Pneumonia Severity Index (PORT Score) — Evidence-based risk stratification for community-acquired pneumonia
The Pneumonia Severity Index (PSI) Calculator, also known as the PORT Score, is a validated clinical prediction tool used by emergency physicians, hospitalists, and intensivists worldwide to assess the severity of community-acquired pneumonia (CAP) and guide safe, evidence-based disposition decisions. Based on the landmark study by Fine and colleagues published in the New England Journal of Medicine in 1997, the PSI stratifies patients into five risk classes (Class I through Class V) with corresponding 30-day mortality rates ranging from 0.1% to over 30%. Community-acquired pneumonia is one of the most common and potentially life-threatening infectious diseases encountered in emergency medicine, affecting millions of patients annually worldwide. Determining which patients are safe for outpatient treatment versus those who require inpatient hospitalization or intensive care unit admission is one of the most consequential clinical decisions in emergency and primary care settings. The PSI provides a structured, objective framework for that decision, replacing purely subjective clinical gestalt with an evidence-based, validated scoring system. The original PSI derivation cohort included 14,199 hospitalized patients with community-acquired pneumonia across 5 teaching hospitals. It was subsequently validated on 38,039 additional inpatients and a third cohort of 2,287 inpatients and outpatients combined. This extensive validation across diverse populations gives the PSI strong external validity, and it has since been incorporated into the major Infectious Diseases Society of America (IDSA) and American Thoracic Society (ATS) guidelines for CAP management. Our PSI Calculator implements the complete two-step algorithm. In Step 1, patients aged 50 years or younger with none of the five comorbid conditions and no abnormal physical examination findings are automatically stratified to the lowest-risk Class I, which carries only a 0.1% 30-day mortality rate and supports outpatient management on oral antibiotics. This auto-assignment recognizes that young, healthy patients with normal vital signs rarely experience severe pneumonia outcomes regardless of laboratory values. Patients who do not meet Class I criteria proceed to Step 2, where a detailed point score is calculated using 20 variables spanning patient demographics, comorbid conditions, physical examination findings, and laboratory and radiographic data. Age itself contributes directly to the score (one point per year for men, age minus ten for women), reflecting the well-established association between advancing age and pneumonia mortality. Five comorbidities carry point values: neoplastic disease (+30), liver disease (+20), congestive heart failure (+10), cerebrovascular disease (+10), and renal disease (+10). Five physical examination findings are scored: altered mental status (+20), respiratory rate 30 or more per minute (+20), systolic blood pressure below 90 mmHg (+20), abnormal temperature (below 35 degrees Celsius or above 39.9 degrees Celsius) (+15), and pulse 125 or more beats per minute (+10). Six laboratory and radiographic findings round out the score: arterial pH below 7.35 (+30), blood urea nitrogen 30 mg/dL or above (+20), serum sodium below 130 mmol/L (+20), glucose 250 mg/dL or above (+10), hematocrit below 30% (+10), and pleural effusion on chest radiograph (+10). Oxygenation is assessed either as PaO2 below 60 mmHg on arterial blood gas or SpO2 below 90% on pulse oximetry (+10). The resulting total point score determines the risk class: Class II (70 points or less, 0.6% mortality, outpatient), Class III (71 to 90 points, 0.9-2.8% mortality, observation or short-stay admission), Class IV (91 to 130 points, 8.2-9.3% mortality, inpatient hospitalization), and Class V (above 130 points, 27-31% mortality, inpatient with ICU consideration). An important clinical caveat is that the PSI was specifically designed to identify patients safe enough for outpatient treatment, not to identify those who need the most aggressive care. Because age contributes heavily to the score, young patients with physiologically severe pneumonia may receive deceptively low scores. Conversely, elderly patients with multiple comorbidities may score highly even with mild pneumonia. Clinicians should always integrate PSI results with the full clinical picture, including the patient's ability to tolerate oral medications, availability of reliable home support, functional status, and other social factors that may necessitate admission regardless of calculated risk class.
Understanding the Pneumonia Severity Index
The PSI (PORT Score) is a 20-variable prediction rule that stratifies community-acquired pneumonia patients into five risk classes with validated 30-day mortality rates. It guides clinicians toward safe, evidence-based admission and discharge decisions.
The Two-Step Algorithm
The PSI uses a two-step approach. Step 1 automatically assigns Class I to patients aged 50 or younger with no comorbidities and no abnormal vital signs — these patients have less than 0.1% 30-day mortality. All other patients proceed to Step 2, where a detailed point score is calculated from demographics, comorbidities, physical exam findings, and lab/radiology data. This design makes the tool efficient: most young, healthy patients are classified in seconds without requiring laboratory tests.
Risk Classes and Mortality Rates
Class I carries approximately 0.1% 30-day mortality and supports outpatient oral antibiotic therapy. Class II (score 70 or less) carries 0.6% mortality — also appropriate for outpatient management. Class III (71-90 points) has 0.9-2.8% mortality and may be managed with brief hospitalization or careful outpatient monitoring. Class IV (91-130 points) has 8.2-9.3% mortality and requires inpatient hospitalization. Class V (above 130 points) carries 27-31% mortality and warrants immediate inpatient admission with consideration of intensive care unit level care.
PSI vs CURB-65: Which to Use
The PSI and CURB-65 are complementary tools. The PSI uses 20 variables and is better at identifying patients safe for outpatient discharge (higher sensitivity for low-risk identification). CURB-65 uses only 5 variables and is faster at bedside. CURB-65 has higher specificity for mortality (74.6% vs 52.2%). The ATS/IDSA 2019 guidelines prefer the PSI for CAP severity assessment. CURB-65 may be useful in resource-limited settings where laboratory results are unavailable. For identifying patients who need ICU-level care, the PSI has demonstrated superior performance.
Clinical Limitations and Pearls
The PSI was designed to identify low-risk patients, not to replace clinical judgment for high-risk ones. Key limitations include age bias — young patients with physiologically severe disease may receive low scores because age contributes directly to the point total. Social and functional factors such as inability to take oral medications, lack of reliable home support, active substance use, or cognitive impairment may indicate admission even for low-scoring patients. The PSI also predates aggressive sepsis screening; high-scoring patients should be evaluated for SIRS and sepsis criteria. Lab-dependent inputs (pH, BUN, PaO2) may not always be immediately available.
Formulas
For male patients, age in years is added directly to the point total. Each positive comorbidity, exam finding, and lab abnormality adds its designated point value.
For female patients, 10 points are subtracted from the age contribution, reflecting the lower age-adjusted mortality risk observed in the original derivation cohort.
Patients meeting all three Step 1 criteria bypass the point scoring algorithm entirely and are automatically assigned to the lowest-risk class.
Reference Tables
PSI Risk Class Scoring and Mortality
| Risk Class | Point Score Range | 30-Day Mortality | Recommended Disposition |
|---|---|---|---|
| Class I | Step 1 auto-assign | 0.1% | Outpatient oral antibiotics |
| Class II | ≤ 70 points | 0.6% | Outpatient oral antibiotics |
| Class III | 71–90 points | 0.9–2.8% | Brief observation or outpatient |
| Class IV | 91–130 points | 8.2–9.3% | Inpatient hospitalization |
| Class V | > 130 points | 27–31% | Inpatient + consider ICU |
PSI Point Values by Variable
| Variable | Points | Category |
|---|---|---|
| Age (male) | +1 per year | Demographics |
| Age (female) | +1 per year − 10 | Demographics |
| Nursing home resident | +10 | Demographics |
| Neoplastic disease | +30 | Comorbidity |
| Liver disease | +20 | Comorbidity |
| CHF | +10 | Comorbidity |
| Cerebrovascular disease | +10 | Comorbidity |
| Renal disease | +10 | Comorbidity |
| Altered mental status | +20 | Physical Exam |
| Respiratory rate ≥ 30/min | +20 | Physical Exam |
| Systolic BP < 90 mmHg | +20 | Physical Exam |
| Temperature < 35°C or > 39.9°C | +15 | Physical Exam |
| Pulse ≥ 125 bpm | +10 | Physical Exam |
| Arterial pH < 7.35 | +30 | Lab/Radiology |
| BUN ≥ 30 mg/dL | +20 | Lab/Radiology |
| Sodium < 130 mmol/L | +20 | Lab/Radiology |
| Glucose ≥ 250 mg/dL | +10 | Lab/Radiology |
| Hematocrit < 30% | +10 | Lab/Radiology |
| PaO₂ < 60 mmHg or SpO₂ < 90% | +10 | Lab/Radiology |
| Pleural effusion | +10 | Lab/Radiology |
Worked Examples
Young Healthy Patient — Class I Auto-Assignment
Step 1 check: Age ≤ 50? Yes (32 years old)
Any comorbidities? No (no neoplastic, liver, CHF, cerebrovascular, or renal disease)
Any abnormal vital signs? No (RR < 30, SBP ≥ 90, temp 35–39.9°C, pulse < 125, mental status normal)
All three Step 1 criteria met → auto-assign Class I
Elderly Patient with Comorbidities — Class IV
Age: +72 points
Nursing home: +10 points
CHF: +10 points, Renal disease: +10 points
RR ≥ 30: +20 points, SBP < 90: +20 points
Temperature 39.2°C (within 35–39.9°C): +0 points
Pulse < 125: +0 points, Mental status normal: +0 points
BUN ≥ 30: +20 points, Sodium < 130: +20 points
Glucose < 250: +0, Hematocrit ≥ 30%: +0, SpO₂ ≥ 90%: +0, No effusion: +0
Total: 72 + 10 + 10 + 10 + 20 + 20 + 20 + 20 = 182 points
Middle-Aged Patient — Class III (Borderline)
Step 1: Age > 50 → proceed to Step 2
Age: +58 points
No comorbidities: +0
Pulse ≥ 125: +10 points
All other physical exam findings normal: +0
All labs normal (pH ≥ 7.35, BUN < 30): +0
Total: 58 + 10 = 68 points
How to Use the PSI Calculator
Enter Demographics
Input the patient's age in years and select biological sex. For females, the PSI automatically subtracts 10 points from the age contribution to account for lower age-adjusted mortality risk. Check if the patient is a nursing home resident (+10 points). These demographics form the foundation of the Step 1 Class I auto-assignment check.
Check Comorbid Conditions
Select all applicable comorbidities from the list: neoplastic disease (+30), liver disease (+20), congestive heart failure (+10), cerebrovascular disease (+10), and renal disease (+10). If the patient is 50 or younger AND no comorbidities are present AND no abnormal vital signs — the system will automatically assign Class I without requiring laboratory values.
Record Physical Exam and Labs
Check any abnormal findings on physical examination: altered mental status (+20), respiratory rate 30 or more per minute (+20), systolic BP below 90 mmHg (+20), abnormal temperature (+15), and pulse 125 or more bpm (+10). Then enter applicable laboratory findings including arterial pH, BUN, sodium, glucose, hematocrit, oxygenation (choose PaO2 or SpO2), and pleural effusion on X-ray.
Review Risk Class and Disposition
After calculating, review the total PSI score, risk class (I-V), 30-day mortality rate, and evidence-based disposition recommendation. The visual score breakdown chart shows the contribution from each input category. Always integrate the PSI result with clinical judgment, patient-specific social and functional factors, and local resources before making a final disposition decision.
Frequently Asked Questions
What is the PSI/PORT Score and who developed it?
The Pneumonia Severity Index (PSI), also known as the PORT Score (Patient Outcomes Research Team), is a validated clinical prediction rule developed by Michael Fine and colleagues, published in the New England Journal of Medicine in 1997 (volume 336, pages 243-250). It was derived from 14,199 hospitalized patients with community-acquired pneumonia (CAP) across five teaching hospitals and subsequently validated on over 38,000 additional patients. The PSI assigns patients to one of five risk classes based on 20 variables covering demographics, comorbid conditions, physical examination findings, and laboratory and radiographic data. The American Thoracic Society (ATS) and Infectious Diseases Society of America (IDSA) include it in their CAP management guidelines as the preferred severity scoring system.
Why might a young patient with severe pneumonia have a low PSI score?
This is the most important clinical limitation of the PSI. Because age contributes directly to the point score (one point per year for males), a 25-year-old male starts with only 25 base points. Even with several positive findings like tachypnea (+20), hypotension (+20), and low arterial pH (+30), the total score could still fall in Class III or lower. This age bias means the PSI can systematically underestimate severity in younger patients who are truly physiologically compromised. Clinicians should always override the PSI when clinical presentation suggests severity that the score doesn't capture. Some experts recommend using clinical parameters like ARDS criteria, septic shock, or bilateral involvement to supplement PSI in young patients.
What is the difference between PSI Class I and Class II, and why does Class I not require a point score?
Class I is a special auto-assigned category for patients who meet all three Step 1 criteria simultaneously: age 50 years or younger, no comorbid conditions (no neoplastic disease, liver disease, CHF, cerebrovascular disease, or renal disease), and no abnormal vital signs (no altered mental status, respiratory rate below 30, systolic BP 90 or above, temperature 35 to 39.9 degrees Celsius, and pulse below 125 bpm). These patients have demonstrated very low mortality risk regardless of laboratory values, so Step 2 point scoring is not needed. Class II covers patients who either do not meet Step 1 criteria or are older than 50, with a calculated point score of 70 or less. Both classes support outpatient management, but Class II requires that a score calculation confirm the low-risk designation.
Is SpO2 an acceptable substitute for PaO2 in the PSI calculation?
Yes. The oxygenation criterion in the PSI was originally defined as PaO2 below 60 mmHg on arterial blood gas (ABG). However, because ABG may not always be immediately available or clinically indicated, an SpO2 below 90% on pulse oximetry is widely accepted as a clinically equivalent alternative for PSI scoring purposes. This is supported by multiple implementations including EBMCalc, IMPACT, and several academic medical center calculators. When ABG is available, use PaO2; when it is not, SpO2 below 90% is the appropriate substitute. Note that the two measurements can occasionally diverge in patients with hemoglobin abnormalities such as methemoglobinemia or carbon monoxide exposure, where SpO2 may be falsely normal.
When should I use PSI versus CURB-65 for pneumonia severity assessment?
Both scores have specific strengths. The PSI excels at identifying patients who are safe for outpatient discharge — it has higher sensitivity for identifying low-risk patients, which means fewer false positives for admission. The ATS and IDSA 2019 guidelines prefer the PSI over CURB-65 for this reason. CURB-65 (Confusion, Urea, Respiratory rate, Blood pressure, age 65 or over) uses only 5 variables, making it faster and feasible when laboratory results are unavailable. CURB-65 has higher specificity for predicting mortality (74.6% vs 52.2% for PSI), meaning fewer healthy patients are falsely flagged as high risk. A practical approach: use the PSI as the primary tool when full laboratory data is available and you are considering outpatient discharge; use CURB-65 as a rapid bedside screen or when lab data is delayed. Importantly, they should supplement, not replace, clinical judgment.
What disposition is appropriate for PSI Class III patients?
Class III (score 71-90 points) represents a low-moderate risk group with a 30-day mortality of 0.9 to 2.8% — meaningfully higher than Class I-II but far below Class IV-V. The evidence does not mandate hospitalization for Class III, and many patients are appropriately managed as outpatients. The preferred approach depends on individual patient factors: if the patient can reliably take oral antibiotics, has adequate home support, no functional limitations, can follow up within 24 hours, and has no concerning social factors, outpatient management is reasonable with careful arrangement of next-day follow-up. If any of these conditions are not met — inability to tolerate oral medications, unreliable follow-up, significant comorbidities not fully captured by the score, or clinician concern — a brief observation admission (24 hours) or short-stay unit placement is appropriate.