🧬 CURB-65 Pneumonia Score

Not every pneumonia needs a hospital bed, and not every mild-looking one is safe at home. CURB-65 sorts community-acquired pneumonia into severity bands from five bedside criteria: Confusion, Urea, Respiratory rate, Blood pressure, and age 65. The 0-5 score predicts 30-day mortality and guides the site-of-care decision. Check the criteria above.

How to Use This Score

Check each criterion present at assessment. The score, mortality estimate, and care setting appear.

The Five Criteria and Mortality

Lim and colleagues derived CURB-65 in 2003 from over a thousand pneumonia patients, choosing criteria that predicted death independently and were available at the bedside. Each criterion captures a failure mode: confusion (brain perfusion), urea (dehydration/kidney), tachypnea (respiratory compensation), hypotension (circulatory failure), age (reserve). The mortality figures come from the validation cohorts and are approximate; real outcomes vary with treatment and comorbidity.

The site-of-care bands follow the mortality curve's shape: 0-1 carries low enough risk for outpatient antibiotics with follow-up, 2 sits at the inflection where short admission or hospital-at-home makes sense, and 3+ has mortality that justifies inpatient care, with 4-5 prompting ICU consideration. CRB-65 (without urea) exists for settings without labs and performs nearly as well.

Worked Example

A 70-year-old with respiratory rate 32 and BUN 24, alert, blood pressure 130/80: age (1) + tachypnea (1) + urea (1) = 3, about 14 percent 30-day mortality, hospital admission. A 45-year-old with the same vitals but no confusion and normal urea: tachypnea alone = 1, about 2.7 percent, outpatient with close follow-up. Same pneumonia physiology, different host reserve, different disposition. The score's real work is the second case: it quantifies why the young patient goes home.

What the Score Misses

CURB-65 does not see hypoxemia: a patient can score 1 while needing 4 liters of oxygen, and oxygen requirement overrides the score toward admission. It does not see comorbidities, frailty, social situation, or ability to return if worsening. Multilobar pneumonia, pleural effusion, and bacteremia raise risk beyond the score. And it was derived for community-acquired pneumonia, not hospital-acquired or aspiration pneumonia. Experienced clinicians use it as the starting point of the disposition decision, then adjust for everything it cannot see. A score of 0 in a patient who looks sick is still a sick patient.

For learners: CURB-65 is worth memorizing because it is short, validated, and used worldwide, and because it teaches the pattern of severity scoring generally: pick independent predictors, weight them simply, validate against outcomes, and tie bands to actions. Our CHA2DS2-VASc calculator and Centor score follow the same design philosophy in cardiology and infectious disease.

Outpatient Pneumonia: Making It Safe

Score 0-1 patients go home with antibiotics (commonly amoxicillin, doxycycline, or a macrolide per local guidance), clear return precautions (worsening breathlessness, confusion, inability to keep fluids down), and follow-up within days. Smoking cessation counseling, flu and pneumococcal vaccination review, and hydration round out the visit. Safety-netting is the intervention that makes outpatient care safe: the patient must know exactly when to come back, and the system must be able to receive them. The score picks the setting; the safety net makes it work.

For informational purposes only, not medical advice. Consult a healthcare professional.