PSI calculator

The Pneumonia Severity Index, also published as the PORT score after the Pneumonia Patient Outcomes Research Team, is a widely taught additive rule for community-acquired pneumonia. This calculator implements the Fine 1997 two-step algorithm: it converts demographics, five comorbidity flags, vital-sign thresholds, and laboratory cutoffs into a point total, then assigns risk class I through V with a published 30-day mortality band and a disposition hint. Optional flags and labs that you omit contribute zero points, which means they are treated as not meeting that criterion.

What the Pneumonia Severity Index measures and how Fine’s two steps work

Community-acquired pneumonia covers a wide range of 30-day risk, from previously healthy adults who recover at home to patients who need inpatient monitoring. Fine and colleagues derived the Pneumonia Severity Index from the PORT cohort and published the rule in the New England Journal of Medicine in 1997 so that clinicians and students could stratify that range with a single, inspectable score instead of an opaque model.

Step 1 asks whether the patient is fifty years of age or younger and has none of five comorbidities—neoplastic disease, liver disease, congestive heart failure, cerebrovascular disease, and renal disease—and none of five examination findings: altered mental status, respiratory rate at least thirty breaths per minute, systolic blood pressure below ninety millimetres of mercury, temperature below thirty-five or at least forty degrees Celsius, and pulse at least one hundred twenty-five beats per minute.

When every Step 1 screen is negative, the original algorithm assigns risk class I without waiting for laboratory points. Step 2 always still computes an additive total: men receive their age in years as demographic points and women receive age minus ten; a nursing-home resident adds ten; comorbidities add thirty, twenty, or ten according to the published table; examination abnormalities add twenty, fifteen, or ten; and laboratory or radiographic findings add thirty for acidemia, twenty for elevated urea or hyponatremia, and ten for hyperglycemia, anemia, hypoxemia, or pleural effusion.

Classes II through V then follow the cutoffs at seventy, ninety, and one hundred thirty points.

Input contract, units, validation, and the shape of a successful response

Sex must be male or female (single-letter m or f is normalized) because the Fine table subtracts ten points for women. Every comorbidity and the nursing-home and pleural-effusion flags are optional booleans that default to false.

Vital signs and laboratories are optional finite numbers: respiratory rate in breaths per minute, systolic blood pressure in millimetres of mercury, temperature in Celsius, pulse in beats per minute, arterial pH, blood urea nitrogen in milligrams per decilitre, sodium in milliequivalents per litre, glucose in milligrams per decilitre, hematocrit in percent, arterial oxygen tension in millimetres of mercury, and oxygen saturation in percent. An omitted vital or lab contributes zero points rather than inventing a normal value, so a score computed from demographics alone is honest about what was scored.

Hypoxemia is awarded at most ten points if either PaO2 is below sixty or saturation is below ninety, never twenty for sending both. Optional output keys are omitted rather than set to null.

How to read risk classes, where the calculator belongs, and what it does not decide

Risk class I is very low risk, with a PORT validation mortality near one tenth of a percent, and the usual teaching disposition is outpatient care when the social setting is safe. Class II remains low risk at about six tenths of a percent and is also commonly taught as outpatient. Class III is moderate risk near nine tenths of a percent and is often framed as observation or a short inpatient stay.

Class IV is high risk near nine point three percent and class V is very high risk near twenty-seven percent; both are taught as inpatient strata, with class V prompting a conversation about higher-acuity monitoring. Those percentages are historical cohort figures from the original validation, not a personal prognosis, and they do not replace CURB-65, ATS/IDSA major-minor criteria, oxygenation trends, or a clinician’s view of comorbidities this table never captures.

Because Step 1 ignores laboratories, a young adult who meets class I screens can still carry lab points in the numeric score; the response reports both the class and the buckets so you can see that tension instead of hiding it. Use the calculator for education, exam keys, research coding of community-acquired pneumonia rows, and product prototypes that need a deterministic PSI step before handing off to licensed systems.

Neighbouring tools such as CURB-65, shock index, or eGFR answer different search intents and should keep their own landings rather than being overloaded here.