KẾT QUẢ BƯỚC ĐẦU ỨNG DỤNG MÔ HÌNH CHẨN ĐOÁN

Abstract

Background: The shift towards minimally invasive, radiation-free imaging is a key paradigm in female infertility evaluation. The "One-stop" model is based on Hysterosalpingo-Foam Sonography (HyFoSy), a two-phase examination — a saline-infusion phase to assess the uterine cavity and a foam-contrast phase to assess tubal patency — allowing simultaneous evaluation of the cavity and tubes in a single visit. However, real-world Vietnamese data on its diagnostic yield, tolerability, and agreement with the surgical reference standard remain limited.

Objectives: (1) To determine the prevalence of utero-tubal abnormalities detected by HyFoSy; (2) to evaluate clinical tolerability and safety via the Number Rating Scale for Pain (NRS) and adverse events; and (3) to assess the agreement between sonographic findings and the surgical/hysteroscopic/histological reference standard among patients who underwent intervention.

Methods: A retrospective cross-sectional descriptive study at Tam Anh General Hospital: 480 cases with baseline characteristics and HyFoSy results (2025 to July 15, 2026) were included to determine prevalence rates, of which 153 cases performed from 04/2026 had their baseline characteristics fully recorded and were assessed for NRS/safety scores; and 35 cases with subsequent interventions were included for comparison.

Results: Among 480 cases, 88.7% showed bilateral tubal patency and 80.4% had normal uterine cavities; the most common lesions were endometrial polyps (9.2%) and unilateral tubal occlusion (10%). In 153 cases, tolerability was excellent, with mean NRS scores of 1.4 ± 1.5, 2.8 ± 1.9, and 0.9 ± 1.3 during the saline phase, the foam-contrast phase, and post-procedure, respectively; 98.7% had no adverse events. Among the 35 intervened, test-positive cases both yielded intervention results that were consistent with the HyFoSy results.

Conclusions: In this exploratory retrospective cohort, the "One-stop" HyFoSy model appeared safe and well tolerated, with high agreement with the reference standard among intervened cases. As the correlation was assessed only in test-positive, intervened cases, diagnostic accuracy (sensitivity, specificity, negative predictive value) could not be estimated; prospective studies with a complete reference standard are warranted.