Article Text

Download PDFPDF

Relative risk rather than absolute risk reduction should be preferred to sensitise the public to preventive actions
  1. Nicolas Chapelle1,2,
  2. Myriam Martel3,
  3. Alan N Barkun3,4,
  4. Marc Bardou5,6
  1. 1Institut des Maladies de l'appareil digestif, CHU Nantes Unité de gastroentérologie, Nantes, Pays de la Loire, France
  2. 2UMR1064, CRTI, Université de Nantes, Faculté de Médecine, Nantes, France
  3. 3Division of Gastroenterology, McGill University Health Centre, Montreal, Québec, Canada
  4. 4Department of Clinical Epidemiology, Biostatistics and Occupational health, McGill University, Montreal, Québec, Canada
  5. 5UMR INSERM 1231, Université de Bourgogne UFR des Sciences de Santé, Dijon, Bourgogne, France
  6. 6INSERM-Centre d'Investigations cliniques 1432 (CIC-1432), CHU Dijon-Bourgogne, Dijon, France
  1. Correspondence to Professor Marc Bardou, INSERM-Centre d’Investigations Cliniques 1432 (CIC 1432), CHU Dijon, Dijon BP 77908, Bourgogne, France; marc.bardou{at}

Statistics from

We thank Lawrence and colleagues1 for their interest in our work,2 about which they raised some comments as the need of expressing results in absolute rather than relative risks.

As they appropriately mentioned in their correspondence, absolute risk is an important parameter for the estimation of the effect of an intervention and must sometimes be preferred to relative risk.

However, when discussing with health professionals and policymakers, using absolute risk reductions, expressed as percentages, may incorrectly lead to an intervention being considered unnecessary. As example, what would be the point of reducing by 30% the occurrence of an event affecting 2% of the population? This is exactly what we were confronted to with the COVID-19 pandemic, when policymakers were criticised for putting in place measures to reduce individual freedoms, which were considered excessive in relation to the perception of risk by the public, for a disease whose overall case fatality is in the 2%–4% range3—exactly the same magnitude as that of colorectal cancer (CRC) incidence (2%) reported by Lawrence1 and colleagues, although the GLOBOCAN data were incorrectly cited.4 Indeed, 2% is the cumulative risk of developing a CRC in the first 74 years of life. However, it is estimated that about 30% of CRC occur in patients of age 75, and that the lifetime cumulative risk for CRC is approximately 4.1% in women and 4.4% in men.5

Therefore, when discussion of a risk, and its reduction, targets health professionals and policymakers, the most meaningful approach is to combine relative risk reduction and absolute number of cases avoided or of lives saved. In their initial submission, Lawrence1 and colleagues did not consider the absolute numbers that are of great concern when actually realising the number of lives which could be potentially saved each year, which reaches around 350 000 worldwide.6

In contradistinction, when interacting with the public at large, we strongly recommend the use of relative risk instead of absolute risk reduction. Indeed, this should be done in order to effectively and convincingly promote health interventions of proven, or strongly suggested, benefits, such as CRC screening.

The recent example of COVID-19 vaccination is illustrative. Preliminary results from mRNA COVID-19 vaccines suggested a relative risk reduction for confirmed COVID-19 cases of around 95% in the vaccinated compared with the placebo group, which has no doubt contributed to driving public adherence to vaccination.7 It may have been much less the case had the absolute risk reduction been discussed, which was around 1% (confirmed COVID-19 cases 1.21% and 0.07% in the non-vaccinated and vaccinated groups, respectively7).

Among all the 369 diseases tracked by the Global Burden of Diseases in 204 countries, CRC is the 15th leading cause in the population aged 50–74 years old and the 13th among patients over 75 worldwide.8 Furthermore, there is a widely reported increase in the incidence of early-onset CRC,9 unlikely to be prevented by existing screening programmes. A 25%–50% relative risk reduction of CRC corresponds to a striking decrease in thousands, even millions, of lives improved or spared,6 with great impact on quality of life and socioeconomic burden. Such a public health perspective helps transparency while convincingly laying out arguments for promoting interventions aimed at reducing the risk of CRC; this was the aim of our published meta-analysis and, we feel, justifies our use of relative risk reduction.



  • Twitter @mbardou

  • Contributors All authors contributed to the preparation and revision of this manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.