[Medical care situation of people with vitiligo : A pilot and feasibility study on guideline-oriented management in German dermatology practices].

Transport safety remains a major global challenge with substantial public health and economic consequences. Road traffic accidents cause over 1.35 million deaths annually (~ 2.5% of global fatalities) and 50 million injuries, ranking as the leading cause of death among individuals aged 5–29 years. The burden is disproportionately concentrated in Southeast Asia and the Western Pacific, accounting for over 50% of global fatalities. This issue is particularly acute in low- and middle-income APEC economies, where road accidents result in economic losses of 3–5% of GDP. These challenges highlight the urgent need for reliable policy frameworks supported by robust analytical tools. Advanced MCDM approaches, capable of synthesizing complex and heterogeneous data, are essential for improving decision reliability and enabling evidence-based transport safety interventions. Although the MCDM framework offers significant potential for addressing complexities related to transport safety engineering, the selection and implementation of appropriate MCDM methodologies are fraught with challenges because of model-related uncertainty, sensitivity, and instability. Subsequently, the reliance on inputs is often subject to variability, further complicating the decision-making process, which emphasizes the need for models that not only ensure stability and reliability but also address issues of sensitivity and uncertainty. Over the years, researchers have developed a plethora of MCDM methods tailored to transport safety applications. Despite the advancements, significant gaps remain in the application of the MCDM methods to transport safety, e.g., their stability and adaptability across diverse contexts, which leads to four key motivations underpinning this study: (1) The lack of universally practicable SPIs at a regional level hinders the ability of policymakers to measure and benchmark progress effectively. (2) Existing MCDM methods are often geographically constrained and lack the flexibility to address the diverse backgrounds of different regions or countries, such as the APEC, which includes 20 countries. (3) Current methods emphasize data aggregation while overlooking key elements of the MCDM process (such as grouping, deconstruction, and decomposition), which are crucial for refining interventions and supporting nuanced decision making. (4) Selecting the most suitable method from a broad range of options in the MCDM database continues to pose a significant challenge for decision makers. This difficulty arises from the sensitivity and uncertainty inherent in the model, which can be influenced by varying the inputs. These factors have a profound effect on the robustness and reliability of policy decisions, highlighting the importance of addressing these issues for effective and defensible policymaking. This gap is exacerbated by the instability of several MCDM methods when used with small-to-medium-sized datasets. Consequently, there is a critical need to develop a systematic and scientifically grounded MCDM framework that can accommodate varying geographical contexts while incorporating the essential phases of the MCDM process (weighting, aggregation, grouping, deconstruction, and decomposition) and prioritizing efficiency, stability, and reliability. This study aimed to address these gaps by developing a hybrid preference function-nested and machine learning-embedded model, referred to as the EXPROM II–K-means with LDA. This study demonstrates that integrating a parameter-free preference modeling mechanism with LDA-enhanced clustering within a unified MCDM framework substantially improves decision reliability by simultaneously reducing parameter sensitivity, enhancing data separability, and stabilizing ranking and grouping outcomes under heterogeneous and uncertain conditions. The real-world applicability of the proposed framework is as follows: