Bmc Med Inform Decis
-
Bmc Med Inform Decis · Dec 2017
The development of a nursing subset of patient problems to support interoperability.
Since the emergence of electronic health records, nursing information is increasingly being recorded and stored digitally. Several studies have shown that a wide range of nursing information is not interoperable and cannot be re-used in different health contexts. Difficulties arise when nurses share information with others involved in the delivery of nursing care. The aim of this study is to develop a nursing subset of patient problems that are prevalent in nursing practice, based on the SNOMED CT terminology to assist in the exchange and comparability of nursing information. ⋯ To support the interoperability of nursing information, a national nursing subset of patient problems based on a terminology (SNOMED CT) has been developed. Using unambiguously defined patient problems is beneficial for clinical nursing practice, because nurses can then compare and exchange information from different settings. A key strength of this study is that nurses were extensively involved in the development process. Further research is required to link or associate nursing patient problems to concepts from a nursing classification with the same meaning.
-
Bmc Med Inform Decis · Dec 2017
Medical subdomain classification of clinical notes using a machine learning-based natural language processing approach.
The medical subdomain of a clinical note, such as cardiology or neurology, is useful content-derived metadata for developing machine learning downstream applications. To classify the medical subdomain of a note accurately, we have constructed a machine learning-based natural language processing (NLP) pipeline and developed medical subdomain classifiers based on the content of the note. ⋯ Our study shows that a supervised learning-based NLP approach is useful to develop medical subdomain classifiers. The deep learning algorithm with distributed word representation yields better performance yet shallow learning algorithms with the word and concept representation achieves comparable performance with better clinical interpretability. Portable classifiers may also be used across datasets from different institutions.