在真实世界诊疗情境下，探讨度普利尤单抗(dupilumad)治疗慢性阻塞性肺疾病(chronic obstructive pulmonary disease, COPD)患者的用药依从性与治疗持续情况，分析影响患者长期坚持治疗的核心驱动因素。

回顾性收集并分析COPD患者的临床病历资料，根据用药规律程度将患者分为依从者36例为观察组，未依从者71例为对照组。以COPD评估测试(COPD assessment test, CAT)作为临床获益的量化评价工具，将CAT评分较基线下降≥2分设定为临床改善的判定标准。运用多因素Logistic回归模型(辅以森林图展示)筛选影响COPD依从性的独立变量，同时借助Kaplan-Meier生存分析法绘制曲线，对比不同疗效亚组间的治疗持续时间差异。

两组COPD患者在基线人口学特征及肺功能指标上具有均衡可比性，年龄、性别分布及基础肺功能无统计学差异(P>0.05)。观察组中CAT评分改善(≥2分)29例(80.56%)高于对照组35例(49.30%)。高依从性群体的临床症状缓解更为显著，CAT评分降幅更大，且COPD患者的依从性与CAT评分改善程度呈显著正相关(P＝0.003)。经多因素Logistic回归校正混杂变量后分析发现，CAT评分改善(≥2分)是维持患者高用药依从性的唯一保护因子(OR＝0.28, 95%CI：0.10~0.72, P＝0.011)。Kaplan-Meier生存曲线进一步证实，达到临床改善治疗持久性显著优于未改善者(Log-rank P＝0.0012)；其中未改善者中位治疗持续时间短至28天，而改善COPD组患者展现出好的治疗延续意愿，其中位治疗持久时间显著延长。

在真实世界的自费医疗模式下，COPD患者是否延续度普利尤单抗治疗，高度依赖于早期主观症状的获益感知。CAT评分的早期改善构成了预测治疗持久性的关键决策信号。临床实践中，应强化生物制剂治疗初期的疗效监测，及时就治疗获益与经济性同患者进行充分沟通，从而优化个体化给药策略，规避无效或低获益治疗的长期延续。

This real-world study aimed to evaluate medication adherence and treatment persistence in chronic obstructive pulmonary disease (COPD) patients receiving dupilumab, and to identify key determinants influencing long-term treatment continuation.

A retrospective analysis was conducted using clinical records of COPD patients. Participants were stratified into adherent and non-adherent groups based on medication regularity. Clinical improvement was measured by the COPD Assessment Test (CAT), with a reduction ≥2 points defined as meaningful benefit. Multivariable logistic regression (illustrated via forest plot) was applied to identify independent predictors of adherence. Kaplan-Meier survival analysis was performed to compare treatment persistence across different response subgroups.

Baseline demographic and lung function characteristics were comparable between groups, with no statistically significant differences in age, sex, or baseline pulmonary function (all P>0.05). Follow-up data indicated that higher adherence was associated with greater symptomatic relief, evidenced by larger reductions in CAT scores, and adherence status positively correlated with CAT improvement (P=0.003). After adjusting for confounders in logistic regression, CAT improvement (≥2 points) remained the only independent protective factor for high adherence (OR=0.28, 95%CI: 0.10~0.72, P=0.011). Kaplan-Meier curves further revealed significantly longer treatment persistence in the clinically improved group compared to the non-improved group (Log-rank P=0.0012). The median treatment duration was only 28 days in the non-improved group, whereas the improved group demonstrated stronger willingness to continue treatment, with markedly prolonged persistence.

In the real-world self-pay healthcare setting, continued use of dupilumab in COPD patients largely depends on early perceived symptomatic benefit. Early improvement in CAT score serves as a critical predictor for treatment persistence. Clinical practice should emphasize close monitoring of initial treatment response and proactive communication with patients regarding efficacy and cost-effectiveness, in order to tailor individualized therapy and avoid prolonged low-benefit treatment.