Recent advancements in artificial intelligence have led to increased interest in predictive modeling across various domains, including medicine. Although numerous metrics have been established for binary classification, the growing adoption of multi-class systems necessitates robust evaluation measures. However, comprehensive simulation studies comparing the performance of existing multi-class metrics under diverse data conditions remain limited. In this study, we first provide a concise overview of commonly used accuracy metrics for multi-class classification. Then, we report a simulation study that systematically evaluates several diagnostic accuracy measures under a wide range of scenarios, including three- and five-class settings, balanced and imbalanced sample sizes, and different distributional assumptions for predictors. We assessed each metric’s performance in terms of bias and 95% confidence interval coverage. Under balanced conditions, most metrics demonstrated stable and unbiased performance, closely approximating the true values. However, under imbalanced conditions, greater bias was observed, with the M-index and polytomous discrimination index exhibiting comparatively more stable performance across various scenarios. The micro-averaged receiver operating characteristic curve area consistently showed higher bias under class imbalance. Finally, we applied these metrics to a glioma tumor grading task using external datasets. This study provides a systematic comparison of commonly used metrics and offers practical guidance for selecting appropriate measures in multi-class classification tasks.

Background: We developed a simplified variant of the ADV score, the AFP-PIVKAII (AP) score for post-transplant hepatocellular carcinoma (HCC) prognosis, which considers only AFP and PIVKA-II levels excluding morphometric tumor size information from the ADV score. This study investigated the prognostic performance of the AP score in predicting HCC recurrence and overall survival (OS) after living donor liver transplantation (LDLT). Methods: We analyzed 843 patients with HCC who underwent LDLT between 2006 and 2015, assessing HCC recurrence and OS in relation to AP score. Results: The median pretransplant AFP and PIVKA-II levels were 12.8 ng/mL and 27 mAU/mL, respectively. The median and mean AP scores were 2.6 log (range: 0.6–9.2 log) and 2.9±1.1 log, respectively. The 5-year time-dependent area under the receiver operating characteristic curve for the AP score in predicting post-transplant HCC recurrence was 0.672 (p<0.001). HCC recurrence and OS curves along AP score intervals of 1.0 log showed statistical differences in accordance with the AP scores (both p<0.001). Using a Youden index J-derived AP score cutoff of 4.0 log, two-tiered groups (ADV <4.0 log vs. ADV ≥4.0 log) showed statistically significant differences in HCC recurrence and OS (both p<0.001). Harrell’s c-indices for AP score with cutoff of 4.0 log and ADV scores with cutoff of 5.0 log regarding HCC recurrence and OS were similar. Conclusion The AP score functions as an integrated surrogate marker for predicting post-transplant outcomes in patients with HCC undergoing LDLT. It may serve as a simplified alternative to the ADV score, particularly in patients with small HCCs.

Purpose: This 6-year post-marketing surveillance (PMS) study was conducted in South Korea to evaluate the real-world safety and effectiveness of eribulin in patients with advanced or metastatic breast cancer previously treated with anthracyclines and taxanes. Materials and Methods: During the study period (17 August 2012 to 16 August 2018), case-report files (CRFs) of patients receiving eribulin were collected. The main study endpoint was to assess the safety of eribulin. Evaluation of the effectiveness of eribulin was an exploratory endpoint. Patients were followed for 1 year after eribulin initiation. Results: CRFs were collected from 64 investigators at 64 sites for 1,079 patients. The safety analysis set (SAS) included 1,001 eribulin recipients; effectiveness was assessed in 244 patients. In the SAS, patients were predominantly female (99.6%), with a median age of 53.0 years, and diagnosed with metastatic breast cancer (92.0%). Eribulin was administered as a median 4th line chemotherapy. A total of 2,124 treatment-emergent adverse events (TEAEs) were reported in 661 patients (66.0%). Neutropenia was the most common TEAE (32.5% of patients), occurring at a median of 9-11 days from initial eribulin administration. Overall response and disease control rates were 31.7% and 95.6%, respectively, and the median duration of eribulin use (time to treatment failure) was 3.0 months. Conclusion This large real-world PMS analysis in patients with advanced or metastatic breast cancer demonstrated the effectiveness of eribulin and found no new safety concerns relative to safety information from prior clinical and real-world studies, and approvals in South Korea and other countries.

Purpose: Although human epidermal growth factor receptor 2 (HER2) positivity is prevalent in microinvasive breast cancer (MIBC), data focused on HER2-positive MIBC are limited. We investigated the clinical course and long-term outcomes of HER2-positive MIBC and evaluated the role of adjuvant chemotherapy. Materials and Methods: The study included patients with curatively resected pT1mi pN0 HER2-positive breast cancer between January 2000 and January 2020. Treatments and survival outcomes, including invasive breast cancer-free survival (IBCFS), distant recurrence-free survival (DRFS), and overall survival (OS) were analyzed. Results: The analysis included 799 female patients. The median age was 51 years (range, 23 to 79 years), and 51.6% (n=412) were premenopausal. Multifocality was confirmed in 17.3% (n=138), and estrogen receptor (ER) positivity in 29.8% (n=238). Adjuvant chemotherapy was administered to 17.5% (n=140), with doxifluridine in 96.4% of cases. One patient (0.1%) received trastuzumab. With a median follow-up of 119.0 months (95% confidence interval [CI], 114.0 to 127.0), the 8-year IBCFS, DRFS, and OS were 91.2% (95% CI, 89.1 to 93.3), 97.5% (95% CI, 96.4 to 98.7), and 98.8% (95% CI, 98.0 to 99.6), respectively. No significant differences were observed between patients with and without adjuvant chemotherapy. The lack of differences in IBCFS by chemotherapy was consistent across subgroups, including pre-/postmenopausal patients, grade 1-2/3 tumors, and ER-negative disease. Conclusion A clinically meaningful proportion of HER2-positive MIBC patients experience IBCFS events with long-term follow-up. Adjuvant chemotherapy did not improve survival, potentially due to the use of an outdated, ineffective regimen. The role of modern adjuvant regimens, particularly those incorporating HER2-targeted therapy, warrants further exploration.

Subarachnoid hemorrhage (SAH) is a stroke subtype with high mortality and poor functional outcomes. Prompt occlusion of a ruptured aneurysm at an early stage is crucial to prevent rebleeding, which can result in even higher mortality and more severe disabilities. The most critical initial decision in SAH management is the choice of treatment method with surgical clipping or endovascular coiling. We aimed to develop an evidence-based clinical guideline to select the optimal initial treatment in patients with SAH. We developed this guideline based on evidence from systematic reviews and meta-analyses via a de novo process. A systematic literature review was conducted across four databases (MEDLINE, Embase, Cochrane, and KoreaMed) to answer two population, intervention, comparison, outcome questions comparing clipping and coiling. The risk of bias was assessed using ROB 2.0 and the Newcastle-Ottawa Scale. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagrams and meta-analyses were generated for functional outcome and mortality. We included six randomized control trials (RCTs) and 58 observational studies. Meta-analysis of RCTs showed that coiling improved functional outcomes compared to clipping (odds ratio [OR], 0.91; 95% confidence interval [CI], 0.86–0.97). No significant mortality difference was observed in RCTs (OR, 1.38; 95% CI, 0.91–2.09), but non-RCTs favored clipping for reduced mortality (OR, 0.77; 95% CI, 0.69–0.86). However, it is difficult to generalize these findings to all clinical situations, as patients with SAH have a highly variable clinical course. Final treatment decision should be tailored to the individual patient’s status, including aneurysm location, morphology, and the expertise available at the treatment center. Such decisions are best made by specialists such as a board-certified physician and should be explained to the patient and their caregivers, along with the rationale for selecting the most appropriate treatment at the given hospital. Korea has many certified endovascular neurosurgeons, cerebrovascular surgeons, and certified cerebrovascular centers. Proper selection of the most suitable treatment method by certified physicians and centers would greatly benefit patient outcomes and healthcare professionals.

Accurate and early identification of stroke and large vessel occlusion (LVO) in emergency settings is essential for improving patient outcomes and ensuring the efficient allocation of medical resources. This clinical practice guideline systematically reviews domestic and international literature and conducts meta-analyses to evaluate the utility and diagnostic accuracy of stroke assessment tools used in prehospital emergency medical services (EMS). We developed a guideline based on evidence from systematic reviews and meta-analyses via a de novo process. A systematic literature review was conducted to evaluate the usefulness of diagnostic EMS assessment tools for diagnosing stroke and LVO. Overall, 70 non-randomized control studies were selected for this study. A meta-analysis was conducted with a subgroup analysis to distinguish between patients with stroke and those with LVO. EMS tools demonstrated high sensitivity but low specificity for diagnosing stroke. In the prehospital setting, using validated EMS stroke assessment tools is recommended for the early identification of stroke and LVO. Upon hospital arrival, stroke specialists should conduct further evaluation and triage to confirm the diagnosis and guide appropriate management. Delays in diagnosing LVO are frequently unacceptable. While experts advocate for the use of EMS assessment tools to facilitate early identification of LVO, these tools alone lack adequate sensitivity. Therefore, further diagnostic evaluations and consultation with stroke specialists upon hospital arrival are recommended.

The mothership (MS) model, where patients are directly transferred to a thrombectomy-capable center, and the drip-and-ship (DS) model, where thrombolysis is initiated at the nearest primary stroke center before transfer for thrombectomy, are the primary transport modes for patients with stroke. We aimed to establish guidelines for selecting the appropriate transfer strategy based on emergent large vessel occlusion (LVO). We developed this guideline based on evidence from systematic reviews and meta-analyses via a de novo process. A systematic literature review was conducted across four databases (MEDLINE, Embase, Cochrane, and KoreaMed) to answer three Population, Intervention, Comparison, and Outcome questions comparing MS and DS models. The risk of bias was assessed using the Newcastle-Ottawa Scale. Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagrams and meta-analyses were generated for functional outcomes, mortality, and successful recanalization. Twenty-six non-randomized controlled studies showed that the MS model improved good functional outcomes by approximately 14% compared with the DS model (odds ratio [OR], 1.14; 95% confidence interval [CI], 1.00–1.30). Fifteen studies reported that mortality in the MS and DS models showed no significant differences (OR, 0.97; 95% CI, 0.84–1.11). Twenty-four studies revealed no significant difference in successful recanalization between the MS and DS models (OR, 0.87; 95% CI, 0.68–1.10). The MS model should be considered first to improve the functional outcome of patients with LVO. However, if thrombectomy cannot be performed immediately after thrombolysis, or if a thrombectomy-enabled hospital is not nearby, the DS model should be considered by stroke specialists depending on transportation time and regional factors. We suggest a mixed approach with the DS model based on specific circumstances or regions to ensure the optimum treatment of patients with acute ischemic stroke (AIS). Appropriate transport for patients with LVO improves the prognosis of AIS.

Background: Studies on whether donor-to-recipient sex match status affects long-term survival after liver transplantation (LT) have yielded contradictory results. This study evaluated whether donor-to-recipient sex match status influenced long-term survival after living donor liver transplantation (LDLT) or deceased donor liver transplantation (DDLT) at a high-volume center. Methods: The study included 6,664 patients who underwent primary LT between January 2000 and December 2022 at our institution. Patients were divided into four groups according to donor-to-recipient sex match status: male-to-male (n=3,427 [51.4%]), male-to-female (n=1,152 [17.3%]), female-to-male (n=1,385 [20.8%]), and female-to-female (n=700 [10.5%]). Results: Regarding clinical characteristics, the four groups differed significantly regarding background liver disease (P<0.001), model for end-stage liver disease score (P<0.001), serum protein induced by vitamin K absence or antagonist II level (P=0.003), presence of concurrent hepatocellular carcinoma (HCC; P<0.001), and type of LT (P=0.003). Overall survival (OS) of all LT recipients did not differ significantly among the groups (P=0.377). Donor-to-recipient sex match status did not affect long-term OS in either LDLT (P=0.176) or DDLT (P=0.220) groups. In addition, sex match status did not significantly influence posttransplant OS among patients who underwent LDLT without HCC (P=0.464), LDLT with HCC (P=0.236), DDLT without HCC (P=0.338), or DDLT with HCC (P=0.818). Conclusions Donor-to-recipient sex match status does not significantly affect posttransplant patient survival or HCC prognosis after LDLT or DDLT.

Olanzapine is a second-generation antipsychotic drug associated with the development of type 2 diabetes. However, the development and remission of diabetes with glutamic acid decarboxylase (GAD) antibody (Ab) positivity related to olanzapine treatment have not been reported. Here, we present the case of a 33-year-old man taking olanzapine who was diagnosed with diabetes with GAD Ab positivity complicated by diabetic ketoacidosis (DKA). After discontinuation of the drug and tight glycemic control, his diabetes remitted, and the GAD Ab became negative spontaneously after 18 months. The classification of diabetes is challenging because of overlapping autoimmune and metabolic features. Olanzapine may have contributed to pancreatic β-cell apoptosis, which could potentially trigger autoimmunity and the development of type 1 diabetes-like features (DKA as the initial presentation, a low C-peptide level, and GAD Ab positivity). The rapid amelioration of hyperglycemia and weight reduction may be responsible for the remission of diabetes after the discontinuation of olanzapine.

Objective: As the demand for community mental health services continues to grow, the need for well-equipped and organized services has become apparent. This study aimed to optimize the roles and infrastructure of mental health services, by establishing, among other initiatives, standardized operating models. Methods: The study was conducted in multiple phases from May 12, 2021, to December 29, 2021. Stakeholders within South Korea and metropolitan mental health welfare centers were targeted, but addiction management support centers, including officials, patients, and their families, were integrated as well. A literature review and survey, focus group interviews, a Delphi survey, and expert consultation contributed to comprehensive revisions and improvements of the mental health service model. Results: The proposed model for community mental health welfare centers emphasizes the expansion of personnel and infrastructure, with a focus on severe mental illnesses and suicide prevention. The model for metropolitan mental health welfare centers delineates essential tasks in areas such as project planning and establishment, community research, and education about severe mental illnesses. The establishment of a 24-hour emergency intervention center was a crucial feature. In the integrated addiction support center model, the need to promote addiction management is defined as an essential task and the establishment of national governance for addiction policies is recommended. Conclusion This study proposed standard operating models for three types of mental health service centers. To meet the increasing need for community care, robust mental health service delivery systems are of primary importance.