[Skip to Content]
[Skip to Content Landing]
Figure.
Participant Flow for Enrollment and Allocation to the Study Groups
Participant Flow for Enrollment and Allocation to the Study Groups

aThe numbers of general practitioners (GPs) excluded for each specific reason are not available.

bMedical reasons for exclusion were personal history (colorectal cancer [CRC] or adenoma larger than 1 cm, mutation of the HNPCC gene, inflammatory bowel diseases), 14.4%; family history (CRC or adenoma larger than 1 cm), 17.4%; recent colonoscopy performed in the context of symptoms, 64.1%; and other reasons, 4.1%. The numbers of patients excluded for each specific reason in each group are not available.

cAdministrative reasons for exclusion were moved to another geographical area, 84.6%; deceased, 3.7%; and other reasons, 11.7%. The numbers of patients excluded for each specific reason in each group are not available.

Table 1.  
Demographic Characteristics of the General Practitioners in Each Group
Demographic Characteristics of the General Practitioners in Each Group
Table 2.  
Demographic Characteristics of the Patients in Each Group
Demographic Characteristics of the Patients in Each Group
Table 3.  
Patient Participation in Fecal Immunochemical Test Cancer Screening (Primary End Point)
Patient Participation in Fecal Immunochemical Test Cancer Screening (Primary End Point)
Table 4.  
Patient Characteristics Related to a Higher Participation in Fecal Immunochemical Test Colorectal Cancer Screening (Secondary End Point)a
Patient Characteristics Related to a Higher Participation in Fecal Immunochemical Test Colorectal Cancer Screening (Secondary End Point)a
Table 5.  
General Practitioner Characteristics Related to a Higher Patient Participation in Fecal Immunochemical Test Colorectal Cancer Screening (Post Hoc Analysis)a
General Practitioner Characteristics Related to a Higher Patient Participation in Fecal Immunochemical Test Colorectal Cancer Screening (Post Hoc Analysis)a
1.
GLOBOCAN 2012: estimated cancer incidence, mortality and prevalence worldwide in 2012. World Health Organization website. http://globocan.iarc.fr. Accessed July 23, 2017.
2.
Howlader  N, Noone  AM, Krapcho  M,  et al. SEER cancer statistics review (CRS) 1975-2014. National Cancer Institute website. https://seer.cancer.gov/csr/1975_2014. Posted April 2017. Accessed July 23, 2017.
3.
Holme  Ø, Bretthauer  M, Fretheim  A, Odgaard-Jensen  J, Hoff  G.  Flexible sigmoidoscopy versus faecal occult blood testing for colorectal cancer screening in asymptomatic individuals.  Cochrane Database Syst Rev. 2013;9(9):CD009259.PubMedGoogle Scholar
4.
Lieberman  D, Ladabaum  U, Cruz-Correa  M,  et al.  Screening for colorectal cancer and evolving issues for physicians and patients: a review.  JAMA. 2016;316(20):2135-2145.PubMedGoogle ScholarCrossref
5.
van der Vlugt  M, Grobbee  EJ, Bossuyt  PM,  et al.  Adherence to colorectal cancer screening: four rounds of faecal immunochemical test-based screening.  Br J Cancer. 2017;116(1):44-49.PubMedGoogle ScholarCrossref
6.
von Karsa  L, Patnick  J, Segnan  N,  et al; European Colorectal Cancer Screening Guidelines Working Group.  European guidelines for quality assurance in colorectal cancer screening and diagnosis: overview and introduction to the full supplement publication.  Endoscopy. 2013;45(1):51-59.PubMedGoogle Scholar
7.
Hol  L, van Leerdam  ME, van Ballegooijen  M,  et al.  Screening for colorectal cancer: randomised trial comparing guaiac-based and immunochemical faecal occult blood testing and flexible sigmoidoscopy.  Gut. 2010;59(1):62-68.PubMedGoogle ScholarCrossref
8.
Senore  C, Inadomi  J, Segnan  N, Bellisario  C, Hassan  C.  Optimising colorectal cancer screening acceptance: a review.  Gut. 2015;64(7):1158-1177.PubMedGoogle ScholarCrossref
9.
Rat  C, Latour  C, Rousseau  R,  et al.  Interventions to increase uptake of faecal tests for colorectal cancer screening: a systematic review.  Eur J Cancer Prev. doi:10.1097/CEJ.0000000000000344PubMedGoogle Scholar
10.
Sabatino  SA, Lawrence  B, Elder  R,  et al; Community Preventive Services Task Force.  Effectiveness of interventions to increase screening for breast, cervical, and colorectal cancers: nine updated systematic reviews for the guide to community preventive services.  Am J Prev Med. 2012;43(1):97-118.PubMedGoogle ScholarCrossref
11.
Martos  C, Crocetti  E, Visser  O, Rous  B; Cancer Data Quality Checks Working Group. A proposal on cancer data quality checks: one common procedure for European cancer registries. Ispra, Italy: European Network of Cancer Registries; 2014.http://www.encr.eu/images/download/A_proposal_on_cancer_data_quality_checks-one_common_procedure_for_European_cancer_registries.pdf. Accessed July 23, 2017.
12.
Zanetti  R, Schmidtmann  I, Sacchetto  L,  et al.  Completeness and timeliness: cancer registries could/should improve their performance.  Eur J Cancer. 2015;51(9):1091-1098.PubMedGoogle ScholarCrossref
14.
Cole  SR, Young  GP, Esterman  A, Cadd  B, Morcom  J.  A randomised trial of the impact of new faecal haemoglobin test technologies on population participation in screening for colorectal cancer.  J Med Screen. 2003;10(3):117-122.PubMedGoogle ScholarCrossref
15.
Federici  A, Giorgi Rossi  P, Borgia  P, Bartolozzi  F, Farchi  S, Gausticchi  G.  The immunochemical faecal occult blood test leads to higher compliance than the guaiac for colorectal cancer screening programmes: a cluster randomized controlled trial.  J Med Screen. 2005;12(2):83-88.PubMedGoogle ScholarCrossref
16.
Hughes  K, Leggett  B, Del Mar  C,  et al.  Guaiac versus immunochemical tests: faecal occult blood test screening for colorectal cancer in a rural community.  Aust N Z J Public Health. 2005;29(4):358-364.PubMedGoogle ScholarCrossref
17.
Birkenfeld  S, Belfer  RG, Chared  M,  et al.  Factors affecting compliance in faecal occult blood testing: a cluster randomized study of the faecal immunochemical test versus the guaiac faecal occult test.  J Med Screen. 2011;18(3):135-141.PubMedGoogle ScholarCrossref
18.
Levi  Z, Birkenfeld  S, Vilkin  A,  et al.  A higher detection rate for colorectal cancer and advanced adenomatous polyp for screening with immunochemical fecal occult blood test than guaiac fecal occult blood test, despite lower compliance rate: a prospective, controlled, feasibility study.  Int J Cancer. 2011;128(10):2415-2424.PubMedGoogle ScholarCrossref
19.
Cole  SR, Smith  A, Wilson  C, Turnbull  D, Esterman  A, Young  GP.  An advance notification letter increases participation in colorectal cancer screening.  J Med Screen. 2007;14(2):73-75.PubMedGoogle ScholarCrossref
20.
Mant  D, Fuller  A, Northover  J,  et al.  Patient compliance with colorectal cancer screening in general practice.  Br J Gen Pract. 1992;42(354):18-20.PubMedGoogle Scholar
21.
Giorgi Rossi  P, Grazzini  G, Anti  M,  et al.  Direct mailing of faecal occult blood tests for colorectal cancer screening: a randomized population study from Central Italy.  J Med Screen. 2011;18(3):121-127.PubMedGoogle ScholarCrossref
22.
Green  BB, Wang  CY, Anderson  ML,  et al.  An automated intervention with stepped increases in support to increase uptake of colorectal cancer screening: a randomized trial.  Ann Intern Med. 2013;158(5 pt 1):301-311.PubMedGoogle ScholarCrossref
23.
Lee  JK, Reis  V, Liu  S,  et al.  Improving fecal occult blood testing compliance using a mailed educational reminder.  J Gen Intern Med. 2009;24(11):1192-1197.PubMedGoogle ScholarCrossref
24.
Aubin-Auger  I, Laouénan  C, Le Bel  J,  et al.  Efficacy of communication skills training on colorectal cancer screening by GPs: a cluster randomised controlled trial.  Eur J Cancer Care (Engl). 2016;25(1)18-16.PubMedGoogle ScholarCrossref
25.
Hewitson  P, Ward  AM, Heneghan  C, Halloran  SP, Mant  D.  Primary care endorsement letter and a patient leaflet to improve participation in colorectal cancer screening: results of a factorial randomised trial.  Br J Cancer. 2011;105(4):475-480.PubMedGoogle ScholarCrossref
26.
Brotons  C, Björkelund  C, Bulc  M,  et al; EUROPREV Network.  Prevention and health promotion in clinical practice: the views of general practitioners in Europe.  Prev Med. 2005;40(5):595-601.PubMedGoogle ScholarCrossref
27.
Aubin-Auger  I, Mercier  A, Lebeau  JP, Baumann  L, Peremans  L, Van Royen  P.  Obstacles to colorectal screening in general practice: a qualitative study of GPs and patients.  Fam Pract. 2011;28(6):670-676.PubMedGoogle ScholarCrossref
28.
Goebel  LJ.  A peer review feedback method of promoting compliance with preventive care guidelines in a resident ambulatory care clinic.  Jt Comm J Qual Improv. 1997;23(4):196-202.PubMedGoogle Scholar
29.
Hudson  SV, Ohman-Strickland  P, Cunningham  R, Ferrante  JM, Hahn  K, Crabtree  BF.  The effects of teamwork and system support on colorectal cancer screening in primary care practices.  Cancer Detect Prev. 2007;31(5):417-423.PubMedGoogle ScholarCrossref
30.
Vinker  S, Nakar  S, Rosenberg  E, Kitai  E.  The role of family physicians in increasing annual fecal occult blood test screening coverage: a prospective intervention study.  Isr Med Assoc J. 2002;4(6):424-425.PubMedGoogle Scholar
31.
Sequist  TD, Zaslavsky  AM, Marshall  R, Fletcher  RH, Ayanian  JZ.  Patient and physician reminders to promote colorectal cancer screening: a randomized controlled trial.  Arch Intern Med. 2009;169(4):364-371.PubMedGoogle ScholarCrossref
32.
Piette  C, Durand  G, Bretagne  JF, Faivre  J.  Additional mailing phase for FIT after a medical offer phase: the best way to improve compliance with colorectal cancer screening in France.  Dig Liver Dis. 2017;49(3):308-311.PubMedGoogle ScholarCrossref
33.
Kiran  T, Wilton  AS, Moineddin  R, Paszat  L, Glazier  RH.  Effect of payment incentives on cancer screening in Ontario primary care.  Ann Fam Med. 2014;12(4):317-323.PubMedGoogle ScholarCrossref
34.
Hvidberg  L, Pedersen  AF, Wulff  CN, Vedsted  P.  Cancer awareness and socio-economic position: results from a population-based study in Denmark.  BMC Cancer. 2014;14:581.PubMedGoogle ScholarCrossref
35.
Hurtado  JL, Bacigalupe  A, Calvo  M,  et al.  Social inequalities in a population based colorectal cancer screening programme in the Basque Country.  BMC Public Health. 2015;15:1021.PubMedGoogle ScholarCrossref
36.
Gonzalez  EC, Ferrante  JM, Van Durme  DJ, Pal  N, Roetzheim  RG.  Comorbid illness and the early detection of cancer.  South Med J. 2001;94(9):913-920.PubMedGoogle ScholarCrossref
37.
Deding  U, Henig  AS, Salling  A, Torp-Pedersen  C, Bøggild  H.  Sociodemographic predictors of participation in colorectal cancer screening.  Int J Colorectal Dis. 2017;32(8):1117-1124.PubMedGoogle ScholarCrossref
38.
van der Vlugt  M, Grobbee  EJ, Bossuyt  PM,  et al.  Adherence to colorectal cancer screening: four rounds of faecal immunochemical test-based screening.  Br J Cancer. 2017;116(1):44-49.PubMedGoogle ScholarCrossref
39.
Pornet  C, Dejardin  O, Morlais  F, Bouvier  V, Launoy  G.  Socioeconomic determinants for compliance to colorectal cancer screening: a multilevel analysis.  J Epidemiol Community Health. 2010;64(4):318-324.PubMedGoogle ScholarCrossref
40.
Woodrow  C, Watson  E, Rozmovits  L, Parker  R, Austoker  J.  Public perceptions of communicating information about bowel cancer screening.  Health Expect. 2008;11(1):16-25.PubMedGoogle ScholarCrossref
Original Investigation
September 5, 2017

Effect of Physician Notification Regarding Nonadherence to Colorectal Cancer Screening on Patient Participation in Fecal Immunochemical Test Cancer Screening: A Randomized Clinical Trial

Author Affiliations
  • 1Department of General Practice, Faculty of Medicine, Nantes, France
  • 2French National Institute of Health and Medical Research, INSERM U1232 - Team 2, Nantes, France
  • 3Association in charge of colorectal cancer screening program, CAPSANTE 44, Nantes, France
  • 4Association in charge of colorectal cancer screening program, AUDACE, La Roche sur Yon, France
  • 5French Health Insurance System, Division of Nantes, Nantes, France
  • 6Cancer registry for Loire-Atlantique and Vendée geographic areas, Nantes, France
  • 7Department of Epidemiology and Biostatistics, Nantes University Hospital, Nantes, France
JAMA. 2017;318(9):816-824. doi:10.1001/jama.2017.11387
Key Points

Question  Does providing general practitioners in France with a list of nonadherent patients improve patient participation in fecal immunochemical test colorectal cancer screening?

Findings  In this randomized clinical trial, providing general practitioners with a list of patients who had not undergone fecal immunochemical test screening resulted in a small but significant increase in patient participation compared with patients who received usual care. Generic reminders with general information did not increase patient participation.

Meaning  Providing general practitioners in France with a list of their nonadherent patients resulted in a modest increase in patient participation in fecal immunochemical test screening.

Abstract

Importance  Increasing participation in fecal screening tests is a major challenge in countries that have implemented colorectal cancer (CRC) screening programs.

Objective  To determine whether providing general practitioners (GPs) a list of patients who are nonadherent to CRC screening enhances patient participation in fecal immunochemical testing (FIT).

Design, Setting, and Participants  A 3-group, cluster-randomized study was conducted from July 14, 2015, to July 14, 2016, on the west coast of France, with GPs in 801 practices participating and involving adult patients (50-74 years) who were at average risk of CRC and not up-to-date with CRC screening. The final follow-up date was July 14, 2016.

Interventions  General practitioners were randomly assigned to 1 of 3 groups: 496 received a list of patients who had not undergone CRC screening (patient-specific reminders group, 10 476 patients), 495 received a letter describing region-specific CRC screening adherence rates (generic reminders group, 10 606 patients), and 455 did not receive any reminders (usual care group, 10 147 patients).

Main Outcomes and Measures  The primary end point was patient participation in CRC screening 1 year after the intervention.

Results  Among 1482 randomized GPs (mean age, 53.4 years; 576 women [38.9%]), 1446 participated; of the 33 044 patients of these GPs (mean age, 59.7 years; 17 949 women [54.3%]), follow-up at 1 year was available for 31 229 (94.5%). At 1 year, 24.8% (95% CI, 23.4%-26.2%) of patients in the specific reminders group, 21.7% (95% CI, 20.5%-22.8%) in the generic reminders group, and 20.6% (95% CI, 19.3%-21.8%) in the usual care group participated in the FIT screening. The between-group differences were 3.1% (95% CI, 1.3%-5.0%) for the patient-specific reminders group vs the generic reminders group, 4.2% (95% CI, 2.3%-6.2%) for the patient-specific reminders group vs the usual care group, and 1.1% (95% CI, −0.6% to 2.8%) for generic reminders group vs the usual care group.

Conclusions and Relevance  Providing French GPs caring for adults at average risk of CRC with a list of their patients who were not up-to-date with their CRC screening resulted in a small but significant increase in patient participation in FIT screening at 1 year compared with patients who received usual care. Providing GPs with generic reminders about regional rates of CRC screening did not increase screening rates compared with usual care.

Trial Registration  clinicaltrials.gov Identifier: NCT02515344

Introduction

Colorectal cancer (CRC) is the third most frequent cancer worldwide,1,2 accounting for 694 000 deaths in 2012 and was the second most lethal cancer after lung cancer.1,2 The five-year survival rate depends on the stage at diagnosis: the rates for the localized stage, local extension to the lymph nodes, and the metastatic stage in 2013 were 90.1%, 70.8%, and 13.1%, respectively.1,2 Screenings based on sigmoidoscopy or fecal tests are associated with a decreased 10-year mortality rate.3 Many countries have implemented CRC screening based on fecal immunochemical tests (FIT).4 In France, asymptomatic individuals aged 50 to 74 years receive a postal invitation to consult their GPs to obtain a FIT testing kit. The GP refers patients with a positive test result to a gastroenterologist for a colonoscopy. Colonoscopy screening is reserved for symptomatic or high-risk individuals on the basis of family history, and GPs serve as gatekeepers for access to colonoscopy.

Patient participation in CRC screening is a major challenge.5 National participation in CRC screening in France was only 29.8% from 2013 to 2014, far below European goals.6 In 2015, the French national health system switched from the fecal occult blood test to FIT because of its the higher sensitivity and specificity.7 Prior work has examined patient-focused strategies to increase participation in CRC screening, such as postal mailing of screening kits, written or telephone reminders, and tailored navigation.8,9 However, research has not sufficiently explored whether reminders directed to GPs increases patient participation.8-10

The objective of this study was to evaluate whether providing GPs with a list of their patients who were at average risk of CRC but had yet to undergo CRC screening would increase patient participation compared with either informing GPs generically about regional-level CRC screening rates or maintaining usual care.

Method
Study Design and Participants

Quiz Ref IDThe study was a cluster randomized clinical trial with 3 parallel groups conducted from July 14, 2015, to July 14, 2016, in 2 areas on the French west coast. General practitioners were clustered within practices to avoid contamination bias stemming from shared tracking mechanisms and communication among GPs within a practice. The study protocol was approved by the human subjects ethics committee of Rennes. A waiver of individual participant-level informed consent was provided by the ethics committee. The full protocol is included in the Supplement.

General practitioners in the Loire-Atlantique and Vendée areas with at least 100 patients older than 16 years on their patient list according to the national health care insurance system were eligible. In October 2014, information about the study was sent to 1501 GPs, who could opt out of participation by contacting the research team.

Since 2015, FIT rather than fecal occult blood test screening has been used in France. FIT screening every 2 years is recommended for patients at average risk of CRC, based on (1) being between the ages of 50 years and 74 years; (2) having neither a personal history of CRC or adenoma larger than 1 cm nor a family history of CRC; and (3) manifesting no symptoms of CRC. Colonoscopy screening is reserved for patients at elevated risk of CRC corresponding to approximately 15% of the population aged 50 years to 74 years.

In the French CRC screening program, local health associations invite patients to participate in the screening if they are eligible for and have yet to complete their FIT screening in the preceding 24 months. The invitation advises them to obtain a FIT kit during a physician’s visit because kits are not mailed directly to patients. Each patient is asked to mail the completed FIT kit with a stool sample in a prepaid envelope. Test results are sent to the patient, the GP, and the local public health association. Patient participation is tracked at 3 and 6 months after the invitation letters are mailed. Patients who have not returned a FIT screening within 3 months are defined as nonadherent. Nonadherent patients may receive as many as 2 written reminders: the first at 3 months and the second at 6 months after the first invitation. Patients with a positive test result are referred to gastroenterology for colonoscopy. Local health associations maintain a tracking system that integrates information from the national health insurance system, FIT laboratories, pathology laboratories, hospital information systems, gastroenterologists, GPs, and the cancer registry, in accordance with European guidelines.6 French GPs receive financial incentives based on the number of CRC screening tests their patients complete, which for the study period was €60 (US $70) for the first 10 completed FIT tests up to an annual maximum incentive of €1030 (US $1205) per year.

Our study focused on the subset of the GP’s average-risk patients who had not completed FIT testing within the 3 months of receiving the first CRC screening invitation. Patients were eligible to participate in our study if they (1) had not returned the FIT and were candidates for the 3-month reminder letter as of July 14, 2015, the start of our study, and (2) were listed as a primary patient of a participating GP. The 3-month reminder letter included a supplemental letter from the study investigators explaining the study. Patients could opt out of participation by contacting the research team. No data related to the study were collected for patients who had opted out.

Intervention and Control Procedures

In July 2015, General practitioners were randomly assigned to 1 of 3 groups (Figure). To avoid contamination bias, the unit of randomization was by medical practice rather than by GP. Randomization was computer generated and stratified over the 2 geographic areas.

Quiz Ref IDThe GPs in the patient-specific reminders group were mailed their first list of patients a few days after the study began (July 2015) and a second updated list was mailed in November 2015. They were informed that they could request another copy if they lost the original list. The GPs were advised of tips that could help them follow up with patients: integrate informatics reminders with each patient file, ensure that medical practice secretaries recontacted each patient on the list, or place the list on their desks. There was no specific requirement, and the GP had the freedom to decide how to use the information.Quiz Ref ID General practitioners in the generic reminders group received CRC screening rates in the local administrative district. They received neither a list of their patients who were nonadherent nor specific instructions or suggestions on how to encourage patients to complete the screening. General practitioners in the usual care group received no information about CRC screening rates. Although the screening program offers a pay incentive to physicians whose patients complete the screening, this incentive was not part of the study, did not change over the study period, and did not differ across groups.

Study End Points

The primary end point was patient participation in FIT CRC screening 1 year after the 3-month reminder had been mailed. Participation was calculated as the proportion of patients who had completed FIT screening among all average-risk patients eligible for CRC screening, which was determined at the beginning of the study. The number of CRCs identified as a result of the FIT screening was also a primary end point. However, these data are not yet available from the national cancer registry.11,12

Secondary end points included screening participation rates in subgroups based on patient characteristics including sex, age, low socioeconomic status (defined as an annual income < €8593 [<US $10 054] for an individual or < €12 889 [US $15 080] for a couple), and presence of a severe chronic disease (defined using a list of 30 severe chronic diseases leading to high costs for patients).13 The economic costs related to each reminder strategy was a planned secondary end point, but these data are not yet available and are not reported in this article.

Post hoc outcomes analyzed whether patient participation to FIT screening was linked to characteristics of the GPs (sex, age, practice location, and activity).

Data Collection

Data regarding patient participation were stored by regional public health authorities, based on recommended procedures implemented for general care.6 As reported previously, local associations overseeing the screening program used a dedicated computerized database that allowed for the collection of data from all partners involved (national health insurance systems, the FIT laboratory, pathology laboratory, hospital information system, gastroenterologists, GPs, patients, the regional cancer registry). Patients who completed the FIT within a year of receiving their first reminder were classified as adherent, whereas other patients were classified as nonresponders. Based on this classification, no data were missing.

Patient and GPs’ characteristics were extracted from the national health care insurance database. This data extraction procedure allowed for complete collection of data related to patient and GPs’ characteristics.

Statistical Analysis

The objective of this analysis was to be as exhaustive as possible in the recruitment of GPs to avoid selection bias. We estimated that at least 1300 GPs would participate among all those eligible, corresponding to 433 in each group. Based on this recruitment, we assumed that participation in CRC screening would be at least 12% higher in the patient-specific reminders group than in the other groups (absolute mean difference, 67% vs 55%; estimated SD, 0.61; bilateral α risk, 5%; statistical power, 80%). Participation in the usual care group was estimated at 55% using the following data. During 2012 and 2013, participation in the fecal occult blood test screening was equal to 38.2% in Western France based on rates measured by the regional public health authority. Based on a literature review,9 we hypothesized that participation in CRC screening would be higher after introduction of the FIT (vs the fecal occult blood test). International studies concluded that using the FIT rather than the fecal occult blood test significantly increased participation from an odds ratio (OR) of 0.86 to 2.14 (median, 1.44).7,14-18 Thus, the study power was based on a projected participation rate of 55% (38.2% × 1.44).

The significance threshold was P < .05. Testing was 2-sided. The study design avoided having missing data. The modified intention-to-treat method was applied and included all randomized GPs. General practitioners who had changed their eligibility criteria (practice cessation, practice reorientation, address modification, or death), either before the randomization or during the study, were excluded from the analysis.

For the primary end point, the GP was the statistical unit of analysis. Each GP was characterized by the ratio of participation of their patients 1 year after participation invitations were sent. This ratio was defined as the number of patients who sent back their FIT screen to the number of patients who at the beginning of the study had not sent back their FIT screen (continuous variable). The results were confirmed with a mixed-model regression analysis adjusted for all GP characteristics as shown in Table 1. Medical practices were considered a between-unit random effect. The randomized group was considered a fixed effect. The random effect accounted for correlations among GPs in the same practice. A global analysis of variance test was performed to compare participation among all groups. If the null hypothesis was rejected, then comparisons between the usual care group with the others would be performed.

Patient factors associated with participation in CRC screening (prespecified analysis provided in the protocol) were assessed with a mixed-general linear model for which the Y was the status of each patient depending on whether he/she participated (logical variable). The patient was the statistical unit. In this model, the GP was considered a within-unit random factor, and the variables were patient characteristics. No selection procedure was performed.

In post hoc analyses, GP factors associated with patient participation in the FIT screening were assessed by a mixed-general linear model for which the Y predicted was the percentage of participation of each GP (continuous variable). General practitioners were the statistical unit. Additionally, medical practices were considered between-random effects, and the fixed variables were GP characteristics. No selection procedure was performed.

Analyses were performed using R 3.20 statistical software. The GLIMMIX and the Mixed Procedures of SAS version 9.4 (SAS Institute Inc) were used to estimate the models including the random effects.

Results
Participants

A written study description was mailed to 1501 GPs identified as eligible for the present study, located in 801 medical practices. One GP opted not to participate (Figure). Between mailing the study introductory letter in October 2014 and randomization in July 2015, 18 GPs were excluded because of practice cessation, practice reorientation, address modification, or death. Another 36 GPs were excluded after the randomization for the same reasons. Among 1482 randomized GPs (mean age, 53.4 years; 576 women [38.9%]), 1446 completed the study and were included in the analysis, 496 of whom were randomized to the patient-specific reminders group; 495, to the generic reminders group; and 455 to the usual care group. Table 1 shows the sex distribution and demographic characteristics of the GPs in each group.

Of the 33 044 patients (mean age, 59.7 years; 17 949 women [54.3%]), 1815 were excluded for medical or administrative criteria, so that the follow-up at 1 year was available for 31 229 (94.5%) (mean age, 60.9 years; 16 919 women [54.4%]): 10 476 patients in the patient-specific reminders group, 10 606 patients in the generic reminders group, and 10 147 patients in the usual care group. Table 2 presents the demographic characteristics of the participants.

Participation in FIT Screening (Primary End Point)

At 1 year, the mean patient participation per GP was 24.8% (95% CI, 23.4%-26.2%) in the patient-specific reminders group, 21.7% (95% CI, 20.5%-22.8%) in the generic reminders group, and 20.6% (95% CI, 19.3%-21.8%) in the usual care group, with between-group differences of 3.1% (95% CI, 1.3%-5.0%) for patient-specific reminders vs generic reminders, 4.2% (95% CI, 2.3%-6.2%) for patient-specific reminders vs usual care, and 1.1% (95% CI, −0.6% to 2.8%) for generic reminders vs usual care (Table 3).

Patient Factors Related to Participation (Secondary End Points)

Patients with a low socioeconomic status and patients who had a chronic disease were less likely to complete the FIT screening (Table 4).

GP Practice Characteristics Related to Participation (Post Hoc Analysis)

The following GP practice characteristics were associated with greater participation in FIT screening of patients who were nonresponders at the beginning of the study: younger physicians, GPs receiving a greater number of medical visits per year, and GPs who had fewer nonresponders on their list at the beginning of the study (Table 5).

Discussion

Quiz Ref IDIn this open, randomized 3-group study, providing GPs with a list of patients who were nonadherent to CRC screening was associated with a modest increase in FIT participation compared with providing GPs with generic reminders about regional CRC screening rates or providing no reminders. These results were consistent across statistical models considering either GPs or patients as the statistical unit after adjustments for GP or patient characteristics. The analysis demonstrated a significant absolute difference of 4.2% between the patient-specific reminders group and usual care group, even though the statistical plan was designed to detect an anticipated difference of more than 12%. The reason this difference could be detected is that the standard deviation of the study (15%) was lower than the value included in the protocol (61%).

The target of this large randomized clinical trial (RCT) was the GP rather than the patient. In most strategies proposed by policy makers, interventions are designed to directly influence patients.8-10 In a previous review focusing on patient participation in CRC screening, Rat et al9 reported the absolute increase of participation observed for the following interventions: mailing of invitation letters to patients (from 6.2% to 7.0%),19,20 postal mailing of kits (from 3.9% to 24.5%),21,22 written reminders (6.7% to 16.4%).22,23 However, few authors have reported the effect of interventions directed at GPs.8-10,24,25 Developing personalized interviews might increase patient participation.10 Furthermore, a training program based on communication skills positively increased patient participation to 12.2%.24 However, these interventions were voluntary; thus, generalizing at a population level might be difficult. According to GPs, lack of time is one of the primary barriers limiting wider promotion of CRC screening during consultation.26,27

Based on a large RCT, these study findings present stronger evidence than a “before vs after” study,28 a cross-sectional study,29 or a small RCT,30 and results confirm the conclusions of a previous smaller trial that targeted physicians.31

Participation in CRC screening increased in the patient-specific reminders group compared with the usual care group. However, the statistical plan was based on the assumption that the participation rate would be higher in all randomized groups because of the switch from the older fecal occult blood test to the FIT in France at the beginning of 2015.7,14-18 One problem was that French authorities decided to interrupt postal mailing of the tests to patients due to the expense of the FIT kits. This modification to the screening program aimed to avoid wasting tests but diminished the positive effect of patient-specific reminders. Piette et al32 demonstrated that interruption of mailing kits to patients decreased the participation rate from 45% to 28%.

In this study, patient participation was higher for younger physicians, possibly because these GPs may have greater concerns about public health and screening procedures.26

Although lack of time has been widely reported as a main barrier to GP involvement in cancer screening,26,27 this study presents original findings. A hypothesis was that physicians with more visits would pay more attention to acute disease and could neglect preventive procedures due to these time constraints. In contrast, the study results show that CRC screening rates were higher for GPs who have busier practices, regardless of the randomized groups. The reason is unclear and requires further investigation. As French authorities have introduced a pay-for-performance grant related to cancer screening, a possible hypothesis is that some busy physicians could also be more sensitive to financial incentives.33

Having a short list of nonadherent patients increased participation rates among nonresponders. One explanation is that targeting nonresponders was easier for GPs who received a shorter list of patients. However, GPs who had shorter lists were probably also the most motivated.

This study provides insight into the factors related to patient participation in CRC screening. Some of these findings are consistent with those of previous studies. Patients with low socioeconomic status had lower participation rates.34,35 Findings regarding patients having a chronic disease provide insight into the impact of comorbidities. Gonzalez et al36 reported that patients who were diagnosed at more advanced stages were more likely to have comorbid diseases. General practitioners may not focus their attention and energy on cancer screening because of concerns about other health issues. Perhaps patients who are not concerned about cancer screening also demonstrate a lack of time for health and prevention in general. Thus, GPs’ greater involvement with these populations may first affect other health issues consistent with risk assessment and disease prevalence.

The literature presents conflicting information regarding whether patient age and sex affect participation in CRC screening.37-39 Recent publications have reported lower participation among younger individuals and male patients37,39 but these findings were not confirmed in our study.

This study has several strengths. First, the inclusion of 1446 GPs and 31 229 patients represents, to our knowledge, the largest randomized study focusing on CRC screening participation.9 Second, the generalizability of these results is important for the following reasons. The intervention was not expensive and did not need to be voluntary. Almost all GPs in the 2 geographic areas were included and participated. Thus, selection bias was not a concern. The effects of the intention-to-treat procedure mean that the patient-specific reminders increased CRC participation regardless of GP attitudes. Third, the increase in patient participation concerned only patients who were not adherent at the beginning of the study, whereas other studies assessed a general participation rate. Fourth, the observed effects might continue to increase if patient-specific reminders were implemented in usual care: (1) the study reported patient participation after a 12-month follow-up, but not all nonadherent patients had a medical consultation during this period, and (2) a hypothesis is that the positive effect might continue to increase in the patient-specific reminders group if an updated list was provided yearly (whereas participation in the usual care group should not be modified).

Limitations

The study also has limitations. First, the main limitation is the small magnitude of the increase in participation. Increasing participation by only 4% (absolute participation increase) might be regarded as disappointing. Second, the design of this study does not allow a clear report of how providing patient-specific reminders to the GPs positively affects CRC screening participation. A hypothesis is that the physicians concentrated their efforts on these nonadherent patients and customized their communication to the personality and characteristics of each patient to enable shared decision making.8,40Quiz Ref ID Third, the generalizability of the study findings might depend on the national setting: generating a list of nonresponders requires good data and the resources to create the list. Even in practices with electronic health records, this might not be possible because tests and procedures are sometimes performed outside of the clinic setting and because results are often entered in ways that cannot be captured electronically.

Conclusions

Providing French GPs caring for adults at average risk of CRC with a list of their patients who were not up-to-date with their CRC screening resulted in a small but significant increase in patient participation in FIT screening at 1 year compared with patients who received usual care. Providing GPs with generic reminders about regional rates of CRC screening did not increase screening rates compared with usual care.

Back to top
Article Information

Corresponding Author: Cédric Rat, MD, PhD, Department of General Practice, Faculty of Medicine of Nantes, 1 rue Gaston Veil, 44035 Nantes, France (cedric.rat@univ-nantes.fr).

Accepted for Publication: August 1, 2017.

Author Contributions: Dr Nguyen had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Rat, Pogu, Bianco, Nanin, Cowppli-Bony, Nguyen.

Acquisition, analysis, or interpretation of data: Rat, Pogu, Le Donné, Latour, Bianco, Nanin, Gaultier, Nguyen.

Drafting of the manuscript: Rat, Le Donné, Bianco, Gaultier, Nguyen.

Critical revision of the manuscript for important intellectual content: Pogu, Latour, Bianco, Nanin, Cowppli-Bony.

Statistical analysis: Bianco, Gaultier, Nguyen.

Obtained funding: Bianco, Nguyen.

Administrative, technical, or material support: Pogu, Le Donné, Latour, Bianco, Nanin.

Study supervision: Rat, Bianco, Nguyen.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported

Funding/Support: Financial support was provided by a grant from the French National Institute of Cancer.

Role of the Funder/Sponsor: The funding source had no role in the design or conduct of the study; the collection, management, analysis, or interpretation of the data; the preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Additional Contributions: We thank Shannon Thorn of the American Journal Experts for editorial support, for which she was compensated. We also thank Bruno Hubert, MD, PhD, French National Institute for Public Health Surveillance, Marc Le Rhun, MD, Nantes University Hospital, and Pierre-Louis Druais, MD, French National College of General Practitioners, who served on the supervisory committee. They were not compensated for their contributions.

References
1.
GLOBOCAN 2012: estimated cancer incidence, mortality and prevalence worldwide in 2012. World Health Organization website. http://globocan.iarc.fr. Accessed July 23, 2017.
2.
Howlader  N, Noone  AM, Krapcho  M,  et al. SEER cancer statistics review (CRS) 1975-2014. National Cancer Institute website. https://seer.cancer.gov/csr/1975_2014. Posted April 2017. Accessed July 23, 2017.
3.
Holme  Ø, Bretthauer  M, Fretheim  A, Odgaard-Jensen  J, Hoff  G.  Flexible sigmoidoscopy versus faecal occult blood testing for colorectal cancer screening in asymptomatic individuals.  Cochrane Database Syst Rev. 2013;9(9):CD009259.PubMedGoogle Scholar
4.
Lieberman  D, Ladabaum  U, Cruz-Correa  M,  et al.  Screening for colorectal cancer and evolving issues for physicians and patients: a review.  JAMA. 2016;316(20):2135-2145.PubMedGoogle ScholarCrossref
5.
van der Vlugt  M, Grobbee  EJ, Bossuyt  PM,  et al.  Adherence to colorectal cancer screening: four rounds of faecal immunochemical test-based screening.  Br J Cancer. 2017;116(1):44-49.PubMedGoogle ScholarCrossref
6.
von Karsa  L, Patnick  J, Segnan  N,  et al; European Colorectal Cancer Screening Guidelines Working Group.  European guidelines for quality assurance in colorectal cancer screening and diagnosis: overview and introduction to the full supplement publication.  Endoscopy. 2013;45(1):51-59.PubMedGoogle Scholar
7.
Hol  L, van Leerdam  ME, van Ballegooijen  M,  et al.  Screening for colorectal cancer: randomised trial comparing guaiac-based and immunochemical faecal occult blood testing and flexible sigmoidoscopy.  Gut. 2010;59(1):62-68.PubMedGoogle ScholarCrossref
8.
Senore  C, Inadomi  J, Segnan  N, Bellisario  C, Hassan  C.  Optimising colorectal cancer screening acceptance: a review.  Gut. 2015;64(7):1158-1177.PubMedGoogle ScholarCrossref
9.
Rat  C, Latour  C, Rousseau  R,  et al.  Interventions to increase uptake of faecal tests for colorectal cancer screening: a systematic review.  Eur J Cancer Prev. doi:10.1097/CEJ.0000000000000344PubMedGoogle Scholar
10.
Sabatino  SA, Lawrence  B, Elder  R,  et al; Community Preventive Services Task Force.  Effectiveness of interventions to increase screening for breast, cervical, and colorectal cancers: nine updated systematic reviews for the guide to community preventive services.  Am J Prev Med. 2012;43(1):97-118.PubMedGoogle ScholarCrossref
11.
Martos  C, Crocetti  E, Visser  O, Rous  B; Cancer Data Quality Checks Working Group. A proposal on cancer data quality checks: one common procedure for European cancer registries. Ispra, Italy: European Network of Cancer Registries; 2014.http://www.encr.eu/images/download/A_proposal_on_cancer_data_quality_checks-one_common_procedure_for_European_cancer_registries.pdf. Accessed July 23, 2017.
12.
Zanetti  R, Schmidtmann  I, Sacchetto  L,  et al.  Completeness and timeliness: cancer registries could/should improve their performance.  Eur J Cancer. 2015;51(9):1091-1098.PubMedGoogle ScholarCrossref
14.
Cole  SR, Young  GP, Esterman  A, Cadd  B, Morcom  J.  A randomised trial of the impact of new faecal haemoglobin test technologies on population participation in screening for colorectal cancer.  J Med Screen. 2003;10(3):117-122.PubMedGoogle ScholarCrossref
15.
Federici  A, Giorgi Rossi  P, Borgia  P, Bartolozzi  F, Farchi  S, Gausticchi  G.  The immunochemical faecal occult blood test leads to higher compliance than the guaiac for colorectal cancer screening programmes: a cluster randomized controlled trial.  J Med Screen. 2005;12(2):83-88.PubMedGoogle ScholarCrossref
16.
Hughes  K, Leggett  B, Del Mar  C,  et al.  Guaiac versus immunochemical tests: faecal occult blood test screening for colorectal cancer in a rural community.  Aust N Z J Public Health. 2005;29(4):358-364.PubMedGoogle ScholarCrossref
17.
Birkenfeld  S, Belfer  RG, Chared  M,  et al.  Factors affecting compliance in faecal occult blood testing: a cluster randomized study of the faecal immunochemical test versus the guaiac faecal occult test.  J Med Screen. 2011;18(3):135-141.PubMedGoogle ScholarCrossref
18.
Levi  Z, Birkenfeld  S, Vilkin  A,  et al.  A higher detection rate for colorectal cancer and advanced adenomatous polyp for screening with immunochemical fecal occult blood test than guaiac fecal occult blood test, despite lower compliance rate: a prospective, controlled, feasibility study.  Int J Cancer. 2011;128(10):2415-2424.PubMedGoogle ScholarCrossref
19.
Cole  SR, Smith  A, Wilson  C, Turnbull  D, Esterman  A, Young  GP.  An advance notification letter increases participation in colorectal cancer screening.  J Med Screen. 2007;14(2):73-75.PubMedGoogle ScholarCrossref
20.
Mant  D, Fuller  A, Northover  J,  et al.  Patient compliance with colorectal cancer screening in general practice.  Br J Gen Pract. 1992;42(354):18-20.PubMedGoogle Scholar
21.
Giorgi Rossi  P, Grazzini  G, Anti  M,  et al.  Direct mailing of faecal occult blood tests for colorectal cancer screening: a randomized population study from Central Italy.  J Med Screen. 2011;18(3):121-127.PubMedGoogle ScholarCrossref
22.
Green  BB, Wang  CY, Anderson  ML,  et al.  An automated intervention with stepped increases in support to increase uptake of colorectal cancer screening: a randomized trial.  Ann Intern Med. 2013;158(5 pt 1):301-311.PubMedGoogle ScholarCrossref
23.
Lee  JK, Reis  V, Liu  S,  et al.  Improving fecal occult blood testing compliance using a mailed educational reminder.  J Gen Intern Med. 2009;24(11):1192-1197.PubMedGoogle ScholarCrossref
24.
Aubin-Auger  I, Laouénan  C, Le Bel  J,  et al.  Efficacy of communication skills training on colorectal cancer screening by GPs: a cluster randomised controlled trial.  Eur J Cancer Care (Engl). 2016;25(1)18-16.PubMedGoogle ScholarCrossref
25.
Hewitson  P, Ward  AM, Heneghan  C, Halloran  SP, Mant  D.  Primary care endorsement letter and a patient leaflet to improve participation in colorectal cancer screening: results of a factorial randomised trial.  Br J Cancer. 2011;105(4):475-480.PubMedGoogle ScholarCrossref
26.
Brotons  C, Björkelund  C, Bulc  M,  et al; EUROPREV Network.  Prevention and health promotion in clinical practice: the views of general practitioners in Europe.  Prev Med. 2005;40(5):595-601.PubMedGoogle ScholarCrossref
27.
Aubin-Auger  I, Mercier  A, Lebeau  JP, Baumann  L, Peremans  L, Van Royen  P.  Obstacles to colorectal screening in general practice: a qualitative study of GPs and patients.  Fam Pract. 2011;28(6):670-676.PubMedGoogle ScholarCrossref
28.
Goebel  LJ.  A peer review feedback method of promoting compliance with preventive care guidelines in a resident ambulatory care clinic.  Jt Comm J Qual Improv. 1997;23(4):196-202.PubMedGoogle Scholar
29.
Hudson  SV, Ohman-Strickland  P, Cunningham  R, Ferrante  JM, Hahn  K, Crabtree  BF.  The effects of teamwork and system support on colorectal cancer screening in primary care practices.  Cancer Detect Prev. 2007;31(5):417-423.PubMedGoogle ScholarCrossref
30.
Vinker  S, Nakar  S, Rosenberg  E, Kitai  E.  The role of family physicians in increasing annual fecal occult blood test screening coverage: a prospective intervention study.  Isr Med Assoc J. 2002;4(6):424-425.PubMedGoogle Scholar
31.
Sequist  TD, Zaslavsky  AM, Marshall  R, Fletcher  RH, Ayanian  JZ.  Patient and physician reminders to promote colorectal cancer screening: a randomized controlled trial.  Arch Intern Med. 2009;169(4):364-371.PubMedGoogle ScholarCrossref
32.
Piette  C, Durand  G, Bretagne  JF, Faivre  J.  Additional mailing phase for FIT after a medical offer phase: the best way to improve compliance with colorectal cancer screening in France.  Dig Liver Dis. 2017;49(3):308-311.PubMedGoogle ScholarCrossref
33.
Kiran  T, Wilton  AS, Moineddin  R, Paszat  L, Glazier  RH.  Effect of payment incentives on cancer screening in Ontario primary care.  Ann Fam Med. 2014;12(4):317-323.PubMedGoogle ScholarCrossref
34.
Hvidberg  L, Pedersen  AF, Wulff  CN, Vedsted  P.  Cancer awareness and socio-economic position: results from a population-based study in Denmark.  BMC Cancer. 2014;14:581.PubMedGoogle ScholarCrossref
35.
Hurtado  JL, Bacigalupe  A, Calvo  M,  et al.  Social inequalities in a population based colorectal cancer screening programme in the Basque Country.  BMC Public Health. 2015;15:1021.PubMedGoogle ScholarCrossref
36.
Gonzalez  EC, Ferrante  JM, Van Durme  DJ, Pal  N, Roetzheim  RG.  Comorbid illness and the early detection of cancer.  South Med J. 2001;94(9):913-920.PubMedGoogle ScholarCrossref
37.
Deding  U, Henig  AS, Salling  A, Torp-Pedersen  C, Bøggild  H.  Sociodemographic predictors of participation in colorectal cancer screening.  Int J Colorectal Dis. 2017;32(8):1117-1124.PubMedGoogle ScholarCrossref
38.
van der Vlugt  M, Grobbee  EJ, Bossuyt  PM,  et al.  Adherence to colorectal cancer screening: four rounds of faecal immunochemical test-based screening.  Br J Cancer. 2017;116(1):44-49.PubMedGoogle ScholarCrossref
39.
Pornet  C, Dejardin  O, Morlais  F, Bouvier  V, Launoy  G.  Socioeconomic determinants for compliance to colorectal cancer screening: a multilevel analysis.  J Epidemiol Community Health. 2010;64(4):318-324.PubMedGoogle ScholarCrossref
40.
Woodrow  C, Watson  E, Rozmovits  L, Parker  R, Austoker  J.  Public perceptions of communicating information about bowel cancer screening.  Health Expect. 2008;11(1):16-25.PubMedGoogle ScholarCrossref
×