Background: The clinical value of postoperative chest radiographs (CXRs) for surgical intensive care unit (ICU) patients is largely unknown. In the present study, we determined the diagnostic and therapeutic efficacy of postoperative CXRs for different surgical subgroups and related their efficacy to the time after ICU admission.
Methods: A prospective, observational study of consecutive postoperative surgical ICU patients was performed during a 10 month period. We restricted our analysis to CXRs obtained within six hours after admission to the ICU. Diagnostic efficacy was defined by the presence of predefined major abnormalities; therapeutic efficacy was defined by predefined actions taken because of any abnormality found on postoperative CXRs.
Results: Of 857 surgical ICU patients, 670 (78%) had a postoperative CXR after admission to the ICU. Of these CXRs, 80 were performed for clinical reasons, and 590 were routinely obtained (i.e., these CXRs were made without a reason other than admission to the ICU itself). The diagnostic efficacy of clinically indicated and routinely obtained CXRs was 18% (14/80) and 13% (79/590), respectively. Of all predefined abnormalities found on CXRs, 60% involved the malposition of invasive devices, such as endotracheal tubes or central venous lines. The therapeutic efficacy of clinically indicated and routinely obtained CXRs was 4% (3/80) and 4% (26/590), respectively. While the diagnostic and therapeutic efficacy of routinely obtained CXRs were not dependent on timing of admission, the diagnostic and therapeutic efficacy of clinically indicated CXRs was higher for CXRs taken closer to the time of ICU admission.
Conclusion: Although the diagnostic efficacy of clinically indicated and routinely obtained postoperative CXRs in surgical ICU patients appears to be significant, their therapeutic efficacy is low.