Electronic supplementary material
The online version of this article (doi:10.1186/1471-2253-14-94) contains supplementary material, which is available to authorized users.
The authors declare that they have no competing interests.
MH, HLV and RJ are the guarantors of integrity of entire study. All authors participated in the study concepts and study design, data acquisition and data analysis and interpretation. MH, PD, FJK and RJ performed the literature research. MH, RJ and MS conducted the statistical analysis. Manuscript drafting and manuscript revision for important intellectual content was performed by all authors. Manuscript editing was performed by all authors. All authors read and approved the final manuscript.
To assess the value of a score-based system which allows standardized evaluation of pulmonary edema on bedside chest radiographs (CXRs) under routine clinical conditions.
Seven experienced readers assessed bedside CXRs of ten patients with an extravascular lung water (EVLW)-value of ≤ 8 mL/kg (range: 4–8 mL/kg; indicates no pulmonary edema) and a series of ten patients with an EVLW-value of ≥ 15 mL/kg (range: 15–21 mL/kg; = indicates a pulmonary edema) with and without customized software which would permit a standardized assessment of the various indications of pulmonary edema. The software provides a score that identifies patients with and without pulmonary edema. EVLW-values were measured instantly after bedside CXR imaging using a pulse contour cardiac output (PiCCO) system and served as a reference standard. The patients were non-traumatic and not treated with diuretics or dobutamine during bedside CXR imaging and the PiCCO measurements. Mean sensitivity, specificity, positive and negative predictive value, the percentage of overall agreement and the free-marginal multirater kappa value was calculated for both the standard and the standardized score-based approach. The net reclassification index was calculated for each reader as well as for all readers.
Evaluation of bedside CXRs by means of the score-based approach took longer (23 ± 12 seconds versus 7 ± 3 seconds without the use of the software) but improved radiologists’ sensitivity (from 57 to 77%), specificity (from 90 to 100%) and the free-marginal multirater kappa value (from 0.34 to 0.68). The positive predictive value was raised from 85 to 100% and the negative predictive value from 68 to 81%. A net reclassification index of 0.3 (all readers) demonstrates an improvement in prediction performance gained by the score-based approach. The percentage of overall agreement was 67% with the standard approach and 84% with the software-based approach.
The diagnostic accuracy of bedside CXRs to discriminate patients with elevated EVLW-values from those with a normal value can be improved with the use of a standardized score-based approach. The investigated system is freely available as a web-based application (accessible via: http://www.radiologie.uk-erlangen.de/aerzte-und-zuweiser/edema).