Sacral neuromodulation (SNM) is a treatment for several pelvic dysfunctions. Since SNM is expensive, the definitive system is implanted after some evaluative steps. Many studies tried to identify factors predicting SNM success in order to identify directly patients responding adequately, but none have proposed a specific tool yet.
This study aimed to analyse the cases addressed to SNM pathway by a tertiary referral centre and identify factors predicting SNM outcomes.

After approval by our Institutional Review Board, we performed a retrospective analysis of data of patients undergoing SNM pathway by our centre from Jan 2010 to Dec 2020. By our center, a Peripheral Nerve Evaluation (PNE) test is performed using a temporary electrode at the S3 foramen which comes out from patient's back and is connected to an external neurostimulator (ENS). If there is a symptom improvement ≥ 50%, the patient is addressed to the implantation of a quadripolar electrode, always connected to an ENS. In case of confirmed symptom improvement, the quadripolar electrode is connected to an implantable pulse generator (IPG). Through this study, we evaluated factors predicting treatment success, from PNE test to IPG positioning (Figure 1). A nomogram was constructed considering factors which were independently associated with the outcomes. The nomogram predictive ability was tested using concordance index and compared with several advanced models obtained through machine learning algorithms. Based on this nomogram, an application software was developed to estimate rapidly the predicted success rate.

The study considered 536 cases undergoing PNE for different indications: overactive bladder (OAB), urinary retention (UR), and chronic pelvic pain (CPP). Univariate analysis highlighted female gender, younger age, and SNM indication (OAB > UR and CPP) were significantly correlated with the treatment success. Considering these variables, we built a nomogram (Figure 2) with a valid predictive ability (concordance index = 0.744). The logistic regression model showed comparable accuracy, sensitivity, and specificity to sophisticated statistical models. Based on the nomogram an application software was successfully built.

Our nomogram with the associated application software could guide clinicians to refer patients straightforward to definitive SNM implant in favourable cases (e.g., young women suffering from OAB) and to perform patient counselling furnishing realistic expectations about this procedure, especially in subjects with low success rates (e.g., old men with UR or CPP). This is in line with the ethical rules of an appropriate informed consent and with a patient-centred care to avoid unrealistic expectations and reduce the risks of medico-legal issues.

This study suggested gender, age and pelvic dysfunction type were factors predicting SNM success. We validated a nomogram which could be used easily in routine clinical practice, also thanks to the application software.

Continence 7S1 (2023) 100864
DOI: 10.1016/j.cont.2023.100864