Aeri Kim

BACKGROUND Population aging has contributing to increase in the number of patients with cancer and chronic progressive illnesses, making hospice and palliative care indispensable for improving end of life care and quality of life. Although home healthcare plays a pivotal role in initiating hospice transition discussions, the appropriate timing remains clinically challenging. Existing prediction approaches rely on structured clinical variables that poorly capture the contextual and communicative factors emerging during consultations. Although free-text consultation records may capture such information, no study has examined whether linguistic markers from these consultations predict subsequent hospice transitions. OBJECTIVE This study examined whether clinical and text-derived linguistic markers from the first home healthcare consultation were associated with subsequent hospice transitions, aiming to identify early signals that may inform hospice referral discussions. METHODS This study used prospectively collected data from 37 patient–caregiver pairs receiving home healthcare at a tertiary hospital in South Korea, yielding 149 recorded consultations between December 2024 and November 2025. The consultations were transcribed using Naver CLOVA Speech, de-identified, and corrected using Gemma 3 (12 B), an on-premise large-language model selected by benchmarking five open-weight models against a human expert-corrected gold standard. A hospice terminology dictionary (4 categories, 34 subcategories, and 1327 terms) was used to derive term occurrence and frequency, and clinical variables were extracted from the hospital clinical data warehouse. Only variables from the first consultation served as candidate predictors; the outcome was a subsequent hospice transition, ascertained from electronic medical records. The analyses used Firth logistic regression, receiver-operating characteristic analysis with DeLong tests, and generalized estimating equations. RESULTS Thirty-seven patients underwent 149 consultations and 10 (27%) transitioned to hospice care. The baseline characteristics did not differ between the groups. The presence of Management of Other Systemic Symptoms terms was lower in the transition group and associated with hospice transition in univariate Firth regression (OR=0.08, 95% CI=0.01-0.54; P=.009). In the primary cancer subgroup (n=28), this association remained significant (OR=0.12, 95% CI=0.01-0.83; P=.03), while cancer status was excluded from the multivariable model due to quasi-complete separation. The combined clinical–text models demonstrated modest discrimination (area under the curve=0.70 in the cancer subgroup) with no significant model differences. Generalized estimating equations analysis showed a markedly lower occurrence in the transition group at baseline (OR=0.09, 95% CI=0.02-0.39; P=.002) and a significant time × group interaction (OR=1.67, 95% CI=1.02-2.72; P=.04). CONCLUSIONS Reduced use of symptom management language at the first consultation was associated with a subsequent hospice transition. Because anticipating the timing of hospice transition is challenging, a signal derived from the initial consultation alone may be more clinically applicable than one requiring repeated encounters. Thus, linguistic markers in consultation texts may serve as early low-burden signals for hospice transition.

Objectives: This study aimed to identify high‐risk factors for type 2 diabetes and develop a machine learning (ML)–based diabetes prediction model using nursing notes from home health care. Design: Retrospective cohort study. Sample: A total 7896 medical records from 1747 patients aged ≥ 20 years who received home health care at a university hospital in South Korea over 10 years. Measurements: The patient's sociodemographic characteristics, diagnosis codes, and clinical history as well as narrative nursing notes detailing nursing assessments and descriptions of the patient's condition were recorded. Logistic regression was used to identify risk factors, and five machine learning models were constructed using 10‐fold cross‐validation. Results: Female sex, depression, hypertension, hyperlipidemia, nursing services such as dressing and nutritional care, and dysuria were identified as prediabetes‐related symptoms extracted from nursing notes using natural language processing (NLP). The Random Forest model demonstrated the highest predictive performance (AUC = 0.985). Conclusions: Nursing documentation is a valuable resource for early diabetes risk screening. High predictive performance was achieved by integrating structured and unstructured data, prediabetes‐related symptoms extracted from nursing notes. This approach enhances nurses’ ability to deliver timely, personalized interventions and improve preventive care outcomes, with significant benefits for patients receiving home care who require continuous support.

Background: Palliative care enhances dignity and quality of life for patients with serious illnesses by managing distressing symptoms and supporting families. However, inadequate awareness and misconceptions often hinder patients and their families from accessing these services. Understanding of public perspectives on palliative and hospice care can facilitate the development of targeted educational resources and awareness campaigns. As social media becomes an important source of health information, analyzing such publicly available web-based sources can yield valuable insights into perceptions of palliative and hospice care. Objective: This study analyzed public perspectives posted on a popular social media platform in South Korea to understand perceptions, challenges, needs, and sentiments related to palliative and hospice care. Methods: Data were collected from Naver Knowledge iN, a popular web-based public forum in South Korea, encompassing 34,501 texts posted between 2002 and 2024. After applying inclusion and exclusion criteria, 9072 relevant perspectives were analyzed. Contextualized topic modeling was used to identify themes, and the optimal model was selected based on coherence, diversity scores, and expert feedback. In addition, multiclass sentiment analysis using a fine-tuned Korean bidirectional encoder representations from transformers (KoBERT) model classified sentiments into 6 categories. The multiclass sentiment model’s performance was evaluated using accuracy, precision, recall, and F1-score. Results: Social media discussions on palliative and hospice care have increased steadily over time. Topic modeling identified 9 themes, with “ethical and legal concerns” and “medical care in hospitals” peaking in recent years, suggesting growing public interest in these areas. “Emotional and psychological support” emerged as the predominant theme, reflecting a significant need for psychosocial assistance among patients and their families. Sentiment analysis revealed that “sadness,” “anxiety,” and “neutral” were common emotions across many topics. Notably, themes such as “emotional and psychological support,” “disease treatment outcomes and prognosis,” “medical care in hospitals,” “financial issues,” and “symptom management” were predominantly associated with “sadness,” while “administrative and volunteer services,” “ethical and legal concerns,” and “nutrition management” were more closely linked with “anxiety.” Conclusions: This study highlights public concerns about palliative and hospice care in South Korea, including ethical dilemmas, caregiving burden, and emotional distress. Findings underscore the need for communication strategies that address informational, emotional, and psychological needs. Culturally sensitive, interactive communication tools, such as artificial intelligence–powered chatbots and public education campaigns, may help dispel misconceptions and promote timely, informed decisions about palliative and hospice care.

Palliative care ensures patients’ dignity and quality of life while supporting families during serious illness. However, barriers such as inadequate awareness hinder access to these services. This study analyzed 9,147 perspectives about palliative and hospice care posted on Naver Knowledge iN, a popular online platform in South Korea, using contextualized topic modeling and multi-class sentiment analysis (SA). Nine major themes were identified. SA showed that “sadness” was the most common emotion, followed by “neutral” and “anxiety.” Notably, negative emotions were closely tied to “emotional and psychological support” theme. The KoBERT model used for SA achieved an accuracy of 0.73 and F1-score of 0.72. These findings highlight the need to address misconceptions and enhance public awareness of palliative care. Recommendations include creating targeted educational resources, implementing proactive screening for emotional distress, and establishing accessible communication platforms.

Palliative and hospice care, traditionally provided in hospitals, are expanding into homes and communities with an increase in the elderly population and number of patients with chronic diseases. Despite the development of various mobile applications, the characteristics and relevance of these technologies in community settings have been insufficiently analyzed, resulting in a lack of clarity regarding their effectiveness. This systematic scoping review aimed to analyze and categorize the characteristic features of mobile applications for palliative and hospice care, and propose key attributes that should be included in community setting based applications. Sixty-six mobile applications for palliative and hospice care were included through mobile platforms (Google Play and the iOS Appstore). Of the total number of palliative and hospice care applications, 44 were intended for use by clinicians and 14 by patients and informal caregivers, respectively The application features can be categorized into eight themes: 1) information and procedures for palliative care and hospice facilities, 2) visiting management, 3) medical record management and documentation, 4) palliative and hospice care guidance and education, 5) care planning management, 6) communication and notification systems, 7) administrative services for facilities, and 8) data management. Determining whether the current applications are clinically effective is challenging. Future palliative care˚uhospice applications should integrate existing disparate functions to provide users with a one-stop solution for palliative and hospice care-related services and allow for care beyond the in-hospital setting, thus providing patients with more end-of-life care options

Background and Objectives: Cognitive function is crucial for older adults’ independence and quality of life. This study investigated the longitudinal relationship between fear of falling and cognitive function using data from the Korean Longitudinal Study of Ageing. Specifically, it examined whether social engagement (regular informal contact and social activities) and physical activity mediate the impact of fear of falling on cognitive function, informing targeted strategies to mitigate cognitive decline. Research Design and Methods: This study analyzed seven waves (2006 to 2018) of a nationally representative longitudinal survey involving 3,453 older adults in Korea. To minimize omitted variable bias, the research applied ordinary least squares models with lagged dependent variables and conducted Sobel-Goodman mediation tests for additional insights. Results: Developing a fear of falling over 2 years is negatively associated with cognitive function. Additionally, although fear of falling negatively impacts social activities and informal contacts, it does not significantly alter physical activity. Mediation analyses show that reduced social activities account for 31% of the cognitive decline associated with fear of falling, compared to 15% for informal social contacts. These findings highlight active social participation as a key factor in protecting cognitive health. Physical activity did not mediate this relationship. Discussion and Implications: Interventions should prioritize enhancing social engagement to mitigate cognitive decline related to fear of falling. Facilitating active community participation can effectively support cognitive health and overall well-being in aging populations.

Background: Daily preferences are a reflection of how adults wish to have their needs and values addressed, contributing to joy and satisfaction in their daily lives. Clinical settings often regard older adults as a uniform group, neglecting the diversity within this population, which results in a shortfall of person-centered care that overlooks their distinct daily care preferences. At the heart of person-centered care lies the imperative to comprehend and integrate these preferences into the care process. Recognizing and addressing gender differences in older adults is critical to customizing care plans, thereby optimizing quality of life and well-being for individuals. This study addresses the need to understand the diverse daily care preferences of adults, particularly among older populations, who represent a growing demographic with unique needs and interests. Objective: This study aims to identify and analyze the key themes and daily care preferences from unstructured adult text narratives with a focus on uncovering gender-specific variations. Methods: This study used 4350 deidentified, unstructured textual data from MyDirectives (MyDirectives, Inc), an interactive online platform. Advanced topic modeling techniques were used to extract meaningful themes, and gender-specific term frequency and distribution were examined to identify gender differences in these elements. Results: The study sample included 2883 women (mean age 63.02, SD 13.69 years) and 1467 men (mean age 67.07, SD 11.73 years). Our analysis identified six major themes: (1) “entertainment” (12.14%, 528/4350), (2) “music” (10.39%, 452/4350), (3) “personal interests and memories” (38.18%, 1661/4350), (4) “intimate relationships” (14.92%, 649/4350), (5) “natural comforts” (16.18%, 704/4350), and (6) “emotional, cultural, and spiritual foundations” (8.18%, 356/4350). Gender differences were evident: women were more likely to express preferences for “personal interests and memories” (40.7% vs 33.3%), “natural comforts” (18.4% vs 11.9%), and “emotional and spiritual foundations” (9.3% vs 6.1%) than men. Men expressed stronger preferences for “entertainment” (18.1% vs 9.1%) and “music” (16.8% vs 7.2%). Common terms across all participants included “dog,” “love,” “friends,” and “book.” Notably, the study revealed significant gender differences in daily care preferences, especially regarding familial relationships and entertainment choices. Conclusions: The findings underscore the importance of recognizing individual daily care preferences in person-centered care, particularly regarding gender. Understanding these preferences is crucial for improving care quality and patient satisfaction, thereby enhancing the overall quality of life for adults receiving care across our health care system.

Aim: To identify the factors affecting the length of stay (LOS) and discharge destination (DD) of home health care (HHC) patients in South Korea.Methods: A retrospective cross‐sectional study was conducted using the electronic health records of 1769 patients from a hospital in South Korea. Data were collected from January 2013 to December 2022. We categorized the independent variables into patient context, structure, and process factors following a modification of Donabedian's model. Hierarchical and multinomial logistic regression analyses were used.Results: The mean length of stay was 26.41 days. Patients were discharged to the following locations: 35.0% continued HHC, 21.0% died, 19% were discharged to their homes, 17.0% were admitted, and 8.0% were sent to other locations. Patients' sex, type of insurance coverage, and primary caregiver as well as the number of nurse visits, HHC admission route, and type of nursing service were predictors of their LOS. Operation history, a high Charlson comorbidity index, the type of insurance coverage, HHC admission route, and certain nursing care services were associated with admission and death as the DD.Conclusions: Process variables (e.g., number of nurse visits, HHC admission route, type of nursing services) have a considerable influence on determining the LOS and DD of HHC patients. This result provides new insights into the use of HHC services and care transitions out of the hospital for patients living in their home, offering evidence to reduce unnecessary readmissions and ensure more effective and efficient HHC.

BACKGROUND Palliative care enhances dignity and quality of life for patients with serious illnesses by managing distressing symptoms and supporting families. However, inadequate awareness and misconceptions often hinder patients and their families from accessing these services. Understanding of public perspectives on palliative and hospice care can facilitate the development of targeted educational resources and awareness campaigns. As social media becomes an important source of health information, analyzing such publicly available web-based sources can yield valuable insights into perceptions of palliative and hospice care. OBJECTIVE This study analyzed public perspectives posted on a popular social media platform in South Korea to understand perceptions, challenges, needs, and sentiments related to palliative and hospice care. METHODS Data were collected from Naver Knowledge iN, a popular web-based public forum in South Korea, encompassing 34,501 texts posted between 2002 and 2024. After applying inclusion and exclusion criteria, 9072 relevant perspectives were analyzed. Contextualized topic modeling was used to identify themes, and the optimal model was selected based on coherence, diversity scores, and expert feedback. In addition, multiclass sentiment analysis using a fine-tuned Korean bidirectional encoder representations from transformers (KoBERT) model classified sentiments into 6 categories. The multiclass sentiment model’s performance was evaluated using accuracy, precision, recall, and F1-score. RESULTS Social media discussions on palliative and hospice care have increased steadily over time. Topic modeling identified 9 themes, with “ethical and legal concerns” and “medical care in hospitals” peaking in recent years, suggesting growing public interest in these areas. “Emotional and psychological support” emerged as the predominant theme, reflecting a significant need for psychosocial assistance among patients and their families. Sentiment analysis revealed that “sadness,” “anxiety,” and “neutral” were common emotions across many topics. Notably, themes such as “emotional and psychological support,” “disease treatment outcomes and prognosis,” “medical care in hospitals,” “financial issues,” and “symptom management” were predominantly associated with “sadness,” while “administrative and volunteer services,” “ethical and legal concerns,” and “nutrition management” were more closely linked with “anxiety.” CONCLUSIONS This study highlights public concerns about palliative and hospice care in South Korea, including ethical dilemmas, caregiving burden, and emotional distress. Findings underscore the need for communication strategies that address informational, emotional, and psychological needs. Culturally sensitive, interactive communication tools, such as artificial intelligence–powered chatbots and public education campaigns, may help dispel misconceptions and promote timely, informed decisions about palliative and hospice care.

BACKGROUND Daily preferences are a reflection of how adults wish to have their needs and values addressed, contributing to joy and satisfaction in their daily lives. Clinical settings often regard older adults as a uniform group, neglecting the diversity within this population, which results in a shortfall of person-centered care that overlooks their distinct daily care preferences. At the heart of person-centered care lies the imperative to comprehend and integrate these preferences into the care process. Recognizing and addressing gender differences in older adults is critical to customizing care plans, thereby optimizing quality of life and well-being for individuals. This study addresses the need to understand the diverse daily care preferences of adults, particularly among older populations, who represent a growing demographic with unique needs and interests. OBJECTIVE This study aims to identify and analyze the key themes and daily care preferences from unstructured adult text narratives with a focus on uncovering gender-specific variations. METHODS This study used 4350 deidentified, unstructured textual data from MyDirectives (MyDirectives, Inc), an interactive online platform. Advanced topic modeling techniques were used to extract meaningful themes, and gender-specific term frequency and distribution were examined to identify gender differences in these elements. RESULTS The study sample included 2883 women (mean age 63.02, SD 13.69 years) and 1467 men (mean age 67.07, SD 11.73 years). Our analysis identified six major themes: (1) “entertainment” (12.14%, 528/4350), (2) “music” (10.39%, 452/4350), (3) “personal interests and memories” (38.18%, 1661/4350), (4) “intimate relationships” (14.92%, 649/4350), (5) “natural comforts” (16.18%, 704/4350), and (6) “emotional, cultural, and spiritual foundations” (8.18%, 356/4350). Gender differences were evident: women were more likely to express preferences for “personal interests and memories” (40.7% vs 33.3%), “natural comforts” (18.4% vs 11.9%), and “emotional and spiritual foundations” (9.3% vs 6.1%) than men. Men expressed stronger preferences for “entertainment” (18.1% vs 9.1%) and “music” (16.8% vs 7.2%). Common terms across all participants included “dog,” “love,” “friends,” and “book.” Notably, the study revealed significant gender differences in daily care preferences, especially regarding familial relationships and entertainment choices. CONCLUSIONS The findings underscore the importance of recognizing individual daily care preferences in person-centered care, particularly regarding gender. Understanding these preferences is crucial for improving care quality and patient satisfaction, thereby enhancing the overall quality of life for adults receiving care across our health care system. CLINICALTRIAL