Severe Community-Acquired Pneumonia (CAP) Admission Order Set: ICU Triage, Antibiotics, Adjuncts, and Day-by-Day Workflow
Overview
Severe community-acquired pneumonia (CAP) is the form requiring intensive care admission and frequently mechanical ventilation, vasopressor support, or both. The IDSA/ATS 2019 definition uses 9 minor criteria (≥3 of which qualify for ICU) plus 2 major criteria (either alone qualifies): septic shock requiring vasopressors and respiratory failure requiring mechanical ventilation. Mortality in severe CAP remains substantial (15–30%) and is highly time-dependent — every hour of delay in appropriate antibiotic therapy worsens outcomes.
This order set captures the first-hour, first-day, and through-discharge interventions: triage scoring (PSI/CURB-65), empirical antibiotic selection by MRSA/Pseudomonas risk profile, adjunctive corticosteroids in selected severe cases (CAPE-COD trial), conservative fluid strategy, mechanical ventilation parameters (lung-protective tidal volume), VTE and stress ulcer prophylaxis, daily de-escalation rounds, and discharge criteria. It is designed for use by hospitalists, intensivists, and emergency physicians admitting adult CAP patients to the ICU or step-down unit.
1. Severity Triage and ICU Admission
IDSA/ATS 2019 ICU admission criteria — both major and minor: ICU admission is required for any patient meeting ONE major criterion (mechanical ventilation OR septic shock requiring vasopressors) OR meeting at least THREE of nine minor criteria. The minor criteria are: respiratory rate ≥30/min, PaO2/FiO2 ≤250, multilobar infiltrates, confusion/disorientation, uremia (BUN ≥20 mg/dL), leukopenia (WBC <4,000), thrombocytopenia (platelets <100,000), hypothermia (core temperature <36°C), and hypotension requiring aggressive fluid resuscitation. SMART-COP and ATS criteria perform similarly to CURB-65 for triage but include lactate and pH which add prognostic discrimination in early sepsis. The single most useful bedside add-on in practice is point-of-care lactate; lactate ≥4 mmol/L in a CAP patient who looks "borderline" by minor criteria alone should escalate triage toward ICU regardless of the formal count. The dominant triage error in community hospital practice is admitting "borderline severe" CAP to floor beds and finding them on the rapid response cart 12-24 hours later — the threshold for ICU should be lower, not higher.
IDSA/ATS 2019 minor and major criteria
ICU admission is mandated when ≥1 major criterion or ≥3 minor criteria are met:
| Type | Criterion |
|---|---|
| Major (any one → ICU) | Septic shock requiring vasopressors |
| Major (any one → ICU) | Respiratory failure requiring mechanical ventilation |
| Minor (≥3 → ICU) | Respiratory rate ≥30/min |
| Minor | PaO2/FiO2 ≤250 |
| Minor | Multilobar infiltrates |
| Minor | Confusion / disorientation |
| Minor | Uremia (BUN ≥20 mg/dL) |
| Minor | Leukopenia (WBC <4,000/μL) |
| Minor | Thrombocytopenia (platelets <100,000/μL) |
| Minor | Hypothermia (core T <36°C) |
| Minor | Hypotension requiring aggressive fluid resuscitation |
PSI vs CURB-65
PSI (Pneumonia Severity Index) classes IV–V and CURB-65 ≥2 identify candidates for hospital admission, but neither replaces the IDSA/ATS criteria for ICU triage. Use CURB-65 (Confusion, Urea >7 mmol/L or BUN >19 mg/dL, Respiratory rate ≥30, Blood pressure SBP <90 or DBP ≤60, age ≥65) as a quick bedside score — 0–1 outpatient candidate, 2 admit to ward, 3–5 consider ICU. PSI is more granular but slower to calculate and better for risk stratification within a study population than at the bedside.
Always Apply Clinical Judgment Above Scores
A young patient with PSI class II but a respiratory rate of 38, multilobar consolidation, and a falling SpO2 needs ICU admission regardless of score. Conversely, a frail older patient with PSI class V and dementia in a stable state may be best managed on the ward with goals-of-care discussions. Scores guide; clinical judgment decides.2. Initial Workup Orders (First Hour)
Time-zero is hospital arrival or ICU transfer. The first-hour bundle for severe CAP mirrors the sepsis bundle when sepsis criteria are met:
| Order | Specifics |
|---|---|
| Vital signs | HR, BP, RR, SpO2 q15min × 4 then q1h; continuous telemetry; core temp q4h |
| IV access | Two large-bore peripheral IVs; arterial line if vasopressor or PaO2 needed; central line if multiple infusions |
| Labs (within 30 min) | CBC with differential, BMP/CMP, magnesium, phosphate, lactate, procalcitonin (baseline), troponin, BNP if HF suspected, ABG, coagulation panel |
| Microbiology (before antibiotics if no delay) | Two sets of blood cultures from different sites; sputum Gram stain and culture; urinary Streptococcus pneumoniae and Legionella antigens; respiratory viral panel (PCR — influenza, RSV, COVID-19); pleural fluid analysis if effusion ≥1 cm on lateral decubitus |
| Imaging | Chest X-ray (PA/lateral if possible); chest CT if cavitation, complication, or diagnostic uncertainty; bedside lung ultrasound for effusion/consolidation |
| Oxygen | Titrate to SpO2 92–96% (88–92% for COPD); start nasal cannula 2–4 L/min; escalate to high-flow nasal cannula 30–60 L/min FiO2 0.4–1.0; non-invasive ventilation if hypercapnic; intubation if persistent hypoxemia, hemodynamic compromise, or progressive failure |
| IV fluid | If sepsis/septic shock: crystalloid 30 mL/kg IV over 3 hours, then conservative strategy; preferred fluid: balanced crystalloid (LR or Plasma-Lyte over 0.9% NaCl) |
| Antibiotics | Within 1 hour for septic shock; within 4 hours regardless — see Section 3 for empirical regimens |
| Vasopressor (if shock persists after fluids) | Norepinephrine 0.05 mcg/kg/min IV titrated to MAP ≥65 mmHg; target central line if continuing >6 hours |
| Other | Foley catheter for hourly UO; NG tube if intubated; head-of-bed ≥30° if intubated (VAP prophylaxis); chlorhexidine oral care |
Pearl — Don't delay antibiotics for cultures
If blood cultures cannot be drawn within 30–45 minutes of arrival in a septic-shock patient, give antibiotics first. The mortality benefit of timely antibiotic administration substantially outweighs the diagnostic yield benefit of pre-antibiotic cultures, particularly because empirical regimens are broad enough to cover the likely pathogens.Pitfall — Excessive crystalloid in CAP
Once initial 30 mL/kg resuscitation is complete and MAP is restored, switch to a conservative fluid strategy. Excess crystalloid in CAP worsens hypoxemia, prolongs ventilation, and increases mortality (FACTT trial in ARDS data extrapolated). Use dynamic measures (passive leg raise, pulse pressure variation) to assess fluid responsiveness rather than chasing CVP targets.3. Empirical Antibiotic Selection
Standard severe CAP regimen
The IDSA/ATS 2019 standard for severe CAP without MRSA or Pseudomonas risk:
- Beta-lactam: Ceftriaxone 1–2 g IV q24h OR cefotaxime 1–2 g IV q8h OR ampicillin-sulbactam 3 g IV q6h
- PLUS Atypical coverage (one of):
- Azithromycin 500 mg IV q24h (preferred — better tolerated)
- Doxycycline 100 mg IV/PO q12h (macrolide alternative)
- Respiratory fluoroquinolone monotherapy (replaces both beta-lactam + macrolide):
- Levofloxacin 750 mg IV q24h OR
- Moxifloxacin 400 mg IV q24h
Atypical coverage is mandatory in severe CAP because Legionella, Mycoplasma, and Chlamydia carry significant mortality without targeted therapy. Macrolides may add anti-inflammatory benefit beyond antimicrobial activity in severe CAP. Avoid empirical fluoroquinolone monotherapy in TB-prevalent regions or in patients at risk of TB because partial fluoroquinolone treatment may obscure subsequent TB diagnosis.
MRSA / Pseudomonas coverage
Add MRSA or Pseudomonas coverage only when locally validated risk factors are present; routine empirical addition without risk factors increases adverse events without improving outcomes (per IDSA 2019 guideline change from 2007).
| Pathogen | Risk factors warranting empirical coverage | Empirical agents |
|---|---|---|
| MRSA | Prior MRSA isolation (especially respiratory); recent IV antibiotics in 90 days; hospitalization in 90 days; gram-positive cocci in clusters on Gram stain; cavitary necrotizing pneumonia; post-influenza pneumonia | Vancomycin 25–30 mg/kg IV loading then AUC-targeted (AUC 400–600 mg·h/L) OR linezolid 600 mg IV q12h |
| Pseudomonas aeruginosa | Prior Pseudomonas isolation; structural lung disease (bronchiectasis, advanced COPD); recent IV antibiotics; immunocompromise; gram-negative bacilli on Gram stain | Replace ceftriaxone with antipseudomonal agent: piperacillin-tazobactam 4.5 g IV q8h (extended infusion) OR cefepime 2 g IV q8h OR meropenem 1 g IV q8h — keep macrolide/doxycycline for atypical |
De-escalation occurs by day 2–3 once cultures and procalcitonin trends are available. If MRSA nasal swab PCR is negative and no MRSA grows from sputum/blood by 48 hours, vancomycin can usually be stopped. If sputum and blood cultures grow only susceptible Streptococcus pneumoniae, narrow to ampicillin or penicillin G monotherapy.
Procalcitonin Guides Antibiotic Duration, Not Initiation
Initiate empirical antibiotics for severe CAP based on clinical syndrome, not procalcitonin (which can be normal in early bacterial pneumonia). Use procalcitonin trends to support de-escalation and discontinuation decisions: a downward trend with clinical improvement supports stopping antibiotics earlier, often at 5–7 days rather than the historical 10–14. A persistently elevated procalcitonin should prompt reassessment for treatment failure or alternative diagnoses.4. Adjunctive Therapy and Supportive Care
Hydrocortisone (CAPE-COD)
The CAPE-COD trial established that adjunctive hydrocortisone reduces 28-day mortality in severe CAP from 11.9% to 6.2% (NNT ~17) when given to patients meeting any of: PaO2/FiO2 <300 with FiO2 ≥0.5 on mechanical ventilation, non-invasive ventilation/HFNC with PaO2/FiO2 <300, PSI score >130, or radiographic multilobar involvement. Exclusions include septic shock at randomization and influenza-confirmed pneumonia.
- Regimen: Hydrocortisone 200 mg IV continuous infusion daily × 4–7 days, then taper over 4–6 days based on clinical course (200 → 100 → 50 mg per day).
- Start within 24 hours of severe CAP recognition; benefit not seen with delayed initiation.
- Glucose control — Steroid-induced hyperglycemia common; target 140–180 mg/dL with insulin infusion. Monitor q4–6h initially.
- Avoid in: confirmed influenza pneumonia (potential harm), aspergillosis, varicella, suspected bacterial endocarditis without source control, recent live vaccine.
Mechanical ventilation strategy
For intubated severe CAP patients, apply lung-protective ventilation (ARDSnet principles) regardless of formal ARDS criteria:
| Parameter | Target / Setting |
|---|---|
| Tidal volume | 6 mL/kg predicted body weight (range 4–8 mL/kg) |
| Plateau pressure | ≤30 cm H2O |
| Driving pressure | ≤15 cm H2O (Pplat − PEEP) |
| PEEP | Lower PEEP/higher FiO2 table for CAP without ARDS; higher PEEP table if PaO2/FiO2 <200 |
| FiO2 / SpO2 target | SpO2 92–96%; lowest FiO2 to maintain |
| Respiratory rate | Titrate to pH ≥7.25; permissive hypercapnia acceptable (PaCO2 50–80, pH ≥7.20) |
| Prone positioning | If PaO2/FiO2 <150 despite 12 hours optimization → 16-h prone sessions per PROSEVA |
| Sedation | Targeted RASS −2 to 0; daily SAT/SBT; analgosedation preferred (fentanyl-based) |
Other supportive care: VTE prophylaxis with enoxaparin 40 mg SC daily (CrCl ≥30) or unfractionated heparin 5000 units SC q8h (CrCl <30) — start within 24 h unless active bleeding. Stress ulcer prophylaxis with pantoprazole 40 mg IV daily for ventilated or coagulopathic patients. Glucose control 140–180 mg/dL. Nutrition — enteral feeding within 24–48 h via NG/OG tube once hemodynamically stable, target 25–30 kcal/kg/day.
Recognize Treatment Failure Early
Failure of clinical improvement by 48–72 hours of appropriate empirical therapy demands urgent re-evaluation. Differential includes: drug-resistant pathogen (MDR bacteria, NTM, fungal), wrong diagnosis (PE, vasculitis, pulmonary edema, drug-induced lung disease, malignancy), complications (empyema, lung abscess, ARDS), inadequate dosing (especially in obesity), or co-infection (viral plus bacterial, secondary bacterial post-influenza). Repeat imaging (CT chest), bronchoscopy with BAL, and infectious diseases consultation.5. Day-by-Day Workflow and De-escalation
| Day | Clinical assessment | Lab / micro / imaging | Antibiotic action |
|---|---|---|---|
| Day 0 (admission) | Triage, ABCs, sepsis bundle, oxygen, intubation if needed | CBC, BMP, lactate, ABG, blood + sputum cultures, urinary antigens, viral PCR, CXR ± CT | Empirical regimen within 4 hours (1 hour if shock) |
| Day 1 | Reassess shock, oxygenation, fluid status; consider hydrocortisone | Daily CBC, BMP, procalcitonin; recheck lactate if elevated | Continue empirical; review urinary antigens (Strep pneumo, Legionella) if rapid result |
| Day 2 | Reassess clinical improvement; SAT/SBT if vented | Sputum culture results often available; blood cultures if positive (preliminary) | De-escalate vancomycin if MRSA nasal PCR negative + no MRSA on cultures; consider narrowing if pathogen identified |
| Day 3 | Reassess for treatment failure if no improvement; lung ultrasound for effusion progression | Final culture sensitivities; procalcitonin trend | Narrow to pathogen-directed therapy; stop atypical coverage if not needed; switch IV→PO if eating, hemodynamically stable, afebrile ≥24h |
| Day 4–5 | Mobilize, wean ventilation, reduce vasopressors | Weekly CBC; LFTs at days 5–7 (especially with linezolid, vancomycin) | Total duration typically 5–7 days for uncomplicated severe CAP if afebrile + stable; longer (10–14 days) for Pseudomonas, MRSA, S. aureus, Legionella, complicated effusion, lung abscess, immunocompromise |
| Day 6+ | Step down from ICU when stable; pulmonary rehab planning | Discharge labs | Complete antibiotic course; ensure oral regimen if discharged before completion |
De-escalation is not optional — it reduces C. difficile, MDR organism selection, and adverse drug events. Document the rationale for any prolonged broad-spectrum coverage in the chart. Daily antimicrobial stewardship rounds in ICUs improve de-escalation rates and shorten duration without increasing failure.
6. Discharge Criteria and Outpatient Transition
Discharge from the ICU and from hospital follows graduated criteria:
ICU → Step-down or Ward (criteria)
- Off vasopressors ≥6–12 hours; MAP ≥65 mmHg without support
- SpO2 ≥92% on FiO2 ≤0.4 (or appropriate baseline for chronic respiratory disease)
- Extubated and tolerating; or stable on chronic ventilator settings (e.g., trach)
- Hemodynamically stable; HR <120; resolving lactic acidosis
- No new organ dysfunction in past 24 hours
Hospital discharge (criteria)
- Afebrile ≥24 hours (some guidelines: 48 hours)
- Vital signs stable: HR <100, RR <24, SBP >90, SpO2 ≥90% on room air or chronic baseline FiO2
- Tolerating oral intake and oral antibiotics
- Mental status at baseline
- Reliable home support and follow-up arranged
- Clear understanding of red-flag symptoms (worsening dyspnea, fever, chest pain) and how to seek care
Discharge Orders
- Antibiotic completion — Switch IV to PO once tolerating oral; common transitions: ceftriaxone+azithromycin → amoxicillin-clavulanate 875 mg PO BID + azithromycin 500 mg PO daily, or levofloxacin IV→PO 750 mg daily. Total course 5–7 days uncomplicated; 10–14 days for Pseudomonas/MRSA/Legionella/complicated.
- Pneumococcal vaccination — PCV20 or PCV15+PPSV23 prior to or at discharge if not already received per ACIP age and risk recommendations.
- Influenza vaccination — In season, prior to discharge.
- Smoking cessation counseling and pharmacotherapy if active smoker.
- Pulmonary rehab referral for any patient with pre-existing chronic lung disease or prolonged ICU stay.
- Outpatient follow-up — Primary care within 7 days; chest imaging at 6 weeks if persistent symptoms or risk factors for malignancy (smoker over 50, abnormal initial imaging beyond consolidation, persistent radiographic abnormality).
When to Call Pulmonology / ID
Pulmonology: persistent radiographic abnormality at 6 weeks despite clinical recovery, suspected malignancy, suspected interstitial lung disease, recurrent CAP in same lobe (obstructive lesion), or chronic bronchiectasis. Infectious Diseases: treatment failure at 48–72h despite appropriate empirical therapy, MDR organism isolated, suspected NTM or fungal pneumonia, immunocompromised host with severe CAP, suspected hospital-acquired/ventilator-associated complication.Frequently Asked Questions
S7. the established Protocol Experience: Severe CAP from Admission to De-escalation
1. Lactate Triages the Borderline Patient
In practice, every CAP patient who meets 2 of 9 IDSA minor criteria (or 1 minor + clinician concern) gets a point-of-care lactate within 15 minutes of triage. Lactate ≥4 mmol/L in this borderline group pushes triage toward ICU regardless of the formal minor count. Experience has shown too many patients with 2 minors + lactate 5.2 stay on the floor for 18 hours before crashing. Conversely, lactate <2 mmol/L with 3 minors but otherwise stable physiology can often be observed on a step-down unit with q1-hour reassessment. Combining the formal criteria with lactate has substantially better discrimination than either alone in experienced hands.
2. Hour-1 Antibiotic with Cultures Drawn First
The IDSA/ATS recommendation is antibiotics within the first hour of severe CAP recognition, AND blood cultures drawn before antibiotic when feasible without delaying the antibiotic. In practice, the ED triage process draws blood cultures from a peripheral site at the same time as the initial IV access — there is essentially zero delay if the team is organized. Experienced practice does NOT accept the "we will get cultures after antibiotic" pathway because culture sensitivity drops substantially within 30 minutes of antibiotic, and identification of the causative organism drives downstream de-escalation. Experienced practice writes standing orders: blood cultures × 2 from separate sites + first antibiotic dose, both within 60 minutes of severe CAP triage call.
3. MRSA and Pseudomonas Risk-Stratified
Empirical coverage for severe CAP in the unit is: ceftriaxone 2 g IV daily + azithromycin 500 mg IV daily for typical severe CAP without risk factors. Experienced practice adds MRSA coverage (vancomycin or linezolid) when ANY of: recent prior MRSA respiratory isolate, cavitary infiltrate on imaging, post-influenza CAP, ESRD on hemodialysis, recent hospitalization or IV antibiotic within 90 days. Experienced practice adds Pseudomonas coverage (substitute piperacillin-tazobactam or cefepime for ceftriaxone) for: structural lung disease (bronchiectasis, advanced COPD on home oxygen, CF), recent hospitalization or IV antibiotic within 90 days, prior Pseudomonas respiratory isolate. The risk-stratified approach avoids both undertreatment and the empirical "everyone gets vanc-cefepime" overuse pattern.
4. Hydrocortisone Post-CAPE COD
The 2023 CAPE COD trial demonstrated mortality benefit for hydrocortisone 200 mg IV daily continuous infusion (or 50 mg IV q6h) for 4 days then taper, in severe CAP requiring high-flow nasal cannula, NIV, or mechanical ventilation. Routinely, experienced practice now incorporates this into severe CAP admission orders when respiratory support meets the trial threshold AND the patient is NOT in septic shock requiring vasopressors (which already has separate corticosteroid evidence — Hydrocortisone Plus Fludrocortisone for Adults with Septic Shock / APROCCHSS protocol). The two evidence bases do not perfectly overlap, but in patients in shock from severe CAP, the steroid is going on either way. Experienced clinicians exclude active fungal pneumonia, untreated TB, and influenza pneumonia (where steroid risk-benefit is murkier).
Illustrative Scenario: Borderline Severe CAP Saved by Lactate
A late-50s man presented to a community hospital ED with cough, fever 38.6°C, productive sputum, and dyspnea for 3 days. Vitals on arrival: RR 28, SpO2 91% on room air, BP 108/62, HR 112, GCS 15. CXR showed bilateral lower lobe infiltrates. Initial CURB-65 was 1 (RR not yet ≥30), and he met only 2 IDSA minor criteria (RR 28 close to threshold, multilobar). The ED team planned floor admission.
On a clinical admission consult the treating clinicians obtained a point-of-care lactate which returned 4.8 mmol/L. They escalated triage to ICU directly, drew blood cultures × 2 + sputum at the same time as IV access placement, and started ceftriaxone 2 g IV + azithromycin 500 mg IV (both delivered within 45 minutes of ED arrival). He had recent hospitalization 6 weeks prior so the treating clinicians added vancomycin for MRSA coverage. Within 4 hours of ICU arrival he required HFNC at 50 L/min and FiO2 0.6, and they added hydrocortisone 200 mg/day continuous infusion per CAPE COD protocol. Blood cultures returned MRSA at 36 hours. He never required intubation, defervesced at 72 hours, and was transitioned to floor at day 4 and discharged at day 8.
Practice Pattern Refinements vs Standard Approach
- Point-of-care lactate within 15 min for borderline severe (2 of 9 minor + clinician concern). Lactate ≥4 escalates to ICU regardless of formal count.
- Hour-1 antibiotic + cultures simultaneous, not antibiotic-then-cultures. Standing orders: blood cx × 2 from separate sites at first IV access placement.
- MRSA coverage if: prior MRSA isolate, cavitary infiltrate, post-influenza, ESRD HD, recent hospitalization/IV abx 90d. Pseudomonas coverage if: structural lung disease, recent hosp/IV abx 90d, prior Pseudomonas isolate.
- Hydrocortisone 200 mg/day continuous infusion × 4 d then taper for severe CAP on HFNC/NIV/mech vent (CAPE COD 2023), separate from septic shock corticosteroid pathway.
- Severe Alcoholic Hepatitis Admission Order Set
- Severe Ulcerative Colitis Flare Admission Order Set
- Myxedema Coma Order Set: Recognition, IV Levothyroxine, Hydrocortisone, and ICU Stabilization
- Sepsis and Septic Shock Admission Order Set: Hour-1 Bundle to ICU Care
- Diabetic Ketoacidosis Admission Order Set: Fluids, Insulin Drip, Electrolyte Replacement
- Spontaneous Bacterial Peritonitis (SBP) Order Set: Diagnostic Paracentesis, Empiric Antibiotics, and Albumin
Primary guideline documents this article draws on. Links resolve via DOI to the publisher; access may require a subscription.
- ATS/IDSA 2019 Guideline: Community-acquired Pneumonia in Adults (American Journal of Respiratory and Critical Care Medicine · 2019)