OBJECTIVE: To evaluate the accuracy of hysteroscopy and transvaginal ultrasonography (TU), based on a histopathological report from endometrial specimens, in diagnosing endometrial pathology in menopausal women with uterine bleeding. METHODS: Four-hundred and nineteen postmenopausal women with uterine bleeding underwent TU, hysteroscopy and endometrial biopsy. Hysteroscopic and sonographic findings have been evaluated on the basis of the final diagnosis established by histologic examination. Sensitivity, specificity and positive predictive value of TU at an endometrial thickness cut-off point of 4 and 8 mm (double layer technique) and of panoramic hysteroscopy have been detected. RESULTS: Normal and abnormal endometrium was found in 222 and in 197 women, respectively. TU showed sensitivity of 95.1%, specificity of 54.8% and positive predictive value of 63.7% at a cut-off limit of 4 mm. With a cut-off limit of 8 mm the corresponding figures were 83.8%, 81.3% and 79.4%. Hysteroscopy demonstrated a sensitivity of 96.5%, specificity of 93.6% and positive predictive value of 92.6%. The combination of the two diagnostic tools showed a 100% sensitivity, 94.8% specificity and 93.3% positive predictive value. CONCLUSIONS: With cut-off limit of 4 mm, TU can be considered the first choice modality of endometrial investigation in women with postmenopausal uterine bleeding to select patients at risk to carry endometrial pathology. Hysteroscopy is a more accurate technique than TU because of better specificity and must be indicated for all patients showing an endometrial strip more than 4 mm. When an endometrial thickness below 4 mm is detected by ultrasound, hysteroscopy may be indicated on clinical background because of the possibility to miss infrequent (2.5% in our series), but relevant endometrial pathologies. Endometrial sampling should follow hysteroscopic view in all cases showing abnormal or suspicious lesions as well as in all cases with irregularly shaped endometrial lining and/or suboptimal endoscopic vision.
OBJECTIVE: To evaluate the accuracy of hysteroscopy and transvaginal ultrasonography (TU), based on a histopathological report from endometrial specimens, in diagnosing endometrial pathology in menopausal women with uterine bleeding. METHODS: Four-hundred and nineteen postmenopausal women with uterine bleeding underwent TU, hysteroscopy and endometrial biopsy. Hysteroscopic and sonographic findings have been evaluated on the basis of the final diagnosis established by histologic examination. Sensitivity, specificity and positive predictive value of TU at an endometrial thickness cut-off point of 4 and 8 mm (double layer technique) and of panoramic hysteroscopy have been detected. RESULTS: Normal and abnormal endometrium was found in 222 and in 197 women, respectively. TU showed sensitivity of 95.1%, specificity of 54.8% and positive predictive value of 63.7% at a cut-off limit of 4 mm. With a cut-off limit of 8 mm the corresponding figures were 83.8%, 81.3% and 79.4%. Hysteroscopy demonstrated a sensitivity of 96.5%, specificity of 93.6% and positive predictive value of 92.6%. The combination of the two diagnostic tools showed a 100% sensitivity, 94.8% specificity and 93.3% positive predictive value. CONCLUSIONS: With cut-off limit of 4 mm, TU can be considered the first choice modality of endometrial investigation in women with postmenopausal uterine bleeding to select patients at risk to carry endometrial pathology. Hysteroscopy is a more accurate technique than TU because of better specificity and must be indicated for all patients showing an endometrial strip more than 4 mm. When an endometrial thickness below 4 mm is detected by ultrasound, hysteroscopy may be indicated on clinical background because of the possibility to miss infrequent (2.5% in our series), but relevant endometrial pathologies. Endometrial sampling should follow hysteroscopic view in all cases showing abnormal or suspicious lesions as well as in all cases with irregularly shaped endometrial lining and/or suboptimal endoscopic vision.
Authors: Megan A Clarke; Beverly J Long; Arena Del Mar Morillo; Marc Arbyn; Jamie N Bakkum-Gamez; Nicolas Wentzensen Journal: JAMA Intern Med Date: 2018-09-01 Impact factor: 21.873