Sequential testing

Two consecutive diagnostic accuracy studies conducted in Scotland as part of the UK pilot screening study investigated whether testing individuals with positive gFOBT tests using an iFOBT could be more effective in selecting those who should receive colonoscopy (Fraser et al. 2006; Fraser et al. 2007) In both studies the two-tier approach gave very high sensitivities of 95–96% with a negative carrying a less than 1% chance of invasive cancer. The odds ratio for iFOBT positive subjects of having cancer was 7.75 (95% CI 1.84–31.4). 

A Chinese study (Li et al. 2006) of 324 subjects who had colonoscopy (mean age 53.5±15.3) showed that an iFOBT following a positive gFOBT had a better specificity for colon cancer detection than gFOBT (94.2% vs. 75.5%), and with similar sensitivity (93.8% and 95.9% vs. 95.9%, p>0.05). 

In a multicentre comparison using different FOBT tests on 554 patients referred for colonoscopy (mean age 59.8±11.7), a combination test with a highly sensitive gFOBT (Hemoccult SENSA) and an iFOBT (FlexSure-FS or Hemeselect-HS, Beckman Coulter Inc. Fullerton, CA, USA) showed slightly reduced sensitivity but significantly fewer false-positive tests than any single test (Greenberg et al. 2000). The specificity of SENSA/FS (95.7%, p=0.03) and SENSA/HS (95.2%, p=0.07) for the detection of colorectal cancer were each greater than that of any individual test. 

Participation rate and choice of test 

Factors that influence participation rate (uptake) are addressed in Chapter 2 (Sect. 2.4, 2.5.1.1 and 2.5.1.2). Whilst many studies have reported the effect on compliance of different test devices and sampling permutations, some of these are contradictory and many reflect local circumstances. Whilst the analytical methodology, gFOBT vs. iFOBT, will not directly influence compliance, the influence of test methodology on the method of sampling, the number of samples required, a requirement for dietary restriction and the improved clinical outcome will all have a bearing on uptake. The magnitude of the influence will depend on local circumstances. Well-conducted randomised trials have clearly demonstrated that better compliance can be achieved using current iFOBTs than with gFOBTs, but the major influencing factor(s) remain a matter of speculation. In his recent paper Grazzini makes the important observation that, in a biennial screening programme looking for a slow growing adenoma, greater compliance over the long term might be more important than a higher detection rate on a single screen (Grazzini et al. 2009).