Philippe Perrin

Short term outcomes of the first round of a pilot colorectal cancer screening programme with guaiac based faecal occult blood test

Objective: To assess both feasibility and short term outcomes of a population based colorectal cancer screening programme using a biennial guaiac based faecal occult blood test (gFOBT). Method: All participants were invited by mail to take part in a screening programme using a non-rehydrated gFOBT. The gFOBTs were first provided by general practitioners (GPs) and then directly mailed to individuals who failed to comply after two invitations. The setting was a French administrative district: Haut-Rhin (710 000 inhabitants). 182 981 residents aged 50–74 years were invited to participate. Results: 19 274 people (10.5%) were excluded from gFOBT screening and 90 706 completed a gFOBT, so that the participation rate was 55.4% of those eligible. 76.5% of the completed gFOBTs were provided by GPs and 15.5% by direct mailing. The gFOBT positivity rate was 3.4%. The positive predictive value was 42.7% for neoplasia (women 30.8%, men 52.5%), 23.6% for advanced adenoma, and 7.6% for cancer. The number of normal colonoscopic procedures (without neoplasia) needed to be performed for each colonoscopy detecting an advanced neoplasia was 1.8, lower in men (1.2) than in women (3.4), and decreasing with age. Detection rates for neoplasia and cancer were 12.8 and 2.3 per 1000 people screened. 206 adenocarcinomas were detected: 47.6% were stage I and 23.8% stage II. The direct cost was estimated at €29.3 per screened person and €13 466 per cancer detected. Conclusions: Participation and diagnostic yield of controlled trials of gFOBT screening are reproducible in the real world at an acceptable cost through an organised population based programme involving GPs.

Measurement of adenoma detection and discrimination during colonoscopy in routine practice: an exploratory study.

BACKGROUND Measuring neoplasia yield is a priority in the quality improvement process for colonoscopy. However, neither the most appropriate quality indicator nor the standard threshold has been established. OBJECTIVE To determine the most appropriate quality indicators to assess the yield of routine colonoscopy. DESIGN Retrospective. SETTING Population-based colorectal cancer screening program in 3 French administrative areas. SUBJECTS One hundred gastroenterologists and their average-risk asymptomatic patients aged 50 to 74 years undergoing colonoscopy for positive guaiac-based fecal occult blood test results. MAIN OUTCOME MEASUREMENTS Comparison of several indicators, mainly the adenoma detection rate (ADR) and polyp detection rate (PDR), the mean number of adenomas per colonoscopy (MNA) and mean number of polyps (MNP) and the proportion of adenomas among polyps (PAP). RESULTS Correlations were good between the ADR and PDR (Pearson coefficient r = 0.88 [95% CI, 0.78-0.94]) and between MNA and MNP (r = 0.89 [95% CI, 0.79-0.94]) (P < .0001 for both). Gastroenterologists were classified as higher or lower detectors in comparison with the lower limit of the 95% confidence interval of the median value for each indicator. The MNP (MNA) provided better discrimination than the PDR (ADR). Concordance between classifications of gastroenterologists according to their MNA and MNP was excellent (κ = 0.89). PAP varied dramatically from 38% to 95% between gastroenterologists and was very poorly correlated with the ADR (r = -0.27 [95% CI, -0.54 to 0.07; P = .11]) and the MNA (r = 0.03 [95% CI, -0.29 to 0.36; P = .88]). LIMITATIONS Some factors influencing the neoplasia yield were not taken into account. CONCLUSIONS The MNP could replace the ADR for the assessment of adenoma detection in routine practice. A separate indicator, PAP, would be necessary to assess adenoma discrimination ability.

Diagnostic accuracy of community pathologists in the interpretation of colorectal polyps

Background and aims Management of patients with endoscopically removed colorectal polyps is generally dependent on pathological evaluation. The aim of this study was to assess the accuracy and clinical impact of pathologic interpretation of colorectal polyps by community pathologists. Methods Two expert gastrointestinal pathologists reviewed the slides of 300 colorectal polyps initially examined by 14 general pathologists. Polyps had been detected by a fecal occult blood test colorectal cancer screening program in Haut-Rhin, a French administrative district. Results Villous histology was overread in 24.8% of cases and high-grade dysplasia in 22.0%. The diagnosis of serrated adenoma was confirmed in 15.7% of cases. The diagnosis of T1 carcinoma was overestimated in seven cases (17.9%) and missed in four. In the screening program, the proportion of correct diagnoses of community pathologists was estimated at 45.3% of polyps, of misclassification without clinical impact at 27.5%, and of misclassification with a theoretical impact on management at 27.2%, leading to over-surveillance in 20.3% of polyps and to unnecessary surgical resection in three individuals. Overall, 37.5% of the pathology reports of malignant polyps were complete, presenting all criteria necessary for therapeutic decision-making. Conclusion Community pathologists exhibited moderate accuracy for interpreting colorectal polyps, with an impact on patient management for around one out of five individuals. Our results confirm the intrinsic poor reliability of the pathologic interpretation of villous histology and high-grade dysplasia and suggest that these advanced pathologic features should be abandoned for clinical use. They illustrate the need for a clarification of the nomenclature of serrated polyps.

The mean number of adenomas per procedure should become the gold standard to measure the neoplasia yield of colonoscopy: A population-based cohort study

BACKGROUND Measuring adenoma detection is a priority in the quality improvement process for colonoscopy. Our aim was (1) to determine the most appropriate quality indicators to assess the neoplasia yield of colonoscopy and (2) to establish benchmark rates for the French colorectal cancer screening programme. METHODS Retrospective study of all colonoscopies performed in average-risk asymptomatic people aged 50-74 years after a positive guaiac faecal occult blood test in eight administrative areas of the French population-based programme. RESULTS We analysed 42,817 colonoscopies performed by 316 gastroenterologists. Endoscopists who had an adenoma detection rate around the benchmark of 35% had a mean number of adenomas per colonoscopy varying between 0.36 and 0.98. 13.9% of endoscopists had a mean number of adenomas above the benchmark of 0.6 and an adenoma detection rate below the benchmark of 35%, or inversely. Correlation was excellent between mean numbers of adenomas and polyps per colonoscopy (Pearson coefficient r=0.90, p<0.0001), better than correlation between mean number of adenomas and adenoma detection rate (r=0.84, p=0.01). CONCLUSION The mean number of adenomas per procedure should become the gold standard to measure the neoplasia yield of colonoscopy. Benchmark could be established at 0.6 in the French programme.

Harms of colonoscopy in a colorectal cancer screening programme with faecal occult blood test: A population-based cohort study

BACKGROUND AND AIMS To assess the harms of colonoscopy in a real world colorectal cancer screening programme with faecal occult blood test. METHODS Retrospective cohort study of all colonoscopies performed in patients aged 50-74 for a positive guaiac-based faecal occult blood test between September 2003 and February 2010 within the screening programme in progress in Alsace (France). Adverse events were recorded through prospective voluntary reporting by gastroenterologists and retrospective postal surveys addressed to persons screened and their general practitioners. RESULTS Of 10,277 colonoscopies, 250 adverse events were recorded, 48 (4.7 ‰, 95% CI 3.4-6.0) of them being moderate or severe, mainly 10 (1.0 ‰, 95% CI 0.4-1.6) perforations and 31 (3.0 ‰, 95% CI 2.0-4.1) bleeding. 91.7% of moderate and severe adverse events were the result of a therapeutic procedure. Of 103 serious adverse events, eight (7.8%) were considered preventable. Gastroenterologists reported 52.2% of moderate and severe adverse events. A mild adverse event or an incident was reported in up to 97.0 ‰ (95% CI 83.2-110.7) colonoscopies. CONCLUSION The harms of colonoscopy were underestimated in all randomized controlled trials on colorectal cancer screening with faecal occult blood test. They are greater in a real world programme, estimated at 7.5 major and 100 minor adverse events per 1000 colonoscopies.

Participation in four rounds of a French colorectal cancer screening programme with guaiac faecal occult blood test: a population-based open cohort study

Objectives Four randomized controlled trials have demonstrated the efficacy of screening using a guaiac faecal occult blood test (gFOBT) on colorectal cancer (CRC) mortality. Whether their results are transposable to the real world is unknown. This study aimed to assess the determinants of participation in the first four rounds of the CRC screening programme using a gFOBT implemented since 2003 in the Haut-Rhin (Alsace) part of the French national programme. Methods We performed a population-based open cohort study of all residents aged 50–74, around 200,000 people. They were invited by mail to participate every other year. The gFOBT kits (Hemoccult II) were first provided by general practitioners, and then directly mailed to persons who failed to comply. Results The uptake decreased significantly across all rounds, from 54.3% to 47.1% (p < 0.0001), mainly in people younger than 60. The proportion of people screened by general practitioners increased significantly from 77.0% in the first round to 84.2% in the fourth (p < 0.01). Overall, 61.3% of the invited population participated at least once, and 14.3% had completed all the four tests. The colonoscopy uptake was around 91%, among the highest ever reported. Conclusions Despite the involvement of general practitioners, the uptake and adherence to repeat testing are modest and deteriorate with time, so that the reduction in CRC mortality in reality will be significantly lower than that in the trials. The benefit-risk balance of the French programme is, at present, less favourable than that shown in the trials.

Su1224 Participation in the First Four Rounds of the Colorectal Cancer Screening Program With Guaiac Fecal Occult Blood Test Conducted in the Haut-Rhin (Alsace, France)

Organized screening programs depend on high participation rates to be effective and efficient. Whether uptake rates obtained in randomized controlled trials on colorectal cancer (CRC) screening with guaiac fecal occult blood test (gFOBT) are transposable in the real world is questionable. gFOBT has several drawbacks, one is the requirement for frequent testing, which may limit compliance. Aim: To assess participation in the first 4 rounds of an organized CRC screening program using gFOBT. Methods: Comparison of the determinants of participation in the first 4 rounds (R1 to R4) of the CRC screening programwith Hemoccult II implemented since 2003 in the Haut-Rhin, a French administrative area. We performed a population-based open cohort study of all average risk residents aged 50 74 years. They were invited to participate every other year. A first letter invited them to visit their general practitioner (GP) for CRC screening. Three recall letters were mailed to all those who had not complied, the second with a gFOBT kit. Results: Main results are presented in the table. The decrease in uptake was similar in men and women and was observed in all age groups except the 70 74 year age group. Overall, after 4 rounds, 242,292 persons had been invited and 34545 (14.3%) had completed 4 tests, 26675 (11.0%) 3 tests, 37193 (15.4%) 2 tests and 50070 (20.7%) 1 test. Overall, 148,483 (61.3%) persons had been screened at least once. Of 86,694 people who were eligible in all rounds, 34,545 (39.8%) had completed all the 4 tests, 8679 (10.0%) 3 tests, 8099 (9.3%) 2 tests, 6552 (7.6%) 1 test, and 28,819 (33.2%) 0 test. The uptake following the 2nd, 3rd and 4th invitation in people who never complied before was 17.0%, 8.8% and 5.5%, respectively. 16% of people who had participated in Rn did not participate in Rn+1. The proportion of people screened by GPs increased significantly from 76.9% in the 1st round to 84.0% in the 4th (p<0.01) while that of people screened by direct mailing of the gFOBT kit decreased significantly from 15.6% to 12.1% (p<0.01). The proportion of uptake following the first 2 invitation letters increased from R1 to R4 from 78.4% to 86.9%, while that following the mailing of the gFOBT kit decreased from 15.3% to 7.8%. The uptake decreased between years 2008 2009 and 2011 2012 from 34.3% to 31.7% in France and from 51.3% to 47.1% in our region. The latter uptake was among the highest achieved in France, where it varied from 7.1% to 51.4% depending on regions. Conclusion: The uptake and adherence to repeat testing are modest and deteriorate with time in the French organized population-based gFOBT CRC screening program, so that the reduction in CRC mortality will be significantly lower than in the trials. Effort is needed to enhance uptake and to reduce inequalities in participation related to sex, age, place of residence and deprivation.

218 Results From the First Four Rounds of the Colorectal Cancer Screening Program With Guaiac Fecal Occult Blood Test Conducted in the Haut-Rhin (Alsace, France)

Whether the results of randomized controlled trials on colorectal cancer (CRC) screening with guaiac fecal occult blood test (gFOBT) are transposable in the real world is questionable. gFOBT has several drawbacks, one is the requirement for frequent testing, which may limit compliance and thereby effectiveness. Aim: To assess the short term outcomes of the four first rounds of a population-based CRC screening program using gFOBT. Methods: Comparison of the outcomes of the four first rounds (R1 to R4) of the organized CRC screening program with Hemoccult II implemented in the Haut-Rhin, a French administrative area, since 2003 (Denis B et al Gut 2007;56:1579-84). All average-risk residents aged 50 74 were invited by mail to participate every other year. Results: Main outcomes are presented in the table. The crude uptake rate decreased from 47.9% to 38.5%. The decrease in adjusted uptake was similar in men and women but was observed in younger age groups (50 64) only and maximal in the 50 54 year age group. 15.8% of people who participated in Rn did not participate in Rn+1. Overall, 56.7% of the target population had completed at least one test and 19.6% four tests. More than 80% of the completed tests were provided by general practitioners. The positive predictive value for advanced neoplasia decreased significantly from 31.4% in R1 to 26.1% in R4 (and detection rate from 9.7‰ to 4.7‰). The rate of stage I colon cancers decreased significantly from 44.9% in R1 to 38.1% in R4. Conclusion: Participation and yield deteriorated with time in our organized population-based gFOBT CRC screening program. This deterioration was not observed in previous randomized controlled trials on gFOBT screening and may question the reproducibility of their effectiveness on the reduction of CRC mortality in the real world. Effort is needed to enhance uptake and to reduce inequalities in participation related to sex, age, place of residence and deprivation.

P.17 Facteurs et motifs de participation au dépistage du cancer colorectal par rectosigmoïdoscopie en France

Objectif Nous avons montre que le depistage du cancer colorectal (CCR) par recto-sigmoidoscopie (RS) etait faisable en France et offrait un meilleur rendement que le depistage par Hemoccult ® (H). But Evaluer les facteurs socio-demographiques et medicaux et les motifs de participation au depistage par RS. Patients et Methodes Enquete telephonique aupres d’un echantillon de la population d’un canton de 55 a 64 ans invitee a participer a une etude proposant un depistage par RS en sus du depistage organise par H. Resultats 202 personnes (100 femmes) ayant realise une RS et 200 (111 femmes) ne l’ayant pas realisee ont repondu. 320 (171 femmes) avaient realise un test H. Les facteurs significativement associes au depistage par RS en analyse multivariee par regression logistique apres remplacement des valeurs manquantes par imputation multiple sont indiques dans le tableau. Quelques facteurs etaient significativement associes au depistage par RS en analyse univariee mais pas en multivariee : âge eleve, etre retraite, avoir un conjoint ayant fait la RS, ne pas etre gene par la crainte d’avoir mal ou d’avoir a signer un consentement eclaire ou d’avoir a se deshabiller, considerer la RS comme utile, fiable, non desagreable, ou non dangereuse, considerer le test Hemoccult ® comme insuffisant et la RS comme ameliorant le depistage du CCR. Enfin, il n’y avait pas d’association significative entre depistage par RS et sexe, nationalite, niveau d’etudes, categorie socioprofessionnelle, couverture sociale, statut marital, tabagisme, antecedent familial de CCR et participation aux depistages des autres cancers. Conclusion Il ne semble pas y avoir de barriere sociodemographique au depistage du CCR par RS en France. La participation au depistage par RS peut y etre amelioree 1) en impliquant fortement les medecins generalistes pour informer et convaincre la population cible et 2) en inscrivant le controle de la douleur comme priorite lors de la realisation des RS et en communiquant sur ce theme aupres de la population et des medecins generalistes. Remerciements, financements, autres Travail realise grâce a une subvention de la Ligue contre le Cancer.