Transcutaneous electrical nerve stimulation in speech therapy rehabilitation of voice and swallowing function in adults—a systematic review

Abstract Introduction In recent years, a number of clinical trials have been published comparing transcutaneous electrical nerve stimulation (TENS) and traditional speech therapy treatment of voice and swallowing functions, but results have been conflicting. Objective Assess the methodological quality of studies and determine whether TENS is an efficient therapeutic strategy for speech therapy treatment of healthy adults or those with dysphonia and/or dysphagia. Methods The databases used were Medical Literature Analysis and Retrieval System Online (MedLine), Biblioteca Virtual em Saúde (BVS), Cochrane Library and Web of Science (ISI Web of Knowledge). The study was conducted between May 2018 and January 2019, in line with Cochrane Handbook guidelines, and included studies on the use of TENS in healthy adults or those with compromised voice and/or swallowing function. Results After the search and extraction of studies, the following were identified: TENS + VOICE: 7 articles; TENS + SWALLOWING: 5 articles. The studies exhibited medium quality and are heterogeneous, making it difficult to determine their effectiveness and the parameters to be used in future research. There were no statistically significant differences between the use of TENS alone or associated with another therapeutic technique for voice. For swallowing function, one study proved better results in cases of associated techniques – TENS + traditional therapy. Discussion Speech therapy should increase the number of studies published and improve their methodological quality, reassessing methodological criteria. Current clinical practice is not grounded in evidence‐based science. Clinical Message the studies analyzed exhibited medium methodological quality; there are variations in the time, number and periodicity of the therapeutic sessions for TENS; there were no statistically significant differences between the use of TENS alone or associated with another therapeutic technique in voice; there were statistically significant differences between the use of TENS associated with traditional therapeutic in swallowing function.

• there were no statistically significant differences between the use of TENS alone or associated with another therapeutic technique in voice; • there were statistically significant differences between the use of TENS associated with traditional therapeutic in swallowing function. specialists to control pain, enhance muscle performance, stimulate wound healing, and improve sensorimotor recovery after different diseases (Blumenfeld et al., 2006;Guirro et al., 2008). It is a safe, noninvasive, simple, inexpensive, and nonpharmacological method (Mansuri et al., 2020;Santos et al., 2016). It seems that neuromotor function can be influenced by one TES parameter: amplitude (Barikroo et al., 2017;Doucet et al., 2012).
In the field of speech therapy, studies using TES remain scarce and, in some cases, are limited to the preliminary or initial therapy phase, where the technique is used to relax laryngeal muscles (Conde et al., 2017;Guirro et al., 2008;Silvério et al., 2015;Siqueira et al., 2017) or for the rehabilitation of mechanical oropharyngeal dysphagia, promoting muscle contraction (Crary & Carnaby, 2014). Several studies have been conducted in this area, showing the increasing use of new speech therapy resources (Guirro et al., 2008;Santos et al., 2016;Stangherlin et al., 2020). For a new therapeutic approach to be accepted, it should provide the same degree of benefits as proven techniques already in use. In recent years, a number of clinical trials have compared TES with traditional speech therapy for voice disorders (Conde et al., 2017;Fowler et al., 2011;Guirro et al., 2008;Silvério et al., 2015;Siqueira et al., 2017) and dysphagia (Blumenfeld et al., 2006;Crary & Carnaby, 2014;Maeda et al., 2017), but these articles reported conflicting results.
TES has been applied in clinical practice in two forms-motor and sensory electrostimulation. Motor electrostimulation or neuromuscular electrical stimulation (NMES) uses low-intensity electrical currents to simulate the passage of a nervous stimulus to the skeletal muscle, promoting involuntary muscle contraction by depolarizing nerve fibers within the region of application (Bhatt et al., 2015;Crary & Carnaby, 2014;Glanz et al., 1996;Humbert et al., 2012;Simonelli et al., 2019). Sensory electrostimulation or transcutaneous electrical nerve stimulation (TENS) has been used to control pain and tension, reduce fatigue, improve local vascularity, help muscle relaxation and as analgesia in the region of application (Conde et al., 2017;Guirro et al., 2008;Mansuri et al., 2020;Santos et al., 2016;Silvério et al., 2015;Sluka & Walsh, 2003;Stangherlin et al., 2020). The two types of stimulus use percutaneous electrodes to transmit waveforms through the skin to stimulate large diameter nerve fibers (Mansuri et al., 2020;Santos et al., 2016;Silvério et al., 2015;Sluka & Walsh, 2003).
Studies show benefits of TENS or NMES in improving swallowing functions (Barikroo et al., 2017;Barikroo et al., 2018;Berretin-Felix et al., 2016;Blumenfeld et al., 2006;Carnaby & Harenberg, 2013), supporting hyolaryngeal elevation Park et al., 2012), reducing treatment sessions and shortening hospital stays (Blumenfeld et al., 2006). Other studies use TENS to increase oropharyngeal sensory input. This stimulus may increase sensory input to the swallowing center of the brain stem, leading to earlier initiation of deglutition and timely protection of the respiratory airway. Research demonstrates that periodical sensory stimulation may induce cortical neuroplasticity (Ortega et al., 2016;Rofes et al., 2014). As such, TES is still used for swallowing rehabilitation (Barikroo et al., 2017;Carnaby & Harenberg, 2013), and is the most widely studied approach to dysphagia rehabilitation (Crary & Carnaby, 2014). However, there is a lack of evidence in the literature (Crary & Carnaby, 2014), highlighting the relatively weak research designs, small studies, and the use of different electrical stimulation parameters for dysphagic patients (e.g., stroke, older healthy adults, progressive diseases, head/neck cancers).
Given that TENS is a new important therapeutic strategy for voice and swallowing functions, a systematic review was conducted to analyze current evidence on TENS obtained in experimental and quasi-experimental studies with healthy, dysphonic and dysphagic adults without other pathologies (e.g., neurological or head and neck cancer). The information in this review may help establish and develop rehabilitation programs aimed at adults with dysphonia and/or dysphagia.

| METHODS
This systematic review was conducted according to the recommendations of the Cochrane Handbook (Higgins et al., 2019). All the stages described were carried out by two independent examiners, with disagreements resolved by a third examiner.
The following databases were surveyed between May 2018 and January 2019: Biblioteca Virtual em Saúde (BVS), Medical Literature Analysis and Retrieval System Online (PubMed/MedLine), Cochrane Library and Web of Science (ISI Web of Knowledge), with no restriction for language or year of publication.
The search for articles in the areas of voice and swallowing function and their association with TENS was performed in three stages: the first involved two searches for articles in the databases.
This search was conducted separately, considering the voice and swallowing areas; the second stage excluded duplicate articles, namely those contained in more than one database, or appearing as different documents in the same research; and the third involved reading and analyzing the texts considering the established inclusion criteria.
The following descriptors were used on the PUBMED and Cochrane Library platforms: 1-("transcutaneous electric nerve stimulation" OR "sensory electrical stimulation" OR "sensory e-stim" OR "sensory transcutaneous electrical stimulation" OR "sensory TES") AND (voice OR dysphonia OR "voice disorders") AND adults; 2-("transcutaneous electric nerve stimulation" OR "sensory electrical stimulation" OR "sensory e-stim" OR "sensory transcutaneous electrical stimulation" OR "sensory TES") AND (deglutition OR "deglutition disorders") AND adults.

| Study selection
The following article eligibility criteria were established (according to the PICO tool): (1) healthy (cases of vocal or deglutition improvement), dysphonic or dysphagic adult sample (neurological, psychiatric, syndromic and diagnosed head and neck cancer cases were excluded to reduce bias); (2) use of TENS as an intervention method; (3) pre and post-intervention comparison; (4) at least one of the following voice or swallowing parameters: results of a specific clinical evaluation (e.g., vocal quality scale or swallowing scale); self-perception scale/ instruments; diadochokinetic performance; acoustic analysis; face and neck muscle activity; and videofluoroscopy; and (5) published reports of experimental and quasi-experimental studies.
The articles were selected independently by two examiners. The titles were read and those that did not meet eligibility criteria were excluded. Next, the abstracts were read and the studies of those that were not excluded were read in their entirety to select the articles included in this review.

| Data extraction and quality assessment
The material was critically assessed considering the methodological quality and the risk of bias of the selected articles, using a standardized scale (Higgins et al., 2019;Sampaio & Mancini, 2017;Shiwa et al., 2011). To understand TENS application for voice and swallowing function, a critical review of the studies with high-quality methodology was carried out, analyzing the objectives, results and methodological standardization used.
The PEDro scale was developed by the Physiotherapy Evidence Database (Sampaio & Mancini, 2017;Shiwa et al., 2011) for use in experimental studies. The maximum score is 10 points and includes assessment criteria of internal validity and presentation of the statistical analysis used. The first item, related to the inclusion criteria, is considered positive or negative, instead of being scored. For the remaining criteria, 1 point is attributed when quality indicators are present and 0 points when they are absent. The scale is composed of the following criteria: (1) eligibility criteria specified (item not scored); (2) random allocation; (3) concealed allocation; (4) groups similar at baseline; (5) subject blinding; (6) therapist blinding; (7) assessor blinding; (8) less than 15% dropouts (measure of at least one primary outcome in 85% of the allocated subjects); (9) intention-to-treat analysis; (10) between-group statistical comparisons (intergroup comparison of at least one primary outcome) (11) point measures and variability data (report of variability measures and parameter estimation of at least one primary variable) (Shiwa et al., 2011).
The scale was used independently and blindly by two researchers, and no disagreements occurred. After the final score was calculated, the following decisions were made (Sampaio & Mancini, 2017;Shiwa et al., 2011): (a) articles with scores of less than 3 were deemed to have low methodological quality and were excluded; (b) articles with scores greater than or equal to 3 were considered eligible for the next stage. da SILVA AND MANGILLI Following the recommendations of the Cochrane Handbook (Higgins et al., 2019), the risk of bias analysis was also applied to the studies selected for this review. The Cochrane tool provides a framework for assessing the risk of bias in a single result from any type of randomized trial. It is structured into domains, that will be identify based on both empirical evidence and theoretical considerations. The tool was also applied independently and blindly by two researchers, and no disagreements occurred.
The following variables were considered to analyze the use of TENS for voice and swallowing functions: study objective, participants, evaluations methods, TES application method (frequency, intensity, duration, and electrode location) and outcomes.

| Study selection
After the search and extraction of studies, the following were identified: (1) TENS + VOICE: 35 studies, 20 of which were excluded for being duplicates. Seven of the 15 remaining articles met the eligibility criteria. Studies that were not experimental (systematic review or case reports) and those involving other diseases (two neurological, one chest and one digestive) were excluded. (2) TENS + SWALLOWING: 37 studies, 13 of which were excluded for being duplicates. Only 5 of the remaining articles met the eligibility criteria. Nine of the excluded studies involved neurologic patients, six were not experimental and four involved other disorders (three digestive and one pain-related).
The flowchart of the review studies is described in Figure 1.

| Data extraction and quality assessment
Analysis of methodological quality did not exclude any of the articles selected. Table 1 presents the PEDro scale score of the articles. Table 2 shows the results of applying the Cochrane tool.
The summary of the variables analyzed in the articles is presented in Chart 1.
All the voice studies (Conde et al., 2017;Fabron et al., 2017;Fowler et al., 2011;Guirro et al., 2008;Santos et al., 2016;Silvério et al., 2015;Siqueira et al., 2017) used only sensory stimulation, and four investigated the effect of TENS on voice considering otolaryngologist assessment to determine vocal lesions. All the studies (Conde et al., 2017;Fabron et al., 2017;Fowler et al., 2011;Guirro et al., 2008;Santos et al., 2016;Silvério et al., 2015;Siqueira et al., 2017) made audio recordings of the participants, with perceptive-auditory assessment by qualified professionals. In addition, for tension/pain, one study applied the Musculoskeletal Pain Questionnaire (Conde et al., 2017) and another the Nordic The results were not statistically significant. It was determined that the TML and TENS provide muscle relaxation and can may be used together or alone to treat dysphonia.
The group that received TML showed more regular diadochokinetic movements in the vocal folds. The group that received TENS showed no change in diadochokinetic parameters The results were not statistically significant. There was no intergroup difference in the acoustic parameters.
A single TENS or LMT session immediately improved vocal quality.
The group that received TENS reported a more positive change in vocal quality than the group that received only LMT The acoustic parameters showed no significant change after the two treatments.
The group that received TENS showed a decline in vocal/laryngeal symptoms in terms of the frequency and intensity of musculoskeletal pain in the neck and shoulders, as well as improved vocal quality.
The group that received LMT exhibited a reduction in throat and posterior neck pain, but no improvement in vocal quality (Continues) Musculoskeletal Symptoms Questionnaire (Silvério et al., 2015). For voice performance, one study (Fabron et al., 2017) also developed a selfassessment protocol where participants rated their perception on a 10-cm visual analogue scale. Another study (Santos et al., 2016) asked subjects if they perceived a change in their voice after the intervention.
Some researchers  used audio recordings to assess only the diadochokinesis of participants and the data were analyzed by a computer program. We also highlight another study (Guirro et al., 2008) that used the Grade, Roughness, Breathiness, Asthenia, Strain, Instability scale to analyze voice parameters and the Superficial Electromyographic assessment to determine the electrical activity of the suprahyoid, sternocleidomastoid and trapezius muscle.
For swallowing function, three studies used only sensory stimulation (Blumenfeld et al., 2006;Maeda et al., 2017;Ortega et al., 2016) and two used both sensory and motor stimuli (Barikroo et al., 2017;Berretin-Felix et al., 2016). The studies (Barikroo et al., 2017;Berretin-Felix et al., 2016;Blumenfeld et al., 2006;Maeda et al., 2017;Ortega et al., 2016) used multidisciplinary assessment and/or the gold standard swallow test (Videofluoroscopy) to determine whether the participants had dysphagia. Different assessment instruments were used: cough latency time, the Functional Oral Intake Scale and Mini-Nutritional Assessment Short Form in one study (Maeda et al., 2017); swallowing pressure data (lingual-palatal and pharyngeal pressures) and pressure timing in two studies (Barikroo et al., 2017;Berretin-Felix et al., 2016); the Swallow Severity Scale in one study (Blumenfeld et al., 2006); and the Eating Assessment Tool, Volume-Viscosity Swallow Test and Penetration-Aspiration Scale (for videofluoroscopy exam) in another (Ortega et al., 2016).
Another study considered 75% of the motor threshold (Ortega et al., 2016) and, finally, the last (Maeda et al., 2017) used an intensity of 3.0 mA.
One study (Blumenfeld et al., 2006) did not specify the number of sessions the individuals underwent, or how many times a day/week. The authors did not report the number of daily interventions or the total average, only that individuals were stimulated until they achieved the goal of the therapeutic treatment plan. One study (Maeda et al., 2017) conducted the largest number of sessions (20), twice a day for five consecutive days, over a two-week period. Two studies (Silvério et al., 2015;Siqueira et al., 2017) held a total of 12 sessions, twice a week for 6 weeks. The other two studies used 10 sessions, 5 days a week for 2 weeks (Ortega et al., 2016) and two or three times a week (Guirro et al., 2008). The greatest number of articles (Barikroo et al., 2017;Berretin-Felix et al., 2016;Conde et al., 2017;Fabron et al., 2017;Fowler et al., 2011;Santos et al., 2016) held only a single stimulation session.

| DISCUSSION
This systematic review studied 12 articles with methodology quality to analyze the current evidence on TENS application as a strategy in the rehabilitation/habilitation of voice and swallowing function.
In relation to the quality of the articles, according to the PEDro scale, none were excluded, that is, all obtained scores above the established threshold and specified the eligibility criteria used.
The articles scored lower in the criteria related to the randomization and blinding of the individuals. Studies that did not randomize displayed bias in subject selection, which may interfere in group comparison. The studies that did not use participant blinding exhibited performance bias, but in those involving electrostimulation, complete blinding can be precluded (Higgins et al., 2019;Sampaio & Mancini, 2017;Shiwa et al., 2011).
There is significant heterogeneity in the studies, even when considering voice and swallowing function individually. Of the 12 articles analyzed, only the results of 2-3 could be compared with one another, precluding meta-analysis to obtain the combined effect of a treatment, since a larger number of homogeneous studies are required.
Skeletal muscle power can be increased using nearly any strategy, provided exercise frequency and load intensity sufficiently exceed the normal or current activation of this muscle (Komi, 1986). Electrical stimulation can manipulate the output pattern of motoneuron activity and combines with muscle activity in the stimulated region, when compared to stimulation using voluntary exercises that gradually and hierarchically activate individual motorneurons (Kitchen, 2001). Furthermore, TENS stimulates the nerve and motor fibers. This activation triggers the descending analgesic systems of an inhibitory character in nociceptive transmission, thereby reducing pain (Santos et al., 2016;Silvério et al., 2015).
In recent years, studies have been carried out to investigate the use of TENS in dysphonic patients (Conde et al., 2017;Fabron et al., 2017;Fowler et al., 2011;Guirro et al., 2008;Santos et al., 2016;Silvério et al., 2015;Siqueira et al., 2017). Musculoskeletal pain reduction and improvements in vocal quality have been reported as the main objectives. This revision show that these studies report positive results in the use of stimulation, but when compared to other traditional therapies, the results did not show greater benefits.
In relation to the results obtained in the voice studies, of the three (Conde et al., 2017;Silvério et al., 2015;Siqueira et al., 2017) that compared the use of TENS with laryngeal manual therapy (LMT), only one , whose main variable was laryngeal diadochokinesis, demonstrated that the use of TENS did not show greater vocal fold movement regularity. The LMT would be a more suitable strategy to treat the diadochokinetic movements of the vocal folds. Another study (Fowler et al., 2011), which aimed to determine the immediate effect of TENS on the voice of healthy individuals, found positive changes in the voice, primarily in fundamental frequency, with no statistical difference.
For swallowing functions, TES is also applied as a direct intervention or an adjuvant treatment to exercise or to thermal-tactile stimulation approaches (Crary & Carnaby, 2014;Lim et al., 2009;Ryu et al., 2009;Simonelli et al., 2019;Sun et al., 2013). Most studies agree that better results are achieved when TES is used as an adjuvant to traditional therapy (Carnaby-Mann & Crary, 2007;Carnaby-Mann & Crary, 2008;Lim et al., 2009;Ryu et al., 2009;Sun et al., 2013). For post-stroke dysphagia patients, a meta-analysis (Chen et al., 2015) showed that both treatmentstraditional and NMESwere more effective in the short term. However, this review indicates insufficient evidence due to the limited number of studies conducted involving individuals with impaired swallowing without other disorders.
With respect to studies that focused on swallowing function in this review, only one (Blumenfeld et al., 2006) compared the use of TENS with traditional speech therapy, demonstrating a significant improvement for both treatment groups, according to the dysphagia classification scale. However, the results were more significant for individuals who received associated treatments. Another study (Maeda et al., 2017) also used the dysphagia classification scale to compare the stimulation and sham groups and found an improvement in dysphagia in individuals who received electrostimulation, albeit not significant. Another study (Ortega et al., 2016) Crary & Carnaby, 2014;Ortega et al., 2016;Park et al., 2009). According to the literature, many questions remain regarding the definition of TES application for dysphagia intervention in adults. It seems that a "one size fits all approach" may be inappropriate. Recent studies controlled scientific rigor, but variability and limited scientific control remain. As such, patient selection, electrode placement, stimulation parameters and exercise programs must be better described (Barikroo et al., 2018;Crary & Carnaby, 2014).
Although the results are encouraging, many aspects require further investigation. The evidence presented is not sufficient to establish TENS as an effective therapeutic approach for speech therapy in cases of voice and swallowing function rehabilitation/habilitation in patients without other associated disorders. It was also not possible to identify technique application patterns, especially in cases of swallowing function. The following are the most limiting factors regarding the level of evidence: the number of study participants; lack of information for randomization; lack of standardized tests for outcomes; and lack of stimulation parameters. More studies are needed to determine standardized application methods and levels of effectiveness.

| CONCLUSION
The following conclusions can be drawn from this review: • The studies analyzed exhibited medium methodological quality, showing difficulty primarily in the anonymous allocation of participants.
• Different TENS parameters are used. The frequency of the stimuli differed for swallowing functions.
• There are different electrode placements within and between the speech therapy areas.
• In relation to voice, the clinical therapy for dysphonia rehabilitation was better described than the techniques used for dysphagia.
• The voice specialty has more parameters to consider during assessment, some measured using complementary examinations, which seems to better explain the effects of TENS.
• In the area of swallowing function there are many nonstandardized assessment instruments to classify the dysphagia scale.
• There are variations in the time, number and periodicity of the therapeutic sessions for voice and swallowing function.
• There were no statistically significant differences between the use of TENS alone or associated with another therapeutic technique for voice. For swallowing function, one study proved better results in cases of associated techniques -TENS + traditional therapy.
• TENS is not the best strategy to analyze the diadochokinetic parameters of voice.
• The use of TENS on voice seems to demonstrate better effects on vocal quality; influence the comfort and stability of vocal emissions; help in muscle relaxation; and change some of the acoustic voice parameters.
• The use of TES in swallowing functions can target sensitivity or motor muscle responses (contraction or relaxation).
• The use of TENS on swallowing seems to improve swallowing function and pressure, and impact sensitivity to coughing and nutritional status.
• The fact that the studies were heterogeneous made it difficult to determine effectiveness and the parameters to use in future research. Speech therapy should increase the number of studies and improve their methodological quality, reassessing their methodological criteria. Current clinical practice is not grounded on evidence-based science.