Commentary: adherence and detection rates in colorectal cancer screening – author's reply


We read with interest the letter by Hilmi et al. suggesting the integration of a patient-orientated score able to stratify asymptomatic subjects according to the risk of advanced neoplasia in colorectal cancer (CRC) screening policies.[1, 2] In detail, the Asia-Pacific score is based on simple risk factors, such as age, sex, smoke and family history, and it is able to select a subgroup of asymptomatic ≥50 years subjects at higher risk of advanced neoplasia.[2]

This suggestion fits-in well-integrated with our observation of a higher efficacy of endoscopy over faecal tests in detecting advanced neoplasia in a primary screening setting.[3] It would appear reasonable, from a population perspective, to recommend a primary endoscopic approach to those ≥50 years subjects at higher risk of advanced neoplasia, such as those with a positive family history for CRC, whereas reserving the less effective, but also less costly, faecal tests to those at lower risk. Alternatively, in the case of limited economic/medical resources, CRC screening could be limited to those at higher risk, who may gain more benefit from the affordable investment.

However, there may be some uncertainty on the validity of such approach. First, despite the logistic regression score has been externally validated in Asian countries, it may be less effective when adopted in Western countries, requiring a country-specific validation prior to its application. Secondly, it would select a population at reduced prevalence of advanced neoplasia for faecal tests, potentially affecting their positive predictive values, when the same cut-off presently recommended would be maintained. As pointed out by the same authors,[1] only a comparative study between a score-based screening policy and a nonstratified approach may unveil the eventual superiority in terms of either efficacy or cost-effectiveness of the former option.


The authors' declarations of personal and financial interests are unchanged from those in the original article.3