← Back to Research Papers

Propagating wellness: understanding the communication mechanisms of yoga and Wuqinxi (five animal frolics) via Lasswell's 5W framework.

Authors: Song Y, Lu L, Luo H
Journal: Frontiers in public health
mental health psychology open access

Abstract

Contemporary psychiatry operates in a substantially different environment from the one Karl Jaspers knew (). As the father of descriptive psychopathology, he aimed to systematically describe rather than explain psychopathology. This approach remains in modern psychiatric practice, though in contrast to Jaspers' time, psychiatrists generally use computers to document their findings. This digital transition has facilitated the use of electronic health records (EHRs) for research, but significant challenges persist. Research using real-world data has been constrained by the incomplete implementation of measurement-based care leading to the variable presence of psychometric instruments in the EHR (). Mental status examination (MSE) data can potentially address this gap by providing a complementary and routinely documented clinical data source that serves as the primary locus of clinically observed psychopathology in the clinical record. However, the potential utility of MSE data is constrained by the realities of clinical practice. Documented MSEs can vary in their quality and comprehensiveness, and the subjective nature of psychopathological assessment can impact interrater reliability (). Using MSE data also does not address the labor-intensive nature of information extraction from unstructured real-world data. Deep learning natural language processing (NLP) technologies enable the efficient extraction of relevant data from unstructured or semi-structured sources. Numerous studies have used NLP techniques to extract or organize symptom data from EHRs focusing either on specific psychiatric disorders or the extraction of transdiagnostic symptoms and signs (–). Fewer studies have focused explicitly on extracting data from the MSE, and schizophrenia patients were commonly studied (, , , , ). These studies have demonstrated how MSE data can be used to identify psychiatric disorder symptoms and signs and adverse outcomes such as increased healthcare resource utilization and costs in naturalistic settings (, ).