To determine the intraindividual impact of radiation dose reduction in abdominal computed tomography (CT) on diagnostic performance in patients with suspected appendicitis. This study was approved by the institutional review board. Seventy-five patients who underwent standard contrast-enhanced abdominal CT for suspected appendicitis between 2004 and 2009 were retrospectively included. Low-dose CT reconstructions with 75%, 50%, and 25% of the original radiation dose level were generated by applying realistic reduced-dose simulation. Two blinded, independent readers assessed image quality, signal-to-noise ratio, and diagnostic confidence on each dataset. Diagnostic accuracy for detection of appendicitis and complications were calculated for each reader. Paired univariate tests were used to determine intraindividual differences. Among 75 subjects included in the analysis (57% female, mean age: 41 ± 18 years), the prevalence of histopathologically confirmed appendicitis was 59%. Signal-to-noise ratio and subjective image quality of 50% and 25% reduced-dose CTs were significantly lower than the reference datasets (all P < .005). Appendicitis was correctly identified in all reference and low-dose datasets (sensitivity: 100%, negative predictive value: 100%). Presence of complications was correctly detected in all reference, 75%, and 50% datasets, but was decreased in 25% datasets (sensitivity: 77.8% and negative predictive value: 97.4%). Diagnostic confidence was high for original and 75% datasets, but significantly lower for 50% and 25% datasets (P < .001). Our results indicate that diagnostic accuracy in abdominal CT acquisitions acquired at 75% and 50% of radiation dose is maintained in patients with suspected appendicitis, whereas further reduction of radiation exposition is associated with decreased diagnostic performance.
Read full abstract