| Summary: | The CIRSmedical is a publicly available error reporting and learning system of the German medical profession for reporting near misses and critical events. To improve learning from the CIRS reports, a systematic analysis of all reported cases would be desirable. This requires in the first instance that quality and completeness of the data are suitable for a systematic analysis. Therefore, in March 2017 we downloaded 100 CIRS cases with the numbers 144893 to 160395 as PDF from CIRSmedical.de and analyzed them considering five quality criteria of IDQA: completeness, value-added, comprehensibility, time/topicality, accessibility. To check the completeness, we determined how often information was missing in the reports. Value-added means if knowledge has been added to a report in the form of comments. For the comprehensibility, we checked if standard terminologies were used. We verified what kind of information existed at the time of analysis and the actuality of the data as well. Accessibility was checked by analyzing the retrieval of case reports. The evaluation of the criterion completeness shows that only one question was answered in all reports. Fifteen questions were answered in less than half the reports. This results from two different input concepts addressing different items. About 50% of the reports had no comments, a professional commentary was found in ases, and three reports had a comment from users. The reported cases are only available as PDF-files. According to our results CIRSmedical has limited suitability for systematic analysis. This is caused by the low completeness and by the limited accessibility of the data sets. One of the major reasons for these findings is that CIRSmedical.de receives data from different CIRS systems which address various questions in the input forms, leading to missing information. The case reports published in CIRSmedical are primarily meant for the analysis of the unique case report. Systematic analyzes are made even more difficult since there exist no interfaces for automatic access. The automatic and systematic analysis would important value and thus the acceptance of CIRSmedical could be improved. Consistent data management with a standardized and structured input can lead to improved data quality resulting in meaningful analysis of CIRS reports.
|
|---|
