Colorectal cancer screening by colonoscopy and trends in disease-specific mortality: a population-based ecological study of 358 German districts.

Abstract

Screening for colorectal cancer (CRC) by colonoscopy was implemented in Germany in 2002. Although the procedure has proven to be effective in reducing disease-specific mortality in numerous clinical studies, its effect at the population level is unclear. We performed an ecological study at the level of 358 German districts, testing the hypothesis that a higher participation rate in screening colonoscopy is associated with a stronger decline in CRC mortality from 2001 to 2012. Information on the use of colonoscopy in each district was extracted from settlement data, used for the remuneration of physicians of the ambulant sector from 2008 to 2011. Yearly mortality rates for each district from 2001 to 2012 were derived from the official mortality statistics. A spatial model was fitted, considering other factors which might influence early detection of CRC (fecal occult blood test (FOBT), diagnostic colonoscopy, material and social area deprivation, and rural-urban disparities). The population-weighted mean annual participation rate during 2008-2011 in screening colonoscopy was 2.0% (range 0.6 to 3.9%). The weighted mean annual percentage change (APC) of CRC mortality was - 2.9% (range - 7.8 to 1.2%). According to the fully adjusted model, a 1% higher annual participation rate in colonoscopy screening was associated with an additional annual change in CRC mortality rate of - 0.34% (p = 0.015). Given an annual 2.0% participation rate, colonoscopy screening attributed 23% to the observed decline. Our real-world data from Germany provide first evidence that colonoscopy screening for CRC is effective in reducing disease-specific mortality at the population level.