Aims For the diagnostic workup of postmenopausal bleeding, histological examination of the endometrium is frequently performed. Failure of endometrial sampling due to insufficient material is often reported but objective criteria for quality assessment of endometrial biopsies are lacking. The aim of the present study is to evaluate the association between the amount of tissue obtained by endometrial sampling and the diagnostic accuracy, and to establish a cut-off level for a minimal amount of tissue required for a conclusive diagnosis.

Methods For this retrospective cohort study, clinicopathological data and Pipelle endometrial sampling slides of 139 patients who were treated with hysterectomy were collected. The surface of evaluable endometrial tissue was measured by means of structured digital assessment. The correlation between the predictive values in relation to the endometrial tissue surface was calculated for different cut-off values.

Results The median endometrial tissue surface was 4.7 mm² (range 0.4–156.4) for benign endometrium, 27.8 mm² (range 0.0–208.4) for premalignant endometrium and 43.8 mm² (range 0.0–223.6) for malignant endometrium. There was a significant association between the endometrial tissue surface and the correctness of diagnosis. A minimal endometrial tissue surface of 35 mm² could be defined, for positive and negative predictive values of 92.6% and 85.7%, respectively.

Conclusions The diagnostic accuracy of Pipelle endometrial sampling is associated with the amount of endometrial tissue surface, with a minimal cut-off value of 35 mm² required to classify an endometrial sample as conclusive. Quantification of endometrial tissue can contribute to standardisation of quality assessment of endometrial samplings.