Focal cystic pancreatic lesions: assessing variation in radiologists' management recommendations

Radiology. 2011 Apr;259(1):136-41. doi: 10.1148/radiol.10100970. Epub 2011 Feb 3.


Purpose: To estimate the prevalence of focal cystic pancreatic lesions (FCPLs) among patients undergoing computed tomographic (CT) or magnetic resonance (MR) imaging at one institution and to examine any variation in radiologists' recommendation practice pattern with regards to FCPLs.

Materials and methods: Institutional review board approval was obtained for this retrospective HIPPA-compliant study. The requirement to obtain informed consent was waived. A cohort of patients with FCPLs was identified from radiology reports by using natural language processing. Patient-specific (ie, age, sex, symptoms, history of pancreatitis), radiologist-specific (ie, years of experience, area of expertise), and FCPL-specific (ie, size, location, septation, calcification, mural nodularity, pancreatic duct involvement, and presence of multiple cysts) variables were obtained. The outcome measure was whether a follow-up study was recommended. A logistic regression model was used to identify relative recommendation rates after controlling for key explanatory variables.

Results: Between January 1 and December 31, 2009, a total of 1067 FCPLs were identified in 765 patients. Prevalence rates ranged from 2.2% at CT to 15.9% at MR imaging. Radiologists recommended a follow-up imaging study in 23.7% of cases of a FCPL. A 2.8-fold difference in the rate of recommendation of further imaging existed across radiologists after controlling for explanatory variables such as lesion-, radiologist-, and patient-specific characteristics. A history of pancreatitis was associated with a nearly two-fold decrease in recommending further imaging.

Conclusion: FCPLs are common, and nearly one-quarter of radiology reports recommend at least one follow-up imaging study. Significant variation exists in the rate of recommendation for further imaging studies by radiologists, even after controlling for key explanatory variables.

MeSH terms

  • Female
  • Humans
  • Magnetic Resonance Imaging / statistics & numerical data*
  • Male
  • Massachusetts / epidemiology
  • Middle Aged
  • Observer Variation
  • Pancreatic Cyst / diagnosis*
  • Pancreatic Cyst / epidemiology*
  • Practice Patterns, Physicians' / statistics & numerical data*
  • Prevalence
  • Referral and Consultation / statistics & numerical data*
  • Reproducibility of Results
  • Risk Assessment
  • Risk Factors
  • Sensitivity and Specificity
  • Tomography, X-Ray Computed / statistics & numerical data*