Random parameter logit model.

BackgroundType 2 diabetes mellitus (T2D) represents a major public health challenge with significant effects on morbidity and mortality. Clinical guidelines provide treatment recommendations, but there is limited understanding of patients’ preferences. This study aimed to elicit preferences for second-line drug treatments for T2D.MethodA Discrete Choice Experiment with a partial-profile design was conducted from August to November 2023, involving German patients with experience in either monotherapy or second-line drug treatment. Participants completed 12 choice tasks, each presenting three alternatives described by attributes: risk of myocardial infarction, risk of stroke, risk of nerve damage, risk of nausea, risk of severe hypoglycemia, weight change, type and frequency of intake, and schedule of intake. Statistical analyses employed the Conditional Logit and Random Parameter Logit models to assess main effects and heterogeneity.ResultsThe study encompassed 583 adult individuals with T2D, evenly divided between the two populations. Key factors influencing choice decisions included risk of nausea, risk of nerve damage, and weight change, with weekly type and frequency of intake risk of myocardial infarction followed. Less impactful but still relevant were risks of stroke, and severe hypoglycemia, while the intake schedule was least significant. Analysis of BMI categories revealed distinct preferences, particularly in weight change, with significant heterogeneity observed among respondents.ConclusionThis study highlights the importance of incorporating patient preferences into clinical decision-making. By quantifying preferences for second-line drug treatments, the study underscores the need for low-risk options that also consider weight change and intake conditions, aligning with the German National Health Care Guideline for T2D objectives for shared decision-making and treatment adherence. Recognizing individual sensitivities to risks and benefits is crucial for tailoring effective T2D treatment strategies. The study bridges clinical findings with patient perspectives, offering valuable insights into clinical practice, consideration for HTA processes, and design of clinical studies.

2型糖尿病（Type 2 diabetes mellitus, T2D）是一项重大公共卫生挑战，对发病率与死亡率均存在显著影响。临床指南虽已提供治疗建议，但学界对患者偏好的认知仍较为有限。本研究旨在明确2型糖尿病二线药物治疗的患者偏好。
方法 本研究于2023年8月至11月开展，采用部分轮廓设计的离散选择实验（Discrete Choice Experiment），纳入曾接受单药治疗或二线药物治疗的德国2型糖尿病患者。受试者需完成12项选择任务，每项任务均呈现3种由以下属性描述的治疗方案：心肌梗死风险、卒中风险、神经损伤风险、恶心风险、严重低血糖风险、体重变化、给药类型与频率，以及给药方案。统计分析采用条件Logit模型与随机参数Logit模型，以评估主效应及群体异质性。
结果 本研究共纳入583名成人2型糖尿病患者，单药治疗与二线药物治疗两个受试群体人数均等。影响选择决策的关键因素依次为恶心风险、神经损伤风险及体重变化，其次为每周给药的类型与频率，以及心肌梗死风险。影响相对较弱但仍具相关性的因素包括卒中风险与严重低血糖风险，而给药方案的影响最小。针对身体质量指数（Body Mass Index, BMI）类别的分析显示，受试者存在显著的偏好差异，尤其体现在体重变化维度，且受访者间存在显著的异质性。
结论 本研究强调了将患者偏好纳入临床决策的重要性。通过量化2型糖尿病二线药物治疗的患者偏好，本研究凸显了研发兼顾体重变化与给药条件的低风险治疗方案的必要性，这与德国《2型糖尿病国家卫生保健指南》中关于共同决策与治疗依从性的目标相一致。识别个体对风险与获益的敏感性，对于制定个性化的高效2型糖尿病治疗策略至关重要。本研究搭建了临床研究结果与患者视角之间的桥梁，可为临床实践、健康技术评估（Health Technology Assessment, HTA）流程的考量以及临床研究设计提供宝贵见解。