Factors associated with the hospitalization of low-risk patients with community-acquired pneumonia in a cluster-randomized trial

J Gen Intern Med. 2006 Jul;21(7):745-52. doi: 10.1111/j.1525-1497.2006.00510.x.


Background: Many low-risk patients with pneumonia are hospitalized despite recommendations to treat such patients in the outpatient setting.

Objective: To identify the factors associated with the hospitalization of low-risk patients with pneumonia.

Methods: We analyzed data collected by retrospective chart review for 1,889 low-risk patients (Pneumonia Severity Index [PSI] risk classes I to III without evidence of arterial oxygen desaturation) enrolled in a cluster-randomized trial conducted in 32 emergency departments.

Results: Overall, 845 (44.7%) of all low-risk patients were treated as inpatients. Factors independently associated with an increased odds of hospitalization included PSI risk classes II and III, the presence of medical or psychosocial contraindications to outpatient treatment, comorbid conditions that were not contained in the PSI (cognitive impairment, history of coronary artery disease, diabetes mellitus, or pulmonary disease), multilobar radiographic infiltrates, and home therapy with oxygen, corticosteroids, or antibiotics before presentation. While 32.8% of low-risk inpatients had a contraindication to outpatient treatment and 47.1% had one or more preexisting treatments, comorbid conditions, or radiographic abnormalities not contained in the PSI, 20.1% had no identifiable risk factors for hospitalization other than PSI risk class II or III.

Conclusions: Hospital admission appears justified for one-third of low-risk inpatients based upon the presence of one or more contraindications to outpatient treatment. At least one-fifth of low-risk inpatients did not have a contraindication to outpatient treatment or an identifiable risk factor for hospitalization, suggesting that treatment of a larger proportion of such low-risk patients in the outpatient setting could be achieved without adversely affecting patient outcomes.

Publication types

  • Randomized Controlled Trial
  • Research Support, N.I.H., Extramural
  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, Non-P.H.S.

MeSH terms

  • Cluster Analysis
  • Community-Acquired Infections / classification
  • Community-Acquired Infections / therapy*
  • Hospitalization / statistics & numerical data*
  • Humans
  • Inpatients
  • Odds Ratio
  • Outpatients
  • Pneumonia / classification
  • Pneumonia / etiology
  • Pneumonia / therapy*
  • Retrospective Studies
  • Risk Assessment