Assessment of clinical and ultrasonographic findings in comparison with pathological results in hysterectomy cases

Background:Hysterectomy is a common gynecological surgery. Hence in this study the validity of clinicaland sonographic findings was compared with pathological results in hysterectomy cases in atraining hospital in Tehran, Iran from 2005 to 2012.Materials and Methods:In this cross-sectional descriptive-analytical study, pathological findings of 300 women underhysterectomy in Javaheri Hospital in Tehran, from 2005 to 2012 were analyzed, with respectto the ultrasonographic and clinical findings of the patients. The inclusion criterion washaving hysterectomy in medical documents. Analysis of data among 300 patients wasperformed using SPSS version 13 software. The used statistical methods included ANOVAand Chi-Square tests. The significance level was considered to be 0.05.Results:The mean age of patients was 46.98 ±7.78 years. Atrophic endometrium (42%), proliferativeendometrium (25.3%), and secretory endometrium (11%), and adenomyosis (10%) were themost common pathological uterine findings and the most common pathological ovarianfindings were simple follicular cysts (25.3%) and complex follicular cysts (10%). 81.3% ofpatients had no abnormality in ultrasonographic findings, 18.3% showed leiomyoma and0.3% had air-fluid level. 42.3% bleeding, 14.3% dysmenorrhea, 0.7% prolapse, and 42%suprapubic pain were present and 0.7% had no clinical finding. There was no significantassociation between pathological findings and age, gravid and abortion count (P > 0.05).There was no significant association between clinical findings and age, children and abortioncount (P > 0.05). There was statistically significant association between pathologicalendometrial findings and US results (P < 0.0001) with 87 percent concordance and in 100percent of endometrial carcinoma and sarcoma and endometrial polyps the sonographyshowed no positive finding and in chronic endometritis and placenta previa showed morepositive results according to sonography. The sensitivity, specificity, and PPV for USfindings compared with endometrial pathological findings as gold standard were 73%, 88%,and 85%, respectively. There was statistically significant association between pathologicalendometrial findings and clinical findings (P < 0.0001) with 84 percent concordance andchronic endometritis showed more asymptomatic status with 20 percent and disorderedproliferative endometrium had most pain with 90.6%, and in endometrial sarcoma andendometrial polyp the dysmenorrhea was most seen clinical finding each one with 100percent and also the bleeding was most common finding in endometrial carcinoma cases with100 percent. The sensitivity, specificity, and PPV for total clinical findings compared withendometrial pathological findings as gold standard was 45%, 40%, and 43%, respectively.The sensitivity, specificity, and PPV for suprapubic pain were 41%, 34%, and 22%,respectively. The sensitivity, specificity, and PPV for bleeding were 24%, 18%, and 29%,respectively. The sensitivity, specificity, and PPV for prolapse were 37%, 28%, and 42%,respectively. The sensitivity, specificity, and PPV for dysmenorrhea were 44%, 30%, and26%, respectively. There was statistically significant association between clinical findingsand US results (P < 0.005) with 90 percent concordance rate and the most common findingwith negative results was bleeding with 85 percent and the lower one was prolapse with 50percent. The patients with endometrial thickening more than 7 mm were more probable tohave myoma with a sensitivity of 85%, specificity of 81%m and PPV of 74%. Also patientswith endometrial thickening more than 7 mm were more probable to have dysmenorrhea witha sensitivity of 65%, specificity of 57%, and PPV of 62%. In the patients with leiomyoma inpathology, bleeding was most common symptom and the myoma was the most commonsonographic finding (P < 0.05) showing a sensitivity of 57%, specificity of 66%, and PPV of65% for sonographic findings and a sensitivity of 44%, specificity of 39%, and PPV of 34%for clinical symptoms. The ovarian findings were not related to symptoms and sonographicfindings (P > 0.05). 64.7% of hysterectomies were for uterine problems and 36.3% were forcauses outside the uterine. 93% of hysterectomies were performed appropriately and 7% wereincorrect.Conclusion:Totally, according to our findings, it may be concluded that there is a significant associationbetween sonographic and clinical findings and pathological findings in patients whounderwent hysterectomy and the sensitivity, specificity, and PPV for US findings comparedwith endometrial pathological findings as gold standard was 73%, 88%, and 85%,respectively and the sensitivity, specificity, and PPV for clinical findings compared withendometrial pathological findings as gold standard was 45%, 40%, and 43%, respectively. Sothe pathological findings would let us for definite diagnosis and interpretation of underhysterectomy cases.