Rapid review
Screening for colorectal cancer: effect on health outcomes
Mapping review
|Updated
The Norwegian Directorate for Health commissioned a summary of the current evidence base regarding health effects of this type of screening.
Download
Key message
There are ongoing discussions about whether to implement a national screening program for colorectal cancer in Norway.
On request from the Norwegian Directorate for Health, The Knowledge Centre at the Norwegian Institute of Public Health prepared a summary of the current evidence base regarding health effects of this type of screening.
We found that:
- A screening program with sigmoidoscopy leads to fewer cases of colorectal cancer, and that fewer die from colorectal cancer
- A screening program using faecal samples (gFOBT) leads to fewer deaths from colorectal cancer, but is unlikely to reduce the incidence of the disease
- The effects of a screening program with colonoscopy is uncertain due to lack of research evidence
- None of the screening methods have been shown to reduce the total number of deaths (for all causes of death seen together)
- The proportion of invited individuals who took part in the screening program varied between 50 and 75%, across the various studies
- A substantial, but variable proportion of those who take part in the screening program are referred to follow-up colonoscopy
- All screening methods er associated with a small, but real risk of serious complications, mainly in connection with follow-up colonoscopy
- There is limited evidence concerning negative consequences of screening
- We cannot state with certainty which screening methods are the most effective
Summary
Background
Colorectal cancer is among the most common cancer types in Norway. Most cases arise from benign precursory stages (adenomas) in the colon, and the prognosis is better with non-metastatic disease. Thus, it may be useful to both detect and remove adenomas, and to detect colorectal cancer early.
Several methods are in use as screening tools for colorectal cancer. This research summary is limited to the most commonly used methods:
- Colonoscopy (flexible tube with camera used to investigate the full colon)
- Sigmoidoscopy (flexible tube with camera to investigate the lower half of the colon)
- Faecal occult blood test (either guiac-based, gFOBT, or with immunological detection of human blood, iFOBT)
The Norwegian National Board for Quality and Prioritisation in Health Care discussed the introduction of a national screening program for colorectal cancer over several meetings in 2010. As recommended by the Board, a randomised trial comparing two screening methods, iFOBT and sigmoidiscopy, was initiated in 2012.
Objective
In this report, we present a summary of the research evidence on the effects of screening for colorectal cancer, based on existing systematic reviews.
Method
Initially we conducted a simple search for relevant systematic reviews, and found a recently published one (March 2016). In order to identify other systematic reviews and recent randomised trials, we conducted a systematic search in several databases for the period October 2015 to June 2016.
After we had finalised our literature search and most of our analyses, a research group from the USA published a systematic review, which we also took into consideration.
We conducted our own analyses to estimate expected health benefits from screening. We used meta-analyses from a 2013 Cochrane review as our starting point. We generated updated effect estimates by adding data from more recently published studies.
Regarding negative consequences and compliance with screening programmes, we report findings from three systematic reviews, supplemented with preliminary findings from ongoing Norwegian studies.
Results
Convincing evidence from several large randomised trials show that screening programs, either with gFOBT or sigmoidoscopy, lead to fewer deaths due to colorectal cancer –14% (95% CI 6 to 22%) and 27% (95 % CI 20 to 34%) relative risk reduction, correspondingly. In addition, sigmoidoscopy reduces the number of new cases of colorectal cancer by 22% (95% CI 18 to 26%), while no such effect has been shown for gFOBT (2% relative risk reduction; 95% CI from 10% reduction to 6% increase). Due to the uncertainty around the effect estimates, we are not able to conclude that one method is better than the other with regards to preventing deaths from colorectal cancer.
We did not identify evidence about health benefits from screening with iFOBT or colonoscopy.
Whether screening for colorectal cancer influences total mortality is unclear. This may be due to insufficient data to detect an effect on all-cause mortality, since deaths due to colorectal cancer constitute a relatively small proportion of all deaths. Alternatively, it may be due to negative effects from screening that counterbalance the reduction in deaths from colorectal cancer. We compared the incidence of deaths from other causes than colorectal cancer, and did not find any indication that screening leads to increased risk of death from other causes.
Screening is associated with a small, but real risk of serious complications, usually in relation to removal of premalignant adenomas during colonoscopy. The evidence is insufficient to conclude regarding other possible negative consequences, e.g. psychological effects.
Discussion
It is not straightforward to predict the health consequences from introducing a screening program in Norway. In 2014, there were 567 deaths from tumours in the colon, rectum and anus in the 55 to 74 years age group. Based on this figure, and applying the effect estimates we have arrived at, screening with sigmoidoscopy or gFOBT will lead to 139 (95% CI 113 to 193) and 79 (95% CI 34 to125) fewer deaths in this age group, per year. If we assume that 70,000 individuals are invited to take part in the national screening program per year, that 55% participate, and that 10% of these are referred after the primary screening, the number of follow up colonoscopies will be 3,850.
There is some uncertainty regarding the extent of negative consequences from participating in this type of screening program.
We cannot state with certainty that one screening method prevents more deaths from colorectal cancer than another.
Conclusion
There is convincing evidence that a screening program will lead to fewer deaths due to colorectal cancer. Screening by means of sigmoidoscopy will also reduce the incidence of the disease. Serious complications due to screening occur, but are rare.