brand logo

Am Fam Physician. 2006;73(8):1425-1428

Clinical Question

What is the role of clinical prediction tools in helping physicians decide on inpatient or outpatient treatment for patients with community-acquired pneumonia (CAP)?

Evidence Summary

CAP often is managed in an outpatient setting, an approach endorsed by evidence-based guidelines from the American Thoracic Society (ATS)1 and the Infectious Diseases Society of America (IDSA).2 However, these guidelines recommend that physicians make an objective risk assessment using a prospectively validated clinical prediction tool to help guide them, at least in part, when deciding on inpatient or outpatient treatment. The most notable of these tools are the Pneumonia Severity Index (PSI) and several variations of the British Thoracic Society (BTS) rule, such as the CURB-65 (Confusion, Urea nitrogen, Respiratory rate, Blood pressure, 65 years of age and older) score.

The PSI (Table 116) was developed from an administrative data set of 14,199 adults and validated by the original investigators in a second group of 2,287 community-based and nursing home patients.3 It was subsequently validated in a number of populations including 158 nursing home patients,6 3,181 patients with CAP at 32 Pennsylvania emergency departments,4 and 1,024 patients at 22 community hospitals.5 In a prospective trial,7 hospitals were randomized to treat patients with CAP using usual care or a PSI-based protocol (i.e., patients presenting to the emergency department with CAP who had a PSI risk class of I, II, or III were treated as outpatients, although physicians used clinical judgment to overrule these criteria in some instances). On average, patients treated using the PSI protocol had greater severity of illness; however, they were less likely to be hospitalized, had shorter hospitalizations, and had similar clinical outcomes compared with patients treated using usual care.7 An online PSI calculator is available at

The rightsholder did not grant rights to reproduce this item in electronic media. For the missing item, see the original print version of this publication.

The CURB-65 score (Table 24,8,9) only includes five variables (compared with up to 20 in the PSI); therefore, it is easier to calculate and interpret at the point of care.8

The authors of the PSI recommend outpatient therapy for patients in PSI risk classes I and II, physician judgment for those in class III, and hospitalization for those in risk classes IV and V.3 The IDSA guideline recommends that physicians consider home therapy for patients in PSI risk classes I, II, and III.2 The BTS guideline recommends that physicians use the CURB-65 or the CRB-65 (which excludes the urea nitrogen value if blood testing is not immediately available) score (Table 24,8,9) when deciding on inpatient or outpatient treatment.9 The ATS guideline recommends that physicians use validated clinical decision rules such as the PSI or the CURB-65 score to support clinical judgment but does not define a recommended cutoff for hospital admission.1 A clinical prediction rule that uses only clinical variables has been developed in nursing home patients; however, it has not been prospectively validated and was based on a retrospective chart review, which is less reliable than prospective data collection.10

Clinical factorPoints
Blood urea nitrogen > 19 mg per dL (6.8 mmol per L)1
Respiratory rate ≥ 30 breaths per minute1
Systolic blood pressure < 90 mm Hg1
Diastolic blood pressure ≤ 60 mm Hg
Age ≥65 years1
Total points:________

All of the guidelines mentioned recommend that physicians use prediction tools to support, not replace, clinical judgment. External factors such as important comorbidities not included in the clinical rules (e.g., human immunodeficiency virus), failure of outpatient oral therapy, and social factors (e.g., a patient’s inability to obtain or reliably take medication) are appropriate considerations when deciding on inpatient or outpatient treatment.11

Applying the Evidence

A 62-year-old man presents with cough, fever, and chills for three days. He has well-controlled hypertension and diabetes but is otherwise healthy. His respiratory rate is 24 breaths per minute and his blood pressure and pulse are in the normal range; he has no signs of confusion. His white blood cell count is 23,000 cells per mm3 (23 × 109 per L) with 80 percent neutrophils, and his blood urea nitrogen is 14 mg per dL (5.0 mmol per L). The patient prefers not to be hospitalized. Is outpatient treatment safe for this patient?

Answer: You calculate the patient’s CURB-65 score rather than the PSI score, because arterial blood gas measurements and radiography are not immediately available. The score is 0, which suggests that it is safe to treat him as an outpatient. Although his white blood cell count is elevated, this risk factor is not included in any of the three validated clinical decision rules.

This guide is one in a series that offers evidence-based tools to assist family physicians in improving their decision-making at the point of care.

This series is coordinated by Mark H. Ebell, MD, MS, deputy editor for evidence-based medicine.

A collection of Point-of-Care Guides published in AFP is available at

Continue Reading

More in AFP

More in PubMed

Copyright © 2006 by the American Academy of Family Physicians.

This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP.  See permissions for copyright questions and/or permission requests.