To compare the accuracy of the End-of-life Assessment Form and common survival prediction scale〔Palliative Prognostic Index (PPI) 、Palliative Performance Scale (PPS) 、Karnofsky Score (KPS) 〕in predicting the survival of patients with advanced malignant tumors, in order to provide a basis for the selection of survival prediction tools for advanced cancer patients.

Patients with advanced malignant tumors admitted to the hospice ward of Linfen Road Community Health Service Center of Jing'an Distirct of Shanghai from April 1, 2018 to February 1, 2020 were retrospectively selected as researchsubjects. At the time of admission, the general information questionnaire, End-of-life Assessment Form, PPI, PPS, KPS were used to evaluate the patient, and the survival time of the patient was observed and recorded (from admission to the date of death) . The survival time of all patients was analyzed by Kaplan-Meier method, and the survival curve was drawn. The Kaplan-Meier method was used to calculate the median survival of patients in different groups of each scale, the log-rank test was used to compare the differences in survival among patients in different groups of each scale, and the survival curves were drawn. Finally, by comparing the predicted survival time and the actual survival time of patients with different score segments of each scale, the accuracy rate of the End-of-life Assessment Form, PPI, PPS and KPS in predicting the survival time of patients with advanced malignant tumors were calculated.

A total of 315 patients with advanced malignant tumors were included in this study, of which 266 (84.4%) patients died during hospitalization and 49 (15.6%) patients were censored (right censored, type Ⅲ censored) . The median survival time of 315 patients was 10.00〔95%CI (8.10, 11.90) 〕d. The median survival time of patients in groups of 20.0~35.0 points, 35.5~45.0 points, 45.5~50.0 points, 50.5~60.0 points and 60.0~100.0 points of the End-of-life Assessment Form were 1.00〔95%CI (0.79, 1.22) 〕d, 5.00〔95%CI (3.92, 6.08) 〕d, 10.00〔95%CI (7.83, 12.17) 〕d, 22.00〔95%CI (18.42, 25.58〕d and 45.00〔95%CI (26.23, 63.77〕d (χ²=360.561, P<0.001) , respectively; The median survival time of patients in groups of 10~20 points, 30~40 points and 50~100 points of KPS were 1.00〔95%CI (0.66, 1.34) 〕d、7.00〔95%CI (5.23, 8.77) 〕d and 30.00〔95%CI (20.87, 39.13) 〕d (χ²=137.280, P<0.001) , respectively; The median survival time of patients in groups of 60%~100%, 30%~50% and 10%~20% of PPS were 35.00〔95%CI (25.90, 44.10) 〕d、12.00〔95%CI (9.66, 14.34) 〕d and 2.00〔95%CI (0.85, 3.15) 〕d (χ²=139.311, P<0.001) , respectively; The median survival time of patients in groups of 0~3.5 points, 4.0~5.5 points, 6.0~7.5 points, 8.0~10.0 points and 10.5~15.0 points of PPI were 33.00〔95%CI (25.39, 40.61) 〕d、12.00〔95%CI (8.15, 15.85) 〕d、6.00〔95%CI (4.72, 7.28) 〕d、3.00〔95%CI (1.76, 4.24) 〕d and 1.00〔95%CI (0.89, 1.11) 〕d (χ²=289.831, P<0.001) , respectively. The accuracy rate of the End-of-life Assessment Form, KPS, PPS and PPI to predict the survival time were 81.27% (256/315) 、57.78% (182/315) 、57.46% (181/315) 、73.65% (232/315) , respectively.

The End-of-life Assessment Form, PPI, KPS and PPS can be used to predict the survival time of advanced malignant tumors, but the End-of-life Assessment Form is superior to PPI, KPS and PPS in predicting the survival time of advanced malignant tumors.

比较临终患者病情评估表与其他常见生存期预测量表〔卡氏评分（KPS）、姑息功能评价量表（PPS）、姑息预后指数（PPI）〕在预测癌症晚期患者生存期中的准确性，旨在为癌症晚期患者生存期预测工具的遴选提供依据。

本研究为回顾性研究，选取2018年4月1日至2020年2月1日入住上海市静安区临汾路街道社区卫生服务中心安宁疗护病房的癌症晚期患者为研究对象。于患者入院时采用一般资料调查表、临终患者病情评估表、KPS、PPS、PPI对其进行评估，并观察、记录患者的生存时间（入院时间至死亡时间，除死亡外的其他生存结局归类为删失）。采用Kaplan-Meier法分析全体患者的生存时间，并绘制生存曲线。通过Kaplan-Meier法计算各量表不同组别患者的中位生存期，通过Log-rank检验比较各量表不同组别患者生存期的差异，并绘制生存曲线。最终，通过比较患者的实际生存时间与临终患者病情评估表、PPI、PPS、KPS预测的生存时间，来判断上述4种量表在不同组别患者及全体患者生存期预测中的准确性。

共纳入315例癌症晚期患者，其中266例（84.4%）于住院期间死亡，49例（15.6%）删失（右删失，Ⅲ型删失）。315例患者的中位生存期为10.00〔95%CI（8.10，11.90）〕d。临终患者病情评估表20.0~35.0、35.5~45.0、45.5~50.0、50.5~60.0、60.5~100.0分组患者的中位生存期分别为1.00〔95%CI（0.79，1.22）〕、5.00〔95%CI（3.92，6.08）〕、10.00〔95%CI（7.83，12.17）〕、22.00〔95%CI（18.42，25.58）〕、45.00〔95%CI（26.23，63.77）〕d（χ²=360.561，P<0.001）；KPS 10~20、30~40、50~100分组患者的中位生存期分别为1.00〔95%CI（0.66，1.34）〕、7.00〔95%CI（5.23，8.77）〕、30.00〔95%CI（20.87，39.13）〕d（χ²=137.280，P<0.001）；PPS 60%~100%、30%~50%、10%~20%组患者的中位生存期分别为35.00〔95%CI（25.90，44.10）〕、12.00〔95%CI（9.66，14.34）〕、2.00〔95%CI（0.85，3.15）〕d（χ²=139.311，P<0.001）；PPI 0~3.5、4.0~5.5、6.0~7.5、8.0~10.0、10.5~15.0分组患者的中位生存期分别为33.00〔95%CI（25.39，40.61）〕、12.00〔95%CI（8.15，15.85）〕、6.00〔95%CI（4.72，7.28）〕、3.00〔95%CI（1.76，4.24）〕、1.00〔95%CI（0.89，1.11）〕d（χ²=289.831，P<0.001）。临终患者病情评估表、KPS、PPS、PPI在全体患者生存期预测中的准确率分别为81.27%（256/315）、57.78%（182/315）、57.46%（181/315）、73.65%（232/315）。

临终患者病情评估表、PPI、KPS、PPS均可用于预测癌症晚期患者生存期，但临终患者病情评估表的预测效果明显优于PPI、KPS、PPS。