BMC Medical Informatics and Decision Making最新文献

Background: Geriatric syndromes (GS) are complex conditions that affect older adults and often require multidisciplinary assessment. Natural language processing (NLP) has emerged as a promising tool for extracting relevant clinical information from unstructured text in electronic health records (EHRs). However, the application of NLP in detecting and monitoring GS remains an evolving area of research. This systematic review explores the role of NLP in the identification and analysis of GS, examining its applications, methodologies, and effectiveness. Furthermore, this review discusses the existing challenges, limitations, and future directions to advance NLP applications in the GS research.

Methods: We conducted a systematic literature search across ten databases to identify studies that applied NLP to GS detection. Articles were screened using predefined inclusion and exclusion criteria, and relevant studies were evaluated for quality using PROBAST. Data were extracted on study characteristics, datasets, annotation processes, NLP approaches, performance metrics, population demographics, and clinical applications. A PRISMA flow diagram was used to illustrate the study selection process.

Results: A total of 65 studies were included, where the majority of the studies used traditional rule-based and machine learning approaches. Publicly available datasets were scarce, and most studies used their private dataset, leading to significant variability in data sources and formats. Annotation methodologies differed across studies, with minimal shared guidelines or standards, making direct comparisons challenging. Performance metrics varied across syndromes, with F1-score, precision, and recall as the most commonly reported. Key challenges included the lack of dataset uniformity, differences in annotation practices, and the absence of external validation.

Conclusion: NLP has shown potential in GS analysis, particularly for the detection of syndromes and epidemiological research. However, the majority of studies only focused on one syndrome, and variability in dataset availability, annotation processes, and model performance present challenges to broader implementation. Future research should focus on improving the comprehensiveness of GS identification, dataset standardisation, enhancing model generalisability, and integrating NLP approaches into clinical workflows.

Background: Our study utilized patient decision aids (PDA) to explore the influence of shared decision making (SDM) on type 2 diabetes patients' intention to use subcutaneous antidiabetic agents.

Methods: A prospective observational comparative study was conducted involving 249 patients with type 2 diabetes who were referred by physicians and subsequently interviewed by pharmacists across different clinics. Patients were classified into two parallel groups based on routine clinical practice. A patient decision aid (PDA) entitled "Type 2 Diabetes: Oral or Subcutaneous Antidiabetic Agents for My Diabetes Control" was developed for this study. Data collection focused on patients' intention to initiate subcutaneous antidiabetic agents, post-test knowledge scores, and satisfaction with clinical visits. These outcomes were compared between the SDM group, which received pharmacist-facilitated shared decision making using a PDA, and the PDA self-completion (control) group, which received usual care and completed the PDA by self-review.

Results: A total of 249 patients were included (SDM group n = 123; control group n = 126). The SDM group had higher baseline disease severity indicators. After adjustment, the SDM group demonstrated higher decision-making (adjusted mean difference 0.59, 95% CI 0.17-1.00), knowledge (0.84, 95% CI 0.59-1.09), and satisfaction (1.86, 95% CI 0.98-2.73) scores compared with controls.

Conclusions: PDA-supported shared decision making was associated with higher short-term decision-making intentions, knowledge, and satisfaction among patients with type 2 diabetes. These findings suggest that PDAs may support patients in considering injectable treatment options, although conclusions regarding treatment initiation or clinical outcomes require further longitudinal evaluation.

Background: Single sign-on (SSO), tap-on-tap-off (TOTO) and virtual desktops (VD) are increasingly being used in hospitals to reduce the burden and risks of password management, but we know little about how these systems are used in practice. We aimed to understand whether, and if so, how and why a TOTO and VD solution supported clinical work in an emergency department (ED).

Methods: Qualitative descriptive design comprising interviews (n = 17) and work observations (~ 7 h) with doctors and nurses in an Australian ED. Data collection and analysis were informed by the unified theory of acceptance and use of technology (UTAUT).

Results: Some participants perceived the TOTO-VD system as useful, however, most questioned the value of the system for ED work. Time to login was relatively long for the needs of the ED, which contributed to the system not being used as intended. Workarounds led to users being automatically logged out mid-task, which further contributed to clinicians' negative experience with the system. There appeared to be limited use for, and potentially new risks introduced with a VD solution, where ED clinicians perceived there to be redundant control of access to computers.

Conclusions: The ED work context presented barriers to TOTO-VD adoption. The system did not support clinical work in an ED and so was worked around by clinicians. We recommend examining and understanding clinician work, including when and how computers are used and what systems are accessed, prior to implementation of any technological solution.

Clinical trial number: Not applicable.

Background: Preeclampsia is a multifactorial hypertensive disorder of pregnancy with heterogeneous clinical presentation and outcomes. Current classification systems inadequately capture the substantial clinical heterogeneity of preeclampsia. This study aimed to identify and characterize distinct phenotypic clusters of the disease using preterm birth-related risk factors.

Methods: This retrospective cohort study included 4,132 singleton pregnancies with preeclampsia between January 2014 and December 2024. We performed unsupervised k-means clustering based on 22 preterm birth-related clinical and biochemical variables. Maternal characteristics and pregnancy outcomes were compared across the identified clusters. Random forest algorithm was employed to evaluate feature importance.

Results: Five distinct clinical clusters were identified. Cluster 1 (N = 740) exhibited metabolic-inflammatory features. Cluster 2 (N = 1090) was characterized by multiple comorbidities, including pre-gestational diabetes (39.2%) and fetal growth restriction (47.8%). Cluster 3 (N = 732) featured a heightened systemic inflammatory response. Cluster 4 (N = 707) presented with the most benign clinical features. Cluster 5 (N = 863) was distinguished by a high prevalence of uterine mechanical factors. These clusters demonstrated significant differences in preterm birth rates and other adverse pregnancy outcomes (P < 0.05). Random forest analysis confirmed the discriminative power of the selected variables.

Conclusion: Cluster analysis reveals five clinically distinct preeclampsia phenotypes with direct implications for preterm birth risk. This refined classification provides a foundation for investigating distinct pathophysiological mechanisms and supports the development of individualized management strategies.