应用计算机决策支持系统制定血液透析患者促红细胞生成素剂量

Computerized decision support system for erythropoietin dosage in hemodialysis patients

目的血液透析患者贫血相关治疗管理目前存在很大的挑战性,探讨是否能通过经计算机决策支持（CDS）系统制定促红细胞生成素使用剂量对改善肾性贫血有帮助。方法选择2013年10月至2014年4月吉林大学第二医院血液透析中心给予促红素治疗持续3个月慢性肾脏病5期的患者564例,其中男性296例,女性268例;年龄22~83岁,平均年龄43.21岁。将患者分为人工组及CDS组,人工组493例,CDS组71例。依据患者干体质量、红细胞压积变化调整促红细胞生成素剂量,监测并观察患者血红蛋白变化,统计两组数据并做统计学分析。结果两组患者年龄主要集中在中年组（45~59岁）。终末期肾病病因主要是糖尿病肾病。两组患者贫血相关生物化学指标变化稳定,无明显差异。经调整中心和基线差异后每周平均促红细胞生成素剂量CDS组比人工组减少4%（率比0.96;95%可信区间0.77~1.18）,差异无统计学意义（P〉0.05）。两组监测的血红蛋白变化幅度很相似（σ2=1.30）。分析两组血红蛋白值调整比值比和未调整比值比,得出人工组平均血红蛋白值为118 g/L（标准误2 g/L）,比CDS组低1.1 g/L（标准误0.4 g/L）（P〈0.001）。在调整两组基线特征的差异后,CDS组比人工组平均血红蛋白值低1.4 g/L（标准误0.5 g/L）。分析线性混合模型得出：CDS组所达到的血红蛋白值中有向100~120 g/L和110~120 g/L增长的趋势,促红细胞生成素使用剂量CDS组比人工组减少4%,差异无统计学意义（P〉0.05）。结论两种促红细胞生成素给药方式对比,两组患者每月血红蛋白值在100~120 g/L波动,治疗效果无明显差别,但CDS组促红细胞生成素总剂量较人工组少。

Objective To explore the erythropoietin（EPO） dosage for improvement of renal anemia by computerized decision support（CDS）, because of anemia management in hemodialysis patients poses significant challenges. Method A total of 564 patients with chronic kidney disease fifth phase who received EPO treatment at the Second Hospital of Jilin University Dialysis Units at least 3-month from October 2013 to April 2014 were enrolled, which included 296 males and 268 females, aged 22-83 years old with mean age of 43.21 years old. All of them were divided into manual group（n = 493） and CDS group（n = 71）,The EPO dosage was adjusted by dry weight and hematocrit changes of patients, then hemoglobin changes of patients were monitored and observed. At last counted and analyzed data of 2 groups. Results The age of patients were mainly on the middle-aged group（45- 59 years old）, diabetes was the main cause of end-stage renal disease（ESRD）. The changes of anemia related biochemical indexes were no significant difference between 2 groups. After adjustment for differences in center and baseline, average weekly EPO dosage in CDS group decreased 4 % than that of manual group（rate 0.96; 95 % confidence interval 0.77- 1.18）, the difference was no statistical significant（P〉0.05）. The monitoring hemoglobin changes of 2 groups was similar（σ2= 1.30）. The adjusted odds and unadjusted odds for hemoglobin value of 2 groups were analyzed, then the results showed that the mean hemoglobin value in manual group was 118[standard error（SE） 2 g/L], which was lower 1.1（SE, 0.4） g/L than that of CDS group（P〈0.001）. But after adjustment for differences in baseline characteristics of 2 groups, CDS group hemoglobin value was lower 1.4（SE, 0.5） g/L than that of manual group. Analyzed linear mixed model concluded that there was the trend increased to 100- 120 g/L and 110- 120 g/L in CDS group, the EPO dosage was lower 4 % than that of CDS group, and there was no statistically significant（P〈0.05）. Conclusion It is demonstrated that comparison of 2 kinds of EPO,the monthly Hb in 2 group patients is 100- 120 g/L, the treatment effect is no significant difference between 2 groups,but EPO dosage in CDS group is lower than that of manual group.