机构地区:[1]首都医科大学附属北京安贞医院、北京市心肺血管疾病研究所核医学科,100029 [2]中国医学科学院、北京协和医学院阜外心血管病医院核医学科 [3]首都医科大学附属北京同仁医院核医学科 [4]中日友好医院核医学科 [5]首都医科大学附属北京朝阳医院核医学科 [6]卫生部北京医院核医学科
出 处:《中华核医学与分子影像杂志》2013年第2期136-140,共5页Chinese Journal of Nuclear Medicine and Molecular Imaging
基 金:北京市科技计划首都临床特色应用研究基金(D101100050010034)
摘 要:目的运用ROC分析方法对比一日法肺V/Q断层显像与传统平面显像在PE诊断中的效能。方法对2010年9月至2011年10月北京市6家三级甲等医疗机构收治的确诊或疑诊急性或亚急性PE,且Wells评分≥2分或血浆D-二聚体〉500μg/L的患者272例,均行一日法肺V/Q平面及断层显像。肺V/Q平面显像结果参照PE诊断前瞻I生研究II(PIOPEDⅡ)标准,分为PE高度可能性(确诊PE)、中度或低度可能性(不能确定)、极低度可能性或肺灌注正常(排除PE),灌注显像结果也分为确诊PE、不能确定(非特定疾病典型表现的多发V/Q异常)、排除PE。全部患者显像后随访6个月。最终以综合分析各影像学检查、实验室检查、临床可能性评估和6个月的随访结果来判定为确诊或排除PE。应用Kappa检验分析2种显像方法结果的一致性,应用X2检验对肺V/Q断层显像和平面显像在PE诊断上的常用评价指标进行比较;运用ROC曲线分析,计算并比较(Z检验)各AUC及其95%CI;比较2种方法不确定诊断率的差异。结果肺V/Q断层显像在诊断PE中的不确定诊断率(1.8%,5/272)仅为平面显像(9.2%,25/272)的1/5。以最终诊断为标准,肺V/Q断层和平面显像对272例PE的诊断指标分别为:Se83.6%(117/140)与78.3%(101/129,X2=1.217,P〉0.05),Sp89.8%(114/127)与94.9%(112/118,X2=2.269,P〉0.05),PPV90.0%(117/130)与94.4%(101/107),NPV83.2%(114/137)与80.0%(112/140),Ac86.5%(231/267)与86.2%(213/247)。2种显像方法诊断PE的符合率为84.9%(231/272,Kappa=0.727,P〈0.05),一致性较好。运用ROC分析2种显像方法的AUC及95%CI,断层显像分别为0.891(0.848—0.933),平面显像分别为0.865(0.819—0.911)(Z=2.65和2.03,均P〈0.05),提示肺V/Q断层显像对PE�Objective - To compare between one-day planar and tomographic pulmonary V/Q ima- ging for the diagnosis of PE using ROC analysis. Methods From September 2010 to October 2011, 272 patients suspected or confirmed of acute/sub-acute PE with Wells scores ≥ 2 or blood D-Dimer values 〉500 μg/L were recruited into this study. Patients' data were collected from six hospitals in Beijing. All patients were followed clinically for at least six months after V/Q imaging. Based on the criteria of prospective investigation of pulmonary embolism diagnosis (PIOPED) study II, the diagnosis by planar pulmonary V/Q scans was categorized as high-probability ( definite PE), moderate or low probability ( suspected PE), and very low probability or normal lung perfusion (no PE). The diagnosis by SPECT V/Q scans was also classi- fied as definite PE, suspected PE and no PE. The final diagnosis was decided by doctors in consensus based on the results of multiple imaging modalities, laboratory tests, clinical data and a six-month follow-up. Kappa test was used to analyze the concordance between planar and tomographic V/Q scans. The difference between the two methods was tested usingx2 test. ROC analysis was used to determine the AUCs. Z test was used to calculate the statistical significance of the difference between the AUCs. Results The uncertain di- agnosis rate was 1.8% (5/272) for V/Q SPECT and 9.2% (25/272) for planar imaging. The diagnostic values for SPECT and planar V/Q scans were respectively: Se 83.6% (117/140) and 78.3% (101/129) (X2 = 1. 217, P 〉 0.05) ; Sp 89.8% (114/127) and 94.9% (112/118) (X2 = 2. 269, P 〉 0.05) ; PPV 90.0% (117/130) and 94.4% (101/107) ; NPV 83.2% (114/137) and 80.0% (112/140) ; Ac 86. 5% (231/267) and 86.2% (213/247). There were no significant differences in Se and Sp between the two methods. The concordance between the two methods was 84.9% (231/272, Kappa = 0. 727, P 〈 0. 05 ). The AUG (95% CI) was 0.891 (0.
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...