Data quality of the German screening colonoscopy registry.

BACKGROUND AND STUDY AIMS: The German screening colonoscopy program is accompanied by a central registry that records the main outcome quality indicators, namely colonoscopy completion rate, adenoma detection rate (ADR), and complication rate. The aim of the present study was to assess the quality of these registry data by comparing them with data from a prospective quality assurance study based on a self-reporting audit and patient feedback of screening colonoscopy. PATIENTS AND METHODS: The completeness of registry information was analyzed by comparing it with prospective data gathered by audit and patient feedback in a previous quality assurance study (ClinicalTrials.gov registration number: NCT00860665) between October 2006 and March 2008. The main outcome parameters were colonoscopy completion rate, ADR, and complication rate. Complications were recorded in three steps in the audit study using case report forms (immediate and subsequent documentation by physicians [CRF-1 and CRF-2], and patient follow-up [CRF-3]), but were documented in the registry without differentiation.
RESULTS: A total of 12 134 individuals (mean age 64.5 years; 47 % men) who underwent screening colonoscopy at 19 private practices in Berlin over the 18-month period were included in the audit study. Patient feedback was obtained for 90.1 %. A total of 12 150 cases had been recorded in the registry by the same private practices during the same period. Colonoscopy completion rate and ADR data were comparable in the audit study and registry (completion rate 98.2 % vs. 97.7 %; ADR 21.0 % vs. 20.5 %). However, compared with the registry data, the complication rate was 3.1-fold higher in the audit (0.46 % vs. 0.15 %; P < 0.001), and double (0.33 % vs. 0.15 %; P < 0.05) when patient feedback was not included.
CONCLUSIONS: Of the screening colonoscopy quality parameters, colonoscopy completion rate and ADR, but not complication rates, were reliably documented in the German national screening colonoscopy registry. Data on complications need to be appropriately standardized and audited in order to be used for credentialing and benchmarking purposes. © Georg Thieme Verlag KG Stuttgart · New York.