CURB-65 — Community-Acquired Pneumonia Severity
상세 가이드 곧 제공 예정
CURB-65 (Pneumonia)에 대한 종합 교육 가이드를 준비 중입니다. 단계별 설명, 공식, 실제 예제 및 전문가 팁을 곧 확인하세요.
The CURB-65 score is a validated clinical severity scoring system for community-acquired pneumonia (CAP) in adults, developed by the British Thoracic Society (BTS) and published by Lim and colleagues in the Thorax journal in 2003. It provides a rapid, bedside-applicable severity assessment that directly guides triage and antibiotic stewardship decisions. CURB-65 stands for the five equally-weighted clinical criteria: Confusion (new-onset disorientation or mental test score ≤8/10); Urea >7 mmol/L (blood urea nitrogen >19 mg/dL); Respiratory rate ≥30 breaths per minute; Blood pressure (systolic <90 mmHg or diastolic ≤60 mmHg); and age ≥65 years. Each criterion scores 1 point; the total ranges from 0 to 5. CURB-65 is most powerfully applied in the context of antimicrobial stewardship: scores of 0–1 predict low 30-day mortality (<3%) and support outpatient oral antibiotic treatment; scores of 2 predict intermediate mortality (~9%) and may support short inpatient stay or supervised community treatment; scores of 3–5 predict high 30-day mortality (17–57%) and require inpatient treatment, with ICU evaluation for scores ≥4. While the Pneumonia Severity Index (PSI) uses 20 variables and provides more precise prognostication, CURB-65 requires only 5 readily available data points, making it the preferred tool at the bedside and in emergency department triage. The two tools are complementary: CURB-65 is efficient for risk-stratification at the point-of-care; PSI provides greater discrimination at high risk and may better identify low-risk patients who can safely be treated at home.
CURB-65 Score = C + U + R + B + 65; C = Confusion (new onset) = 1 point; U = Urea >7 mmol/L (BUN >19 mg/dL) = 1 point; R = Respiratory rate ≥30/min = 1 point; B = Blood pressure (SBP <90 or DBP ≤60 mmHg) = 1 point; 65 = Age ≥65 years = 1 point; Total 0–5; 0–1 = Low risk (outpatient); 2 = Moderate risk (hospital consideration); 3–5 = High risk (inpatient); ≥4 = Consider ICU
- 1Assess for new-onset confusion: use the Abbreviated Mental Test Score (AMTS, 10 questions) — score ≤8/10 is abnormal; or ask simple orientation questions (person, place, time) — any new disorientation scores 1 point. Document whether confusion is pre-existing (baseline) or new — only new confusion counts.
- 2Obtain blood urea nitrogen (BUN/urea) from blood chemistry: urea >7 mmol/L (or BUN >19 mg/dL in US units) scores 1 point. This reflects intravascular dehydration, systemic inflammatory response, and impaired renal perfusion — markers of physiological stress from pneumonia.
- 3Measure respiratory rate by counting breaths over 60 seconds: a rate ≥30 breaths/min scores 1 point. Tachypnoea reflects pulmonary compromise from consolidation, pleural effusion, or systemic sepsis — it is one of the earliest objective markers of respiratory decompensation.
- 4Measure blood pressure by non-invasive cuff: systolic BP <90 mmHg or diastolic BP ≤60 mmHg scores 1 point. Hypotension is a late sign of haemodynamic compromise in pneumonia and is associated with significantly elevated mortality; it may indicate septic shock or tension effusion.
- 5Record the patient's age: age ≥65 years scores 1 point. Older adults have reduced physiological reserve, higher rates of co-morbidity, atypical presentations, and poorer outcomes from pneumonia — this demographic factor is an independent mortality predictor.
- 6Sum the five criterion scores to obtain the CURB-65 total (0–5) and apply the corresponding disposition recommendation: score 0–1 → low risk, consider outpatient oral antibiotics (amoxicillin 500 mg TDS ± macrolide); score 2 → moderate risk, consider hospital assessment or supervised community care; score 3–5 → high risk, admit for IV antibiotics.
- 7Integrate CURB-65 with clinical judgement: a patient scoring 0 who is unable to take oral medication, has oxygen saturations <94%, or has a social situation precluding safe home management should be admitted despite a low score; conversely, a patient scoring 2 who is young and previously well may safely be managed with oral antibiotics under close outpatient follow-up.
CURB-65 of 0 in a young adult with CAP supports same-day discharge with antibiotic prescription and safety-net advice.
None of the five criteria are met; this patient has low physiological derangement and is at low risk of deterioration — hospitalisation and IV antibiotics are not necessary and increase costs and nosocomial infection risk.
Three criteria met (confusion, elevated urea, age ≥65) puts this patient in the high-risk category despite RR and BP being near-normal.
The combination of new confusion, elevated urea, and age ≥65 in a CAP patient indicates significant systemic illness requiring inpatient evaluation and empiric IV antibiotics pending culture results.
CURB-65 of 4–5 has mortality rates of 40–57%; these patients have septic shock from pneumonia and require emergency department resuscitation.
All five CURB-65 criteria are present; the combination of hypotension, severe tachypnoea, high urea, and confusion indicates severe septic shock from pneumonia requiring immediate aggressive resuscitation and critical care involvement.
CURB-65 can underestimate risk in patients with significant comorbidities because it does not incorporate underlying disease burden.
PSI includes 20 variables including comorbidities (COPD, liver disease, heart failure); a CURB-65 of 1 in a COPD patient may misclassify as low risk — clinical context always overrides a simple score.
Emergency department triage nurses use CURB-65 to rapidly stratify CAP patients into admission and discharge pathways, improving patient flow and bed utilisation, enabling practitioners to make well-informed quantitative decisions based on validated computational methods and industry-standard approaches
Antibiotic stewardship programmes use CURB-65 scores to guide IV-to-oral antibiotic switch decisions in patients admitted with CAP, reducing IV antibiotic exposure and hospital length of stay, helping analysts produce accurate results that support strategic planning, resource allocation, and performance benchmarking across organizations
Primary care physicians use CRB-65 (the no-blood-test version) at the point of care to decide whether to admit or treat CAP patients at home with oral antibiotics, allowing professionals to quantify outcomes systematically and compare scenarios using reliable mathematical frameworks and established formulas
Hospital at home (HAH) and community respiratory services use CURB-65 as part of eligibility criteria for supported home treatment of moderate-severity CAP (score = 2) with daily nursing review, supporting data-driven evaluation processes where numerical precision is essential for compliance, reporting, and optimization objectives
Quality improvement teams audit CURB-65 documentation rates in CAP patients as part of BTS/NICE CAP management standards, benchmarking compliance against national quality indicators, which requires precise quantitative analysis to support evidence-based decisions, strategic resource allocation, and performance optimization across diverse organizational contexts and professional disciplines
CURB-65 vs PSI in young patients
In patients <50 years with no comorbidities and CURB-65 = 0, PSI will reliably categorise them as class I–II (very low risk), reinforcing the outpatient treatment decision. The advantage of using both tools in borderline cases (CURB-65 = 1–2 in young patients) is that PSI may provide additional reassurance of safety for home treatment. Young, healthy patients with CURB-65 ≤1 have very low mortality rates and rarely benefit from hospitalisation.
COVID-19 and CURB-65
CURB-65 was applied extensively during the COVID-19 pandemic for severity assessment of COVID pneumonia. While it retains predictive value, COVID pneumonia has distinct characteristics — hypoxia without subjective dyspnoea ('happy hypoxia'), rapid deterioration patterns, and a cytokine storm phase distinct from typical bacterial CAP. SpO₂ monitoring and CRP/ferritin/D-dimer trends are important additional prognostic markers in COVID pneumonia not captured by CURB-65.
Immunocompromised patients
CURB-65 underestimates severity in immunocompromised patients (solid organ transplant recipients, haematological malignancy, HIV/AIDS, prolonged corticosteroid therapy) who may lack typical inflammatory responses despite severe pneumonia. Fever may be absent; confusion may occur at lower severity; laboratory markers may be blunted. Immunocompromised patients with any respiratory symptoms and infiltrate on imaging should be treated as at least moderate severity and promptly discussed with infectious disease teams.
Legionella pneumonia — special considerations
Legionella pneumophila pneumonia (Legionnaires' disease) can present with severe CAP disproportionate to initial CURB-65 score due to its rapid clinical course. Urinary Legionella antigen should be sent in all hospitalised CAP patients. Legionella is not covered by standard beta-lactam antibiotics — macrolide or fluoroquinolone coverage must be added. Hyponatraemia (Na <130 mmol/L) is a distinctive feature of Legionella pneumonia not reflected in CURB-65.
| CURB-65 Score | 30-Day Mortality | Risk Category | Recommended Management |
|---|---|---|---|
| 0 | <1% | Very low risk | Outpatient oral antibiotics; no admission required |
| 1 | 2–3% | Low risk | Outpatient treatment; 48h follow-up; safety-netting |
| 2 | 9% | Moderate risk | Hospital assessment; consider short admission vs supervised community care |
| 3 | 17% | High risk | Admit; IV antibiotics; blood cultures; O₂ monitoring |
| 4 | 40% | Very high risk | Admit; HDU consideration; IV broad-spectrum; ICU assessment |
| 5 | 57% | Severe | ICU assessment; vasopressors likely; immediate resuscitation |
How does CURB-65 compare to the PSI (Pneumonia Severity Index)?
PSI (PORT score) uses 20 variables across 3 assessment categories (demographics, comorbidities, physical findings, and laboratory/radiology results) to stratify patients into classes I–V. It is more sensitive for identifying low-risk patients safe for outpatient treatment and has better prognostic accuracy at the extremes of risk. However, it requires more data collection and is less practical at the bedside. CURB-65 is faster (5 variables), more widely applied in the ED and ward setting, and slightly better at identifying high-severity patients. Both tools complement each other; many guidelines recommend using CURB-65 for triage with PSI for more refined risk stratification.
Does CURB-65 guide antibiotic selection or just admission decisions?
CURB-65 primarily guides severity stratification and site-of-care decisions rather than antibiotic selection. Antibiotic choice is based on likely pathogens, local resistance patterns, prior culture results, and allergy history. BTS guidelines recommend: low severity (CURB-65 0–1) = oral amoxicillin; moderate severity (CURB-65 2) = amoxicillin ± macrolide oral/IV; high severity (CURB-65 3–5) = IV broad-spectrum beta-lactam (co-amoxiclav or 3rd generation cephalosporin) + macrolide or respiratory fluoroquinolone.
What is the 'CRB-65' version and when is it used?
CRB-65 is a modified version that omits blood urea measurement, using only Confusion, Respiratory rate, Blood pressure, and age ≥65 — making it applicable in primary care where rapid blood results are unavailable. CRB-65 scores 0–4; a score of 0 = very low risk (consider home treatment); 1–2 = intermediate risk (hospital referral assessment recommended); 3–4 = high risk (urgent hospital admission). CRB-65 has slightly lower specificity than CURB-65 but is simpler for GP use.
Can CURB-65 be used for hospital-acquired pneumonia (HAP)?
CURB-65 was specifically developed and validated for community-acquired pneumonia. HAP and healthcare-associated pneumonia (HCAP) involve different pathogen profiles (more resistant organisms including Pseudomonas, MRSA, Gram-negative bacteria), different clinical contexts, and different treatment algorithms. CURB-65 should not be the primary scoring tool for HAP; the IDSA/ATS 2016 HAP guidelines recommend clinical assessment for severity rather than a specific scoring tool.
Why is the mortality for CURB-65 = 0 not zero?
CURB-65 is a clinical risk stratification tool, not a mortality prediction algorithm. Even with a score of 0, the 30-day mortality from CAP is approximately 0.7–1.2% — reflecting that pneumonia causes some deaths even in young, healthy, well-perfused patients (e.g., from fulminant viral pneumonia or atypical pathogens). The CURB-65 identifies patients at low enough risk for outpatient treatment, not patients with zero mortality risk.
Should oxygen saturation be incorporated into CURB-65?
Oxygen saturation is not part of the original CURB-65 criteria but many guidelines recommend incorporating SpO₂ as a modifier. BTS guidelines suggest that patients with SpO₂ <94% on room air should be treated as at least moderate severity regardless of CURB-65 score. SMART-COP (an Australian severity tool) incorporates SpO₂ and PaO₂, providing better prediction of the need for intensive respiratory or vasopressor support — a different but complementary endpoint to 30-day mortality.
Is CURB-65 validated in elderly care home patients?
CURB-65 was originally validated in general hospital populations. Its performance in frail, elderly nursing home residents is less certain — the age criterion (≥65) is positive in virtually all nursing home patients, reducing its discriminative power. Modified tools (PSI, SOAR score for frail adults, ResApp) may be more appropriate in this population. Clinical frailty assessment and goals-of-care discussion are particularly important when scoring elderly care home patients with pneumonia.
How does CURB-65 assist antibiotic stewardship?
Antibiotic stewardship programmes use CURB-65 to justify de-escalation of broad-spectrum IV antibiotics in patients with low-severity scores who are improving. A patient admitted with CURB-65 = 2 who defervesces, eats and drinks normally, and has improving blood markers by 48–72 hours can safely step down from IV to oral antibiotics and be considered for early discharge — reducing antibiotic exposure, hospital costs, and nosocomial complication risk.
전문가 팁
CURB-65 and PSI are complementary tools that answer slightly different questions: 'How sick is this patient right now?' (CURB-65) vs 'What is the overall mortality risk across a population of similar patients?' (PSI). In clinical practice, use CURB-65 for rapid triage at the ED or ward level, and use PSI to provide additional confidence for borderline decisions — particularly when considering whether a young patient with CURB-65 = 1–2 can safely be treated at home.
알고 계셨나요?
Pneumonia killed more people in the pre-antibiotic era than any other infectious disease — it was called 'the captain of the men of death' by William Osler in 1901, borrowing John Bunyan's phrase. The discovery of penicillin in 1928 and its clinical use from 1942 onwards transformed pneumonia from a condition with 30% mortality to one where the vast majority of patients recover fully. Despite this, pneumonia remains the leading infectious cause of death worldwide today, killing approximately 4 million people annually.
참고 자료
- ›Lim WS et al. — Defining community acquired pneumonia severity on presentation to hospital: an international derivation and validation study (Thorax 2003)
- ›BTS Guidelines — Management of Community Acquired Pneumonia in Adults (2009 updated 2019)
- ›Fine MJ et al. — Pneumonia Patient Outcomes Research Team (PORT) study — PSI derivation (NEJM 1997)
- ›NICE NG138 — Pneumonia (community-acquired): antimicrobial prescribing (2019)
- ›IDSA/ATS Consensus Guidelines on the Management of CAP in Adults (2007)