Automated identification provides an efficient way to categorize patient safety incidents. Previous studies have focused on identifying single incident types relating to a specific patient safety problem, e.g., clinical handover. In reality, there are multiple types of incidents reflecting the breadth of patient safety problems and a single report may describe multiple problems, i.e., it can be assigned multiple type labels. This study evaluated the abilty of multi-label classification methods to identify multiple incident types in single reports. Three multi-label methods were evaluated: binary relevance, classifier chains and ensemble of classifier chains. We found that an ensemble of classifier chains was the most effective method using binary Support Vector Machines with radial basis function kernel and bag-of-words feature extraction, performing equally well on balanced and stratified datasets, (F-score: 73.7% vs. 74.7%). Classifiers were able to identify six common incident types: falls, medications, pressure injury, aggression, documentation problems and others.
|Title of host publication||MEDINFO 2017|
|Subtitle of host publication||Precision Healthcare through Informatics - Proceedings of the 16th World Congress on Medical and Health Informatics|
|Editors||Adi V. Gundlapalli, Marie-Christine Jaulent, Dongsheng Zhao|
|Place of Publication||Amsterdam|
|Number of pages||5|
|Publication status||Published - 2017|
|Event||16th World Congress of Medical and Health Informatics: Precision Healthcare through Informatics, MedInfo 2017 - Hangzhou, China|
Duration: 21 Aug 2017 → 25 Aug 2017
|Name||Studies in Health Technology and Informatics|
|Conference||16th World Congress of Medical and Health Informatics: Precision Healthcare through Informatics, MedInfo 2017|
|Period||21/08/17 → 25/08/17|
Bibliographical noteCopyright the International Medical Informatics Association (IMIA) and IOS Press 2017. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.
- Machine learning
- Patient safety
- Risk management