About CURB-65 Calculator
7 min read
CURB-65 Calculator: Score Pneumonia Severity, Estimate Mortality Risk, and Guide Hospitalization Decisions
TL;DR: Five binary criteria produce a score from 0 to 5 that predicts 30-day mortality in community-acquired pneumonia. A score of 0 or 1 means outpatient treatment is usually safe (mortality under 3.2%). A score of 3 or higher signals the need for hospital admission, and scores of 4 to 5 warrant ICU consideration with mortality approaching 41.5%. Enter your values above for an instant result.
Table of Contents
- Why Five Data Points Can Predict Pneumonia Outcomes
- Six Clinical Scenarios Where CURB-65 Changes the Plan
- The CURB-65 Formula, Mortality Table, and Disposition Guide
- How to Score a Patient: Step by Step
- Putting the Formula to Work: Two Real-World Examples
- Where Clinicians Go Wrong With CURB-65
- FAQ
- Assumptions and Notes
- Your Next Step
- Further Reading
Why Five Data Points Can Predict Pneumonia Outcomes
Community-acquired pneumonia (CAP) kills roughly 1.5 million adults annually in countries with established healthcare systems, yet most cases resolve safely at home with oral antibiotics. The clinical challenge is sorting the 80% who can go home from the 20% who need a hospital bed, and separating that group from the smaller fraction who need intensive care. Getting this wrong in either direction costs lives or wastes scarce resources.
The CURB-65 score, published by Lim et al. in 2003 in Thorax, distils that triage decision into five bedside observations: Confusion, Urea (blood urea nitrogen above 7 mmol/L), Respiratory rate, Blood pressure, and age 65 or older. Each criterion met adds one point. The score was validated on 1,068 patients across three prospective studies in the UK, New Zealand, and the Netherlands, producing a stepwise mortality gradient from 0.7% at score 0 to 41.5% at scores 4 and 5.
The biological logic is straightforward. Confusion signals end-organ hypoperfusion or septic encephalopathy. Elevated urea reflects renal stress from dehydration or systemic inflammatory response. Tachypnoea above 30 breaths per minute indicates failing gas exchange. Hypotension marks cardiovascular compromise. And age captures the cumulative effect of immunosenescence, reduced pulmonary reserve, and higher comorbidity burden. Individual genetic variation in immune response genes (such as MBL2 and TLR4 polymorphisms) can shift a patient's true risk above or below what any point-based score predicts, but for bedside triage, CURB-65 performs as well as more complex models.
The calculator above does this in about ten seconds.
Six Clinical Scenarios Where CURB-65 Changes the Plan
-
A 72-year-old presents to the emergency department with cough and fever. Her chest X-ray confirms a right lower lobe infiltrate. Without a severity score, the admission decision depends on the attending physician's gestalt, which varies widely. CURB-65 converts subjective impression into a number: if her score is 0 or 1, discharge with oral amoxicillin and 48-hour follow-up is supported by mortality data below 3.2%.
-
A GP is deciding whether to refer a 58-year-old man with pneumonia. He has a respiratory rate of 32 and a systolic BP of 85 mmHg but is otherwise alert with normal renal function. That is already 2 points, placing him in the 13% mortality band. Guidelines from the British Thoracic Society recommend hospital assessment at a score of 2, and this patient meets the threshold on physiological criteria alone.
-
An inpatient care team is escalating a pneumonia patient to the ICU. CURB-65 scores of 4 or 5 carry 30-day mortality near 41.5%. When a ward patient's confusion deepens and urea climbs above 7 mmol/L, recalculating the score provides an objective trigger for intensivist involvement rather than waiting for overt clinical deterioration.
-
A rural clinic with limited lab access needs a rapid triage tool. The CRB-65 variant drops urea entirely and uses only 4 bedside criteria, producing a score from 0 to 4 that requires zero blood work. Studies show CRB-65 retains about 90% of the discriminatory power of the full score, and the calculator supports both versions.
-
A hospital quality team is auditing pneumonia admission rates. Benchmarking the proportion of CURB-65 score 0 and 1 patients who were admitted reveals potential over-hospitalization. National audit data from the UK suggest that roughly 40% of CAP admissions have low-severity scores, representing a significant opportunity to reduce bed days without increasing readmission risk.
-
A medical student is learning clinical decision rules for the first time. CURB-65 is one of the simplest validated prediction tools in medicine: 5 binary inputs, no weighting differences, and a clear disposition algorithm. Understanding it builds the foundation for more complex scoring systems like APACHE II (which uses 12 physiological variables) and the Pneumonia Severity Index (which uses 20).
The CURB-65 Formula, Mortality Table, and Disposition Guide
Each criterion met scores 1 point. The total determines risk and recommended disposition.
CURB-65 Score = Confusion (1 if present)
+ Urea > 7 mmol/L (1 if true)
+ Respiratory Rate >= 30 breaths/min (1 if true)
+ Blood Pressure: Systolic < 90 OR Diastolic <= 60 mmHg (1 if true)
+ Age >= 65 years (1 if true)
Score range: 0 to 5
CURB-65 Scoring Criteria Reference
| Criterion | Abbreviation | Threshold | Points |
|---|---|---|---|
| Confusion | C | New mental confusion (AMT score ≤ 8) | 1 |
| Urea / BUN | U | > 7 mmol/L (> 19.6 mg/dL) | 1 |
| Respiratory Rate | R | ≥ 30 breaths/min | 1 |
| Blood Pressure | B | Systolic < 90 or Diastolic ≤ 60 mmHg | 1 |
| Age | 65 | ≥ 65 years | 1 |
CURB-65 Mortality Risk by Score
| Score | 30-Day Mortality | Risk Category |
|---|---|---|
| 0 | ~0.7% | Low |
| 1 | ~3.2% | Low |
| 2 | ~13.0% | Moderate |
| 3 | ~17.0% | High |
| 4 | ~41.5% | Very High |
| 5 | ~41.5% | Very High |
Disposition Recommendations
| Score | Recommended Action |
|---|---|
| 0–1 | Outpatient treatment with oral antibiotics; consider social circumstances |
| 2 | Short inpatient stay or closely supervised outpatient care |
| 3 | Hospital admission, assess for intensive care |
| 4–5 | Hospital admission, strong consideration for ICU transfer |
Limitations: CURB-65 does not account for comorbidities like COPD, immunosuppression, multilobar involvement on imaging, or parapneumonic effusion. Patients with these features may need escalation regardless of a low score. The Pneumonia Severity Index (PSI) captures more variables but requires lab results and imaging data that are not always immediately available.
How to Score a Patient: Step by Step
-
Assess mental status. Ask orientation questions or use the Abbreviated Mental Test (AMT). A score of 8 or below on the AMT, or any new-onset confusion compared to baseline, scores 1 point. Pre-existing cognitive impairment complicates this: compare to the patient's documented baseline, not to a healthy reference.
-
Check blood urea nitrogen. A serum urea above 7 mmol/L (or BUN above 19.6 mg/dL in US units) scores 1 point. If lab results are pending and the clinical picture requires immediate triage, use CRB-65 (without urea) as a preliminary score and recalculate when results return.
-
Count the respiratory rate. Measure over a full 60 seconds while the patient is unaware you are counting. Rates of 30 breaths per minute or higher score 1 point. Supplemental oxygen can mask the severity of respiratory compromise; note the FiO2 at the time of measurement.
-
Measure blood pressure. Systolic below 90 mmHg or diastolic at or below 60 mmHg scores 1 point. Use a properly sized cuff. A single reading may not reflect the patient's true haemodynamic state if they are anxious or in pain; repeat after 5 minutes of rest if the initial reading is borderline.
-
Record the patient's age. Age 65 or older scores 1 point. This threshold captures the steep increase in pneumonia mortality observed in epidemiological data: 30-day case fatality roughly triples between age groups 50 to 64 and 65 to 79.
-
Sum the five criteria. The total is the CURB-65 score. Match it against the mortality and disposition tables above.
-
Apply clinical judgement on top of the score. A non-obvious insight: a CURB-65 of 1 in a patient with multilobar infiltrates and a lactate of 4 mmol/L is not a safe discharge. The score is a starting point for disposition, never the final word.
Putting the Formula to Work: Two Real-World Examples
Example 1: 48-Year-Old Shift Worker With Cough and Fever
A 48-year-old overnight warehouse worker presents to his GP with 4 days of productive cough, fever of 38.6 C, and pleuritic chest pain. He is alert and oriented. Vitals: respiratory rate 22, blood pressure 125/78. Blood work shows urea at 5.4 mmol/L.
Scoring:
| Criterion | Value | Met? | Points |
|---|---|---|---|
| Confusion | Alert, oriented | No | 0 |
| Urea | 5.4 mmol/L | No (≤ 7) | 0 |
| Respiratory Rate | 22 breaths/min | No (< 30) | 0 |
| Blood Pressure | 125/78 mmHg | No | 0 |
| Age | 48 years | No (< 65) | 0 |
| Total | 0 |
Result: CURB-65 = 0. Estimated 30-day mortality is approximately 0.7%. He is a candidate for outpatient management with oral antibiotics (amoxicillin 500 mg three times daily for 5 days per BTS guidelines). His GP schedules a 48-hour telephone review and instructs him to return if breathlessness worsens or fever persists beyond 72 hours of treatment.
Example 2: 78-Year-Old Retired Teacher With Worsening Breathlessness
A 78-year-old retired teacher is brought to the emergency department by her daughter. She has been increasingly confused over the past 24 hours. Vitals: respiratory rate 34, blood pressure 82/55. Urea returns at 11.2 mmol/L. Chest X-ray shows bilateral lower lobe consolidation.
Scoring:
| Criterion | Value | Met? | Points |
|---|---|---|---|
| Confusion | New onset confusion | Yes | 1 |
| Urea | 11.2 mmol/L | Yes (> 7) | 1 |
| Respiratory Rate | 34 breaths/min | Yes (≥ 30) | 1 |
| Blood Pressure | 82/55 mmHg | Yes (Sys < 90 and Dia ≤ 60) | 1 |
| Age | 78 years | Yes (≥ 65) | 1 |
| Total | 5 |
Result: CURB-65 = 5. Estimated 30-day mortality is approximately 41.5%. She requires immediate hospital admission with strong consideration for ICU. The team initiates intravenous co-amoxiclav plus clarithromycin, orders blood cultures and arterial blood gas, and contacts the intensivist for assessment. Fluid resuscitation targets a mean arterial pressure above 65 mmHg.
Where Clinicians Go Wrong With CURB-65
Ignoring pre-existing confusion in elderly patients. Baseline dementia does not automatically mean the confusion criterion is met. The question is whether mental status has worsened from the patient's baseline. Scoring 1 for a patient with stable, longstanding dementia inflates the score. Compare to documented cognitive baselines or collateral history from carers.
Using mg/dL without converting. US labs report BUN in mg/dL. The CURB-65 threshold is urea > 7 mmol/L, which equals BUN > 19.6 mg/dL. Applying the 7 threshold directly to a mg/dL value misses nearly every positive result, since normal BUN runs 7 to 20 mg/dL. Always confirm which unit the lab is reporting.
Counting respiratory rate for only 15 seconds and multiplying. A 15-second count of 7 breaths extrapolates to 28, but the actual 60-second count might be 31 due to irregular breathing patterns common in pneumonia. A full 60-second count at rest is the standard. The difference between 28 and 31 flips the criterion from 0 to 1.
Relying on the score alone when red flags are present. A CURB-65 of 1 does not override clinical findings like oxygen saturation below 92%, bilateral infiltrates, or lactate above 4 mmol/L. The British Thoracic Society guidelines explicitly state that CURB-65 should be used alongside clinical judgement, not as a replacement. Roughly 10% of patients with low CURB-65 scores still require admission based on non-CURB factors.
Forgetting that diastolic threshold is "less than or equal to" 60. A diastolic reading of exactly 60 mmHg scores 1 point. Clinicians sometimes apply a strict "less than 60" rule and miss this boundary case. The original Lim et al. validation used ≤ 60 for the diastolic cutoff.
Omitting the score entirely in young, "well-appearing" patients. Pneumonia kills roughly 2,500 adults under 65 annually in the UK alone. A 45-year-old with a respiratory rate of 32, confusion, and systolic BP of 88 scores 3 on CURB-65 despite being well under the age threshold. Skipping the calculation because a patient "looks young and healthy" misses physiological severity that the score captures.
Assumptions and Notes
- Margin of error. CURB-65 is a population-level prediction tool. Individual outcomes vary based on comorbidities, pathogen virulence, time to antibiotic administration, and immune status. The reported mortality percentages (0.7% to 41.5%) are averages from the Lim et al. 2003 validation cohort of 1,068 patients across three countries.
- Professional disclaimer. This calculator is an educational and clinical decision-support tool. It does not replace physician assessment, imaging findings, laboratory trends, or patient-specific factors. All treatment decisions should involve a qualified healthcare professional.
Your Next Step
The hard part of pneumonia triage is not the arithmetic. It is acting on the number. A score of 0 should give a clinician confidence to send a patient home with clear safety-netting instructions. A score of 4 should accelerate the conversation with the ICU team, not trigger a "wait and see" approach. Calculate the score, then let it do what it was designed to do: move the disposition decision from intuition to evidence.