| Objective:To compare the diagnostic value of different Imaging diagnostic methods include hysterosalpingography, two-dimensional transvaginalsonography, three-dimensional transvaginal sonography, two-dimensional hy-sterosonography and three-dimensional hysterosonography for intrauterineadhesions.Method:English database include Pubmed, Cochrane library, EMBASEand Chinese Database include Wanfangdata, CKNI, CBM are searched for theliteratures related to the methods mentioned above for diagnosing IUA. Twodifferent systematic reviewers independently screen the literatures according tothe inclusion and exclusion criteria which pre-established and then determine theincluded literatures to Meta analysis after assessing their quality. Sensitivity,specificity, positive likelihood ratio, negative likelihood ratio, diagnostic oddsratio, area under curve, Q index are pooled and summary receive operatingcharacteristic curve is drawn by Meta-disc software for the included studies.Compare the diagnostic value of the above imaging diagnostic methods for IUAis by the Q value’s Z-examination in the end. Result:In accordance with the inclusion and exclusion criteria, thereare totally24studies being analysed.8out of them are about HSG,11for2DTVS,5for3DTVS.10for2DSHG and4to left for3DSHG. A totalof9482patients involved,1014of them confirmedly diagnosed to be IUApatients. The included patients for the different methods are1297,2257,3130,1262and1536, respectively, in addition, IUA patients are130,510,268,89and19,respectively. The pooled diagnostic odds ratio are133.46[95%CI(32.460-548.74)],207.76[95%CI(41.82-1032.11)],276[95%CI(37.09-2059.67)],399.16[95%CI(90.82-1754.32)] and1322.7[95%CI(169.48-10323.9)],respectively; the pooled sensitivity are0.877[95%CI(0.808-0.928)],0.680[95%CI(0.637-0.720)],0.909[95%CI(0.886-0.942)],0.879[95%CI(0.798-0.936)]and0.842[95%CI(0.604-0.966)],respectively;the pooled specificity are0.932[95%CI(0.916-0.945)],0.990[95%CI(0.984-0.994)],0.999[95%CI(0.997-1.000)],0.994[95%CI(0.988-0.998)] and1.000[95%CI(0.998-1.000)], respectively;thepooled positive likelihood ratio are14.643[95%CI(6.179-34.700)],37.554[95%CI(11.582-121.766)],35.009[95%CI(3.829-320.072)],69.927[95%CI(22.154-220.714)] and336.92[95%CI(79.080-1435.5)], respectively; the poolednegative likelihood ratio are0.196[95%CI(0.061-0.631)],0.238[95%CI(0.019-0.518)],0.098[95%CI(0.043-0.220)],0.234[95%CI(0.122-0.451)] and0.217[95%CI(0.100-0.472)], respectively; the pooled area under curve are0.9649,0.9595,0.9686,0.9797and0.9952. When compare the diagnostic accuracybetween HSG,2DTVS,3DTVS,2DSHG and3DSHG, P<0.05makes a big statistical significance; The accuracy of pairwise comparisons to theremaining diagnostic imaging methods all show P>0.05, suggesting thatthe difference is not statistically significant.Conclusion:HSG is not applicable for IUA diagnosis due to its maximumrate of misdiagnosis,while2DTVS is not suitable for IUA screening because of its highest rate of missed diagnosis.With a high degree of sensitivity andspecificity and the lowest rate of misdiagnosis,3DTVS can be used as theprimary means of screening IUA. The diagnostic accuracy of3DSHG was betterthan the other imaging diagnostic methods,so it can be recommended for theconfirmed diagnosis and classification for IUA. |