Community-Acquired Pneumonia

Facts and Figures

Physical Exam

Criteria for diagnosis

Specific organisms

Drug Classes
Mechanism of action

Site of care
Assessment of response

Literature cited
Complete bibliography



Site of Care
  • having made the diagnosis of CAP, the 1st step in treatment must assess the severity of illness and a decision made as to whether the patient may be treated as an outpatient or if hospitalization is required

  • it is estimated that approximately 75% of patients with CAP may be treated as outpatients

  • several studies have attempted to identify risk factors that predict increased mortality; Canadian and ATS guidelines defined severe CAP as those cases that included respiratory failure, respiratory rate > 30/min, sepsis or shock, or the presence of extensive infiltrates 20

  • the IDSA guidelines for hospitalization are based on Fine's prediction rule 21 which takes a two-step approach in deciding the appropriate site of care (see IDSA treatment flowchart)

  • according to the rule, patients can be stratified into five risk classes (I-V) using a point system based on 19 variables (Table 5)

    TABLE 5. Prediction model scoring system 22
    Patient characteristics
    Points assigned *
    Demographic factors
    age: males
    age: females
    nursing home resident

    age (in years)
    age (in years) - 10
    Comorbid illnesses
    neoplastic disease
    liver disease
    congestive heart failure
    cerebrovascular disease
    renal disease

    Physical examination findings
    altered mental status
    respiratory rate > 30/min
    SBP < 90 mm Hg
    temperature < 35 C or > 40 C
    pulse > 125/min

    Laboratory findings
    pH < 7.35
    BUN > 10.7 mmol/L
    sodium < 130 mEq/L
    glucose > 13.9 mmol/L
    hematocrit < 30
    PO2 < 60 mm Hg**
    pleural effusion

    * the total risk score is obtained by summing the age in years (age - 10 for females) and the points for each applicable characteristic
    ** oxygen saturation < 90% was also considered abnormal

  • Step 1 attempts to identify patients in Class I by using information obtained from the H&P which is the criteria most commonly available in the primary care setting; this approach is intended to allow a quick assessment of patients who are candidates for outpatient treatment

  • the assessment algorithm in Step 1 considers patient age (> 50 years?), comorbid conditions present (neoplastic disease, CHF, cerebrovascular disease, renal or liver disease?) and physical findings (altered mental status, pulse 125/min, RR 30/min, SBP < 90 mm Hg, temperature < 35 C or 40 C?); a "no" response to all of these questions would place the patient in Class I (Table 6); a reply of "yes" to any of the above would place the patient in Classes II-V and further evaluation based on the above scoring system would be used to determine risk class

    TABLE 6. Stratification of risk score 23
    Risk class
    Based on
    < 70 points
    71-90 points
    > 131

  • patients in Risk Classes I-III have a low risk of mortality (0.1-0.4%, 0.6-0.7%, and 0.9-2.8% respectively) and may be candidates for outpatient care

  • Step 2 uses the scoring chart (Table 5) to stratify patients into risk classes II-V; hospitalization is recommended for patients in classes IV and V; patients in class III may be treated as outpatients or inpatients depending on clinical judgment

  • problems with the prediction model include the possibly time-consuming requirement of calculating a score based on 19 variables and the fact that social factors, which often make outpatient care impractical, are not considered

  • the prediction model is intended to be used as an aid in deciding site of care but is not meant to replace clinical judgment


20) Mandell LA, Community-acquired pneumonia: etiology, epidemiology, and treatment, Chest, 1995, 108:35S-42S.

21) Fine MJ, et al., A prediction rule to identify low-risk patients with community-acquired pneumonia, NEJM, 1997 Jan 23, 336(4):243-50.

22) Fine MJ, et al., A prediction rule to identify low-risk patients with community-acquired pneumonia, NEJM, 1997 Jan 23, 336(4):243-50.

23) Fine MJ, et al., A prediction rule to identify low-risk patients with community-acquired pneumonia, NEJM, 1997 Jan 23, 336(4):243-50.

There's no place like home... Top of page
© Copyright 1999, All rights reserved. L. España. Rev. 3/99   3/30/99