Performance of ultrasound and rates of guideline-concordant care in a diverse postmenopausal bleeding cohort
Professional society guidelines endorse the use of ultrasound to evaluate an initial episode of postmenopausal bleeding, with endometrial sampling reserved for patients with an endometrial thickness exceeding 4 mm. However, the data supporting this ultrasound-based approach are derived from cohort studies with minimal racial and ethnic diversity. This study aimed to evaluate rates of adequate endometrial visualization, prompt guideline-concordant care, and identify risk factors for lack of timely indicated sampling in a multi-institutional cohort of patients who underwent ultrasound to evaluate postmenopausal bleeding. Patients aged ≥45 years who underwent transvaginal ultrasound for an initial coding-based diagnosis of postmenopausal bleeding were eligible for this retrospective cohort study. Ultrasound findings of the endometrium were classified as thin (endometrium ≤4 mm), thick (endometrium >4 mm), or inadequate (endometrium not completely visualized). Associations between clinicodemographic factors, ultrasound findings, and follow-up patterns were analyzed. We defined care as guideline-concordant if patients with a thin endometrium on the index ultrasound did not undergo biopsy, or if those with a thick or inadequately visualized endometrium received indicated endometrial sampling within 3 months after the index ultrasound. We analyzed outcomes using univariable and multivariable logistic regressions. A total of 3614 patients were included in the study (48.4% non-Hispanic White, 28.5% non-Hispanic Black, 9.6% Hispanic, 3.0% Asian, and 10.4% other/missing). Rates of inadequate ultrasound examination varied between racial/ethnic cohorts (non-Hispanic White 5.9%, non-Hispanic Black 18.8%, Hispanic 12.7%, and Asian 9.2%). Uterine fibroids were present in 71.5% (737/1031) of non-Hispanic Black patients, as opposed to 43.3% (757/1750) of non-Hispanic White patients, and were strongly associated with lower odds of adequate ultrasound (odds ratio, 0.29 [95% confidence interval, 0.23-0.38]; P<.001). Rates of endometrial sampling were 53.5% (209/391) after an inadequate ultrasound and 75.3% (1329/1765) after a thickened endometrium finding. Guideline-concordant care rates were 85.7% for non-Hispanic White (1500/1750), 78.0% for non-Hispanic Black (804/1031), 83.6% for Hispanic (290/347), and 82.7% for other/unknown (402/83) patients. Compared with non-Hispanic White patients, non-Hispanic Black patients were less likely to receive guideline-concordant care (adjusted odds ratio, 0.64 [95% confidence interval, 0.52-0.78]). Non-Hispanic Black patients with postmenopausal bleeding have a higher rate of inadequate ultrasound examination due to fibroids and are less likely to receive prompt guideline-concordant care. An ultrasound-based workup of postmenopausal bleeding risks delays in endometrial cancer diagnoses and may contribute to racial disparities in outcomes. A universal biopsy approach for postmenopausal bleeding may be considered to decrease variation in and improve guideline-concordant care.
