Winter is Coming: Are You Using the Right Pneumonia Score?

Clinical Scenario

Earl, proudly hosting his 75th birthday party, blew out the candles with such force (and cough) that half the frosting—and likely half the room—were coated in suspicious aerosols. Two days later, he arrived in the ED febrile and confused, with a respiratory rate of 32, SpO₂ 88% on room air, and systolic BP hovering around 90 mmHg. Chest X-ray showed multilobar infiltrates, and his heart rate was 115, raising concern for significant pneumonia severity. His family thought it was just “birthday excitement,” but your clinical decision tools think otherwise…

Don’t just diagnose pneumonia, predict who might acutely decompensate.

Summary

CURB-65: Fast, familiar, and great for admission decisions - but its simplicity can overlook some high-risk, ICU-bound patients.

PSI/PORT: In-depth and data-heavy - best for mortality risk and disposition decisions, but may miss early deterioration.

SMART-COP: Designed to detect who will crash - flags need for respiratory support and ICU-level care that other scores may miss.

Tool Comparison

	CURB-65	PSI/PORT	SMART-COP
Purpose	Predicts mortality risk and need for hospitalization in community-acquired pneumonia (CAP).	Estimates 30-day mortality risk in CAP; guides site-of-care decisions (outpatient vs. inpatient).	Predicts need for intensive respiratory or vasopressor support (IRVS) in CAP.
Use When...	At initial presentation, especially in ED or outpatient settings for quick triage.	At diagnosis of CAP when more comprehensive data are available (labs, comorbidities, vitals).	In hospitalized CAP patients to identify those likely to need ICU-level care.
Scoring	5 variables: Confusion, Urea >7 mmol/L (BUN >19 mg/dL), RR ≥30, BP <90 systolic or ≤60 diastolic, Age ≥65.	20 variables including demographics, comorbidities, vitals, and labs; total 0–395 points.	8 variables including SBP, multilobar infiltrates, albumin, RR, tachycardia, confusion, oxygenation, pH.
Performance	Simple and quick; good mortality prediction but less granular than PSI.	Highly accurate for mortality prediction; validated in >40,000 patients.	Strong predictor for need of IRVS. Better identifies severe disease than CURB-65 or PSI, especially in younger patients.
Workup	Score 0–1: Outpatient likely safe. 2: Consider admission. ≥3: High risk, hospitalize.	Class I–II: Outpatient. III: Brief inpatient or observation. IV–V: Inpatient or ICU.	Score ≥3: High risk for IRVS; consider ICU evaluation or early critical care support.
Common Pitfalls	Does not account for comorbidities or oxygenation. Includes points for confusion and blood urea nitrogen, which in the acutely ill elderly patient, could be due to a variety of factors. May underestimate severity in younger patients.	Complex and time-consuming, as it requires many data points. May underestimate severe pneumonia in an otherwise young healthy patient. Any patient over 50 years of age is automatically classified as risk class 2. Performed slightly worse than CURB-65 in prediction of mortality and need for intensive care.	Requires ABG or O₂ data; may overestimate severity in patients with transient hypoxemia. Does not estimate mortality. Variables not included, such as comorbidities, functional status, frailty, and physician gestalt, may still recommend ICU admission.

NOTE: None of the current scores include acute phase inflammatory markers such as procalcitonin, which may improve risk stratification.

Case Resolution

Earl’s week started with cake and candles but progressed to crackles and confusion. CURB-65 suggested he stay in the hospital, PSI/PORT reminded us that 75-year-olds with confusion and multilobar pneumonia cannot be discharged with a goodie bag, and SMART-COP waved an ICU flag. Thanks to early recognition and escalation, he recovered quickly (and now coughs away from desserts!).