Concentric needle electrodes are superior to perineal surface-patch electrodes for electromyographic documentation of urethral sphincter relaxation during voiding

Authors


Sangeeta T. Mahajan, Division of Female Pelvic Medicine and Reconstructive Surgery, Departments of Obstetrics and Gynecology and Urology, University Hospitals of Cleveland, MAC 7015, 11100 Euclid Avenue, Cleveland OH44106, USA.
e-mail: sangeeta.mahajan@uhhs.com

Abstract

OBJECTIVE

To compare interpretations of electromyographic (EMG) recordings from perineal surface patch electrodes (PSPEs) to those from urethral concentric needle electrodes (CNEs) during voiding.

PATIENTS AND METHODS

Consenting women underwent urodynamic testing with a 30 G, 3.8 cm CNE at the 12 o’clock position in the striated urethral sphincter muscle, and with PSPEs placed at the 2 and 10 o’clock positions around the anus. Pressure-flow studies were conducted with simultaneous input from both EMG electrodes. Representative, de-identified paper copies of EMG signals were assembled by chronology and electrode type. Six examiners unaware of the patient details were asked to determine if the tracings were interpretable and whether there was quiescence of the urethral sphincter motor unit during voiding. The agreement between the interpretations of each tracing was assessed using McNemar and κ statistics.

RESULTS

Twenty-two women undergoing urodynamic testing for incontinence (16), voiding dysfunction (two) or urinary retention (four) participated in this study. CNE tracings were consistently more interpretable than PSPE tracings (mean 89% vs 67%). When tracings were interpretable, a significantly higher percentage of CNE EMG tracings (mean 79%) had urethral sphincter motor unit quiescence than PSPE EMG tracings (mean 28%). The κ values for agreement among the reviewers’ interpretations were highly variable and none were statistically significant. Reviewers unanimously agreed on only 12 of the 44 tracings, and 11 of these showed quiescence when using a CNE.

CONCLUSIONS

CNEs are more often interpretable than PSPEs for determining motor unit quiescence during voiding. CNE EMG appears to have greater clinical utility for central reading than PSPEs.

Abbreviations
EMG

electromyography(ic)

CNE

concentric needle electrode

PSPE

perineal surface patch electrode.

INTRODUCTION

A urodynamic evaluation of urinary storage and emptying is common in women with urinary incontinence or voiding dysfunction. Normal urination is characterized by urethral relaxation followed by detrusor contraction [1]. Urethral relaxation may be documented by urethral pressure measurements or, more typically, using electromyography (EMG). EMG of the urethral sphincter during filling and voiding cystometrography remains the best available means to document relaxation of the striated urethral sphincter during voiding. Although various techniques are available to obtain EMG readings, i.e. surface patch, wire or concentric needle electrodes (CNEs), urethral CNEs remain the standard for measuring neuromuscular activity due to their ability to isolate electrical activity from specific muscle fibres within a 0.5-mm radius of the tip.

Difficulties associated with correct needle placement, patient discomfort and the requirement for limited patient mobility during testing have made urethral CNE less favoured in clinical settings. The presumed comparability of results, ease of electrode placement and excellent patient tolerance have resulted in the preferred use of (perineal) surface patch electrodes (PSPEs) to document urethral sphincter activity, and urodynamic test results interpreted as though CNE and PSPE EMG recordings are interchangeable.

Although PSPEs are undeniably more comfortable than CNEs, their reliability remains unconfirmed. Studies showing the separate innervation of the levator ani muscles and urethral sphincter suggest that perineal measurements of levator ani will not accurately reflect striated urethral sphincter muscle activity [3]. Vereecken and Verduyn [4] reported a dissociation of urethral and anal sphincter activity while simultaneously recording EMG activity of the anal sphincter, levator muscles and external urethral sphincter using wire electrodes, particularly in individuals with lower urinary tract pathology. Furthermore, volume conduction is an inevitable artefact of PSPEs, making external urethral sphincter EMG activity impossible to differentiate from the activity of other musculature in the region.

The aim of the present study was to compare the ability of independent reviewers to interpret EMG recordings from PSPEs and CNEs recorded to document external urethral sphincter activity during pressure-flow studies.

PATIENTS AND METHODS

Twenty-two women undergoing urethral sphincter EMG with CNE as part of their routine clinical urodynamic testing were enrolled in the present trial, approved by the Institutional Review Board. A 30 G, 3.8 cm CNE was placed ≈ 5 mm ventral to the urethral meatus into the striated urethral sphincter muscle and connected to a pre-amplifier on the electrodiagnostic instrument. The placing of the CNE was confirmed by auditory findings and by the morphology of the motor unit action potential. Two PSPEs (Neotrode®, ConMed Corp., Utica, NY, USA) were then placed at the 2 and 10 o’clock positions around the anus, and connected to a second pre-amplifier on the electrodiagnostic instrument. The electrodes were then covered with tape to prevent them getting wet. The electrodiagnostic instrument processed inputs from the PSPE and CNE, and the two EMG tracings were visualized simultaneously on separate channels. Recording parameters included a bandwidth of 20–10 000 Hz, sensitivity of 50–100 µV and sweep speed of 10 ms per division. The Viking IVp electrodiagnostic instrument (Nicolet Instrument Corporation, Madison, WI, USA) was used to continuously record EMG activity throughout the entire urodynamic study.

Multi-channel urodynamic testing was conducted with the patients seated in a birthing chair reclining at 45°, using 8 F dual-tip microtip transducers (Millar Instruments Inc., Houston, TX, USA) placed in the bladder, urethra and either vagina or rectum. A model 6000 (LifeTech Inc., Houston, TX, USA) multichannel urodynamic system processed and displayed the recorded data. The bladder was filled retrogradely with room-temperature saline at 80 mL/min. Representative paper copies of signals from both types of EMG electrodes were printed at (i) maximum cystometric capacity before an attempt to void, (ii) during an attempt to void, and (iii) after (ii), when the subject was no longer attempting to void.

De-identified paper copies of EMG signals from both electrode types at each of the three times were assembled by electrode type and chronology to create two series of EMG tracings per patient, one showing PSPE EMG outputs and the other CNE outputs (e.g. Figs 1 and2). The series of three tracings were labelled as: ‘prevoid’, ‘voiding’, and ‘postvoid’, regardless of whether or not the study participant was able to void. Tracing sets were randomly intermixed and assembled into a packet for review by examiners who were unaware of patients’ clinical status. The electrode type used to obtain each EMG tracing set was not indicated. In all, 44 tracing sets were thus created for the 22 patients.

Figure 1.

Examples of de-identified paper copies of CNE EMG signals from two different patients before, during and after voiding, assembled by electrode type and chronology. a, an example of a tracing set determined to be interpretable by all examiners; b, another tracing set which two examiners determined to be uninterpretable, two examiners thought it showed motor unit quiescence, and two thought it did not.

Figure 2.

Examples of de-identified paper copies of PSPE EMG signals from two different patients before, during and after voiding, assembled by electrode type and chronology. a, an example of a tracing set determined to be interpretable by all six examiners; b, a tracing set from a different patient which four examiners found to be uninterpretable and two thought to be interpretable, but not showing motor unit quiescence.

Six reviewers familiar with use of EMG examined tracings during urodynamic testing, including four urogynaecologists, one female urologist, and one physiatrist board-certified in EMG. Three of the reviewers were external to Loyola Medical Center. All examiners were unaware of the patient's clinical or urodynamic findings. Each examiner was asked to determine if tracings were interpretable (according to their practice), and if so, whether the tracing appeared to document motor unit quiescence with attempts to void. All examiners completed their observations independently and were unaware of the other reviewers’ interpretations. The data were analysed using McNemar and κ statistics to determine the agreement between EMG tracings and examiner interpretations of each tracing.

RESULTS

Of the 22 women who participated in this study, the indications for urodynamic testing included incontinence in 16, voiding dysfunction in two and urinary retention in four. Fourteen of the patients were able to void during the study; eight were unable to void, including four with urinary retention, one with voiding dysfunction and three with no previous voiding complaints. Medical comorbidities, previous urogynaecological procedures and urodynamic diagnosis were not correlated to the interpretability of EMG tracings.

CNE EMG tracings were more often rated as interpretable than PSPE EMG tracings (Table 1); a mean (range) of 89 (73–100)% of CNE EMG tracings were judged to be interpretable by the examiners. All examiners agreed that PSPE EMG tracings were more often uninterpretable, at 67 (9–96)%.

Table 1.  Comparison of interpretations of tracings by examiner
Electrode/resultExaminer
123456
Total interpretable (N = 44), n393618274342
CNE EMG
Interpretable, %:
 Quiescence826864779686
 No quiescence1818 9 0 5 9
 Not interpretable 0142723 0 5
Number with quiescence171413162018
PSPE EMG
Interpretable, %:
 Quiescence3214 9185541
 No quiescence4664 0274155
 Not interpretable23239155 5 5
Number with quiescence 8 4 3 51310
Both EMG tracing sets interpretable, n (%) of 2217 (77)14 (64)2 (9) 7 (32)21 (96)20 (91)
N tracings sets with agreement of both 8 4 2 313 9
κ value agreement between interpretations of both 0.16 0.16N/AN/A 0.12 0.13

When tracings were interpretable, CNE tracings showed motor unit quiescence more frequently than PSPEs (Table 1). Of interpretable CNE EMG tracings, 79 (64–96)% had motor unit quiescence; PSPE EMG tracings did not show motor unit quiescence as frequently, at 28 (9–55)%, even when interpretable.

When both the CNE and PSPE EMG tracings for a given patient were interpretable, the intra- and inter-reviewer variation in assessments of motor unit quiescence was large (Table 1). Reviewers found that 9–96% of patients had interpretable tracing sets for both electrode types. When both tracings were interpretable for a given patient there was significant variation in the rates of agreement as to whether quiescence did or did not occur in each of the two tracing sets for the 22 patients. Of patients with two interpretable EMG tracing sets, there was agreement between individual examiner interpretations of quiescence for both tracings in 54 (29–100)% of patients. The κ values for agreement between interpretations of CNE and PSPE EMG tracings between individual reviewers could not be determined for two reviewers and were highly variable for the remaining examiners (mean 0.14). None of the κ values for agreement were statistically significant (P > 0.05).

There was unanimous agreement between external reviewer interpretations of individual EMG tracings in only 12 of the 44 tracing sets. Of these 12 tracings, 11 were from a CNE; all 12 showed motor unit quiescence.

DISCUSSION

Barrett [6] first described the use of disposable infant surface electrocardiogram electrodes during urodynamic testing to determine perineal relaxation, as an indirect indicator of simultaneous external urethral sphincter relaxation during voiding. Using observational data, Barrett showed grossly similar findings on SPE EMG testing compared to peri-anal wire electrodes, previously the standard method of determining perineal relaxation. A later study by Barrett and Wein [7] examined changes in PSPE EMG with uroflowmetry, and concluded that PSPEs adequately detect perineal relaxation when attempting to diagnose neurological voiding dysfunction.

Although patients better tolerate PSPE EMG electrodes, their reliability when compared to concurrent standard CNEs is doubtful. Volume conduction results in the compounding of motor unit signals from all the muscles of the perineal body, makes signals from an individual muscle difficult to identify. Given the differing innervation of the levator ani muscles and the external urethral sphincter it is unlikely that EMG activity of the levator ani muscles recorded with PSPEs adequately reflects motor unit activity of the external urethral sphincter.

In the present study CNE tracings were more often rated as interpretable, and showed motor unit quiescence more often, with greater inter-reviewer agreement when interpreted by six reviewers. Significantly, of the 12 tracings with unanimous reviewer agreement on findings, 11 were obtained using a CNE. Even when PSPE tracings were interpretable they were less likely to show the motor unit quiescence documented by simultaneous CNE EMG.

The present study may be limited by several important factors. First, there were relatively few patients (22), possibly inadequate to detect a significant correlation between reviewer interpretations of the two electrode types. It is possible that assessing more patients might have resulted in different clinical findings. Second, the six experts who reviewed the tracing samples determined the outcome of the study. Although all the reviewers had clinical and research experience with the use of urethral EMG, it is possible that selecting different expert examiners might have altered the study outcome. Third, we relied on expert reviewers to apply their definitions of ‘interpretable’ and ‘presence of motor unit quiescence’ while reviewing EMG tracings, as no established criteria exist for interpretability and there has been no agreement on what constitutes urethral EMG quiescence during voiding studies. Formal study interpretation guidelines defining interpretability and urethral EMG quiescence may have yielded different results. Also, the choice of filter bandwidth may affect both types of EMG tracings, with a lower bandwidth possibly improving the interpretability of PSPE EMG. It is also possible that a comparison of live readings rather than static images would have yielded different results. Another limitation is that reviewers may have had previous opinions about the superiority or inferiority of the different electrode types, and when reviewing an EMG may have had a conscious or unconscious bias about the quality of the tracing based upon the nature of the tracing they believed they were reviewing.

The present results suggest that PSPEs are more frequently judged to be uninterpretable by independent reviewers. Although this may relate to decreased clinical usefulness, the study did not analyse the interpretation by the clinician performing the urodynamic testing and clinical interpretation. However, we recommend the use of a CNE if documenting EMG quiescence during voiding is clinically or scientifically important. This study confirms the clinical impression that although PSPEs decrease patient discomfort with EMG testing, they are not an adequate alternative to the standard CNE.

CONFLICT OF INTEREST

None declared.

Ancillary