The Pneumonia Severity Index turns 20 routine findings — age and sex, nursing-home residence, five comorbidities, five exam abnormalities, and seven lab or radiographic results — into a 30-day mortality risk class for adults with community-acquired pneumonia. It is one of the most extensively validated pneumonia risk tools and is used mainly to identify low-risk patients who can safely be treated as outpatients.
How the PSI is scored
The rule was derived from more than 14,000 patients and validated on over 40,000 (Fine MJ et al., N Engl J Med 1997). Each of the 20 items carries a fixed integer weight; the weights are summed to a total that maps to Risk Class II–V. Age dominates the score because it contributes its full value (men) or its value minus 10 (women). The two heaviest single items are neoplastic disease and an arterial pH below 7.35, each worth 30 points.
This calculator implements the transparent point table and reports the derivation-cohort 30-day mortality for each class.
Risk Class I is a screening shortcut, not a low score
The single most common mistake in implementing the PSI is assigning Class I to any patient with a low point total. Class I is instead a screening step that runs before scoring: a patient younger than 50 with none of the five comorbidities and none of the five exam abnormalities is Class I. Crucially, nursing-home residence and every lab or imaging finding are excluded from this screen — so a healthy 45-year-old with an abnormal lab is still Class I, while a 55-year-old with zero risk factors is Class II, not Class I. This calculator applies the screen first and shows the point total for reference.
How it guides disposition and its limits
The PSI is strongest at confirming who is safe for outpatient care: Classes I–III carry low 30-day mortality and often support outpatient or brief-observation management, while Classes IV–V generally warrant admission, and Class V often ICU-level care. The score is a decision aid, not a decision — it does not capture oxygenation trends, stable social support, drug interactions, or the trajectory of illness, and it should never be the sole basis for an admission or discharge. Many clinicians pair it with a fast bedside tool such as CURB-65.