This article evaluated the inter-rater reliability of assessing characteristics of liver lesions and overall staging of hepatocellular carcinomas using various staging systems (OTPN, AASLD, and LI-RADS).
10 blinded readers reviewed 100 liver MRI cases assessing lesions for HCC criteria (arterial hyper-enhancement, washout, and pseudocapsule). Readers were also asked to assign an OTPN, AASLD and LI-RADS score (ver. 2013.1) based on their findings.
5 abdominal imaging staff at an OPTN approved liver transplant center (6-11 years of experience) and 5 “novice” radiology residents were selected to be a part of the study. 10 cases were used for training but not included as part of the study results.
High Kappa scores were achieved for the finding of arterial hyper-enhancement (0.67). Moderate scores for wash-out and the presence of a pseudocapsule (0.48 and 0.52 respectively).
As for staging moderate agreement was noted for OPTN staging (0.53) and fair agreement for LI-RADS and AASLD (0.35 and 0.39).
Overall agreement amongst experts was significantly higher in comparison to that of novice residents (P<.001). Interestingly (yet not surprising) agreement between experts for LI-RADS and AASLD was fair to moderate (0.43 and 0.46) in comparison to OPTN (0.64).
PICO Analysis:
Population: Radiologists
Intervention: The effect of experience for interpretation and accurate staging of hepatocellular carcinomas.
Comparison: “experts” versus “novice” residents
Outcome: There is higher consistency and accuracy amongst experts reading liver MRI. There is moderate agreement amongst “experts” for relatively complex classification systems (LI-RADS and AASLD) in comparison to high agreement for OPTN
Advancement in Knowledge:
1. Experience plays a role in the interpretation of Liver MRI
2. There is only moderate agreement amongst “experts” in staging hepatocellular carcinoma using LI-RADS and AASLD.
Reference:
Pubmed Full Text
No comments:
Post a Comment