Abandoning daily routine chest radiography in the intensive care unit: meta-analysis

Radiology. 2010 May;255(2):386-95. doi: 10.1148/radiol.10090946.


Purpose: To systematically examine whether abandoning daily routine chest radiography would adversely affect outcomes, such as mortality and length of stay (LOS), and identify a subgroup in which daily routine chest radiography might be beneficial.

Materials and methods: This was a meta-analysis of clinical trials that examined the effect of abandoning daily routine chest radiography in adults in intensive care units (ICUs). Studies were identified through searches of MEDLINE, Cochrane Database, Database of Abstracts of Reviews of Effects, Biological Abstracts, and CINAHL. The results were expressed as odds ratios (ORs) or weighted mean difference (WMD) along with their 95% confidence intervals (CIs).

Results: Eight studies with a total of 7078 patients were identified. A pooled analysis revealed that the elimination of daily routine chest radiography did not affect either hospital or ICU mortality (OR, 1.02;[95% CI: 0.89, 1.17; P = .78 and OR, 0.92; 95% CI: 0.76, 1.11; P = .4, respectively). There was no significant difference in ICU LOS (WMD = 0.19 days; 95% CI: -0.13, 0.51; P = .25), hospital LOS (WMD = -0.29 days; 95% CI: -0.71, 0.13; P = .18), and ventilator days (WMD = 0.33 days; 95% CI: -0.12, 0.78; P = .15) between the on-demand and daily routine groups. Regression analyses failed to identify any subgroup in which performing daily routine chest radiography was beneficial.

Conclusion: Systematic but unselective daily routine chest radiography can likely be eliminated without increasing adverse outcomes in adult patients in ICUs. Further studies are necessary to identify the specific patient population that would benefit from daily routine chest radiographs.

Publication types

  • Meta-Analysis

MeSH terms

  • Evidence-Based Medicine
  • Hospital Mortality
  • Humans
  • Intensive Care Units / organization & administration*
  • Length of Stay
  • Outcome Assessment, Health Care
  • Practice Patterns, Physicians'*
  • Radiography, Thoracic / statistics & numerical data*
  • Regression Analysis
  • Respiration, Artificial / statistics & numerical data