Analysis of risk factor domains in psychosis patient health records
Abstract Background Readmission after discharge from a hospital is disruptive and costly, regardless of the reason. However, it can be particularly problematic for psychiatric patients, so predicting which patients may be readmitted is critically important but also very difficult. Clinical narrative...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2019-10-01
|
| Series: | Journal of Biomedical Semantics |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1186/s13326-019-0210-8 |
| _version_ | 1819225392438837248 |
|---|---|
| author | Eben Holderness Nicholas Miller Philip Cawkwell Kirsten Bolton Marie Meteer James Pustejovsky Mei-Hua Hall |
| author_facet | Eben Holderness Nicholas Miller Philip Cawkwell Kirsten Bolton Marie Meteer James Pustejovsky Mei-Hua Hall |
| author_sort | Eben Holderness |
| collection | DOAJ |
| description | Abstract Background Readmission after discharge from a hospital is disruptive and costly, regardless of the reason. However, it can be particularly problematic for psychiatric patients, so predicting which patients may be readmitted is critically important but also very difficult. Clinical narratives in psychiatric electronic health records (EHRs) span a wide range of topics and vocabulary; therefore, a psychiatric readmission prediction model must begin with a robust and interpretable topic extraction component. Results We designed and evaluated multiple multilayer perceptron and radial basis function neural networks to predict the sentences in a patient’s EHR that are associated with one or more of seven readmission risk factor domains that we identified. In contrast to our baseline cosine similarity model that is based on the methodologies of prior works, our deep learning approaches achieved considerably better F1 scores (0.83 vs 0.66) while also being more scalable and computationally efficient with large volumes of data. Additionally, we found that integrating clinically relevant multiword expressions during preprocessing improves the accuracy of our models and allows for identifying a wider scope of training data in a semi-supervised setting. Conclusion We created a data pipeline for using document vector similarity metrics to perform topic extraction on psychiatric EHR data in service of our long-term goal of creating a readmission risk classifier. We show results for our topic extraction model and identify additional features we will be incorporating in the future. |
| first_indexed | 2024-12-23T10:08:52Z |
| format | Article |
| id | doaj.art-21cc23baec5a447ea9d864d8e56c4a7b |
| institution | Directory Open Access Journal |
| issn | 2041-1480 |
| language | English |
| last_indexed | 2024-12-23T10:08:52Z |
| publishDate | 2019-10-01 |
| publisher | BMC |
| record_format | Article |
| series | Journal of Biomedical Semantics |
| spelling | doaj.art-21cc23baec5a447ea9d864d8e56c4a7b2022-12-21T17:51:00ZengBMCJournal of Biomedical Semantics2041-14802019-10-0110111010.1186/s13326-019-0210-8Analysis of risk factor domains in psychosis patient health recordsEben Holderness0Nicholas Miller1Philip Cawkwell2Kirsten Bolton3Marie Meteer4James Pustejovsky5Mei-Hua Hall6Psychosis Neurobiology Laboratory, McLean Hospital, Harvard Medical SchoolPsychosis Neurobiology Laboratory, McLean Hospital, Harvard Medical SchoolPsychosis Neurobiology Laboratory, McLean Hospital, Harvard Medical SchoolPsychosis Neurobiology Laboratory, McLean Hospital, Harvard Medical SchoolBrandeis University Department of Computer ScienceBrandeis University Department of Computer SciencePsychosis Neurobiology Laboratory, McLean Hospital, Harvard Medical SchoolAbstract Background Readmission after discharge from a hospital is disruptive and costly, regardless of the reason. However, it can be particularly problematic for psychiatric patients, so predicting which patients may be readmitted is critically important but also very difficult. Clinical narratives in psychiatric electronic health records (EHRs) span a wide range of topics and vocabulary; therefore, a psychiatric readmission prediction model must begin with a robust and interpretable topic extraction component. Results We designed and evaluated multiple multilayer perceptron and radial basis function neural networks to predict the sentences in a patient’s EHR that are associated with one or more of seven readmission risk factor domains that we identified. In contrast to our baseline cosine similarity model that is based on the methodologies of prior works, our deep learning approaches achieved considerably better F1 scores (0.83 vs 0.66) while also being more scalable and computationally efficient with large volumes of data. Additionally, we found that integrating clinically relevant multiword expressions during preprocessing improves the accuracy of our models and allows for identifying a wider scope of training data in a semi-supervised setting. Conclusion We created a data pipeline for using document vector similarity metrics to perform topic extraction on psychiatric EHR data in service of our long-term goal of creating a readmission risk classifier. We show results for our topic extraction model and identify additional features we will be incorporating in the future.http://link.springer.com/article/10.1186/s13326-019-0210-8Natural language processingRisk predictionMachine learningElectronic health recordPsychotic disorders |
| spellingShingle | Eben Holderness Nicholas Miller Philip Cawkwell Kirsten Bolton Marie Meteer James Pustejovsky Mei-Hua Hall Analysis of risk factor domains in psychosis patient health records Journal of Biomedical Semantics Natural language processing Risk prediction Machine learning Electronic health record Psychotic disorders |
| title | Analysis of risk factor domains in psychosis patient health records |
| title_full | Analysis of risk factor domains in psychosis patient health records |
| title_fullStr | Analysis of risk factor domains in psychosis patient health records |
| title_full_unstemmed | Analysis of risk factor domains in psychosis patient health records |
| title_short | Analysis of risk factor domains in psychosis patient health records |
| title_sort | analysis of risk factor domains in psychosis patient health records |
| topic | Natural language processing Risk prediction Machine learning Electronic health record Psychotic disorders |
| url | http://link.springer.com/article/10.1186/s13326-019-0210-8 |
| work_keys_str_mv | AT ebenholderness analysisofriskfactordomainsinpsychosispatienthealthrecords AT nicholasmiller analysisofriskfactordomainsinpsychosispatienthealthrecords AT philipcawkwell analysisofriskfactordomainsinpsychosispatienthealthrecords AT kirstenbolton analysisofriskfactordomainsinpsychosispatienthealthrecords AT mariemeteer analysisofriskfactordomainsinpsychosispatienthealthrecords AT jamespustejovsky analysisofriskfactordomainsinpsychosispatienthealthrecords AT meihuahall analysisofriskfactordomainsinpsychosispatienthealthrecords |