Artikel
Automatically ICD code-based search in routine data for detecting adverse events in neurosurgery. Is the sensitivity of this method comparable to the “gold standard” morbidity & mortality conference?
Suche in Medline nach
Autoren
-
Lotta Zech
- Klinik für Neurochirurgie, Campus Lübeck, Universitätsklinikum Schleswig-Holstein, Lübeck
-
Volker M. Tronnier
- Klinik für Neurochirurgie, Campus Lübeck, Universitätsklinikum Schleswig-Holstein, Lübeck
-
Matteo M. Bonsanto
- Klinik für Neurochirurgie, Campus Lübeck, Universitätsklinikum Schleswig-Holstein, Lübeck
| Veröffentlicht: | 2. Juni 2015 |
|---|
Gliederung
Text
Objective: Morbidity and mortality conferences are a well – known and accepted tool to register quality in the health care system. Since 2006 the Department of Neurosurgery in Luebeck uses weekly M&M conferences for screening and registering adverse events. The registered data are further analyzed in comparison to the routine hospital information system (HIS) data.
Method: Based on routine data, the electronic surgical procedure catalogue is used for detecting all operations at the Department of Neurosurgery. During the M&M conference, these operations are screened for adverse events and registered in an independent database. In addition a search in routine data for specific ICD codes defined as surgical associated complications due to the Code (G 96.0 / G 97.8 and T 81 / T 89) from the hospital information system was performed.
Results: The data showed that during the period between 2006 - 2013 8018 operations were performed and 1349 complications were registered (16,8 %) during the M& M conference. The adverse events were categorized into three main predefined groups and their subcategories:
- 1.
- Neurosurgical complications (12,2 %; n = 979): not expected by the natural course of the disease but empirically known from literature as adverse event.
- 2.
- Neurosurgical complicated courses (4,2 %; n = 333): defined as complications expected by the pathology and/or the specific location of the lesion.
- 3.
- Medical complications (2,5 %; n = 202): non-surgical complications that require additional diagnostic and/or subsequent medical treatment.
Out of n = 1349 complications registered during M&M conferences only n = 482 were identified by routine data; thus 64,3 % could not be ascertained via routine data. Likewise there was a gap between detecting CSF fistula during M&M conferences (n = 152) in comparison to identifying CSF fistula by routine data (n = 90); 40,8 % were not singled out via routine data.
Conclusions: Screening adverse events based on operations performed and discussed in M&M conferences proved to be a reliable method to detect adverse events. The examples above underline the advantages of screening adverse events with M&M conferences compared to using routine data from the hospital information system. This aspect should carefully be considered due to the rising number of hospital quality reports based on hospital information system data.
