The Pneumonia Severity Index, usually called PSI or the PORT score, was built to answer one question: which patients with community-acquired pneumonia can safely be treated at home? It was derived and validated in the PORT cohort in the 1990s and remains the most sensitive tool for identifying low-risk patients, which is why it is still embedded in guideline pathways despite being more cumbersome than CURB-65.
What goes into the score
PSI is unusual in having a two-step structure. Step one is a screening rule: a patient aged 50 or under with no comorbidity, normal vital signs, and normal mental status is Class I and needs no further scoring. Everyone else proceeds to step two, which assigns points for age, nursing-home residence, five comorbid conditions (neoplasm, liver disease, congestive heart failure, cerebrovascular disease, renal disease), physical examination findings (altered mental status, respiratory rate ≥30, systolic BP <90, temperature <35°C or ≥40°C, pulse ≥125), and laboratory values (pH, BUN, sodium, glucose, hematocrit, PaO₂).
Reading the class
The total maps to five classes. Class I and II are low risk and generally suitable for outpatient care. Class III is the ambiguous middle and often warrants a short observation stay. Class IV and V are high risk and should be admitted, with Class V considered for intensive care. The point thresholds are ≤70 for Class II, 71–90 for III, 91–130 for IV, and above 130 for V.
Evidence and use
PSI was validated in a large prospective cohort and has been shown to reduce unnecessary admissions when used systematically. Its weakness is that it is more complex than CURB-65 and requires laboratory data, so it is less useful in the emergency department at triage. Many guidelines recommend using PSI to rule out admission and CURB-65 to rule in severity, treating the two as complementary rather than competing. Neither score replaces clinical judgment, and neither should be used to deny admission to a patient who looks unwell.
Source and disclaimer
Source: Fine MJ et al., "A prediction rule to identify low-risk patients with community-acquired pneumonia", New England Journal of Medicine, 1997; and the Infectious Diseases Society of America / American Thoracic Society CAP guideline. Educational use only — not a substitute for clinical assessment, and not for diagnostic or treatment decisions.