Background: Approximately 60% of patients who have a stroke are discharged to their homes, after which they require ongoing community-based rehabilitation to support long-term recovery. Transitional care services (TCS) serve as an intervention aimed at reducing hospital readmissions by maintaining treatment continuity, either when patients transition between different levels of care within the same setting or when they move to a new care environment. Objectives: The purpose of this review was to evaluate the efficacy of TCS in reducing readmission rates, enhancing functional outcomes, improving self-efficacy, and boosting the quality of life for patients who had a stroke while also alleviating caregiver burden. Methods: To review randomized clinical trials performed between January 2014 and December 2023, we adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for reporting systematic reviews. We established a search strategy combining “transitional care,” “stroke,” “cerebrovascular disorders,” “patient discharge,” and “continuity of patient care” and used the PubMed, Embase, Cochrane Library, and CINAHL databases. We selected readmission rate and functional status as the primary outcomes and quality of life, self-efficacy, and care burden as secondary outcomes. We conducted a meta-analysis for each outcome indicator where quantitative synthesis was possible. For the meta-analysis, we used RevMan 5.4 and performed subgroup analyses according to intervention intensity, postdischarge intervention method, and intervention duration. Results: In the meta-analysis for each outcome indicator, we observed that TCS was effective in reducing the readmission rate of individuals living with stroke and improving their functional status. Additionally, patients with stroke who received TCS showed better quality of life, improved self-efficacy, and reduced caregiver burden compared to those who received usual care. Subgroup analyses revealed that patients with stroke who received moderate or higher intensity of TCS for 2 months demonstrated improvements in functional status compared to those who received lower intensity or 3-month interventions. Discussion: This is the first known systematic review to synthesize studies on the efficacy of TCS for patients with stroke over the past 10 years, examining outcome indicators such as readmission rates and functional status. Based on subgroup analysis, we derived an effective TCS strategy (e.g., intensity and duration) to improve functional status.

Background/Objectives: Gastric cancer is a leading cause of cancer-related mortality, particularly in East Asia, with a notable burden in Republic of Korea. This study aimed to construct and develop machine learning models for the prediction of gastric cancer mortality and the identification of risk factors. Methods: All data were acquired from the Korean Clinical Data Utilization for Research Excellence by multiple medical centers in South Korea. A total of 23,717 gastric cancer patients were divided into two groups by cause of mortality (all-cause of 2664 and disease-specific of 1620) and investigated. We used comprehensive data integrating clinical, pathological, lifestyle, and socio-economic factors. Cox proportional hazards analysis was conducted to estimate hazard ratios for mortality. Five machine learning models (random forest, gradient boosting machine, XGBoost, light GBM, and cat boosting) were developed to predict mortality. The models were interpreted by SHAP, one of the explainable AI techniques. Results: For all-cause mortality, the gradient-boosting machine learning model demonstrated the highest performance with an AUC-ROC of 0.795. For disease-specific mortality, the light GBM model outperformed others, achieving an AUC-ROC of 0.867. Significant predictors included the AJCC7 stage, tumor size, lymph node count, and lifestyle factors such as smoking, drinking, and diabetes. Conclusions: This study underscores the importance of integrating both clinical and lifestyle data to enhance mortality prediction accuracy in gastric cancer patients. The findings highlight the need for personalized treatment approaches in the Korean population and emphasize the role of demographic-specific data in predictive modeling.

Colorectal cancer (CRC) prognosis varies significantly, yet conventional statistical models struggle to capture the complex, non-linear interactions among clinical variables. Furthermore, most predictive models are based on Western populations, limiting their applicability to Korean patients. This study aimed to develop an explainable AI (XAI) model for CRC mortality prediction using a nationwide Korean cohort to provide clinically actionable insights. We conducted a retrospective cohort study using the Korean Cancer Public Library Database. A total of 9,069 patients with CRC were analyzed for all-cause mortality (1,878 deaths) and 8,589 patients for CRC-specific mortality (1,398 deaths). Four ML algorithms-support vector machine, random forest, XGBoost, and LightGBM-were constructed. We employed explainable AI techniques, including SHapley Additive exPlanations (SHAP), to quantify the contribution of each predictor and ensure model interpretability. All models showed good discrimination (AUC: 0.82-0.94). LightGBM was presented as the best-optimized model with an AUC of 0.824 [95% CI 0.80-0.85] in all-cause mortality. For CRC-specific mortality, LGB again yielded the AUC of 0.867 [95% CI 0.84-0.89]. SHAP revealed tumor stage and carcinoembryonic antigen as top mortality predictors across ages. Metabolic markers (e.g., hypertension, cholesterol) and liver enzymes were more predictive in younger patients. We developed the first interpretable machine learning model that accurately predicts CRC survival in a nationwide Korean cohort. Age-specific risk factors identified by SHAP not only support personalized care but also advance the application of precision oncology in Asian settings.

Background: Cancer survivors frequently face persistent nutrition-related challenges after treatment. Mobile health tools may extend access to dietary self-management support beyond clinic settings, but feasibility and preliminary effects remain insufficiently characterized in this population. Objective: This study aims to evaluate the feasibility, user engagement, and preliminary effects of a 4-week mobile nutrition app on dietary behavior and quality of life (QoL) among cancer survivors, and to explore whether higher engagement is associated with greater improvements. Methods: A single-arm, prospective pilot feasibility study was conducted at a tertiary cancer center in Korea. Participants used a mobile nutrition app that provided dietary feedback and self-monitoring features. In-app log data were analyzed to determine engagement metrics (session frequency, duration, and gap regularity) using an elbow-based 10-minute session threshold. Primary outcomes included the Nutrition Quotient for Adults (NQ-2021) and European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) scores, measured before and after 4 weeks. Nonparametric paired analyses assessed changes, and exploratory correlations examined relationships between engagement and outcomes. Results: Among 27 enrolled participants, 24 cancer survivors (88.9%) completed the intervention and postassessment; the majority were female (17/24, 70.8%), with a mean age of 58.5 (SD 8.7) years. Breast cancer was the most common diagnosis (11/24, 45.8%), and most participants reported no diet-related adverse effects (20/24, 83.3%) and stable body weight during the study period. Participants averaged 2.3 app sessions per day and a median cumulative use of 177.5 minutes. Retention was 88.9%, and median adherence to daily self-monitoring exceeded 85%. The NQ Moderation domain improved significantly (mean 76.6, SD 17.5 → mean 81.0, SD 13.7; P =.02), while Balance and Dietary Behavior showed positive trends (mean 63.7, SD 16.1 → mean 65.5, SD 13.3, P =.14; mean 64.9, SD 17.3 → mean 67.1, SD 15.9, P =.10). In QoL outcomes, appetite loss decreased (mean 17.9, SD 22.0 → mean 7.7, SD 14.6; P =.03) and global health status increased modestly (mean 68.5, SD 21.4 → mean 72.9, SD 20.1; P =.08). Higher engagement correlated with improved moderation ( r =0.46; P =.02) and reduced appetite loss ( r =–0.42; P =.04). Exploratory subgroup analyses suggested stronger effects among participants aged 60 years and older (ΔNQ Moderation +7.9; P= .04) and those with longer cancer survivorship (>3 years; P= .047). No adverse events were reported. Conclusions: This pilot feasibility study demonstrates high user engagement, satisfactory retention, and preliminary improvements in nutritional behavior and QoL among cancer survivors using a mobile nutrition app. These findings indicate the feasibility of a larger controlled trial to confirm the app’s effectiveness and explore long-term adherence strategies.

Kwon OB, et al. J Korean Med Sci. 2026 Dec;41:e325. https://doi.org/10.3346/jkms.2026.41.e325

Objectives: Ischemic stroke is influenced by long-term metabolic and renal deterioration; however, many risk prediction frameworks rely on single time-point measurements. We examined whether multi-period patterns in national health screening indicators are associated with incident ischemic stroke in Korea.Methods: Using customized National Health Insurance Service data with 3 biennial screenings (P1: 2013–2014; P2: 2015–2016; P3: 2017–2018), we identified incident ischemic stroke during 2019–2023 (Korean Standard Classification of Diseases-7 I63). After applying eligibility criteria and excluding individuals with missing screening values, we performed 1:1 propensity score matching on sex, 1-year age strata, and insurance type (97 454 matched pairs; n=194 908). Multi-period indicators included waist circumference increase ≥10%, sustained blood pressure ≥130/80 mmHg, sustained fasting glucose ≥126 mg/dL, proteinuria progression, and creatinine elevation in ≥2 periods (sex-specific thresholds). Associations were evaluated using conditional logistic regression; a comparator model used P3-only indicators.Results: In the multi-period model, stroke was associated with waist circumference increase ≥10% (odds ratio [OR], 1.05; 95% confidence interval [CI], 1.01 to 1.08), sustained blood pressure ≥130/80 mmHg (OR, 1.34; 95% CI, 1.31 to 1.37), sustained fasting glucose ≥126 mg/dL (OR, 1.66; 95% CI, 1.60 to 1.73), creatinine elevation in ≥2 periods (OR, 1.08; 95% CI, 1.06 to 1.10), and proteinuria progression (OR, 1.36; 95% CI, 1.32 to 1.39). In the P3-only model, all single-time-point indicators were associated with incident stroke (ORs, 1.08 to 1.47).Conclusions: Multi-year patterns in metabolic screening indicators were associated with incident ischemic stroke. Repeated health screening measurements may complement single time-point assessments and support continuous risk-factor monitoring and patient-centered prevention.