Back to the list
Congress: ECR25
Poster Number: C-20435
Type: Poster: EPOS Radiologist (educational)
Authorblock: D. I. Krimli, S. Shihata, A. A. Felemban, M. I. Alsayed, O. H. Alaslani, B. A. Bannan, Z. Alsabban, Z. Zia; Jeddah/SA
Disclosures:
Danah Ismail Krimli: Nothing to disclose
Salah Shihata: Nothing to disclose
Abdulbari Abdulkaliq Felemban: Nothing to disclose
Mohammed Ibrahim Alsayed: Nothing to disclose
Ohoud Hameed Alaslani: Nothing to disclose
Badr Abdulmajied Bannan: Nothing to disclose
Zehour Alsabban: Nothing to disclose
Zergham Zia: Nothing to disclose
Keywords: Abdomen, Cardiovascular system, Emergency, CT, MR, Ultrasound, Computer Applications-Detection, diagnosis, Quality assurance
Findings and procedure details

1. Discrepancy Rates Between Provisional and Finalized Reports of 1,367 emergency studies, 70 (5.1%) exhibited discrepancies between provisional resident reports and finalized consultant interpretations. These discrepancies were distributed across imaging modalities as follows:  

      • CT: 56/874 (6.4%)  
      • MRI: 4/89 (4.5%)  
      • US: 10/402 (2.5%)  

 

2. Clinical Significance Determined by RadPeer Reviewer Scores  

The clinical significance of each discrepancy was independently assessed by two board-certified radiologists using the ACR 2016 RadPeer scoring system, which categorizes discrepancies based on their potential impact on patient management:  

  • 2a/3a: Non-significant discrepancies (e.g., stable postoperative changes, incidental findings).  
  • 2b/3b: Clinically significant discrepancies (e.g., missed acute infarctions, pulmonary embolism).  

Key Findings:  

  • 3% (45/70) of discrepancies were graded as clinically significant (2b/3b) by the reviewers.
  • 7% (25/70) were classified as non-significant (2a/3a).

Subgroup Analysis by Body Region:  

  • Neuroimaging: Of 30 discrepancies, 36.7% (11/30) were graded as significant (e.g., 8 missed acute infarctions, 3 venous thromboses).
  • Chest/Abdomen: Of 40 discrepancies, 50% (20/40) were graded as significant (e.g., 3 appendicitis, 4 bowel obstructions).

Critical Note:  

Clinical significance was determined solely by the reviewers’ application of RadPeer criteria to the discrepancies identified between provisional and final reports. The scoring reflects the reviewers’ consensus on the potential impact of the discrepancy—not the mere existence of a discrepancy.

 

3. Inter-Reviewer Agreement in RadPeer Scoring

Disagreements between reviewers in categorizing clinical significance occurred in 20% of cases (14/70):  

  • Neuroimaging: 30% disagreement (9/30), primarily in grading subtle findings (e.g., ischemic changes scored as 3b vs. 2a).
  • Chest/Abdomen: 12.5% disagreement (5/40), often involving severity thresholds (e.g., appendicitis graded as 3b vs. 2a).

Agreement Analysis:  

  • Full agreement (1/2a/3a/2b/3b): 80% of cases (56/70).
  • Partial agreement (e.g., 2a vs. 3a): 12% of cases (8/70).
  • Full disagreement (e.g., 2b vs. 3a): 8% of cases (6/70).
GALLERY