Keywords:
Patterns of Care, Outcomes, Endoscopy, Ultrasound, Conventional radiography, CT, Gastrointestinal tract, Colon, Abdomen
Authors:
D. Tzias, C. Alexakis, P. Patel, S. Moodie; London/UK
DOI:
10.1594/ecr2013/C-0401
Methods and Materials
Our study was conducted in Epsom General Hospital,
a District General Hospital in Surrey,
UK.
Data was collected using the electronic patient records(EPR) and patient archive and communication system(PACS),
the DI-COM system in our hospital.
All patients referred for colonoscopy based on abnormal CT imaging,
over a twelve-month period,
were identified retrospectively.
Endoscopic findings were recorded.
The corresponding CT scan report (that lead to endoscopic referral) was reviewed for each patient,
noting the luminal pathologies.
Furthermore,
two assessors independently scored each CT report with a “confidence factor” (low/high confidence),
based on the degree of certainty expressed in the report for the specific luminal abnormality.
For example,
when an abnormality was described as "possibly" or "maybe" present,
the CT report was scored as "low" confidence; if described as "likely" then the report would score "high" confidence.
Fisher’s exact test was used to compare the percentage of agreement between radiological and colonoscopic reports for each of the 2 confidence groups.