Introduction The current BCSP pathway recommends radiological examination of the colon for people with a positive faecal occult blood test who are unable to undergo colonoscopy. The proportion of people undergoing radiological examination and polyp/cancer yield is unknown.
Methods All patients undergoing lower gastrointestinal investigation following a positive faecal occult blood test within the English national Bowel Cancer Screening Programme (BCSP) in the first 4 years of the programme (August 2006–July 2010) were identified. The number, percentage, demographics and co-morbidity (as defined by ASA grade) of people having CT colonography, barium enema, and plain abdominal CT as the first investigation were recorded and variability between centres was assessed. Use of radiology and yield of cancer and high risk polyps were also recorded, and compared to colonoscopy. Outliers were determined using Tukey limit methods.
Results Use of radiological tests as a first line investigation increased steadily with age from 0.99% in those aged <60 years to 6.04% in those aged >74 years. Radiological tests were used in more women than men (2.65% vs 2.35%, p<0.01). Radiological investigation increased with co-morbidity from 1.94% in people graded ASA 1 to 38.36% in ASA 4 (p<0.001). Cancer and high risk polyp detection rates for all first-line investigations are shown below. Detection rates for radiological tests were lower in this older, co-morbid sub-population than found for colonoscopy. There was considerable variation in the use of radiology between centres (0.3% to 9.1%), not related to age or co-morbidity. Two centres had a very low percentage of people having radiology tests and three very high.
Conclusion The number of people having radiology tests as an alternative to colonoscopy in the BCSP is highly variable across England but is associated with increasing age and co-morbidity. Cancer and high risk polyp detection rates appear lower in this sub-population compared to colonoscopy yield. Accuracy of radiology data input and examination of the factors contributing to these data requires further investigation and analysis.
Competing interests None declared.