Compliance to company based colorectal cancer screening in Germany using fecal immunochemical test (FIT) - Results of Almost Half a Million FIT tests

Background Despite the existence of the statutory early cancer detection program in Germany and the removal of financial barriers, which is frequently reported in the literature to be the main obstacle in screening, uptake of colorectal cancer screening remains quite low. The campaign for colorectal cancer screening in German companies reported in this article started in 2010. It was initiated because of the low compliance with opportunistic public colorectal cancer screening efforts. Its goal is to improve participation by offering an organized screening program using a simple test (FIT). Methods An offer for company employees is publicized through posters, company newsletters and the intranet. The difference between the positivity rates of those who returned the kits within 20 days and later than 20 days was assessed using the Z-test. The average time between a positive result and colonoscopy was estimated using the Poincaré plot method. The positive predictive values were calculated for carcinomas, advanced adenomas or any lesions. The sensitivity and specificity of immoCare-C published by Vogel et al. and Hundt et al. were used to derive the confidence intervals for the positive likelihood ratio (for carcinoma and any kind of adenoma). Results A total of 312,147 kits were returned and analyzed (return rate 70.2%). 5.6% gave a positive result. The PPV for cancer aged between 55 and 74 was 4.6% for men and women (95% CI: 2.38%-6.76% and 1.28%7.99%, respectively), but 22% for men (95% CI: 17.93%-26.65%) and 8% for women (95%CI: 3.63%12.26%) for advanced adenomas. The PPV for any lesion was higher for those with familial risk (49.3%) and 42.6% for those without familial risk (95% CI: 40.2%-45.0%), but with overlapping confidence intervals. Conclusions The reported sample is not representative. Although, offering CRC screening in companies may be an effective way of increasing uptake in the target population. Differences in the test performance between men and women need further evaluation. DOI : 10.14302/issn.2471-7061.jcrc-17-1624 Corresponding author: Claudia Pieper, Institute for Medical Informatics, Biometry and Epidemiology, Head of Project Group Health Care Epidemiology, University Hospital of Essen, Hufelandstraße 55, 45122 Essen, Germany. Fax: +49-201-92239-333, Phone: +49-201-92239-239, e-mail: claudia.pieper@uk-essen.de Running title: Company-based colorectal cancer screening in Germany using FIT

Furthermore, the repeated longitudinal adherence to FOBT, which is crucial for its effectiveness (4), has been reported to be quite lower than one-time uptake (5)(6)(7).
Reasons to decline colonoscopy as an invasive screening go from "overstating the benefits" to "downstream effects and potential harms". Overall, opponents argue that most people undergoing screening are neither identified as having cancer nor protected from its developing (8).
A screening test must be well accepted.
However, uptake of screening could be improved by offering methods that are more acceptable. Adler et al. (2014) reported that of the 63% who refused colonoscopy, 97% accepted an alternative non-invasive screening method (FOBT) or a minimally invasive technique (blood test) (9). This refers to both men and women. Despite the fact that colonoscopy is the "golden standard" screening tool for colorectal cancer and its precursor lesions, its actual benefit is diminished due to low uptake (10).
In recent years, the FIT (fecal immunochemical test), which is also known as iFOBT (immune fecal occult blood test), has emerged as an alternative to the guaiacbased test. Despite the lack of evidence of its effect on CRC mortality, it has shown better sensitivity for detection of carcinomas and adenomas, and acceptable specificity (11)(12)(13) over gFOBT, whereby the latter was shown to reduce mortality in a large meta-analysis (14).
FIT has also been demonstrated to be more acceptable to the target population than gFOBT. In randomized studies, the response to invitations offering FIT screening was found to be up to 16% higher than that for gFOBT. This was attributed to the more acceptable collection method, the smaller number of fecal samples, and the lack of dietary restrictions (15)(16)(17)(18)(19). In addition, due to its quantitative nature, FIT hemoglobin cut-off levels can be tailored to the risk population and colonoscopy resources available.
In a randomized control trial (RCT) comparing FIT to colonoscopy as a primary screening tool, one-time FIT was non-inferior to colonoscopy in the detection of colorectal cancer, but had lower sensitivity for adenomas. This was viewed by the author to be a possible disadvantage in terms of CRC prevention as a goal of screening but also as a potential advantage in terms of reducing overdiagnosis and the accompanying unnecessary colonoscopy costs and complications. The difference in mortality will be assessed at the end of the ongoing 10-year trial. Overall, the colonoscopy complication rates were significantly lower in the FIT group than in the colonoscopy group, which is an important ethical consideration in terms of the "first do no harm" principle (10). The point has also been raised that large scale implementation of the FIT as a screening method may improve health care equity (20).
Various studies have found that participation in fecal immunochemical testing is higher than in colonoscopy procedures, which could improve the overall diagnostic yield beyond the mere sensitivity and specificity of the test (9, 10, 21).

Methods
An offer for company employees to have an FIT

Laboratory Analysis
All test kits were analyzed at the Care diagnostica laboratory on the day they were received using immoCARE-C tests with a cut-off value of 50 ng Hb/ml buffer.

Statistical Analysis
Return rates were monitored for the duration of

Results
We received a response from 444,888 individuals during the time of this study.  50% of the completed kits were returned within the first 24 days and 90% were returned within 81 days of when they were ordered ( Figure 2).
Of the total kits analyzed, 5.6% gave a positive result. The test positivity rates were higher for men, except in the below 40 age groups ( Figure 3) (p-value for total difference <0.0001).
The number of FIT requests, return rates, and the number and percentage of those who tested positive is shown in Table 1 by age and sex.
Those who returned the kits within 20 days were more likely to have a positive result than those who took longer to return them, with average positivity rates of 6% and 5.4%, respectively (p-value < 0.001).      (27).

Conclusions
Offering CRC screening in companies may be an effective way of increasing uptake in the target population. Differences in the test performance between men and women need further evaluation.   For the draft of the manuscript, the authors from outside the laboratory only took note of descriptive numbers and rates for further statistical analysis. No data sets, either anonymized or encoded, were referred.

List of Abbreviations
Therefore, it was not necessary to obtain approval by the Ethical Committee of the University of Duisburg-Essen.

Consent for publication
Availability of data and material In spite of anonymity, the datasets analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Funding
The study was funded by the CARE diagnostica Laborreagenzien GmbH. The analysis and the draft of the manuscript was funded by own resources of the corresponding authors' affiliation.

Authors' contributions
CP drafted the manuscript and performed the statistical analysis. EA performed the statistical analysis under the supervision of CP. DG participated in the data collection and the statistical analysis.
All authors have critically reviewed the content and have approved the final version submitted for publication.