Predicting Success Using Response from Sacral Neuromodulation Test Stimulation for Urgency Incontinence

Objectives: Currently, 50% reduction in urgency incontinence episodes (UUIE) is arbitrarily used to determine who receives an internal pulse generator (IPG) after stage I lead implantation for sacral neuromodulation (SNM). However, with this cutoff one-third of women receiving an IPG note inadequate symptom control by 24months. A higher post-lead implant cutoff of 75% reduction in UUIE has been suggested. The aim of this study was to determine an ideal percent reduction in UUIE 2 weeks after stage 1 SNM that predicts success at 24months and compare the accuracy to prior cutoffs.

Methods: Data from a multicenter trial evaluating outcomes of SNM over 24months were utilized to determine an ideal cutoff for percent reduction in UUIE after stage 1 SNM. Receiver operating characteristic curves, maximizing sensitivity and maintaining a minimum specificity of 0.25, were used. The primary outcome was treatment success at 24months defined as Patient Global Impression of Improvement (PGI-I) of 1-3 (very much better to little better), consistent with the parent trial. Secondary outcomes were treatment success at 6 and 12months. All accuracy statistics were internally validated using bootstrapping.

Results: This analysis included 149 women who received an IPG and completed a 2 week post-stage 1 diary. The mean age was 63yo with 7.5% Hispanic, 85% white. Women had a median of 4.7 (IQR 3.7-6.0) daily UUIE at baseline. Over the 24months, 15% received additional treatment for UUI (primarily OAB medications) and 7% had their SNM removed. The median reduction in UUIE 2 weeks post-stage 1, in this population who received an IPG, was 86% (IQR 70-100%). At 24months, 62% of women met treatment success, while 70% and 77% did so at 12 and 6months, respectively. The ideal cutoff for treatment success was 72% (95% CI 64-76%) reduction in UUIE at 24months, 71% (95% CI 63-82%) at 12months and 76% (95% CI 72-86%) at 6months. These cutoffs did not meaningfully vary when treatment success was defined as PGI-I of 1-2 (very much better to much better). Accuracy statistics are presented in Table 1 with the concordance indices < 0.53 and accuracy < 0.60 for all cutoffs at 24 months.

Conclusions: Percent reduction in UUIE after stage 1 is a poor predictor of treatment success at 24months. Using a cutoff of 71-76% improves specificity, while decreasing sensitivity over the traditional 50% reduction in UUIE. For patients preferring to maximize success a cutoff of 71-76% is likely beneficial, whereas 50% may be preferred for highly refractory patients willing to risk no improvement. Therefore, percent reduction in UUIE cutoff for IPG should be optimized to the patient’s treatment goals. These data demonstrate a clear need to incorporate factors other than reduction in UUIE when deciding to proceed with IPG implantation.