To the Editor We have read with great interest the retrospective population-based cohort study by Sullivan et al.¹ They found an association of early palliative care (received 0 to 30 days after cancer diagnosis) with decreased survival (adjusted hazard ratio, 2.13; 95% CI, 1.97-2.30) in patients with advanced lung cancer.

The authors used the propensity score method for reduction of selection bias among groups. The propensity score included factors such as the Charlson Comorbidity Index score and the Functional Comorbidity Index score. However, these parameters do not represent the true performance status of patients. Performance status is one of the major factors in cancer treatment decision-making and is assessed using the Eastern Cooperative Oncology Group scale² or the Karnofsky scale.³ In addition, performance status is also associated with specialists’ decisions to deliver palliative care. Therefore, patients with greater functional impairment are more likely to receive palliative care sooner.

致编辑我们非常感兴趣地阅读了Sullivan等人的基于人群的回顾性队列研究。¹他们发现，晚期肺癌患者的早期姑息治疗（在癌症诊断后0至30天接受治疗）与生存率降低（校正风险比为2.13； 95％CI为1.97-2.30）相关。

作者使用倾向评分法来降低群体之间的选择偏见。倾向评分包括诸如查尔森合并症指数评分和功能性合并症指数评分等因素。但是，这些参数不能代表患者的真实状态。表现状态是癌症治疗决策中的主要因素之一，可以使用东部合作肿瘤小组²量表或卡诺夫斯基量表进行评估。³此外，绩效状态还与专家提供姑息治疗的决定有关。因此，功能障碍较大的患者更有可能较早接受姑息治疗。