RÉSUMÉ
BACKGROUND: The study aimed to evaluate the rate of endometrial sampling (ES) failure, predictive factors of success, and reliability as diagnostic methods of Endosampler versus Novak. METHODS: A retrospective single-center study was carried out with all patients who underwent ES via Endosampler or Novak in 2020 and 2021. Demographic data, personal background, and histopathologic results were evaluated. RESULTS: Eighty-six patients underwent ES by Novak and 90 by Endosampler. The failure rate of ES was 43.2% with lower values for Endosampler (33.3% vs. 53.5%, P<0.05). Age, biopsy device, menopausal status, indication for biopsy, and amount of sample collected were predictive factors of failure. Analyzing each device, Endosampler was only affected by menopausal status. Only 50% in Novak and 62.5% in the Endosampler group of endometrial neoplasia cases were detected by these methods. Analyzing the performance for endometrial neoplasia (EN), we obtained higher values of sensitivity and accuracy for Endosampler (62.5% vs. 50.0% and 83.3% vs. 72.7%), respectively. CONCLUSIONS: In our study, the failure rate obtained was in line with other previous studies. Menopausal status, age, type of biopsy device, indication for biopsy, and amount of sample collected affected ES performance. Analyzing diagnostic performance for EN, we found that these methods have better reliability for positive results than for negative ones, which may indicate the need for further evaluation in cases of high clinical suspicion. In short, we obtain a higher rate of success rate in Endosampler devices and better performance in diagnosing EN, which is the major objective of an ES.
