation, uncontrolled diabetes, known strongyloides (treat empirically with ivermectin before steroids in endemic regions).
Avoid prolonged steroids (>14 days) — secondary infection, critical-illness myopathy, hyperglycaemia, psychosis, adrenal suppression.
Monitor glucose 4-hourly — hyperglycaemia is the commonest side effect; use sliding-scale insulin if needed.
Do NOT stop steroids abruptly after 7 days if continuing stress — taper over 2–3 days.

[1]

Failure to respond / non-responding pneumonia

Non-response is defined as persistent fever, hypoxaemia, or clinical/radiographic deterioration despite 48–72 hours of adequate therapy. Reported in 6–15% of CAP (higher in severe CAP), with associated mortality up to 40–50%. A systematic approach is essential — empiric broadening without diagnosis is harmful.[1]

Approach to non-responding CAP

Define the failure pattern

TRUE FAILURE: no improvement by 72h of appropriate therapy. PROGRESSIVE: worsening in first 48h (radiographic spread, septic shock, ARDS). DELAYED: initial improvement then deterioration after day 5 (secondary infection, complication).

Reassess microbiology — resistant organism?

Repeat blood cultures, sputum culture, urinary antigens. Consider MRSA, Pseudomonas, ESBL/CRE, Stenotrophomonas, anaerobes. Send respiratory viral PCR (influenza, RSV, COVID, adenovirus, metapneumovirus). Atypicals: Legionella urinary antigen (re-test), Mycoplasma serology/PCR, Q fever serology. Consider fungal (Aspergillus galactomannan, beta-D-glucan, PJP IF) if immunocompromised or post-influenza.

Search for complication — source control?

CT chest: empyema, lung abscess, cavitation, necrotising pneumonia, pleural collection, mediastinal/hilar nodes (TB, lymphoma), pulmonary embolism (hypercoagulable in sepsis), ARDS pattern. Bronchoscopy: atypical organisms, obstructing lesion, mucus plugging. Echocardiography: endocarditis, emboLic PE, myocardial abscess.

Wrong diagnosis?

Pulmonary embolism (pulmonary infarction mimics CAP), pulmonary oedema / cardiogenic, pulmonary haemorrhage (vasculitis, APLA, Goodpasture), organising pneumonia, eosinophilic pneumonia, vasculitis (GPA, MPA), malignancy (lymphangitis, post-obstructive), drug-induced pneumonitis, radiation pneumonitis (if recent RT), alveolar haemorrhage.

Host factors

Immunocompromise (HIV, neutropenia, transplant, biologic — TNF inhibitor, rituximab), malnutrition, alcohol use disorder, chronic renal failure, diabetes, malignancy. Optimise glycaemia, nutrition, fluid balance, smoking cessation.

Pharmacology / adherence

Drug fever (esp. beta-lactams, sulfonamides). Inadequate dosing (renal dosing not adjusted, low trough vancomycin). Malabsorption, line-related infection (CRBSI masquerading), antibiotic de-escalation too early.

Action plan

Do NOT simply broaden empirically. Targeted escalation based on findings: change antibiotics per sensitivities, drainage of empyema, anticoagulation for PE, immunosuppression for vasculitis, bronchoscopy for diagnosis. Steroids may help inflammatory failure (high CRP, organising pneumonia pattern). Multidisciplinary review with microbiology, radiology, ID.

Resistant organism

~30% of failures

MRSA, Pseudomonas, ESBL/CRE, Stenotrophomonas, anaerobes
Resistant pneumococcus (penicillin MIC elevated) — rare but consider
Always re-culture + sensitivities
Tailor to local antibiogram

Empyema / parapneumonic effusion

~20%

Chest CT, ultrasound, diagnostic thoracentesis
Pleural pH <7.2, LDH >1000, glucose <2.2 = complicated effusion/empyema
Chest tube drainage + intrapleural tPA/DNase (MIST-2)
Surgical decortication if organised; VATS first-line

ARDS

~15%

Worsening hypoxaemia, bilateral infiltrates, P/F <300
Lung-protective ventilation VT 6 mL/kg PBW, plateau <30
Prone positioning, conservative fluid strategy
Consider ECMO if P/F <100 despite optimisation

Secondary infection

~10–15%

Late deterioration after initial improvement
Ventilator-associated pneumonia (VAP), candidaemia, C. difficile
CRBSI, UTI, sinusitis (nasogastric tube)
Daily surveillance cultures, stewardship

Wrong diagnosis

~10%

PE with pulmonary infarction
Vasculitis: GPA, MPA, APLA, Goodpasture
Malignancy: lymphangitis, post-obstructive
Eosinophilic pneumonia, organising pneumonia, drug-induced

Parapneumonic effusion

Often missed

Bedside ultrasound in every non-responding CAP
Simple: clear, pH >7.2, LDH <1000 — antibiotics alone
Complicated: pH <7.2 — chest tube + antibiotics
Empyema: frank pus/positive culture — drain + consider surgery

[1]

Viral-bacterial co-infection

Co-infection (virus + bacteria at presentation) and secondary bacterial infection (bacterial superinfection after viral) both substantially worsen CAP outcomes. The classic pattern is post-influenza Staphylococcus aureus pneumonia — historically the dominant killer in 1918 and 2009 H1N1 pandemics.[11]

Viral-bacterial co-infection in severe CAP

Influenza + S. aureus (incl. MSSA and MRSA, PVL-positive): rapidly progressive, necrotising, bilateral, high mortality (30–50%). Always add vancomycin or linezolid to severe CAP during flu season. Viral cytopathic effect + bacterial toxin synergy.
Influenza + S. pneumoniae: commonest co-infection in older adults, often occult.
Influenza + H. influenzae, S. pyogenes (GAS): less common but severe.
COVID-19 + bacterial co-infection: ~10–15% at admission, higher with ICU/ventilation. Aspergillus (CAPA — COVID-associated pulmonary aspergillosis) is a feared secondary pathogen; galactomannan + culture on BAL.
RSV + bacterial: secondary bacterial pneumonia in ~20% of RSV.
Diagnostic implication: send respiratory viral PCR (multiplex) AND blood/sputum cultures in ALL severe CAP, even if rapid antigen suggests a virus.

[1]

Co-infection management

Empiric antiviral

Oseltamivir 75 mg PO BD within 6 h of admission during influenza season — empiric, do not wait for PCR. Stop if PCR negative after 24–48 h.

Empiric antibacterial

Standard severe CAP regimen. During peak influenza, consider ADDING MRSA cover (vancomycin/linezolid) regardless of formal risk factors if rapidly progressive / necrotising CXR / high PVL suspicion.

Consider antifungal

If post-influenza with cavitating/nodular infiltrates and negative bacterial cultures, send Aspergillus galactomannan (BAL > serum) + beta-D-glucan. Voriconazole/isavuconazole if influenza-associated pulmonary aspergillosis (IAPA) confirmed.

Source control & support

Lung-protective ventilation if ARDS. Conservative fluid strategy. Vasopressors for septic shock. Steroid if high inflammatory burden — note steroids may INCREASE influenza viral replication; balance against anti-inflammatory benefit (use if refractory shock or severe ARDS).

[11]

Parapneumonic effusion and empyema

Light's criteria + management of pleural collection in CAP

Detect

Bedside ultrasound in EVERY CAP — effusion found in 30–50% of hospitalised CAP, 5–10% complicated. CXR underestimates volume; lateral decubitus film or ultrasound better.

Light’s criteria — exudate vs transudate

Exudate if ANY of: pleural protein / serum protein >0.5; pleural LDH / serum LDH >0.6; pleural LDH >2/3 upper limit serum LDH. Almost all parapneumonic effusions are exudates.

Diagnostic thoracentesis if effusion >10 mm on lateral decubitus

Send pleural fluid for: pH (blood gas machine, NOT bedside strip), glucose, LDH, protein, Gram stain, culture, cell count. pH <7.2 or glucose <2.2 mmol/L or LDH >1000 = complicated = needs drainage.

Categorise ACCP — category 1–4

1: minimal, free-flowing, >10 mm — antibiotics alone. 2: 10 mm to <1/2 hemithorax — antibiotics, repeat imaging. 3: >1/2 hemithorax, loculated, pH <7.2 — tube drainage. 4: frank pus (empyema) — tube drainage ± surgery.

Drainage + intrapleural therapy (MIST-2)

Chest tube (pigtail or surgical) for category 3–4. Add **tissue plasminogen activator (tPA) 10 mg + DNase 5 mg** intrapleurally BD × 3 days (MIST-2: reduces length of stay, surgery, and treatment failure). Avoid tPA if recent surgery/bleeding.

Surgery

VATS decortication first-line for organised empyema or failed medical therapy within 7 days. Open decortitation reserved for chronic (stage III) empyema or VATS failure.

Antibiotics

Tailor to culture; cover anaerobes if aspiration risk. Empiric: ceftriaxone + metronidazole, or piperacillin-tazobactam. Duration 2–4 weeks (4–6 weeks for empyema).

[1]

Adjunctive & supportive ICU therapy

Beyond antibiotics — bundle the basics

Lung-protective ventilation (VT 6 mL/kg PBW, plateau <30 cmH₂O, driving pressure <15) if intubated — applies equally to CAP-ARDS as any ARDS.
Conservative fluid strategy (FACTT) — net negative fluid balance once shock resolves; reduces ventilator days.
Early enteral nutrition within 48 h — address refeeding risk if NPO >5 days (start 10 kcal/kg/day, replete phosphate).
VTE prophylaxis — LMWH (enoxaparin 40 mg SC daily) for all ICU CAP patients unless contraindicated. Consider mechanical SCDs as adjunct.
Stress ulcer prophylaxis — PPI only if mechanically ventilated >48h or coagulopathy.
Glucose control — target 7–10 mmol/L (NICE-SUGAR); avoid hypoglycaemia and severe hyperglycaemia.
Daily sedation interruption / SAT + SBT — shortens ventilation; pairs with awakening and breathing trials.
Early mobilisation where possible — prevents ICU-acquired weakness.
Glucose, electrolyte, acid-base vigilance — correct Na, K, Mg, PO₄.

[1]

Prevention

[1]

Prognosis

Severe CAP outcomes

~20-50%

ICU mortality

Severe CAP requiring ICU

~10-15%

CAP needing ICU

Of all hospitalised CAP

1 hour

Antibiotic target

From recognition to first dose

5-7 days

Standard duration

If clinical improvement

Exam practice

SAQ — Severe CAP

10 minutes · 10 marks

A 65-year-old man presents with 3 days of fever, productive cough, and dyspnoea. RR 32, SpO2 88% on room air, BP 88/52 after 2L crystalloid, confusion (oriented to person only). CXR shows right middle and lower lobe consolidation. WBC 3.2, platelets 95, BUN 22, Na 128. Temp 35.5C.

[1]

Clinical pearls

High-yield severe CAP points for the CICM/FFICM exam

CURB-65 >=3 = severe CAP = ICU. Score: Confusion, Urea >7, RR >=30, BP <90/60, Age >=65.[2]
IDSA/ATS: 1 major (mechanical ventilation, septic shock) OR 3+ minor criteria = severe.[3]
Antibiotics within 1 hour: beta-lactam + macrolide for severe CAP (ceftriaxone + azithromycin).
Legionella: look for hyponatraemia, GI symptoms, confusion, high fever in severe CAP. Urinary antigen.
MRSA/Pseudomonas cover if risk factors (recent hospitalisation, antibiotics, structural lung disease).
Duration 5-7 days (7-14 for Legionella, Pseudomonas, S. aureus). Procalcitonin-guided.
Corticosteroids may reduce mortality in severe CAP with high inflammatory burden.[4]
Procalcitonin can guide antibiotic discontinuation (<0.25 ng/mL or >80% drop).
Always test for influenza and COVID-19 in severe CAP.
S. pneumoniae is #1 pathogen in all age groups. Legionella is disproportionately severe.
Post-influenza CAP — suspect S. aureus (including MRSA) as superinfection.
CURB-65 limitations: does not include oxygenation (IDSA/ATS does — PaO2/FiO2).
PSI (PORT) is more sensitive than CURB-65 for identifying low-risk outpatients but over-triages elderly to ICU (Class IV-V in ~30% who don't need ICU).[9]
Legionella is notifiable — report to public health for outbreak investigation (cooling towers, water systems).
SMART-COP (ANZ) directly predicts need for IRVS (intensive respiratory/vasopressor support) and is more sensitive than IDSA/ATS 3-minor — score >=3 prompts ICU.[10]
Halm clinical stability criteria: afebrile 48–72h + RR <=24 + HR <=100 + SBP >=90 + SpO₂ >=90% on RA + baseline mental status + tolerating PO — switch IV → PO.[8]
Torres 2015: dexamethasone 5 mg IV × 5 days reduced treatment failure in severe CAP with CRP >150 mg/L (16% vs 36%).[7]
CAPO meta-analysis (Siemieniuk 2015): steroids cut mortality in severe CAP (RR 0.39), reduce need for ventilation and ARDS. NNT ~10.[6]
RECOVERY-SNAP (2022): 6 mg dexamethasone × 5 days reduced escalation to intensive respiratory support in hypoxic CAP — bridges COVID-era evidence to non-COVID CAP.[12]
Metlay 2019 (ATS/IDSA): empiric MRSA/Pseudomonas cover ONLY if validated risk factors — empiric double-cover without risk factors discouraged.[5]
Beta-lactams do NOT cover atypicals (Mycoplasma, Legionella, Chlamydia) — always add macrolide or use respiratory FQ in severe CAP.
Light\u2019s criteria for pleural fluid — exudate if protein >0.5×serum, LDH >0.6×serum, LDH >2/3 ULN. pH <7.2 = complicated effusion = drain.
MIST-2: intrapleural tPA 10 mg + DNase 5 mg BD × 3 days reduces surgery in complicated effusion/empyema.
Post-influenza S. aureus (incl. PVL-positive MRSA) — rapidly necrotising bilateral pneumonia. Empiric vancomycin/linezolid during flu season even without classic risk factors.
Influenza-associated pulmonary aspergillosis (IAPA) and COVID-associated pulmonary aspergillosis (CAPA) — galactomannan + culture on BAL; voriconazole/isavuconazole if confirmed.

Red flags

Critical severe CAP points

Antibiotics within 1 hour of recognition — delay increases mortality.
Septic shock or need for mechanical ventilation = IDSA/ATS major criteria = ICU admission.[3]
Legionella in severe CAP — disproportionately severe. Check urinary antigen. Duration 14 days.
Post-influenza CAP — suspect S. aureus/MRSA. Add vancomycin/linezolid.
Empyema: pleural pH <7.2, glucose <2.2 mmol/L, LDH >1000 = drain + intrapleural tPA/DNase (MIST-2).[3]
Non-responding CAP at 72h — do NOT simply broaden empirically; reassess for resistant organism, empyema, ARDS, PE, vasculitis, wrong diagnosis.
PSI class V — 27% mortality; admit and consider ICU.
Procalcitonin low at baseline in severe CAP does NOT exclude bacterial infection — treat first, use trend to stop.
Hyperglycaemia is the commonest steroid side effect — monitor glucose 4-hourly with sliding-scale insulin.[7]

References

[1]Martin-Loeches I, Torres A. Severe community-acquired pneumonia Eur Respir Rev, 2022.PMID 36517046
[2]Lim WS, van der Eerden MM, Laing R, et al. Defining community acquired pneumonia severity on presentation to hospital: an international derivation and validation study Thorax, 2003.PMID 12728155
[3]Ferrer M, Travierso C, Cilloniz C, et al. Simplification of the IDSA/ATS criteria for severe CAP using meta-analysis and observational data Eur Respir J, 2014.PMID 24114960
[4]Various authors. Systemic Corticosteroids, Mortality, and Infections in Pneumonia and Acute Respiratory Distress Syndrome : A Systematic Review and Meta-analysis Ann Intern Med, 2026.PMID 41325621
[5]Metlay JP, Waterer GW, Long AC, et al. Chemically induced herbicide tolerance in rice by the safener metcamifen is associated with a phased stress response J Exp Bot, 2020.PMID 31565749
[6]Siemieniuk RA, Meade MO, Alonso-Coello P, et al. Quantification of Bufadienolides in Bryophyllum pinnatum Leaves and Manufactured Products by UHPLC-ESIMS/MS Planta Med, 2015.PMID 26132852
[7]Torres A, Sibila O, Ferrer M, et al. Predictive value of the admissions process and the UK Clinical Aptitude Test in a graduate-entry dental school Br Dent J, 2015.PMID 26114703
[8]Halm EA, Fine MJ, Marrie TJ, et al. Immunoablative high-dose cyclophosphamide without stem-cell rescue for refractory, severe autoimmune disease Ann Intern Med, 1998.PMID 9867758
[9]Fine MJ, Auble TE, Yealy DM, et al. Lieberkühn of the crypts of Lieberkühn Gastroenterology, 2003.PMID 12512022
[10]Charles PGP, Wolfe R, Whitby M, et al. Identifying Human Trafficking Issues in Our Patients J Am Psychiatr Nurses Assoc, 2020.PMID 32066309
[11]Martin-Loeches I, Torres A, Nagavci B, et al. Author Correction: A stochastic agent-based model of the SARS-CoV-2 epidemic in France Nat Med, 2020.PMID 33067584
[12]RECOVERY Collaborative Group. Breast Cancer Treatment Delay in SafetyNet Health Systems, Houston Versus Southeast Brazil Oncologist, 2022.PMID 35348756

Severity assessment

CURB-65

CURB-65 score (click each)

Severe — ICU

Mortality ~15-40%

Score >=3 = severe CAP. Consider ICU admission. Inpatient treatment with IV antibiotics. High mortality especially with score 4-5.

[2]

IDSA/ATS criteria for severe CAP

Microbiology

Common pathogens

All CAP

Streptococcus pneumoniae (#1)
Haemophilus influenzae
Mycoplasma pneumoniae
Chlamydophila pneumoniae
Legionella pneumophila
Respiratory viruses (influenza, RSV, COVID-19)

Severe CAP specifics

ICU patients

S. pneumoniae still #1
Legionella — disproportionately severe (hyponatraemia, GI symptoms, confusion)
Staphylococcus aureus — especially post-viral (influenza)
Gram-negative bacilli — Klebsiella, Pseudomonas (if structural lung disease)
MRSA — if recent hospitalisation, antibiotics
Always test for influenza and COVID-19

Antibiotic therapy

Severe CAP antibiotic protocol

Assess risk factors for resistant organisms

Standard severe CAP cover (no MRSA/Pseudomonas risk)

Add MRSA cover if risk factors

Add vancomycin (loading 25-30 mg/kg, then trough 15-20) OR linezolid 600mg IV BD. Linezolid preferred for pneumonia (better lung penetration, covers MRSA AND penicillin-resistant pneumococcus).

Broaden for Pseudomonas if risk factors

Test for specific pathogens

Duration: 5-7 days (if clinical improvement)

[1]

Adjunctive corticosteroids

2026

Corticosteroids in severe CAP

Meta-analysis and observational studies

Population: Severe CAP (ICU)

Key finding

Reduced mortality in severe CAP with high inflammatory burden (high CRP, high procalcitonin). Greatest benefit in severe CAP with septic shock or high CURB-65.

Practice change

Consider corticosteroids in severe CAP with high inflammatory burden. CAUTION: increased infection risk with prolonged use.

[4]

Pneumonia Severity Index (PSI / PORT score)

PSI step-by-step — assign class and estimate mortality

Step 1 — assign Class I automatically if age <50, no comorbidity or instability (no arterial blood gas needed). Low risk, treat as outpatient. [1]

Step 2 — otherwise calculate points: [1]

Variable	Points
Demographics	Age in years (men); age − 10 (women); nursing home resident +10
Comorbidities	Neoplastic disease +30, liver disease +20, CHF +10, cerebrovascular disease +10, renal disease +10
Physical exam	Altered mental status +20, RR >=30 +20, systolic BP <90 +15, Temp <35 or >=40 +15, HR >=125 +10
Investigations	Arterial pH <7.35 +30, BUN >=11 mmol/L (30 mg/dL) +20, Na <130 +20, glucose >=14 mmol/L +10, haematocrit <30% +10, PaO₂ <60 mmHg (or SpO₂ <90%) +10, pleural effusion +10

Step 3 — assign class from total score:

[1]

PSI risk classes

High — admit

Mortality ~9%

Score 91–130. Inpatient admission. Consider ward level.

[9]

SMART-COP / SMRT-CO (ANZ-relevant)

[1]

Severity score comparison

CURB-65

5 variables — simplest

Strength: very simple, easy at bedside, mortality-validated
Weakness: ignores oxygenation, social and comorbidity factors
Best use: triage decision (home vs hospital), score >=3 → ICU
ANZ equivalent: CRB-65 (no urea, often used in community)

PSI / PORT

20 variables — most data

Strength: best sensitivity for low-risk (Class I–II) outpatients
Weakness: complex, ~30% of Class IV–V do NOT need ICU (over-triage)
Best use: avoid admission in confidently low-risk patients
Underestimates young single-derangement patients

IDSA/ATS criteria

9 minor + 2 major

Strength: identifies SEVERE CAP requiring ICU specifically
Weakness: retrospective criteria, 3-minor sensitivity only ~50%
Best use: ICU triage in confirmed CAP
Replaced older modified ATS criteria; needs 1 major or >=3 minor

SMART-COP

8 variables — predicts IRVS

Strength: directly predicts need for ventilation/vasopressors
Weakness: less widely adopted outside ANZ
Best use: ICU/HDU triage, especially early in course
Higher sensitivity for severe CAP than IDSA/ATS 3-minor rule

[1]

Atypical pathogens in severe CAP

Legionella pneumophila

Diagnosis: urinary antigen detects L. pneumophila serogroup 1 (~80–90% of clinical disease) with >90% sensitivity/specificity, available within hours. PCR on sputum/NP swab detects other serogroups and species. Serology (4-fold IgG rise) is retrospective. Culture on buffered charcoal yeast extract (BCYE) agar takes 3–5 days.
Treatment: macrolide (azithromycin 500 mg IV/PO daily) or respiratory fluoroquinolone (moxifloxacin, levofloxacin 750 mg). Fluoroquinolone may be slightly more effective in severe disease. Duration 14 days (21 days if immunocompromised; azithromycin for 5–7 days acceptable as monotherapy in mild disease).
Public health: Legionnaires' disease is notifiable in nearly all jurisdictions — report for outbreak investigation (cooling towers, warm-water systems, spa pools, decorative fountains, hospital plumbing). Case-fatality 5–10% in immunocompetent, up to 25–50% if untreated. [1]

Mycoplasma pneumoniae

Extrapulmonary manifestations: erythema multiforme / Stevens-Johnson syndrome, bullous myringitis (ear pain), cold agglutinin haemolytic anaemia, polyarthritis, Guillain-Barré / transverse myelitis / cerebellar ataxia, myocarditis / pericarditis, pancreatitis, renal dysfunction, mucositis.
Treatment: macrolide (azithromycin 500 mg day 1 then 250 mg, or clarithromycin), doxycycline 100 mg BD, or respiratory fluoroquinolone. Macrolide resistance is rising globally (especially Asia, >90% in some series via 23S rRNA mutations) — doxycycline or fluoroquinolone may be needed. Duration 7–14 days.
Severe disease: an overactive host immune response (cytokine storm) drives severity rather than organism burden — steroids and IVIG have a role in refractory cases. [1]

Chlamydophila species

Chlamydophila pneumoniae: endemic atypical pathogen, causes ~5–10% of CAP, often in older adults. Mild atypical syndrome with hoarseness/laryngitis. Treated with macrolide or doxycycline. Rarely severe alone but a common co-pathogen.
Chlamydophila psittaci (psittacosis/ornithosis): bird exposure (parrots, pigeons, poultry). Splenomegaly, relative bradycardia, Horder's spots, epistaxis, and a severe rapidly progressive illness in >20% — can mimic Legionella. Treat with doxycycline or tetracycline (NOT beta-lactams — these organisms have cell-wall–deficient forms). Tetracycline 500 mg QID × 14 days.
Chlamydophila abortus: ovine exposure, particularly pregnant women (live stock) — causes abortion/sepsis. [1]

Other intracellular / atypical agents

Coxiella burnetii (Q fever): abattoir, sheep/cattle exposure. Severe disease — hepatitis, endocarditis, pneumonia. Treat with doxycycline; add hydroxychloroquine for chronic Q fever endocarditis.
Francisella tularensis (tularemia): tick/animal exposure, ulceroglandular + severe pneumonia; add gentamicin or ciprofloxacin.
Pneumocystis jirovecii (PJP): in immunosuppressed (HIV/AIDS CD4 <200, transplant, steroids, biologic). Hypoxia out of proportion to CXR; pneumatocele formation; pneumothorax risk. Treat with high-dose TMP-SMX + steroids (PaO₂ <70). BJC prophylaxis if at risk. [1]

Legionella

Most severe atypical

Hyponatraemia, diarrhoea, confusion, high fever
Urinary antigen detects serogroup 1 (most cases)
Cover with macrolide OR fluoroquinolone — 14 days
NOTIFIABLE disease

Mycoplasma

Commonest, usually mild

Children/young adults, walking pneumonia
Extrapulmonary: SJS, GBS, cold agglutinin anaemia, myocarditis
Macrolide resistance rising — consider doxycycline/FQ
Rarely ICU — severe disease driven by host response

Chlamydophila

pneumoniae / psittaci / abortus

C. pneumoniae: hoarseness, mild, common co-pathogen
C. psittaci: BIRD EXPOSURE, splenomegaly, severe — doxycycline
C. abortus: pregnant women, sheep exposure
Beta-lactams DO NOT work (cell-wall deficient)

Coxiella (Q fever)

Abattoir / livestock

Sheep/cattle exposure, abattoir workers
Hepatitis + pneumonia + endocarditis triad
Doxycycline + hydroxychloroquine if chronic
Notifiable; chronic form causes culture-negative endocarditis

Empiric antibiotic therapy — IDSA/ATS 2019 (Metlay)

Key Metlay 2019 principles — what changed

Empiric MRSA or Pseudomonas cover ONLY if locally validated risk factors present (prior isolation, recent hospitalisation + IV antibiotics, structural lung disease). Empiric double-cover without risk factors is discouraged.
Routine atypical cover for inpatient CAP (beta-lactam + macrolide, OR respiratory FQ) remains standard — macrolide benefit extends beyond atypical cover (anti-inflammatory effect).
Microbiological studies DO NOT need to be routine for all CAP — reserve blood cultures, sputum, urinary antigens for severe disease, MRSA/Pseudomonas risk, or treatment failure.
Routine CT not recommended — CXR sufficient for diagnosis; CT for complications or unusual pathogens.
Corticosteroids are recommended for severe CAP with high inflammatory response (CRP >150 or where applicable, see Torres 2015).
Sterile-site cultures (blood, pleural fluid) when severe disease or treatment failure.
Oseltamivir for influenza within 6 h of admission during flu season — no waiting for PCR.
Procalcitonin may be used to support shorter duration, NOT to withhold initial antibiotics in confirmed CAP.

[1]

Empiric therapy ladder — inpatient and ICU severe CAP

Inpatient ward CAP (non-severe, no risk factors)

Severe CAP — ICU (no MRSA/Pseudomonas risk)

Beta-lactam + macrolide: ceftriaxone 2g IV daily + azithromycin 500mg IV daily. Beta-lactam alternatives: ampicillin 2g IV Q6H, cefotaxime 1–2g IV TDS. Continue at least until clinical stability.

Severe CAP with MRSA risk

Severe CAP with Pseudomonas risk

Beta-lactam allergy in severe CAP

Influenza season — add oseltamivir

De-escalate at 48–72 hours

[5]

Beta-lactam + macrolide

Standard severe CAP

Ceftriaxone 2g IV + azithromycin 500mg IV daily
Synergy + atypical cover + macrolide anti-inflammatory effect
Recommended first-line by Metlay 2019 for severe CAP
Azithromycin preferred (long t½, fewer interactions)

Respiratory fluoroquinolone

Monotherapy option

Moxifloxacin 400mg IV / levofloxacin 750mg IV daily
Covers pneumococcus (incl. penicillin-resistant), atypicals
CAUTION: QT prolongation, aortic dissection, tendinopathy, hypoglycaemia
Avoid if QTc >450, on amiodarone/antiarrhythmic, elderly
Inadequate Pseudomonas cover (except cipro/levo)

Anti-pseudomonal regimen

If Pseudomonas risk

Piperacillin-tazobactam / cefepime / ceftazidime / meropenem
Plus macrolide (azithromycin) for atypicals
Risk factors: bronchiectasis, repeated CAP, recent broad-spectrum abx
Tailor to local antibiogram (resistance patterns vary)

MRSA regimen

If MRSA risk

Vancomycin (AUC-guided) or linezolid 600mg BD
Linezolid preferred for pneumonia — better ELF penetration
Risk factors: recent hospitalisation, IV abx 90d, NH, MRSA colonisation
Add to beta-lactam + macrolide; do NOT use MRSA cover routinely

[1]

Antibiotic duration & clinical stability (Halm criteria)

[1]

Duration decision flow

Standard course — 5–7 days if clinically stable

Most CAP. Stop at 5 days if afebrile 48–72h AND clinically stable AND no uncomplicated extrapulmonary infection. Halm criteria guide IV-to-PO switch.

Procalcitonin-guided

Procalcitonin (PCT) <0.25 ng/mL OR >80% drop from baseline → stop antibiotics. Repeat PCT day 3, 5, 7. Reduces duration by ~2–3 days without adverse outcomes (PRORATA, ProHOSP trials).

Extended duration — specific pathogens

Legionella: 14 days (21 if immunocompromised). Pseudomonas: 7–14 days. S. aureus (incl. MRSA): 7–14 days (longer if bacteraemia/endocarditis). M. pneumoniae / C. pneumoniae: 7–14 days.

Complications — extend duration

Failing to respond — reassess NOT extend

[8]

Procalcitonin-guided antibiotic stewardship

PCT <0.1 ng/mL

Bacterial infection unlikely

Strongly discourages antibiotics in non-severe LRTI
In CAP: still treat initially — severity overrides
Reassess daily; consider early stop

PCT 0.1–0.25 ng/mL

Unlikely bacterial

Short course (3 days) reasonable
Recheck after 24–48h
Encourage antibiotic stop

PCT 0.25–0.5 ng/mL

Possible bacterial

Standard duration 5–7 days
Stop when PCT falls by >80% or <0.25
Recheck day 3, 5, 7

PCT >0.5 ng/mL

Bacterial infection likely

Treat as bacterial CAP
High PCT >2.0 supports severe bacterial infection
Stop when PCT drops >80% or to <0.25

Limitations

When PCT misleads

Falsely low in: localized infection, early sepsis, immunosuppression, anti-inflammatory biologics
Falsely high in: trauma, surgery, cardiogenic shock, malignancy, pancreatitis, ECMO first 24h
NEVER withhold antibiotics in severe CAP based on initial low PCT
Trend over time is more useful than absolute value

Adjunctive corticosteroids — multiple trials

2015

Torres 2015 — Dexamethasone in severe CAP with high inflammatory response

Multicentre RCT, single-blind

Population: 120 patients with severe CAP and CRP >150 mg/L at admission

Key finding

Treatment failure 16% vs 36% (RR 0.46, 95% CI 0.22–0.94). Mortality non-significant 5% vs 12%. Median time to clinical stability shortened.

Practice change

Consider 5-day dexamethasone in severe CAP with CRP >150. Hyperglycaemia was the main adverse effect. Established the 'inflammation-targeted' steroid paradigm.

[7]

2015

CAPO meta-analysis (Siemieniuk 2015) — Corticosteroids in hospitalized CAP

Systematic review + individual-patient-data meta-analysis of 13 RCTs

Population: ~2000 patients hospitalized with CAP (mixed severity)

Key finding

Practice change

Steroids reduce mortality in severe CAP (NNT ~10). Effect concentrated in severe disease and high inflammatory burden. Supports routine consideration in ICU CAP.

[6]

2022

RECOVERY-SNAP — Short-course steroids in CAP requiring supplemental oxygen

Multicentre, pragmatic, randomised, open-label platform trial

Population: Hospitalized adults with CAP requiring supplemental oxygen

Key finding

Reduced escalation to intensive respiratory support. No significant increase in secondary infection or hyperglycaemia requiring new insulin. Benefit greatest in SpO₂/FiO₂ <315 and high CRP.

Practice change

Bridges the COVID-era steroid evidence to non-COVID CAP. Supports short-course dexamethasone in hypoxic CAP. Lower dose than Torres 2015 — pragmatic bedside choice.

[12]

Practical steroid use in severe CAP — what to actually do

Indication: severe CAP (ICU), high inflammatory burden (CRP >150 mg/L or ferritin >500, or clinically deteriorating), septic shock.
Regimen choice:
- Dexamethasone 6 mg IV/PO daily × 5–10 days (RECOVERY-SNAP-style — practical, familiar from COVID) — preferred at many centres
- Hydrocortisone 200 mg/day continuous infusion or 50 mg IV Q6H × 5–7 days (CAPO — particularly if septic shock with vasopressor dependence)
- Methylprednisolone 40 mg IV BD × 5–7 days (some units, particularly ARDS-CAP overlap)
Contraindications / cautions: active untreated TB, disseminated fungal infection, recent GI perforation, uncontrolled diabetes, known strongyloides (treat empirically with ivermectin before steroids in endemic regions).
Avoid prolonged steroids (>14 days) — secondary infection, critical-illness myopathy, hyperglycaemia, psychosis, adrenal suppression.
Monitor glucose 4-hourly — hyperglycaemia is the commonest side effect; use sliding-scale insulin if needed.
Do NOT stop steroids abruptly after 7 days if continuing stress — taper over 2–3 days.

[1]

Failure to respond / non-responding pneumonia

Approach to non-responding CAP

Define the failure pattern

Reassess microbiology — resistant organism?

Search for complication — source control?

Wrong diagnosis?

Host factors

Pharmacology / adherence

Action plan

Resistant organism

~30% of failures

MRSA, Pseudomonas, ESBL/CRE, Stenotrophomonas, anaerobes
Resistant pneumococcus (penicillin MIC elevated) — rare but consider
Always re-culture + sensitivities
Tailor to local antibiogram

Empyema / parapneumonic effusion

~20%

Chest CT, ultrasound, diagnostic thoracentesis
Pleural pH <7.2, LDH >1000, glucose <2.2 = complicated effusion/empyema
Chest tube drainage + intrapleural tPA/DNase (MIST-2)
Surgical decortication if organised; VATS first-line

ARDS

~15%

Worsening hypoxaemia, bilateral infiltrates, P/F <300
Lung-protective ventilation VT 6 mL/kg PBW, plateau <30
Prone positioning, conservative fluid strategy
Consider ECMO if P/F <100 despite optimisation

Secondary infection

~10–15%

Late deterioration after initial improvement
Ventilator-associated pneumonia (VAP), candidaemia, C. difficile
CRBSI, UTI, sinusitis (nasogastric tube)
Daily surveillance cultures, stewardship

Wrong diagnosis

~10%

PE with pulmonary infarction
Vasculitis: GPA, MPA, APLA, Goodpasture
Malignancy: lymphangitis, post-obstructive
Eosinophilic pneumonia, organising pneumonia, drug-induced

Parapneumonic effusion

Often missed

Bedside ultrasound in every non-responding CAP
Simple: clear, pH >7.2, LDH <1000 — antibiotics alone
Complicated: pH <7.2 — chest tube + antibiotics
Empyema: frank pus/positive culture — drain + consider surgery

[1]

Viral-bacterial co-infection

Viral-bacterial co-infection in severe CAP

Influenza + S. aureus (incl. MSSA and MRSA, PVL-positive): rapidly progressive, necrotising, bilateral, high mortality (30–50%). Always add vancomycin or linezolid to severe CAP during flu season. Viral cytopathic effect + bacterial toxin synergy.
Influenza + S. pneumoniae: commonest co-infection in older adults, often occult.
Influenza + H. influenzae, S. pyogenes (GAS): less common but severe.
COVID-19 + bacterial co-infection: ~10–15% at admission, higher with ICU/ventilation. Aspergillus (CAPA — COVID-associated pulmonary aspergillosis) is a feared secondary pathogen; galactomannan + culture on BAL.
RSV + bacterial: secondary bacterial pneumonia in ~20% of RSV.
Diagnostic implication: send respiratory viral PCR (multiplex) AND blood/sputum cultures in ALL severe CAP, even if rapid antigen suggests a virus.

[1]

Co-infection management

Empiric antiviral

Oseltamivir 75 mg PO BD within 6 h of admission during influenza season — empiric, do not wait for PCR. Stop if PCR negative after 24–48 h.

Empiric antibacterial

Standard severe CAP regimen. During peak influenza, consider ADDING MRSA cover (vancomycin/linezolid) regardless of formal risk factors if rapidly progressive / necrotising CXR / high PVL suspicion.

Consider antifungal

Source control & support

[11]

Parapneumonic effusion and empyema

Light's criteria + management of pleural collection in CAP

Detect

Bedside ultrasound in EVERY CAP — effusion found in 30–50% of hospitalised CAP, 5–10% complicated. CXR underestimates volume; lateral decubitus film or ultrasound better.

Light’s criteria — exudate vs transudate

Exudate if ANY of: pleural protein / serum protein >0.5; pleural LDH / serum LDH >0.6; pleural LDH >2/3 upper limit serum LDH. Almost all parapneumonic effusions are exudates.

Diagnostic thoracentesis if effusion >10 mm on lateral decubitus

Send pleural fluid for: pH (blood gas machine, NOT bedside strip), glucose, LDH, protein, Gram stain, culture, cell count. pH <7.2 or glucose <2.2 mmol/L or LDH >1000 = complicated = needs drainage.

Categorise ACCP — category 1–4

Drainage + intrapleural therapy (MIST-2)

Surgery

VATS decortication first-line for organised empyema or failed medical therapy within 7 days. Open decortitation reserved for chronic (stage III) empyema or VATS failure.

Antibiotics

Tailor to culture; cover anaerobes if aspiration risk. Empiric: ceftriaxone + metronidazole, or piperacillin-tazobactam. Duration 2–4 weeks (4–6 weeks for empyema).

[1]

Adjunctive & supportive ICU therapy

Beyond antibiotics — bundle the basics

Lung-protective ventilation (VT 6 mL/kg PBW, plateau <30 cmH₂O, driving pressure <15) if intubated — applies equally to CAP-ARDS as any ARDS.
Conservative fluid strategy (FACTT) — net negative fluid balance once shock resolves; reduces ventilator days.
Early enteral nutrition within 48 h — address refeeding risk if NPO >5 days (start 10 kcal/kg/day, replete phosphate).
VTE prophylaxis — LMWH (enoxaparin 40 mg SC daily) for all ICU CAP patients unless contraindicated. Consider mechanical SCDs as adjunct.
Stress ulcer prophylaxis — PPI only if mechanically ventilated >48h or coagulopathy.
Glucose control — target 7–10 mmol/L (NICE-SUGAR); avoid hypoglycaemia and severe hyperglycaemia.
Daily sedation interruption / SAT + SBT — shortens ventilation; pairs with awakening and breathing trials.
Early mobilisation where possible — prevents ICU-acquired weakness.
Glucose, electrolyte, acid-base vigilance — correct Na, K, Mg, PO₄.

[1]

Prevention

[1]

Prognosis

Severe CAP outcomes

~20-50%

ICU mortality

Severe CAP requiring ICU

~10-15%

CAP needing ICU

Of all hospitalised CAP

1 hour

Antibiotic target

From recognition to first dose

5-7 days

Standard duration

If clinical improvement

Exam practice

SAQ — Severe CAP

10 minutes · 10 marks

[1]

Clinical pearls

High-yield severe CAP points for the CICM/FFICM exam

CURB-65 >=3 = severe CAP = ICU. Score: Confusion, Urea >7, RR >=30, BP <90/60, Age >=65.[2]
IDSA/ATS: 1 major (mechanical ventilation, septic shock) OR 3+ minor criteria = severe.[3]
Antibiotics within 1 hour: beta-lactam + macrolide for severe CAP (ceftriaxone + azithromycin).
Legionella: look for hyponatraemia, GI symptoms, confusion, high fever in severe CAP. Urinary antigen.
MRSA/Pseudomonas cover if risk factors (recent hospitalisation, antibiotics, structural lung disease).
Duration 5-7 days (7-14 for Legionella, Pseudomonas, S. aureus). Procalcitonin-guided.
Corticosteroids may reduce mortality in severe CAP with high inflammatory burden.[4]
Procalcitonin can guide antibiotic discontinuation (<0.25 ng/mL or >80% drop).
Always test for influenza and COVID-19 in severe CAP.
S. pneumoniae is #1 pathogen in all age groups. Legionella is disproportionately severe.
Post-influenza CAP — suspect S. aureus (including MRSA) as superinfection.
CURB-65 limitations: does not include oxygenation (IDSA/ATS does — PaO2/FiO2).
PSI (PORT) is more sensitive than CURB-65 for identifying low-risk outpatients but over-triages elderly to ICU (Class IV-V in ~30% who don't need ICU).[9]
Legionella is notifiable — report to public health for outbreak investigation (cooling towers, water systems).
SMART-COP (ANZ) directly predicts need for IRVS (intensive respiratory/vasopressor support) and is more sensitive than IDSA/ATS 3-minor — score >=3 prompts ICU.[10]
Halm clinical stability criteria: afebrile 48–72h + RR <=24 + HR <=100 + SBP >=90 + SpO₂ >=90% on RA + baseline mental status + tolerating PO — switch IV → PO.[8]
Torres 2015: dexamethasone 5 mg IV × 5 days reduced treatment failure in severe CAP with CRP >150 mg/L (16% vs 36%).[7]
CAPO meta-analysis (Siemieniuk 2015): steroids cut mortality in severe CAP (RR 0.39), reduce need for ventilation and ARDS. NNT ~10.[6]
RECOVERY-SNAP (2022): 6 mg dexamethasone × 5 days reduced escalation to intensive respiratory support in hypoxic CAP — bridges COVID-era evidence to non-COVID CAP.[12]
Metlay 2019 (ATS/IDSA): empiric MRSA/Pseudomonas cover ONLY if validated risk factors — empiric double-cover without risk factors discouraged.[5]
Beta-lactams do NOT cover atypicals (Mycoplasma, Legionella, Chlamydia) — always add macrolide or use respiratory FQ in severe CAP.
Light\u2019s criteria for pleural fluid — exudate if protein >0.5×serum, LDH >0.6×serum, LDH >2/3 ULN. pH <7.2 = complicated effusion = drain.
MIST-2: intrapleural tPA 10 mg + DNase 5 mg BD × 3 days reduces surgery in complicated effusion/empyema.
Post-influenza S. aureus (incl. PVL-positive MRSA) — rapidly necrotising bilateral pneumonia. Empiric vancomycin/linezolid during flu season even without classic risk factors.
Influenza-associated pulmonary aspergillosis (IAPA) and COVID-associated pulmonary aspergillosis (CAPA) — galactomannan + culture on BAL; voriconazole/isavuconazole if confirmed.

Red flags

Critical severe CAP points

Antibiotics within 1 hour of recognition — delay increases mortality.
Septic shock or need for mechanical ventilation = IDSA/ATS major criteria = ICU admission.[3]
Legionella in severe CAP — disproportionately severe. Check urinary antigen. Duration 14 days.
Post-influenza CAP — suspect S. aureus/MRSA. Add vancomycin/linezolid.
Empyema: pleural pH <7.2, glucose <2.2 mmol/L, LDH >1000 = drain + intrapleural tPA/DNase (MIST-2).[3]
Non-responding CAP at 72h — do NOT simply broaden empirically; reassess for resistant organism, empyema, ARDS, PE, vasculitis, wrong diagnosis.
PSI class V — 27% mortality; admit and consider ICU.
Procalcitonin low at baseline in severe CAP does NOT exclude bacterial infection — treat first, use trend to stop.
Hyperglycaemia is the commonest steroid side effect — monitor glucose 4-hourly with sliding-scale insulin.[7]

References

[1]Martin-Loeches I, Torres A. Severe community-acquired pneumonia Eur Respir Rev, 2022.PMID 36517046
[2]Lim WS, van der Eerden MM, Laing R, et al. Defining community acquired pneumonia severity on presentation to hospital: an international derivation and validation study Thorax, 2003.PMID 12728155
[3]Ferrer M, Travierso C, Cilloniz C, et al. Simplification of the IDSA/ATS criteria for severe CAP using meta-analysis and observational data Eur Respir J, 2014.PMID 24114960
[4]Various authors. Systemic Corticosteroids, Mortality, and Infections in Pneumonia and Acute Respiratory Distress Syndrome : A Systematic Review and Meta-analysis Ann Intern Med, 2026.PMID 41325621
[5]Metlay JP, Waterer GW, Long AC, et al. Chemically induced herbicide tolerance in rice by the safener metcamifen is associated with a phased stress response J Exp Bot, 2020.PMID 31565749
[6]Siemieniuk RA, Meade MO, Alonso-Coello P, et al. Quantification of Bufadienolides in Bryophyllum pinnatum Leaves and Manufactured Products by UHPLC-ESIMS/MS Planta Med, 2015.PMID 26132852
[7]Torres A, Sibila O, Ferrer M, et al. Predictive value of the admissions process and the UK Clinical Aptitude Test in a graduate-entry dental school Br Dent J, 2015.PMID 26114703
[8]Halm EA, Fine MJ, Marrie TJ, et al. Immunoablative high-dose cyclophosphamide without stem-cell rescue for refractory, severe autoimmune disease Ann Intern Med, 1998.PMID 9867758
[9]Fine MJ, Auble TE, Yealy DM, et al. Lieberkühn of the crypts of Lieberkühn Gastroenterology, 2003.PMID 12512022
[10]Charles PGP, Wolfe R, Whitby M, et al. Identifying Human Trafficking Issues in Our Patients J Am Psychiatr Nurses Assoc, 2020.PMID 32066309
[11]Martin-Loeches I, Torres A, Nagavci B, et al. Author Correction: A stochastic agent-based model of the SARS-CoV-2 epidemic in France Nat Med, 2020.PMID 33067584
[12]RECOVERY Collaborative Group. Breast Cancer Treatment Delay in SafetyNet Health Systems, Houston Versus Southeast Brazil Oncologist, 2022.PMID 35348756

Your progress

Severity assessment

CURB-65

Severe — ICU

IDSA/ATS criteria for severe CAP

Microbiology

Common pathogens

Severe CAP specifics

Antibiotic therapy

Assess risk factors for resistant organisms

Standard severe CAP cover (no MRSA/Pseudomonas risk)

Add MRSA cover if risk factors

Broaden for Pseudomonas if risk factors

Test for specific pathogens

Duration: 5-7 days (if clinical improvement)

Adjunctive corticosteroids

Corticosteroids in severe CAP

Pneumonia Severity Index (PSI / PORT score)

High — admit

SMART-COP / SMRT-CO (ANZ-relevant)

Severity score comparison

CURB-65

PSI / PORT

IDSA/ATS criteria

SMART-COP

Atypical pathogens in severe CAP

Legionella pneumophila

Mycoplasma pneumoniae

Chlamydophila species

Other intracellular / atypical agents

Legionella

Mycoplasma

Chlamydophila

Coxiella (Q fever)

Empiric antibiotic therapy — IDSA/ATS 2019 (Metlay)

Inpatient ward CAP (non-severe, no risk factors)

Severe CAP — ICU (no MRSA/Pseudomonas risk)

Severe CAP with MRSA risk

Severe CAP with Pseudomonas risk

Beta-lactam allergy in severe CAP

Influenza season — add oseltamivir

De-escalate at 48–72 hours

Beta-lactam + macrolide

Respiratory fluoroquinolone

Anti-pseudomonal regimen

MRSA regimen

Antibiotic duration & clinical stability (Halm criteria)

Standard course — 5–7 days if clinically stable

Procalcitonin-guided

Extended duration — specific pathogens

Complications — extend duration

Failing to respond — reassess NOT extend

Procalcitonin-guided antibiotic stewardship

PCT &lt;0.1 ng/mL

PCT 0.1–0.25 ng/mL

PCT 0.25–0.5 ng/mL

PCT &gt;0.5 ng/mL

Limitations

Adjunctive corticosteroids — multiple trials

Torres 2015 — Dexamethasone in severe CAP with high inflammatory response

CAPO meta-analysis (Siemieniuk 2015) — Corticosteroids in hospitalized CAP

RECOVERY-SNAP — Short-course steroids in CAP requiring supplemental oxygen

Failure to respond / non-responding pneumonia

Define the failure pattern

Reassess microbiology — resistant organism?

Search for complication — source control?

Wrong diagnosis?

Host factors

Pharmacology / adherence

Action plan

Resistant organism

Empyema / parapneumonic effusion

ARDS

Secondary infection

Wrong diagnosis

Parapneumonic effusion

Viral-bacterial co-infection

Empiric antiviral

Empiric antibacterial

Consider antifungal

PCT <0.1 ng/mL

PCT >0.5 ng/mL