Following a pilot phase in 23 metropolitan departments, organized colorectal cancer screening was rolled out nationwide starting in 2008. Since 2010, this program has covered all French departments, with the exception of Mayotte (which became a French department on March 31, 2011). Screening is coordinated at the local level by departmental or interdepartmental management structures (a total of 89 management structures). The participation data, as well as the exclusion rates and the rates of people with a positive test result, for the 2013–2014 periods presented here are therefore, for the most recent year, those derived from screening based on the guaiac test.

Materials and Methods

The specifications for cancer screening programs published in the Official Journal on December 21, 2006 (Appendix to No. 295) stipulate that individuals aged 50 to 74 must be invited to undergo colorectal cancer screening every two years. The indicators are thus calculated over a two-year period, during which the entire population of a department is invited to participate in screening.

Calculation of the participation rate

The participation rate is the ratio of the number of people screened to the target screening population (men and women aged 50 to 74), from which individuals excluded from screening for medical reasons are subtracted. It is defined as:

The number of people screened—that is, men and women aged 50 to 74 who underwent a screening test during the evaluation period (here, from January 1, 2013, to December 31, 2014)—and the number of people excluded from screening are provided by the administrative bodies via a standardized annual questionnaire.

The average eligible population is obtained by calculating the arithmetic mean of the INSEE target population of people aged 50–74 for the calendar years 2013 and 2014, from which the average number (arithmetic mean) of people aged 50–74 excluded from the program for medical reasons during the two calendar years (2013 and 2014).

At the national level and for the purpose of comparing departments, the denominators used to calculate participation rates are the population data provided by INSEE (2007–2042 projections, central scenario), calculated identically across the entire country. From this population, individuals who are temporarily or permanently excluded from screening for medical reasons are subtracted. The definition of medical exclusions is specified in the specifications (see above).

Calculation of the exclusion rate

The exclusion rate is the ratio of the number of people temporarily or permanently excluded from the organized colorectal cancer screening program (see above) to the INSEE target population for screening (men and women aged 50 to 74).

The number of excluded individuals is reported by the administrative units via a standardized annual questionnaire.

The average eligible population is obtained by calculating the arithmetic mean of the INSEE target population of people aged 50–74 for the calendar years 2013 and 2014, from which the average number (arithmetic mean) of people aged 50–74 excluded from the program for medical reasons during those two calendar years (2013 and 2014).

Calculation of the percentage of positive tests

This is the ratio of the number of people with a positive test result to the number of people with a test result that can be analyzed by the screening test reading centers.

This information is provided by the administrative units via a standardized annual questionnaire.

Age groups

A breakdown into 5-year age groups was used.

The 50–54 age group includes individuals who turned 50 during the year under review, even if they had not yet reached that age at the time of screening; the 70–74 age group includes individuals who had turned 75 at the time of screening, generally due to a delay in taking the test after receiving the invitation sent up to age 74.

Standardization of indicators

Participation rates, exclusion rates, and the rate of people with a positive test result are calculated by department, by region, and for France as a whole. They are standardized by age and sex relative to the 2009 French population (INSEE projection 2007–2042, central scenario). This standardization allows, by applying the same age and sex structure (that of the 2009 French population) to geographic units, for comparing results between these units (departments or regions) and for studying changes over time in the various indicators examined.

Benchmarks

The European benchmark recommends a participation rate of the target population of at least 45% for the screening program to be cost-effective. This benchmark applies to all European countries that have implemented an organized screening program.

Results

Participation

Over the period from January 1, 2013, to December 31, 2014, the population of people aged 50 to 74 is estimated at 18.4 million, of whom at least 16 million were invited to participate in the program. Nearly 5 million underwent a test, representing a participation rate of 29.8% (see table). This rate remains below the minimum acceptable European target of 45% participation and far behind the desirable rate of 65%^¹. Women participate in the program at a higher rate than men (31.5% versus 27.9%, respectively), regardless of the age group observed (see table). The participation rate increases with age among both men and women (see table). It rises among men from 23.8% for those aged 55–59 to 36.1% for those aged 70–74, and among women from 27.8% for those aged 50–54 to 37.1% for those aged 70–74.

As in previous years, participation rates vary by department (see table). Only the departments of Haut-Rhin and Saône-et-Loire have participation rates above the recommended minimum threshold (46.6% and 47.5%, respectively), and five departments (Yonne, Ardennes, Ille-et-Vilaine, Bas-Rhin, Loir-et-Cher, and Isère) have rates close to this threshold (40–42%). In total, 51 departments have a participation rate above 30%. By contrast, the rate is particularly low (<20%) in the departments of Corsica, Paris, French Guiana, Hauts-de-Seine, and Lozère. These differences can be partly explained this year by the varied implementation of the national directive of October 6, 2014, which urged organizations managing organized cancer screening programs to prepare for the discontinuation of gaïac test distribution starting in early November 2014. Therefore, spatiotemporal comparisons of participation rates for 2013–2014 must, even more so this year, be interpreted with great caution. Taking into account the time lags in sending invitations to those invited to participate in the program through survival analysis methods could be of significant interest in this context. However, this approach cannot be applied given the aggregated nature of the participation data reported for the evaluation.
Estimating the proportion of the eligible population² that was screened appears to be an alternative, and allows the estimated participation rate for a given period to be weighted by the coverage rate of the invitation program (the proportion of the eligible population invited during the same period). Over the 2013–2014 period, the proportion of the eligible population that underwent screening is estimated at 32.1% (relative difference: –8.3% compared to the 2011–2012 period). It varies considerably by region and department: see map and table, ranging from 4.3% in Corsica to 60.4% in Guadeloupe. Compared to the previous period, the proportion of the eligible population screened increased in 23 departments, ranging from +0.3% in Finistère to +88.2% in Guadeloupe, while it decreased in 73 departments.

Exclusion

Certain individuals are excluded from screening, either permanently for medical reasons (personal or family history of adenomas or colorectal cancer, familial adenomatous polyposis, chronic inflammatory bowel disease, hereditary non-polyposis colorectal cancer, or Lynch syndrome) and referred for more specific follow-up, or temporarily due to symptoms or because they underwent a colonoscopy within the past five years with normal results. These exclusions must be regularly documented and as comprehensive as possible to ensure the quality of the program. This exclusion rate is 12.6% (+4.2% compared to the 2011–2012 period), representing just over 2 million people excluded. It is slightly lower among men (12.4%) than among women (12.8%). This rate increases significantly with age, from 6.8% among men aged 50–54 to 20.2% among men aged 70–74, and from 7.7% to 18.5% among women in the same age groups, respectively. There are significant regional disparities in exclusion rates, ranging from 4.7% in Corsica and French Guiana (+23.7% and +42.7%, respectively, compared to 2011–2012) to 19.4% (+28.5% compared to 2011–2012) for Bas-Rhin: see map and table. Exclusions from the program are more frequent in departments that have participated in the program for a longer period (14.1% for pilot departments that joined the program before 2008, compared to 12% for non-pilot departments), since a higher proportion of individuals with previous cancerous or precancerous lesions or who had a normal colonoscopy within the past 5 years were identified during screenings. However, these departmental disparities in the medical exclusion rate also highlight the heterogeneity in the quality of data collection across departments. The number of excluded individuals remains poorly documented in some departments, particularly due to the poor performance of their systems for collecting data on medical exclusions. At the departmental level, there is a variety of usable data sources to document medical exclusions, but these are not always accessible to administrative bodies. This situation highlights the urgent need to optimize the collection of information on medical exclusions to ensure the quality of the program, by defining a national strategy to better guide and support the professionals involved.

Positive Tests

Nearly 5 million tests were conducted, of which just over 103,700 were positive, with a positivity rate in line with expectations (2.2%) according to European recommendations. This positivity rate is higher among men (2.5%) than among women (1.9%), and increases with age (see table). The positive test rate has decreased from 2.4% to 2.2%, representing an 8.3% decrease compared to the previous period (2011–2012). This decrease is consistent with the repeated screening campaigns, which make it possible to identify and exclude individuals with precancerous or cancerous colorectal lesions from the target screening population, and to refer them for more specific follow-up.
There is significant variability in the positive test rate across departments, with values ranging from 0.7% in Martinique to 5.7% in Hérault (see map and table). The positive test rate also remains high in French Guiana (3.9%), Calvados (4.2%), and Var (4.5%). It increased between the two periods in 23 departments, particularly sharply in Calvados (from 3.3% to 4.2%), Var (from 3.5% to 4.5%), and Hérault (from 3.3% to 5.7%). The rate of positive tests, however, decreased during the same period in just over two-thirds (69) of the departments. This particularly sharp decline in the departments of Martinique (-41.7%), Vaucluse (-44%), and Côte-d’Or (-48.8%), is consistent with the repeated screening campaigns that identify and exclude individuals with precancerous or cancerous colorectal lesions from the target population of organized screening, and refer them to more specific follow-up by their physician.

Conclusion

More than five years after the nationwide rollout of the organized colorectal cancer screening program, participation remains uneven across departments and still falls far short of European standards for organized colorectal cancer screening. However, the effectiveness of a screening program depends heavily on the uptake by the target population. Addressing inequalities in access to and use of screening programs (Objective 1 of the 2014–2019 Cancer Plan) requires making more effective, easier-to-use tests available nationwide to increase acceptance among the target population. The improved performance of this test is also likely to increase the confidence and involvement of general practitioners, which is known to be a key factor in encouraging patient participation in the program. With this in mind, the Directorate General for Health, in its decree of September 23, 2014, approved the replacement of the Hemoccult® II test with an immunological test (OC-Sensor®), which is more sensitive for detecting advanced adenomas and cancers. Its automated reading provides greater reliability and reproducibility. The transition to the immunological test is scheduled for April 2015.