Cholangiocarcinoma: Causes of Failure in Detection in Biliary Brushing Specimens.
Frank N Moore, Sharon B Sams, Stephen S Raab. University of Colorado at Denver Health Science Center, Aurora
Background: Our study aims to determine the causes of error in detecting cholangiocarcinoma in bile duct brushings by performing root cause analysis of false-negative bile duct brushing cases to assess (1) specimen and (2) sampling quality. Determining root cause allows for the design of quality improvement initiatives to detect high-grade dysplasia and cholangiocarcinoma.
Design: We performed a 5 year retrospective review of all false-negative bile duct brushings in patients who were diagnosed with either high-grade dysplasia (n=17) or cholangiocarcinoma (n=15) in bile duct biopsies obtained at the time of bile duct brushing. The mean age of patients was 66 years old (range = 43-82 years old). Original bile duct brushing specimens were re-screened by 2 board certified pathologists and one senior resident and reclassified accordingly. Root cause analysis was performed on each case to determine specimen quality and sampling quality. Quality specimens were defined as specimens without processing artifact or obscuring components. Quality sampling was defined as specimens with identifiable tumor.
Results: The frequency of a positive bile duct brushing for suspicious and malignant bile duct biopsies was 19% (n=64). Re-screened sensitivity and specificity of bile duct brushing for either high-grade dysplasia or cholangiocarcinoma were 43% and 86.5% respectively. In 59.3% of cases, failure in detection was seen with a poor quality specimen and poor sampling quality; a system-related error. In 25% of cases, failure in detection was seen with a poor quality specimen and adequate sampling; a combination of system-related and pathologist-related error. In 9.3% of cases, failure in detection was seen with an excellent quality of specimen and poor sampling; a system-related error. In 6.3% of cases a cause of failure was seen with an excellent specimen quality and adequate sampling; a pathologist-related error. System-related error was due to poor sampling, processing, and lack of quality control methods. Pathologist-related error was due to cognitive error.
Conclusions: The causes of false negative bile duct brushing were multifactorial and included system and pathologist-related error. Poor specimen quality and poor sampling quality appeared to have the most significant role in false-negative bile-duct brushing. Pathologist-related error may benefit from consensus by a second pathologist as a meaningful quality improvement measure and warrants further investigation.
Category: Quality Assurance
Monday, February 28, 2011 1:00 PM
Poster Session II # 219, Monday Afternoon