Table 3.

Area under the receiver operating characteristic curve (AUC) of the rest HARP ratio and the Valsalva HARP ratio in the prediction of successful trials versus those in which there was dislodgment/failure to relieve POP symptoms. HARP ratio = levator hiatal area to ring pessary size

Trials	Parameter (cm)	AUC (95% CI)	p value	Best cut-offa
All successful trials or those in which there was dislodgment/failure to relieve POP symptoms (n = 126)	Rest HARP ratio	0.63 (0.53–0.73)	0.017	HAval/pessary size ≤ 5.00 (sensitivity 0.68, specificity 0.67)
	Valsalva HARP ratio	0.67 (0.58–0.77)	0.001
Trials of women without complete avulsion (n = 77)	Rest HARP ratio	0.65 (0.51–0.78)	0.039	HArest/pessary size ≤ 2.94 (sensitivity 0.59, specificity 0.63)
	Valsalva HARP ratio	0.59 (0.45–0.73)	0.222
Trials of women with complete avulsion (n = 49)	Rest HARP ratio	0.56 (0.39–0.72)	0.497	HAval/pessary size ≤ 5.13 (sensitivity 0.79, specificity 0.72)
	Valsalva HARP ratio	0.79 (0.65–0.92)	0.001

Bold indicates the stastistically significant parameters

Sensitivity (i.e., of all successful trials, percentage that the model predicts as successful)

Specificity (i.e., of all trials in which there was dislodgment/failure to relieve POP symptoms, percentage that the model predicts as trials in which there was dislodgment/failure to relieve POP symptoms)

HAval maximal Valsalva maneuver, HArest hiatal area at rest

^aBest cut-off in the prediction of successful trials