Sacral Neuromodulation (SNM) is an established therapy for treatment of overactive bladder and non-obstructive urinary retention. SNM is surgically implanted, placing therapy contacts near the sacral nerve with a pulse generator to power and control therapy parameters. Sensory and motor thresholds for stimulation-induced responses are thought to reflect the proximity of the therapy contacts to the target nerve and are important for clinical troubleshooting of the therapy [1,2]. Equivalent motor thresholds are also used in preclinical testing as essential ‘physio-markers’ of relative stimulation impacts on the targeted nerve [3]. Our study aims to quantify the presence of acute testing outcomes, recorded therapy settings, and specific efficacy documentation from routinely captured clinical and surgical patient records.

We conducted a retrospective study of sensory and motor testing results and related therapy parameters for SNM patients receiving therapy implantation and therapy between January 2016 and December 2019. Patient records were identified using surgical and clinical schedules for SNM patients that were implanted or undergoing followup care at the University of Minnesota Clinic and Surgical Center (Minneapolis, MN). For each visit relating to SNM, we transcribed surgical notes and clinical records related to SNM sensory or motor testing, device therapy settings, therapy efficacy or other notes related to SNM therapy. Sensory threshold is the lowest stimulation setting (e.g. amplitude) at which a paresthesia or similar percept is produced and reported by a patient [1,2].  Motor threshold is the lowest stimulation amplitude that produces an observed muscle contraction, such as a bellows or toe movement, or other detectable motor response. We also captured any data relating to patient diagnosis and clinical signs and descriptions of their condition, gender and age. Once recorded, we summarized data for each patient according to visit type and variables as described. We did not seek out test outcomes, therapy parameters or patient-specific information that may have been stored in corporate databases, clinical programming devices, the implanted devices, themselves, or other informal records.

Sensory and motor testing notes were assessed using six criteria: 1) Qualitative mention of testing outcome (e.g. “good”), 2) inclusion of the motor or sensory amplitude of stimulation (e.g. value), 3) patient reported (sensory) or observed (motor) anatomical location of the response, 4) additional quantified non-amplitude threshold stimulation parameters, such as pulse-width or frequency, and 5) inclusion of specific electrical contacts tested, 6) discomfort documented during testing. 

A total of 40 patients were included in the study and reported surgical and clinical visits for SNM from June 2017 - November 2019.

The study included 27 females (68%) and 13 males with a mean age of 58.6y (SD=15.6).  Primary clinical symptoms included urgency (n=30/40, 75%), frequency (n=27, 68%), urinary retention (18%) and OAB (10%).  
For the clinical-surgical visits evaluated (127), we recorded 25 percutaneous surgeries (PNE), 6 stage 1 implants (Stage1), 33 stage 2 implants (Stage2), and 8 device removals. Follow up visits included 15 PNE follow-ups, 25 stage 2 follow-ups and 15 nonsurgical follow-ups. The mean number of SNM-related visits per patient was 3.7 (range=1-26). 

For sensory threshold testing, qualitative outcomes were present in patient records for 80% and 67% of patients during PNE and Stage1 implantation visits, respectively, but quantitative thresholds were included for less than 50% of these visits (Fig 1). Anatomical sensory locations were recorded for 68% of PNE visits but for almost no other visits.  Postoperative sensory test values were observed less frequently (<35% Stage; <10% Stage 2) or were completely absent (no sensory anatomy locations were observed).  Sensory testing data were also lacking in most follow-up visit records (<10% included quantitative sensory thresholds). No sensory thresholds included other parameters of stimulation or the stimulation contacts tested.

For motor threshold testing, qualitative observations were observed for nearly 90% and 100% of patients during PNE and Stage1 implantation visits, respectively (Fig 2). But again quantitative data were included for less than 70% and 90% of these visits. Anatomical motor location was recorded for >80% of the PNE implantations and Stage 1 visits but were much smaller for Stage 2 visits (<40%) and not present for follow-up visits. Similar to sensory testing, motor testing data did not include adequate detail of other test stimulation parameters or the specific lead contacts tested. No surgical visits for device removal included motor or sensory testing.

Patient records also lacked adequate information regarding therapeutic stimulation parameters being delivered. Only 6 of 40 patients (15%) included records relating to therapy settings. Only 8 of 119 (7%) possible visits included any information regarding therapy settings. When a setting was included, the most common mention was as a device program (e.g. ‘Program 1’), but not as a quantitative value, such as amplitude, and the information typically lacked electrical units, such as mA or V. No therapy records included specific mention of the lead contacts programmed to deliver therapy.  There was also no inclusion of other therapy parameters such as pulse-width, frequency or duty cycle. In general, patient records did not include information that would allow parameter-based differentiation of efficacy (e.g. across parameter or testing outcomes).  Using patient records alone, it would be impossible to replicate an effective therapy parameter or avoid an ineffective parameter.

Finally, specific efficacy descriptions or quantification relating to SNM device settings or measurements were also lacking in the patient records. Of the 95 visits that could have included efficacy description or notes regarding efficacy of delivered therapy, only 28 (29%) of the visit records included at least a qualitative mention of efficacy. Only 18 records (16%) included some level of quantitative symptoms such as ‘50% reduction’ or ‘70% improvement in symptoms’. Only 20 (50%) of the 40 patients had records that included quantitative descriptions of efficacy. For 16 (80%) of these 20 patients, the quantification described improved symptoms.

For our single institution, information regarding SNM threshold testing, therapy delivery settings or efficacy detail were inconsistently captured during patient care. While additional stimulation parameter data may exist in implanted devices or corporate servers, these data storage locations do not facilitate clinical utilization to guide patient care or independent clinical research. Similarly, efficacy or other device data may exist in nursing notes or other locations that are inaccessible for directly interpreting device data.  For patient records that included sensory and motor thresholds, few records included quantitative stimulation amplitudes, therapy delivery contacts or other parameters to allow a transparent understanding of patient therapeutic opportunities or facilitate improvements to therapy technology through design inputs.

Current patient records do not adequately document acute SNM stimulation testing, therapy parameters or efficacy details at our single center. Stimulation thresholds, including quantitative parameters and contacts used, are thought to be important tools for improving therapy in individual patients [1,2] and are routinely captured and valuable in preclinical models [3]. Increasing the completeness and quality of these clinical data are important for making meaningful therapy adjustments and as essential design inputs needed to improve implantable neurotechnologies. There is a significant opportunity to address these critical needs through increased quality of documentation or improvements in threshold testing, therapy parameter documentation and efficacy tracking.

Goldman HB, Lloyd JC, Noblett KL, et al. International Continence Society best practice statement for use of sacral neuromodulation. Neurourol Urodyn. 2018;37(5):1823-1848.
Cohen B, Tunuguntla H, Gousse A. Predictors of success for first stage neuromodulation: motor versus sensory response. J Urol. 2006;175:2178Y2181.
Zirpel L, Su X, Wotton J, Reinking N, Steiner A, Nelson DE, Brink TS. Correlation of sacral nerve lead targeting and urological efficacy: motor mapping, electrode position, and stimulation amplitude. Transl Androl Urol. 2016;5(S2):AB291.