DIAGNOSTIC ACCURACY OF HYSTEROSCOPY AND AGREEMENT WITH HISTOPATHOLOGY IN WOMEN WITH ABNORMAL UTERINE BLEEDING: A CROSS-SECTIONAL STUDY IN BASRA, IRAQ

Abnormal uterine bleeding is frequent and sometimes disabling with structural, hormonal, systemic and malignant causes. Hysteroscopy can be used to directly inspect the uterine cavity and to biopsy a target, however it cannot be used to replace histopathology in the diagnosis of endometrial hyperplasia, atypia or endometrial carcinoma. Aim to report hysteroscopic and pathological results in women with AUB and to compare overall the correlation between hysteroscopy and pathological results. This was a single-center cross-sectional diagnostic-accuracy study of women of childbearing age presenting with abnormal uterine bleeding and who were treated by hysteroscopy in Basra Maternity and Children Hospital, Basra, Iraq. The index test was hysteroscopy, and reference standard was histopathology. Descriptive data, overall sensitivity, specificity, predictive values, accuracy and Cohen kappa were determined. The analitic cohort in the Results included 125 women; the time of ending the study and the completeness of the histopathological verification needs author verification. Mean age was 40.3 ± 9.7 years. The most common presentation was heavy menstrual bleeding (41.6%). Both hysteroscopy (29.6%) and histopathology (32.0%) showed the most frequent finding endometrial polyps. Based on the reported overall 2 × 2 table, hysteroscopy had a sensitivity of 94.6% (95% CI 88.0%–97.7%), specificity of 81.2% (64.7%–91.1%), positive predictive value of 93.6% (86.8%–97.0%), negative predictive value of 83.9% (67.4%–92.9%), and accuracy of 91.2% (84.9%–95.0%) for detecting any non-normal histopathological finding. Agreement was substantial (κ = 0.77; p < 0.001). Estimates specific to lesions are re-calculable from source data. There was good overall agreement between hysteroscopy and histopathology in this selected hospital cohort, and it was helpful in detecting intracavitary abnormalities. Histopathological examination cannot be omitted for the confirmation of hyperplasia, atypia and malignancy. The estimates of the various lesions made for diagnosis should be checked again prior to submission of the findings for publication.