Diagnostic value of chest radiographs in bedridden patients suspected of having pneumonia

Am J Med. 2010 Jan;123(1):88.e1-5. doi: 10.1016/j.amjmed.2009.09.012.

Abstract

Objectives: To assess the diagnostic value of the chest radiograph for the diagnosis of pneumonia in bedridden patients, using non-contrast-enhanced high-resolution chest computed tomography (CT) as the gold standard.

Methods: We prospectively evaluated bedridden patients hospitalized with moderate to high clinical probability of pneumonia. Chest radiographs were interpreted in a blinded fashion by 3 observers and classified as definite, normal, or uncertain for pneumonia. Chest CT was obtained within 12 hours of chest radiograph. We applied Bayesian analysis to assess the accuracy of chest radiograph in the diagnosis of pneumonia.

Results: In a 5-month period, 58 patients were evaluated, 31 (53%) were female. Their chest radiographs were interpreted as negative, uncertain, or positive for pneumonia in 31 (53%), 15 (26%), and 12 (21%) patients, respectively, while CT confirmed pneumonia in 11 (35%), 10 (67%), and in 10 (83%). The sensitivity of the chest radiograph to diagnose pneumonia was 65%, the specificity was 93%, the positive and negative predictive values were, respectively, 83% and 65%, while the overall accuracy was 69% (95% confidence interval, 50%-79%).

Conclusions: In bedridden patients with suspected pneumonia, a normal chest radiograph does not rule out the diagnosis, hence, a chest CT scan might provide valuable diagnostic information.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Age Distribution
  • Aged
  • Aged, 80 and over
  • Bed Rest / adverse effects*
  • Female
  • Follow-Up Studies
  • Hospitals, University
  • Humans
  • Incidence
  • Inpatients
  • Male
  • Middle Aged
  • Pneumonia / diagnostic imaging*
  • Pneumonia / epidemiology
  • Pneumonia / etiology
  • Probability
  • Prospective Studies
  • Radiography, Thoracic / methods*
  • Risk Assessment
  • Sex Distribution
  • Tomography, X-Ray Computed / methods*
  • Young Adult