Connie Rees

103 Table 4.3. Diagnostic Accuracy of Individual Clinical and MRI variables Histopathological adenomyosis diagnosis Sensitivity Specificity PPV NPV PLR NLR Overall accuracy Radiology report 21.8% 91.5% 63.0% 63.7% 2.6 0.9 63.6% Intoxications Smoking Medical history 14.1% 57.3% 31.4% 57.3% 0.3 1.5 51.3% History of curettage 12.8% 89.7% 45.5% 60.7% 1.2 1.0 59.0% History of caesarean section 21.8% 45.3% 37.8% 58.2% 0.4 1.7 51.5% Regular cycle 35.9% 13.7% 40.6% 66.7% 0.4 4.7 47.3% Hormonal medication 44.9% 49.6% 37.2% 57.4% 0.9 1.1 47.7% Endometriosis 56.4% 61.5% 49.4% 67.9% 1.5 0.7 59.5% Uterine Fibroids Symptoms 34.6% 45.3% 29.7% 51.0% 0.6 1.4 41.0% Dysmenorrhoea 65.4% 54.7% 50.0% 71.1% 1.4 0.6 59.9% AUB 34.6% 65.8% 41.5% 60.6% 1.0 1.0 54.2% Chronic pain 37.2% 63.2% 41.4% 60.7% 1.0 1.0 53.6% Subfertility 16.7% 90.6% 59.1% 62.4% 1.8 0.9 62.0% Dyschezia 21.8% 81.2% 45.9% 61.3% 1.2 1.0 58.3% Dyspareunia MRI Features 29.5% 65.0% 37.1% 58.5% 0.8 1.1 51.6% Mean JZmax ³12 mm 78.2% 56.4% 63.5% 82.5% 1.8 0.4 72.2% Mean JZdiff ³5 mm 92.3% 23.9% 51.8% 90.3% 1.2 0.3 58.8% Mean JZ/MYO ³0.4 76.9% 16.2% 50.8% 76.0% 0.9 1.4 55.2% HSI foci 70.5% 89.7% 90.2% 82.0% 6.8 0.3 84.7% MRI = Magnetic Resonance Imaging; PPV = Positive Predictive Value; NPV = Negative Predictive Value; PLR = Positive Likelihood Ratio; NLR = Negative Likelihood Ratio; AUB = Abnormal Uterine Bleeding; JZmax = Maximal Junctional Zone Thickness; JZdiff = Junctional Zone Differential; JZ/MYO = Junctional Zone to Myometrium Ratio; HSI foci = High Signal Intensity foci. Logistic Regression Analysis A univariate and multivariate logistic regression analysis was done to compare the predictive factors in the original study and external validation study. The results of both univariate and multivariate regression analysis are summarised in Table 4.4. In the univariate regression analysis the following variables showed p values <0.10: BMI, endometriosis, uterine fibroids, dysmenorrhea, subfertility, mean JZ, maximal JZ, JZ differential, adenomyosis focus signal

RkJQdWJsaXNoZXIy MTk4NDMw