中国5所医院红细胞输注情况病例对照研究

Investigation of RBC transfusion of five hospitals in China

目的调查参加研究的5所医院红细胞输注指征,输血前后血红蛋白（Hb）水平变化以及与红细胞输注相关的其他临床因素,对临床红细胞输注提供科学依据。方法对5家医院的血库信息系统查询2013年全年住院患者中首次红细胞输注病例,并从中随机选择1000例收集相关数据,病例不满1 000例的医院的全部病例纳入研究;将所有数据按照人口统计学特征、并发症、是否手术、红细胞特征以及Hb水平进行区分。根据主要疾病诊断将病例属于6大诊断组的病例分出,选择在同一时间段入院且在相同诊断组的未输血病例为作为对照,进行临床队列研究,对输注前Hb水平及其他红细胞输注相关临床因素进行统计学分析。结果从约5 000例红细胞输注患者中选择出属于6个诊断组的病例1 356例,对照病例1 201例,外科患者（病例数/对照数）包括骨科手术组（312/216）;普外科手术组（367/332）;内科患者包括实体瘤（230/232）;消化道出血（212/221）;心血管疾病（87/77）;血液系统恶性肿瘤（100/101）。Hb水平与手术患者和内科患者的红细胞输注均具有相关性。手术患者中,Hb为（70—90） g/L患者组相对于Hb〉100 g/L患者组,输注红细胞的OR值（95%可信区间）为27. 0（13. 9—52. 7）,内科患者中该OR值为71. 5（37. 8—135. 4）。外科患者红细胞输注只与红细胞输注前Hb水平相关;内科患者中主要诊断、不同医院以及年龄〉75岁等均是进行红细胞输注的独立因素。心血管疾病患者组相对于实体瘤患者组进行红细胞输注的OR值为2. 5（1. 2—4. 9）,血液系统恶性肿瘤患者组与消化道出血患者组相对于实体瘤患者组进行红细胞输注的OR值分别为0. 4（0. 2—0. 6）和0. 5（0. 3—0. 8）。其中46%的血液系统恶性肿瘤患者红细胞输注前的Hb〈60 g/L。结论红细胞输注前Hb水平是与红细胞输注最相关的因素。其他相关临床因素包括年龄、心血管疾病、主要诊断等仅与内科患者的红细胞输注相关,与外科患者红细胞输注无关。比较西方国家红细胞输注标准[输血前Hb：（70—80） g/L]与中国红细胞输注标准（输血前Hb：60 g/L）的临床研究将为科学合理的红细胞输注标准的制定提供依据。

Objective To investigate the hemoglobin level and other risk factors related to RBC transfusion in 5 Chinese hospitals in 2013, and to provide scientific basis for RBC transfusion. Methods We identified all patients who received atleast one RBC transfusion and then randomly choose 1000 patients from each of five hospitals and characterize the underlying demographics, co-morbidities, surgical procedures （if performed） , RBC characteristics, and pre-transfusion hemo- globin concentrations. The transfused patients （cases） were placed into six diagnostic groups based on underlying disorders. Un-transfused patients were randomly selected as controls foreach group matched for the same underlying diagnosis. We assessed the pre-transfusion hemoglobin trigger and other clinical characteristics associated with transfusion. We used logistic regression to assess independent risk factors associated with trans- fusion. The risk factors evaluated were age, gender, underlying diagnosis, cardiovascular disease, hospital, and pre-transfu- sion hemoglobin level. Results From the ~ 5000 patients who received transfusion, we chose 6 diagnostic groups：Surgical： Orthopedic surgery （ case/controls ; 312/216） , general surgery （367/332） ; Medical ： solid tumor （ 230/232）, gastrointesti- nal bleeding （ 212/221 ）, cardiovascular （ 87/77）, and hematological malignancies （ 100/101 ）. Pre-transfusion hemoglobin was associated with the highest risk of transfusion in both surgical and medical patients.For example, compared to 100 g/L, the odds ratio （OR） （ 95% confidence interval） for pre-transfusion hemoglobin of 70-79 g/L in surgical patients was 27. 0 （ 13.9-52. 7）, and in medical patients was 71.5 （37.8-135.4）. In surgical patients only pre-transfusion hemoglobin was associated with transfusion. In medical patients, besides pre-transfusion hemoglobin threshold, the independent factors asso- ciated with transfusion were primary diagnosis, hospital and age 75＋. Compared with patients with solid tumor, the odds of transfusion were greater in those with underlying cardiovascular disease OR = 2. 5 （ 1.2-4. 9）, and less frequent in those with hematological malignancies, OR= 0. 4 （0. 2-0. 6） and gastrointestinal bleeding, OPt= 0. 5 （0. 3-0. 8）. The pre- transfusion hemoglobin was less than 60 g/L in 46% of patients with hematological malignancies. Conclusion The hemoglo- bin level before RBC infusion is the most important factor related to RBC infusion. Other clinical factors, including age, car- diovascular disease and primary disease, were only related to the RBC transfusion in medical patients, but not to the RBC transfusion in surgical patients. Comparison of the standard of RBC infusion in Western countries （hemoglobin： 70--80 g/l） and Chinese RBC Infusion Standard （hemoglobin： 60g/1 before transfusion） will provide a basis for scientific and rational formulation of RBC infusion Standard.