Elsevier

The Lancet

Volume 348, Issue 9040, 30 November 1996, Pages 1472-1477
The Lancet

Articles
Randomised controlled trial of faecal-occult-blood screening for colorectal cancer

https://doi.org/10.1016/S0140-6736(96)03386-7Get rights and content

Summary

Background

There is growing evidence that faecal-occult-blood (FOB) screening may reduce colorectal cancer (CRC) mortality, but this reduction in CRC mortality has not been shown in an unselected population-based randomised controlled trial. The aim of this study was to assess the effect of FOB screening on CRC mortality in such a setting.

Methods

Between February, 1981, and January, 1991, 152 850 people aged 45–74 years who lived in the Nottingham area of the UK were recruited to our study. Participants were randomly allocated FOB screening (76 466) or no screening (controls; 76 384). Controls were not told about the study and received no intervention. Screening-group participants were sent a Haemoccult FOB test kit with instructions from their family doctor. FOB tests were not rehydrated and dietary restrictions were imposed only for retesting borderline results. Individuals with negative FOB tests at the first screening, together with those who tested positive but in whom no neoplasia was found on colonoscopy, were invited to take part in further screening every 2 years. Screening was stopped in February, 1995, by which time screening-group participants had been offered FOB tests between three and six times. Screening-group participants who had a positive test were offered full colonoscopy. All participants were followed up until June, 1995. The primary outcome measure was CRC mortality.

Findings

Of the 152 850 individuals recruited to the study, 2599 could not be traced or had emigrated and were excluded from the analysis. Thus, there were 75 253 participants in the screening group and 74 998 controls. 44 838 (59·6%) screening-group participants completed at least one screening. 28 720 (38·2%) of these individuals completed all the FOB tests they were offered and 16118 (21·4%) completed at least one screening but not all the tests they were offered. 30 415 (40·4%) did not complete any test. Of 893 cancers (20% stage A) diagnosed in screening-group participants (CRC incidence of 1·49 per 1000 person-years), 236 (26·4%) were detected by FOB screening, 249 (27·9%) presented after a negative FOB test or investigation, and 400 (44·8%) presented in non-responders. The incidence of cancer in the control group (856 cases, 11% stage A) was 1·44 per 1000 person-years. Median follow-up was 7·8 years (range 4·5–14·5). 360 people died from CRC in the screening group compared with 420 in the control group–a 15% reduction in cumulative CRC mortality in the screening group (odds ratio=0–85 [95% CI 0·74–0·98], p=0026).

Interpretation

Our findings together with evidence from other trials suggest that consideration should be given to a national programme of FOB screening to reduce CRC mortality in the general population.

Introduction

Colorectal cancer (CRC) is the second commonest cause of death from malignant disease in England and Wales, and resulted in about 16 000 deaths in 1993.1 Although there have been advances in the management of symptomatic CRC, there has been little overall reduction in CRC mortality during the past 30 years. Tumour stage is an important determinant of outcome; 24–28% of patients have metastatic disease at presentation and the tumour is confined to the bowel wall in only 6–10% (Dukes' stage A).2, 3, 4 Early diagnosis before the development of symptoms may be an effective way of reducing CRC mortality.

Tumours diagnosed as a result of screening by faecal-occult-blood (FOB) testing are known to include a higher proportion at a less advanced stage than those presenting symptomatically.5, 6 FOB tests are also cheap,7 safe, and acceptable to the population.8 Three case-control studies have shown that FOB screening led to a reduction in the risk of death from CRC,9, 10, 11 but, because of the self-selection bias, these findings must be viewed with caution. The efficacy of screening by FOB tests should be tested by comparison of disease-specific mortality among individuals who are offered screening with unscreened controls in the setting of a randomised controlled trial. Three European randomised controlled trials of population screening for CRC by FOB tests have confirmed that the test has a high rate of compliance and that CRC can be detected at an earlier stage.6, 12, 13 The Minnesota study14 of FOB screening in a volunteer population, reported a significant reduction in disease-specific mortality after annual screening by FOB tests, with a non-significant reduction in the group offered biennial screening. However, in studies of health-conscious volunteers, control-group mortality tends to be lower and compliance higher than in the general population,15 so their findings may not give a realistic estimate of the effect in an unselected population.

In this randomised controlled trial we assessed the effect of biennial screening by FOB tests on CRC mortality in the general population of the Nottingham area of the UK. We report here the initial CRC mortality data.

Section snippets

Methods

We recruited individuals for the pilot study between February, 1981, and June, 1983,16 and for the main study between February, 1985, and January, 1991. Individuals who lived in the Nottingham area of the UK were identified according to the general practice at which they were registered. Family Health Service Authority and general practice registers were used to compile a list of men and women aged 50–74 (45–74 years in the pilot study) in each general practice. Family doctors at each practice

Results

The trial profile shows participant numbers throughout the study (figure 1). Of the 152 850 individuals recruited into the study and randomised to screening (76 466) and control groups (76 384), 2599 (1·7%) could not be traced by the Office of Population Censuses and Surveys or had emigrated and were, therefore, excluded from the analysis. Thus, of the remaining 150 251 study participants, 75 253 were in the screening group and 74 998 were unscreened controls. The groups were well matched in

Discussion

It is encouraging that almost 60% of screening-group participants were screened at least once because such compliance was achieved without population education or a statement of definite benefit in the invitation for screening. Compliance would probably improve if a well-organised national screening programme for CRC was introduced in the UK.24, 25

For CRC screening by FOB test to be effective in the general population, a balance between sensitivity and specificity must be achieved. We were not

References (30)

  • JV Selby et al.

    Effect of fecal occult blood testing on mortality from colorectal cancer: a case control study

    Ann Intern Med

    (1993)
  • H Saito et al.

    Reduction in risk of mortality from colorectal cancer by fecal occult blood screening with immunochemical hemagglutination test. A case-control study

    Int J Cancer

    (1995)
  • J Wahrendorf et al.

    Effectiveness of colorectal cancer screening: results from a population-based case-control evaluation in Saarland, Germany

    Eur J CancerPrev

    (1993)
  • O Kronborg et al.

    Initial mass screening for colorectal cancer with fecal occult blood test

    Scand JGastroenterol

    (1987)
  • J Kewenter et al.

    Screening and rescreening for colorectal cancer in 27 700 subjects

    Cancer

    (1988)
  • Cited by (2498)

    • Principles of cancer screening

      2024, Surgery (United Kingdom)
    View all citing articles on Scopus
    View full text