Home > DARE Reviews > Screening for colorectal cancer: a...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet]. York (UK): Centre for Reviews and Dissemination (UK); 1995-.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet].

Screening for colorectal cancer: a targeted, updated systematic review for the U.S. Preventative Services Task Force

EP Whitlock, JS Lin, E Liles, TL Beil, and R Fu.

Review published: 2008.

Link to full article: [Journal publisher]

CRD summary

This review evaluated the screening accuracy and harms of colonoscopy and flexible sigmoidoscopy for colorectal cancer in average-risk populations. The authors concluded that colonoscopy was sensitive in detecting larger lesions (over 10 mm), but had a higher risk for complications than flexible sigmoidoscopy. The conclusions are probably reliable, but under reporting of study quality may warrant a cautious interpretation.

Authors' objectives

To evaluate the effectiveness, diagnostic accuracy, and harms of colonoscopy and flexible sigmoidoscopy as screening methods for colorectal cancer in the community setting.


PubMed, DARE, Cochrane Database of Systematic Reviews, National Guideline Clearing House, Institute of Medicine, NICE and HTA (Health Technology Assessment) databases were searched for relevant studies. Search dates spanned from January 2000 to January 2008. The bibliographies of relevant articles and reviews were screened, and experts were contacted to locate additional studies.

Study selection

Eligible for inclusion in the review were randomised controlled trials (RCTs), clinical controlled trials (CCTs), observational studies, and screening registries containing community-based participants aged 40 years or older. The interventions of interest were flexible sigmoidoscopy and colonoscopy. The included interventions were single application colonoscopy conducted after various computed tomography (CT) technologies (where segmental un-blinding was used to recheck CT colonography-located lesions missed on the first-pass colonoscopy) and simulated flexible sigmoidoscopy.

The effectiveness outcomes of interest were mortality (all cause, or colorectal cancer specific), although no studies were found. Safety outcomes of interest were those requiring hospitalisation and/or leading to death, for example: perforation, major bleeding, severe abdominal symptoms, and cardiovascular events.

Diagnostic studies were eligible if they included colonoscopy as the reference standard, and reported sensitivity and specificity per-person, supplemented by per-polyp analyses (miss rates), advanced neoplasia, or adenomas by size. Case control studies; those that did not address average risk populations, or failed to report sub-group analysis in this respect; and those with an incompletely applied or inadequate reference standard were excluded.

Two reviewers selected studies for inclusion in the review.

Assessment of study quality

Two reviewers assessed the quality of included studies, using an adapted version of the United States Preventive Services Task Force criteria for effectiveness studies. The QUADAS (Quality Assessment of Diagnostic Accuracy Studies Assessment) tool was used for diagnostic accuracy studies. Poor quality studies were excluded.

Data extraction

Data for screening test performance were extracted on per-person sensitivity and specificity supplemented by per-polyp analyses (miss rates), with 95% confidence intervals (CI), where available. Data on complication rates were also extracted.

Data extraction was carried out by one reviewer and verified by another reviewer. Disagreements were resolved by discussion.

Methods of synthesis

A narrative synthesis was provided in the presence of heterogeneity. Statistical heterogeneity was assessed using the Q and I2 statistics. Where there was no heterogeneity (e.g. in the analysis of harms), a fixed-effect meta-analysis was applied to estimate the pooled complication rate. A random-effects model was also used in the exploration of heterogeneity. Sensitivity analyses were performed to examine the influence of specific study and patient characteristics.

Results of the review

Nine studies were included in the analysis. The quality of studies was reported to be fair.

Accuracy and harms of colonoscopy (three out of 17 studies): In the first study (n=1,233 average-risk patients) there was no statistically significant difference between the sensitivity of colonoscopy (per-person detection rate of adenomas 6, 8 or 10 mm or larger) and that of CT colonography. In the second study (n=96 average-risk patients), colonoscopy missed 10% of polyps 6 mm or larger, but did not miss any polyps measuring 10 mm or larger. In the third study (n=452 asymptomatic, average-risk patients), 77% of neoplasms measuring 10 mm or larger were detected, and one out of five colorectal lesions were found that were previously detected by CT colography. Pooled data from 12 studies (n=57,742 patients) found 2.8 total serious complications (perforations, haemorrhage, diverticulitis, cardiovascular events, severe abdominal pain and death) per 1,000 procedures (95% CI: 1.5 to 5.2). There was no statistically significant heterogeneity. When the analysis was limited to studies conducted in the United States (seven studies), the rate of complications was reduced to 2.5 per 1,000 procedures (95% CI: 1.0 to 6.1).

Accuracy and harms of simulated flexible sigmoidoscopy (six out of eight studies): Simulated sigmoidoscopy was used in six large cohort studies (n=14,938 average-risk patients). The pooled sensitivity (using biopsy or visual inspection to determine colonoscopy referral) was 58 to 75% (based on a small number of lesions). Estimated sensitivity for advanced neoplasia was 72 to 86%. Pooled data from six studies (n=126,985 patients) found 0.34 serious complications per 1,000 procedures (95% CI: 0.06 to 1.9). There was no statistically significant heterogeneity. Polypectomy was conducted during the majority of flexible sigmoidoscopy examinations.

It was not possible to estimate the influence of age on complication rates in either analysis.

Authors' conclusions

Colonoscopy was sensitive in detecting colorectal cancer lesions 10 mm or larger, and may be more sensitive than CT colography in detecting for smaller adenomas. Colonoscopy was associated with higher risks for immediate complications than flexible sigmoidoscopy.

CRD commentary

The review question was clear and supported by detailed inclusion criteria which are potentially reproducible. The search strategy included some relevant sources for published studies, but there was no apparent attempt to locate unpublished material. Publication bias was not assessed. Appropriate validity assessment tools were used to assess the quality of effectiveness and diagnostic studies. However, the results of this were not given in detail, making it difficult to verify the reported global assessment. The reported review process demonstrated attempts to minimise errors and bias. Heterogeneity was taken into account in the proposed methods of synthesis. The authors' conclusions reflected the results from a small number of included studies. The conclusions are probably reliable, but under reporting in relation to study quality may warrant a cautious interpretation.

Implications of the review for practice and research

Practice: The authors stated that emphasis should be placed on quality standards for the implementation of any operator-dependent colorectal cancer screening test.

Research: The authors stated that frequent reconsideration of available evidence and updating of recommendations is warranted.


The Agency for Healthcare Research and Quality, contract HHSA-290-2007-10057-I-EPC3, task order 3.

Bibliographic details

Whitlock E P, Lin J S, Liles E, Beil T L, Fu R. Screening for colorectal cancer: a targeted, updated systematic review for the U.S. Preventative Services Task Force Annals of Internal Medicine 2008; 149(9): 638-658. [PubMed: 18838718]

Other publications of related interest

Whitlock EP, Lin J, Liles E, et al. Screening for colorectal cancer: an updated systematic review. Rockville, MD, USA. Agency for Healthcare Research and Quality, Evidence Synthesis; No. 65, Part 1. 2008.

Zauber AG, Landsorp-Vogelaar I, Knudson AB, et al. Evaluating test strategies for colorectal cancer screening - age to begin, age to stop, and timing of screening intervals: a decision analysis of colorectal cancer screening for the U.S. preventive services task force from the cancer intervention and surveillance modelling network (CISNET). Rockville, MD, USA: Agency for Healthcare Research and Quality. Evidence Synthesis; Number 65, Part 2. 2009.

Indexing Status

Subject indexing assigned by NLM


Age Factors; Aged; Colorectal Neoplasms /diagnosis /prevention & control; Decision Support Techniques; Early Diagnosis; Female; Humans; Male; Mass Screening /adverse effects /methods /standards; Middle Aged; Time Factors



Database entry date


Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.

CRD has determined that this article meets the DARE scientific quality criteria for a systematic review.

Copyright © 2014 University of York.

PMID: 18838718