**3.**OutcomeofNMESandPESinterventionsforpeoplewithoropharyngealdysphagia.































**Table 3.** *Cont.*


 NMES is at motor stimulation level unless explicitly mentioned. Data included in meta-analyses. Notes. ASHA-NOMS–American speech-language-hearing association national outcome measurement system; BASOFF—behavioural assessment scale of oral functions in feeding; BI—Barthel index; CDS—clinical dysphagia scale; CNS—central nervous system; CP—cerebral palsy; CT—computed tomography; CVA—cerebrovascular accident; DD—degree of dysphagia; A-DHI–Arabic dysphagia handicap index; DOSS—dysphagia outcome and severity scale; DSRS–dysphagia severity rating scale; DT—dysphagia therapy; EAT-10—eating assessment tool-10; EES—electrokinesiographic/electromyographic study of swallowing; EQ-5D—European Quality of Life Five Dimension; FDS—functional dysphagia scale; FOIS—functional oral intake scale; FEDSS—fiberoptic endoscopic dysphagia severity scale; FEES–fiberoptic endoscopic evaluation of swallowing; HNCI–head neck cancer inventory; IADL—instrumental activities of daily living; ICH–intracranial haemorrhage; ICU—intensive

care unit; LPM—laryngeal-pharyngeal mechanogram; MASA—Mann assessment of swallowing ability; MDADI—M.D. Anderson dysphagia inventory; LCD—laryngeal closure duration; LOS—length of stay; MBS—modified barium swallow; MBSImp–modified barium swallow impairment profile; MEG–magnetoencephalography; MMSE—mini-mental state exam; MEP–motor evoked potentials; MRI—magnetic resonance imaging; mRS–modified rankin scale; MS—multiple sclerosis; NEDS—neurological examination dysphagia score; NIHSS–national institutes of health stroke scale; NIHSS—National Institutes of Health Stroke Scale; NMES—neuromuscular electrical stimulation; NS—not significant; OD— oropharyngeal dysphagia; OPSE—oropharyngeal swallow efficiency; OST—oral sensorimotor treatment; OTT—oral transit time; PAS—penetration–aspiration score; PED EAT-10 pediatric eating assessment tool-10;PES–pharyngeal electrical stimulation; PESO—pharyngoesophageal segment opening; PPS–performance status scale; PTT—pharyngeal transit time; RMT—resting motor threshold; RSST—repetitive saliva swallowing test; rTMS—repetitive transcranial magnetic stimulation; SAH—subarachnoid haemorrhage; SAPP—swallowing activity and participation profile; SDQ—swallowing disturbance questionnaire; sEMG—surface electromyography; SFS—swallow function score; SHEMG— electromyographic activity of the submental/suprahyoid muscles complex; SI—similarity index; SLT–speech and language therapist; SSA—standardised swallowing assessment; SWAL-QOL—swallowing quality of life; TBI—traumatic brain injury; tDCS—transcranial direct current stimulation; TOR-BSST—Toronto bedside swallowing screening test; UES—upper esophageal sphincter; UPDRS—unified Parkinson's disease rating scale; VAS—visual analogue scale; VFSS—videofluoroscopic swallowing study; VVS-T—volume viscosity swallow test; WST—water swallow test.

*Peripheral Neurostimulation Interventions*. Across the 42 included studies, 30 studies reported on NMES and eight studies reported on PES. Four studies used another type of neurostimulation (i.e., rTMS) in addition to NMES or PES, either within the same group or different treatment groups.

*Participants* (Table 2). The 42 studies included a total of 2281 participants (mean 54.3; SD 39.1). The sample sizes ranged from the smallest sample of 16 participants [60,61] to the largest sample of 162 participants [36]. By intervention type, samples were characterized as follows: NMES total 1706, mean 56.9, SD 38.9, range 18–135; PES total 410, mean 51.3, SD 49.0, range 16–162; and combined neurostimulation total 165, mean 41.3, SD 19.3, range 18–64. The mean age of participants across all studies was 61.8 years (SD 15.3), with one study reporting age range only (65–93 years) [61]. Participant mean age across all studies ranged from 4.2 years [54] to 84.4 years [39]. The mean age of participants by intervention group was: NMES 60.9 years (SD 16.9), PES 64.7 years (SD 11.9), and combined neurostimulation 63.8 years (SD 6.4).

Across all studies, 61.0% (SD 13.5) of participants were male and one study did not report gender distribution [30]. Percentage of males by intervention group was NMES 62.6% (SD 14.0), PES 56.7% (SD 9.6), and other/combined 65.4% (SD 12.3). Most studies included stroke patients (*n* = 31), while three studies included mixed populations [28,41,43] and one study reported OD without further underlying medical diagnosis [39]. Other diagnoses by intervention group were: Parkinson's disorder (*n* = 2) [32,46], cerebral palsy (*n* = 2) [50,54], and head and neck cancer (*n* = 2) [36,48] in NMES; and multiple sclerosis (*n* = 1) [63] in PES.

Across the 42 studies, VFSS was most frequently used to confirm participant's diagnosis of OD (*n* = 31), whereas six studies used FEES [49,53,54,60,64,65]. Several of these studies combined instrumental assessment with either a screen (*n* = 2) [58,65] or clinical assessment (*n* = 6) [49,50,53–55,68]. One study used either clinical assessment or VFSS [50]. One study used a single screen [56], three studies used clinical assessment only [35,38,59], and one study used both [33]. The studies were conducted across 14 countries, with studies most frequently conducted in Korea (*n* = 11), China (*n* = 7), the UK (*n* = 7), Spain (*n* = 4), Italy (*n* = 2), Turkey (*n* = 2), and Germany (*n* = 2).

*Outcome Measures* (Table 2). Outcomes measures varied greatly across all studies included in the review, covering several domains within the area of OD. The Penetration Aspiration Score was the most reported outcome measure (PAS; 18 studies), followed by Functional Oral Intake Scale (FOIS; 12 studies), Functional Dysphagia Scale (FDS; 5 studies), Dysphagia Severity Rating Scale (DSRS; 5 studies), Swallowing Quality of Life questionnaire (SWAL-QOL; 4 studies), and Dysphagia Outcome and Severity Scale (DOSS; 3 studies).

*NMES Intervention* (*n* = 30: Tables 2 and 3). In total, 22 studies included two study arms or groups, whereas eight studies included three groups [31–34,38,40,55,57]. All but five NMES studies [29,39,43,53,54] combined neurostimulation with simultaneous DT consisting of a wide range of behavioural interventions (e.g., head and body positioning, bolus modification, oromotor exercises, or swallow manoeuvres). Six studies included a NMES only group without DT [29,33,38,39,43,55], with five of these studies using NMES at motor stimulation level [29,33,38,43,55] and one study using NMES at sensory stimulation level [39]. An additional seven studies included a treatment arm with NMES at sensory stimulation level combined with DT [32,44–46,53,54,57]. All other participants in NMES groups received stimulation at motor level. Five studies compared different NMES electrode positions [28,34,40–42] and seven studies included a sham stimulation group [36,39,48,50,52–54].

Control groups included mostly sham NMES stimulation and/or DT. Only one study included a control group receiving neither DT nor NMES [30], and one study included usual care across different healthcare settings as the comparison group [51].

*PES Intervention* (*n* = 8: Tables 2 and 3). All eight studies compared PES to a sham version of the treatment [58–65]. None of the studies included other treatment groups (e.g., DT) or control groups (e.g., usual care or no treatment).

*Combined Neurostimulation Interventions* (*n* = 4: Tables 2 and 3). Three studies in the combined intervention group compared three different treatments. Of these, one study compared PES, paired associative stimulation (PAS) and rTMS [68], a second study compared DT, rTMS combined with DT, and NMES combined with DT [67], and a third study compared rTMS, PES and capsaicin stimulation [66]. A fourth study combined NMES stimulation with sham rTMS or rTMS stimulating different hemispheres (ipsilesional, contralesional or bilateral) [69].

#### *5.3. Risk of Bias Assessment and Methodological Quality*

The tau values from the Begg and Mazumdar rank correlation were 0.101 (two-tailed *p* = 0.589) and < 0.000 (two-tailed *p* > 0.999) for NMES and PES, respectively. The NMES meta-analysis incorporates data from 16 studies, which yielded a *z*-value of 4.107 (two-tailed *p* < 0.001). The fail-safe N is 55 indicating 55 'null' studies need to be located and included for the combined two-tailed *p*-value to exceed 0.050. Therefore, there would need to be 3.4 missing studies for every observed study for the effect to be nullified. The PES metaanalysis incorporates data from five studies yielding a *z*-value of 1.156 (two-tailed *p* < 0.248). Since the combined result is not statistically significant, the fail-safe N (which addresses the concern that the observed significance may be spurious) is not relevant. Both of these procedures (i.e., Begg and Mazumdar rank correlation and fail-safe N) indicate the absence of publication bias.

Figures 2 and 3 present, respectively, the risk of bias summary per domain for all included studies combined and for individual studies. The majority of studies had low risk of bias with very few exceptions.

**Figure 2.** Risk of bias summary for all included studies (*n* = 42) in accordance with RoB-2.


**Figure 3.** Risk of bias summary for individual studies (*n* = 42) in accordance with RoB-2.

#### **6. Meta-Analysis: Effects of Interventions**

*6.1. Neuromuscular Electrical Stimulation (NMES) Meta-Analysis*

Eleven studies were included in the NMES meta-analysis [28,29,34,37,40,42,45,47,49,51,55], of which six studies included two or three different intervention groups [28,34,40,42,45,55]. A total of 20 studies were excluded from meta-analysis for the following reasons: in three

studies, OD diagnosis was not confirmed by instrumental assessment (VFSS or FEES); five studies provided insufficient data for meta-analyses; and, twelve studies were excluded to reduce heterogeneity: six studies including subject populations with medical diagnoses other than stroke (i.e., children with cerebral palsy, head and neck cancer patients, patients with Parkinson's disease, and elderly), five studies because of outcome measures (e.g., kinematic or biomechanical variables in VFS recordings), and one study using sensory NMES stimulation.

*Overall within-group analysis* (Figure 4). A significant, large pre-post intervention effect size was calculated using a random-effects model (*z*(17) = 6.477, *p <* 0.001, Hedges' *g* = 1.272, and 95% CI = 0.887–1.657). Pre-post intervention effect sizes ranged from 0.000 to 3.826. In 13 of the 18 NMES intervention groups, effect sizes were large (Hedges' *g* > 0.8), indicating that NMES accounted for a significant proportion of standardized mean difference for these studies. Between-study heterogeneity was significant (*Q*(17) = 106.7, and *p* < 0.001), with *I* 2 showing that heterogeneity accounted for 84.1% of variation in effect sizes across studies.

**Figure 4.** Neuromuscular electrical stimulation (NMES) within intervention group pre-post metaanalysis [28,29,34,37,40,42,45,47,49,51,55]. Note. Refer to Table 2 for explanation of the subgroups.

*Overall between-group analysis* (Figure 5). A significant, small post-intervention betweengroup total effect size in favour of NMES was calculated using a random-effects model (*z*(8) = 2.589, *p* = 0.010, Hedges' *g* = 0.433, and 95% CI = 0.105–0.760). Between-study heterogeneity was significant (*Q*(8) = 18.0, and *p* = 0.021), with *I* <sup>2</sup> showing that heterogeneity accounted for 55.6% of variation in effect sizes across studies.

Between-subgroup analyses. Subgroup analyses (Table 4) were conducted to compare diagnostic groups. Treatment effects were highest (moderate) for stroke patients, while other groups showed no significant effect sizes. For all other subgroup analyses, only stroke patients were included to improve homogeneity between studies. Subgroup analyses between studies compared intervention types (NMES, NMES + DT), time between pre- and post-intervention measurement, outcome measures, total stimulation times, electrodes configurations, pulse durations, and pulse rates (Table 4). NMES as an adjunctive treatment to DT showed significant, moderate positive treatment effects, whereas NMES alone showed non-significant effects. Effect sizes comparing time between pre- and posttreatment measurements showed no clear results. Although no effects could be identified at 2 weeks, a significant, positive effect size was found at 7 weeks. When comparing effect sizes based on outcome measures, the only significant effect found was a significant, large effect size for oral intake. The non-significant effects sizes for visuoperceptual evaluation of

instrumental assessment ranged between negligible negative to moderate positive effects. Total stimulation time subgroup analyses showed significant, moderate positive treatment effects for longer stimulation times (>100 min). Shorter stimulation times did not result in significant effects. Comparisons for electrode configurations showed significant, moderate positive effects sizes for infrahyoid configuration. Electrode configuration based on patients' characteristics, including OD outcome scores, indicated non-significant moderate effects, whereas both suprahyoid combined with infrahyoid and suprahyoid configurations resulted in negligible effects. Final comparisons between studies using different pulse durations did not suggest a linear relationship, whereas pulse rate comparisons indicated that studies using higher frequencies showed increased significant, positive moderate effect sizes.


**Figure 5.** NMES between group post meta-analysis [29,37,40,47,49,51,55]. Note. Refer to Table 2 for explanation of the subgroups.

#### *6.2. Pharyngeal Electrical Stimulation (PES) Meta-Analysis*

Five studies using PAS in adult stroke patients were included in the metaanalyses [58,62,65,66,68]. Three studies were excluded from meta-analyses for the following reasons: overlap in participant population between studies, insufficient data for meta-analyses, and no confirmation of OD diagnosis prior to treatment.

Overall within-group analysis. The pre-post intervention effect sizes for the included studies ranged from 0.265 (small effect) [66] to 0.802 (large effect) [62], with an overall moderate effect size of 0.527 (Figure 6). As one study, however, did not provide PAS data for all included participants [65], a sensitivity analysis was conducted for both PAS and DSRS, indicating minimal differences in effect sizes.

**Figure 6.** PES within intervention group pre-post meta-analysis [58,62,65,66,68].


**Table 4.** Between subgroup meta-analyses for NMES and pharyngeal electrical stimulation (PES) comparing intervention groups of included studies.

Note. \* Significant.

Overall between-group analysis. A non-significant post-intervention between-group total effect size in favour of PES was found using a random-effects model (*z*(4) = 0.718, *p* = 0.473, Hedges' *g* = 0.099, and 95% CI = −0.170–0.368), suggesting no improvement in PAS outcomes following PES neurostimulation (Figure 7). Between-study heterogeneity was non-significant (*Q*(4) = 1.8, and *p* = 0.766).


**Figure 7.** PES between group post meta-analysis [58,62,65,66,68].

Between-subgroup analyses. Subgroup analyses were conducted (Table 4) comparing total stimulation time between studies, favouring shorter stimulation times (*z*(1) = 0.940, *p* = 0.347, Hedges' *g* = 0.300, and 95% CI = −0.325–0.925).

#### **7. Discussion**

This study (Part I) aimed to determine the effects of PES and NMES in people with OD without excluding populations based on medical diagnoses. To base findings on the highest level of evidence, only RCTs were included. This systematic review and meta-analysis were conducted using PRISMA procedures as a guide.

#### *7.1. Systematic Review Findings*

When comparing RCTs in pharyngeal and neuromuscular electrical stimulation (i.e., PES and NMES), various methodological problems became apparent. Some studies did not define OD or used divergent definitions, whereas other studies applied different inclusion criteria. Most studies included patients with confirmed OD by instrumental assessment, but several studies used screening, patient self-report or clinical assessments instead. Consequently, participant characteristics may differ widely between studies. Despite most studies included stroke patients, meta-analysis comparing diagnostic groups other than stroke was possible for NMES, however this could not be conducted for PES.

Furthermore, the great variety in outcome measures also restricted comparisons by meta-analysis. As heterogeneity between studies indicates that no estimated overall effect by meta-analysis should be determined, combining studies targeting different domains within the area of OD will have similar implications. For instance, meta-analyses based on both patients' self-reported health-related quality of life and visuoperceptual evaluation of instrumental assessments would very likely lead to inappropriate estimated overall effects. Thus, to reduce heterogeneity between outcome measures, some studies were excluded from the meta-analysis. This strong focus on reducing heterogeneity between studies when performing meta-analysis also implies that data other than the authors' primary outcomes may have been preferably included in this analysis. For example, the primary outcome for Dziewas, Stellato, Van Der Tweel, Walther, Werner, Braun, Citerio, Jandl, Friedrichs, Nötzel, Vosko, Mistry, Hamdy, McGowan, Warnecke, Zwittag and Bath [59] and Suntrup, Marian, Schröder, Suttrup, Muhle, Oelenberg, Hamacher, Minnerup, Warnecke and Dziewas [64] was readiness for decannulation, which was considered too different from outcomes in the other included studies.

All eight PES studies compared neurostimulation with sham stimulation. However, among the 30 NMES studies, the comparison group variably consisted of usual care, DT, another dysphagia treatment or a combination of treatments. In contrast to PES studies

that did not include any DT groups, most NMES studies combined neurostimulation with simultaneous DT. However, DT consisted of a wide range of behavioural interventions, using different treatment dosages, timings, and durations. Moreover, DT was referred to by many different names and acronyms (e.g., dysphagia training, behavioural intervention, classic treatment, or standard care). This suggest that care should be taken with the use of DT as an overarching term to group many different behavioural interventions to estimate overall effect sizes in meta-analyses.

Furthermore, RCTs are characterised by random allocation of participants to intervention groups and blinding or masking the nature of treatment for participants. However, in neurostimulation studies, blinding is frequently not feasible and participants may identify what treatment arm they have been assigned to (e.g., the presence of neurostimulation equipment, the experience of active stimulation). Also, since neurostimulation thresholding in PES is frequently applied in all groups to mask treatment assignment, patients receiving sham stimulation would still have been exposed to a certain level of neurostimulation during thresholding. Those studies not using thresholding in sham groups (e.g., [59,64]) might show larger treatment effect differences when comparing neurostimulation versus sham stimulation.

#### *7.2. NMES*

When considering meta-analyses for NMES, the highest effect sizes were found for stroke populations. As existing reviews in NMES [10,12,18,19] excluded other patient populations, no comparisons could be made between clinical populations. In addition, only two reviews conducted meta-analyses [18,19] selecting studies using different inclusion criteria (e.g., excluding comparison groups with active treatment components [18] or excluding chronic stroke patients [19]). Reviews may also prefer different outcome data for meta-analyses, especially in the case of RCTs using a large battery of assessments. As such, total numbers of included studies vary per review, but comparisons between reviews may be falsely estimated due to differences in methodology.

In this systematic review, a wide range in effect sizes was found in NMES RCTs depending on outcome measures used. However, oral intake scales showed highest effects sizes when compared to visuoperceptual evaluation of instrumental assessment or clinical assessment. This might be explained by NMES treatment usually taking place over consecutive weeks, in contrast to other neurostimulation techniques (e.g., PES or rTMS) that may be restricted to limited sessions over a few days only.

The great heterogeneity between DT groups also impeded comparisons between NMES only, NMES plus DT, and DT-only groups. No RCTs provided adequate DT group data to be included in the meta-analysis. For NMES groups, only two studies were included. As a result, information about the effects of DT is lacking. The negligible effect sizes found for NMES without DT were based on only two studies and the moderate effect sizes for combined NMES and DT were based on a total of seven studies.

Most studies performed NMES at motor stimulation level, whereas only a few studies included a group receiving NMES at sensory stimulation level. As none of these latter studies could be included in meta-analyses, no further details are available on comparisons between effect sizes for sensory versus motor stimulation. Also, terminology was confusing as sensory stimulation was sometimes referred to as sham stimulation [39].

NMES studies showed marked variation in the technical parameters and protocols applied. When comparing electrode configurations, both hyoid and combined hyoid and suprahyoid configurations showed negligible effects, whereas infrahyoid configurations resulted in moderate effects. A study using patient-dependent configurations showed promising results as well [55]. However, it remained unclear which criteria were used to decide on individual configurations. Furthermore, reporting on many technical parameters proved to be either incomplete or unclear for several studies (e.g., data on pulse duration, pulse rate, or stimulation time). As technical parameters may depend on medical device manufacturers, comparisons between brands may be warranted. For example, when

considering pulse duration, a clear distinction in effect sizes is found between one study using a lower pulse rate—indicating a negative effect size—versus eight studies using higher pulse rates with moderate effect sizes.

#### *7.3. PES*

Compared to NMES, fewer PES studies were identified and thus a more limited metaanalysis was conducted. RCTs included stroke populations, except for one study that included patients with multiple sclerosis [63]. All studies compared active PES with sham treatment in stroke patients and used mostly visuoperceptual evaluation of radiographic recordings of the swallowing act as an outcome measure. Meta-analysis identified a non-significant post-intervention between-group total effect size in favour of PES. This finding seemed in line with findings by Chiang, Lin, Hsiao, Yeh, Liang and Wang [19], but this comparison is limited as it is based on only two studies. Additionally, Bath, Lee and Everton [18] reported that PES studies did not show an effect for many outcome measures (e.g., post-treatment proportions of participants with dysphagia, swallowing ability, penetration and aspiration scores or nutrition). However, in contrast to previous reviews, Cheng, Sasegbon and Hamdy [7] found a significant, moderate effect size in favour of PES when conducting meta-analysis. Again, inclusion criteria between reviews differed. For example, two studies [59,64] were excluded from meta-analysis in this review as well as the reviews by Chiang, Lin, Hsiao, Yeh, Liang and Wang [19] and Bath, Lee and Everton [18], but were included in the review by Cheng, Sasegbon and Hamdy [7]. This may have impacted the overall effect size as both PES studies showed significant treatment effects.

#### *7.4. Moderators*

Differences between NMES and PES studies made comparisons between RCTs difficult and hindered meta-analyses. Studies used different participant inclusion criteria in relation to underlying medical diagnoses or chronicity of stroke and used a large variety of outcome measures covering different domains within the area of OD. Outcome measures may also lack responsiveness, thus lack sensitivity to change during treatment. Moreover, studies varied significantly in technical parameters of neurostimulation. The number of studies and participants restricted the ability of statistical analyses to consider how each variable may have impacted the effects of neurostimulation.

Studies frequently neglected to report on potential moderators of stimulation effects in sufficient detail. For example, stroke severity and OD severity are inextricably linked and may moderate stimulation effects, yet only very few studies provided data on stroke severity. Similar problems occur when the chronicity of a stroke is not reported or the possibility of spontaneous recovery is ignored. This is especially true during NMES treatment, which may span a period of several weeks. In addition, no consensus was reached regarding the optimal moment for outcome measurement. Consequently, in this review, between-subgroup meta-analyses were conducted using post-intervention data only, so that the possibility of spontaneous recovery during the intervention period was taken into consideration.

#### *7.5. Limitations*

Despite a rigorous reviewing process following PRISMA guidelines and the use of RoB 2 to reduce bias, this review is subject to some limitations. Only RCTs published in English were included in this current study. Thus, some RCTs may have been excluded based on language criteria when their findings could have contributed to the current metaanalysis. Furthermore, meta-analyses included mostly stroke studies, thereby not providing effect sizes for other diagnostic patient populations. However, the main limitation of this review originates from the high degree of heterogeneity between studies, making comparisons across studies challenging. As such, generalisations and meta-analyses should be interpreted with care.

#### **8. Conclusions**

Meta-analyses for RCTS in NMES found a significant, large pre-post intervention effect size and significant, small post-intervention between-group effect size in favour of NMES. For PES studies, the meta-analyses showed a significant, moderate effect size for pre-post intervention, whereas overall between-group analysis did not result in significant treatment effects. Based on these results, NMES seems to have a more promising outcome compared to PES. However, only careful generalisations and interpretations of these metaanalyses can be made due to the NMES studies showing high heterogeneity in protocols and experimental variables, including potential moderators, and featuring inconsistent methodological reporting.

There is a need for more RCTs with larger sample sizes in addition to the standardisation of protocols and guidelines for reporting. These changes would better facilitate comparisons of studies and help to determine intervention effects more definitively. Delphi studies involving international experts might allow for a consensus to be reached, thus supporting future research, comparability and generalisability.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm11030776/s1, Table S1: PRISMA 2020 for Abstracts Checklist; Table S2: PRISMA 2020 Checklist.

**Author Contributions:** Conceptualization: R.S., R.C., A.-L.S., L.B., S.H. Formal analysis: R.S., R.C. Methodology: R.S., R.C. Project administration: R.S., R.C. Validation: R.S., R.C. Writing—review & editing: R.S., R.C., A.-L.S., L.B., S.H., B.J.H., L.R., S.W.-G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **Neurostimulation in People with Oropharyngeal Dysphagia: A Systematic Review and Meta-Analysis of Randomised Controlled Trials—Part II: Brain Neurostimulation**

**Renée Speyer 1,2,3,\*, Anna-Liisa Sutt 4,5, Liza Bergström 6,7, Shaheen Hamdy 8, Timothy Pommée 9, Mathieu Balaguer 9, Anett Kaale 1,10 and Reinie Cordier 2,11**


**Abstract:** *Objective.* To assess the effects of brain neurostimulation (i.e., repetitive transcranial magnetic stimulation [rTMS] and transcranial direct current stimulation [tDCS]) in people with oropharyngeal dysphagia (OD). *Methods.* Systematic literature searches were conducted in four electronic databases (CINAHL, Embase, PsycINFO, and PubMed) to retrieve randomised controlled trials (RCTs) only. Using the Revised Cochrane risk-of-bias tool for randomised trials (RoB 2), the methodological quality of included studies was evaluated, after which meta-analysis was conducted using a random-effects model. *Results.* In total, 24 studies reporting on brain neurostimulation were included: 11 studies on rTMS, 9 studies on tDCS, and 4 studies on combined neurostimulation interventions. Overall, within-group meta-analysis and between-group analysis for rTMS identified significant large and small effects in favour of stimulation, respectively. For tDCS, overall withingroup analysis and between-group analysis identified significant large and moderate effects in favour of stimulation, respectively. *Conclusion.* Both rTMS and tDCS show promising effects in people with oropharyngeal dysphagia. However, comparisons between studies were challenging due to high heterogeneity in stimulation protocols and experimental parameters, potential moderators, and inconsistent methodological reporting. Generalisations of meta-analyses need to be interpreted with care. Future research should include large RCTs using standard protocols and reporting guidelines as achieved by international consensus.

**Keywords:** deglutition; swallowing disorders; RCT; intervention; repetitive transcranial magnetic stimulation; transcranial direct current stimulation; rTMS; tDCS

**Citation:** Speyer, R.; Sutt, A.-L.; Bergström, L.; Hamdy, S.; Pommée, T.; Balaguer, M.; Kaale, A.; Cordier, R. Neurostimulation in People with Oropharyngeal Dysphagia: A Systematic Review and Meta-Analysis of Randomised Controlled Trials—Part II: Brain Neurostimulation. *J. Clin. Med.* **2022**, *11*, 993. https://doi.org/10.3390/ jcm11040993

Academic Editor: Jacopo Galli

Received: 7 December 2021 Accepted: 11 February 2022 Published: 14 February 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Oropharyngeal dysphagia (OD) or swallowing problems is highly prevalent among stroke patients, people with progressive neurological diseases, patients with head and neck cancer, and in frail older persons [1,2]. Prevalence estimates of OD may vary depending on underlying medical diagnoses, but have been reported as high as 80% in stroke and Parkinson's disease [3], and 70% in oncological populations [4]. OD is associated with dehydration, malnutrition, aspiration pneumonia, and increased mortality [5–7], but also leads to decreased health-related quality of life [8].

Treatment and management of OD may vary widely. However, apart from traditional compensatory and rehabilitative strategies including diet modifications, postural adjustments, oromotor training and swallow manoeuvres [9], recent studies report on the possible beneficial effects of non-invasive brain stimulation. Brain neurostimulation aims to modulate cortical excitability and include techniques such as repetitive transcranial magnetic stimulation (rTMS) and transcranial direct current stimulation (tDCS). rTMS uses electromagnetic induction resulting in depolarisation of postsynaptic connections, whereas tDCS uses direct electrical current shifting the polarity of nerve cells [10]. Neurostimulation protocols may vary greatly per study, including different neurostimulation sites, frequencies, stimulation duration and number of different outcome measures are used to objectify treatment effects, and individual responses to stimulation are highly variable [10–12].

Aspiring to improved treatment efficacy in OD management, non-invasive brain stimulation has achieved growing interest over the past decade. Several reviews have been published on rTMS and tDCS [10,12–18], each publication having different inclusion and exclusion criteria and methodology. All previous reviews targeted brain neurostimulation interventions in post-stroke populations except for one review that included patients with acquired brain injury [16]; to date, all reviews on brain stimulation set criteria based on medical diagnoses. Moreover, not all reviews performed meta-analysis [14] and as several neurostimulation trials have only been published recently, earlier reviews will have identified fewer studies.

This is the second paper (Part II) of two companion papers on treatment effects of neurostimulation in people with OD. The first systematic review (Part I) reported on the effects of pharyngeal electrical stimulation (PES) and neuromuscular electrical stimulation (NMES).

The aim of this systematic review (Part II) is to determine the effects of brain neurostimulation (i.e., rTMS and tDCS) in people with OD without excluding populations based on medical diagnoses. Only randomised controlled trials (RCTs) will be included being the highest level of evidence. Meta-analyses will be conducted to summarise results and report on possible moderators of treatment effects.

#### **2. Methods**

The methodology and reporting of this systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement and checklist (Supplementary Tables S1 and S2) [19,20]. Adhering to the PRISMA statement and checklist ensures essential and transparent reporting of systematic reviews. The protocol for this review was registered with PROSPERO, the international prospective register of systematic reviews (registration number: CRD42020179842).

#### *2.1. Information Sources and Search Strategies*

An electronic database search for extant literature was conducted on 6 March 2021, using the following four databases: CINAHL, Embase, PsycINFO, and PubMed. Publications dates included in the search were 1937–2021, 1902–2021, 1887–2021, and 1809–2021, respectively. Generally, search strategies consisted of combinations of terms related to 'dysphagia' and 'randomised controlled trial'. Both subject headings (e.g., MeSH and Thesaurus terms) and free text terms were used to search databases. The full list of electronic search strategies

for each database can be found in Table 1. To identify literature not found utilising these strategies, the reference lists of eligible articles were checked.

**Table 1.** Search strategies.


#### *2.2. Inclusion and Exclusion Criteria*

To be eligible for inclusion in this systematic review, studies had to meet the following criteria: (1) participants had a diagnosis of oropharyngeal dysphagia; (2) the study included non-invasive neurostimulation interventions aimed at reducing swallowing or feeding problems; (3) the study included a control group or comparison intervention group; (4) participants were randomly assigned to one of the study arms or groups; and (5) the study was published in English language.

Interventions such as non-electrical peripheral stimulation (e.g., air-puff or gustatory stimulation), pharmacological interventions and acupuncture, were considered out of scope of this review, thus were excluded. Invasive techniques and/or those that did not specifically target OD (e.g., deep-brain stimulation studies after neurosurgical implementation of a neurostimulator) were also excluded. Conference abstracts, doctoral theses, editorials, and reviews were excluded.

#### *2.3. Systematic Review*

*Methodological Quality and Risk of Bias.* The Revised Cochrane risk-of-bias tool for randomised trials (RoB 2) [21] was used to assess the methodological quality of the included studies. The RoB 2 tool identifies domains to consider when assessing where bias may have been introduced into a randomised trial: (1) bias arising from the randomisation process; (2) bias due to deviations from intended interventions; (3) bias due to missing outcome data; (4) bias in measurement of the outcome; and (5) bias in selection of the reported result. For each domain, a series of signalling questions are answered to give a judgement (i.e., "low risk of bias", "some concerns", or "high risk of bias"), which can then be assessed in aggregate to determine a study's overall risk of bias [21].

*Data Collection Process.* Data were extracted from the included studies using a data extraction form created for this purpose. This form allowed for extraction of data under several categories, relevant to meta-analyses, including participant diagnosis, inclusion and exclusion criteria, sample size, age, gender, intervention goal, intervention agent/delivery/dosage, outcome measures, and treatment outcomes.

*Data, Items and Synthesis of Results.* Titles and abstracts of included studies were reviewed for eligibility by two independent reviewers. Next, the same two reviewers assessed the selected original articles at a full-text level to determine their eligibility. To ensure rating accuracy, a random selection of one hundred records were scored and discussed over two consecutive group sessions prior to rating the remaining records. Any disagreement

between the first two reviewers was resolved by consulting a third reviewer. Assessment of methodology study quality followed an equivalent process. None of the reviewers had formal or informal affiliations with any of the authors of the included studies.

Extracted data were extrapolated and synthesised within the following categories to allow for comparison: participant characteristics, inclusion criteria, intervention conditions, outcome measures and intervention outcomes. Effect sizes and significance of findings were used to assess treatment outcomes.

#### *2.4. Meta-Analysis*

Using the extracted data, effect sizes were compared for the following: (1) pre-post outcome measures of OD and (2) mean difference in outcome measures from pre- to postintervention scores between neurostimulation and comparison controls. Control groups either received no treatment, sham stimulation and/or traditional dysphagia therapy (DT; e.g., compensatory and rehabilitative strategies including diet modifications, postural adjustments, oromotor training and swallow manoeuvres). Only studies using instrumental assessment (e.g., videofluoroscopic swallow study [VFSS] or fiberoptic endoscopic evaluation of swallowing [FEES]) to confirm OD were included.

When selecting what data points to extract, data collected using outcome measures based on visuoperceptual evaluation of instrumental assessment were preferred over clinical non-instrumental assessments. Oral intake measures were only included if no other clinical data were available, whereas screening tools and patient self-report measures were excluded entirely. When selecting outcome measures for meta-analyses, reducing heterogeneity between studies was given priority. Consequently, measures other than the authors' primary outcomes may have been preferred if these measures contributed to greater homogeneity.

Comprehensive Meta-Analysis Version 3.3.070 [22] software was used to complete the meta-analysis, allowing comparison of sample size, effect size, group means and standard deviations of pre- and post-measurements. In the case that no parametric data were available, the reported non-parametric data (i.e., medians, interquartile ranges) were converted into parametric data for meta-analysis purposes. Studies with multiple intervention groups were analysed separately for each experimental-control comparison. If studies included the same participants, only one study was included in the meta-analysis. Where reported data were insufficient, attempts were made to contact authors of individual studies and request additional data.

Using Comprehensive Meta-Analysis, a random-effects model was used to calculate effect sizes. This was due to variations in participant characteristics, sampling, interventions, and measurement, which suggested a low likelihood that studies would have similar true effects. Heterogeneity was estimated using the *Q* statistic to determine the spread of effect sizes about the mean and *I* <sup>2</sup> was used to estimate the ratio of true variance to total variance. *I* 2-values of less than 50%, 50% to 74%, and higher than 75% denote low, moderate, and high heterogeneity, respectively [23]. Effects sizes were generated using the Hedges' *g* formula for standardised mean difference with a confidence interval of 95%. Effects sizes were interpreted using Cohen's *d* convention as follows: *d* ≤ 0.2 as no or negligible effect; 0.2 < *d* ≤ 5 as small effect; 0.5 < *d* ≤ 0.8 as moderate effect; and *d* > 0.8 as large effect [24].

Forest plots of effect sizes for OD outcome scores were generated for both types of neurostimulation (i.e., rTMS and tDCS): (1) pre-post neurostimulation and (2) neurostimulation interventions versus comparison groups. Subgroup analyses were conducted to compare effect sizes as a function of different moderators and neurostimulation types including: outcome measures, total treatment duration, total neurostimulation time, and stimulation characteristics (e.g., pulse range, stimulation current, and stimulation site). To take into consideration the possibility of spontaneous recovery during the intervention period, only between-subgroup meta-analyses were conducted using post-intervention data.

Utilising Comprehensive Data Analysis software, publication bias was evaluated as per the Begg and Muzumdar's rank correlation test and the Fail-safe N test. Begg and Muzumdar's rank correlation test provides information on the rank correlations between standardised effect size and the ranks of their variances [25]. In addition to a tau value, a two-tailed *p* value is also generated. Where the analysis results in a value of zero, it can be concluded that there is unlikely to be an association between the effect size and ranks of variance. Conversely, the closer to one the tau or *p* values, the more likely there is to be an association between the effect size and ranks of variance. Therefore, high standard error would be connected to higher effect sizes if publication bias was the result of asymmetry. If larger effects are represented by low values, tau would be over zero; conversely tau would be negative if larger effects are represented by high values.

The Fail-safe N test is a calculation of the quantity of studies with zero effect size that could be incorporated into the meta-analysis prior to the result losing statistical significance, that is, the quantity of excluded studies that would result in the effect being nullified [26]. Results should be treated with care where the fail-safe N is relatively small, however, when it is large, conclusions can be confidently drawn that the treatment effect, while potentially raised by the removal of some studies, is not nil.

#### **3. Results**

#### *3.1. Study Selection*

A total of 8059 studies were retrieved through the subject heading and free text searches (CINAHL: *n* = 239, Embase: *n* = 4550, PsycINFO: *n* = 231, and PubMed: *n* = 3039). Following removal of duplicates at a title and abstract level (*n* = 1113), a total of 6946 records remained. A total of 261 original articles were assessed at a full-text level, with articles grouped according to type of intervention. At this stage, no studies were excluded based on type of intervention (e.g., behavioural intervention, neurostimulation). Of these, 58 articles on neurostimulation were identified that satisfied the inclusion criteria. Four additional studies were found through reference checking of the included articles. This process resulted in a final number of 24 included studies. Figure 1 presents the flow diagram of the overall reviewing process according to PRISMA.

**Figure 1.** Flow diagram of the reviewing process according to PRISMA.

#### *3.2. Description of Studies*

Tables 2 and 3 report detailed descriptions of all included studies. Table 2 includes data on study characteristics including methodological study quality, inclusion and exclusion criteria, and details on participant groups. Information is provided for all study groups (control and intervention groups), medical diagnosis, sample size, age and gender. Table 3 reports on intervention characteristics, including goals, intervention components, outcome measures, intervention outcomes, as well as main conclusions.

*Brain stimulation Interventions* (Table 2). Across the 24 included studies, eleven studies reported on rTMS and nine studies reported on tDCS. Four studies used another type of neurostimulation (i.e., NMES) in addition to rTMS, either within the same group or over different treatment groups.

*Participants* (Table 2). The 24 studies included a total of 728 participants (mean 30.3; SD 13.4). The sample sizes ranged from the smallest sample of 14 participants [27] to the largest sample of 64 participants [28]. By intervention type, samples were characterised as follows: *rTMS* total 280, mean 25.5, SD 7.6, range 15–40; *tDCS* total 283, mean 31.4, SD 14.6, range 14–59; and combined neurostimulation total 165, mean 41.3, SD 19.3, range 18–64. The mean age of participants across all studies was 64.6 years (SD 5.8), ranging from 51.8 years [29] to 74.9 years [27]. By intervention group, the mean age of participants was: *rTMS* 63.6 (4.8), *tDCS* 66.2 (SD 6.9), and combined neurostimulation 66.5 years (SD 4.4).

Across all studies 59.6% (SD 12.7) participants were male and two studies did not report gender distribution [29,30]. Percentage of males by intervention group was *rTMS* 61.9% (SD 12.8), *tDCS* 57.5% (SD 10.9), and other/combined 65.4% (SD 12.3). Most studies included stroke patients (*n* = 21), with other diagnoses by intervention group reported as: presbyphagia due to central nervous system disorder (*n* = 1) [31] in *tDCS*; Parkinson's disorder (*n* = 1) [30] and brain injury (*n* = 1) [32] in *rTMS*. All 24 studies used VFSS to confirm participants' diagnosis of OD. The studies were conducted across 12 countries, with the highest number of studies conducted in Korea (*n* = 6), Egypt (*n* = 4), China (*n* = 3), Italy (*n* = 2) and Japan (*n* = 2).

*Outcome Measures* (Table 3). Outcomes measures varied greatly across all studies included in the review, covering several domains within the area of OD. The Penetration Aspiration Score (PAS) was the most reported outcome measure (8 studies), followed by the Dysphagia Outcome and Severity Scale (DOSS; 7 studies), Functional Oral Intake Scale (FOIS; 3 studies) and Degree of Dysphagia (DD; 3 studies).

*rTMS Intervention (n = 11:* Tables 2 and 3*).* All but one of the rTMs studies [33] compared rTMS stimulation with sham rTMS. One single study compared rTMS with rTMS combined with DT, and DT only [33]. Three more studies included three arms; two studies compared rTMS using different frequencies versus sham rTMS [32,34], and one study compared bilateral and unilateral rTMS versus sham rTMS [35].

*tDCS Intervention (n = 9:* Tables 2 and 3*).* Eight studies compared tDCS with sham tDCS [27,29,31,36–41], and one study compared tDCS with theta-burst stimulation (TBS) [31]. All but one study (31) combined both study arms with DT. In one study both groups received simultaneous catheter balloon dilatation in addition to DT [40].

*Combined Neurostimulation Interventions (n = 4: see* Tables 2 and 3*).* Three studies in the combined intervention group compared three different treatments. Of these, one compared rTMS, PES and paired associative stimulation (PAS) [42], a second compared DT, rTMS combined with DT, and NMES combined with DT [43], and a third compared rTMS, PES and capsaicin stimulation [44]. A fourth study combined NMES stimulation with sham rTMS or rTMS stimulating different hemispheres (ipsilesional, contralesional or bilateral) [45].




**Group Descriptives (Mean** ± **SD)**

**Procedure, Delivery and Dosage per Intervention**

Procedure:

•

• 10 days (5 days per week) followed by 5

booster sessions every month for 3 months

• 10 trains of 20 Hz stimulation, each lasting

for 10 s with intertrain interval of 25 s.

Intensity set at 90% of the RMT.

> •

Stimulation to cortical area: first dorsal

interosseous (hand area) for each

hemisphere. Location identified from where

rTMS elicited MEPs of the highest

amplitude.

•

Both hemispheres stimulated, one at a time

during each session.

Sham group:

•

Similar parameters producing the same

noise, but with the coil rotated away from

scalp

rTMS or sham (Magstim 200)

**Group a**

**Age, Gender, Medical Diagnoses** Treatment group: Age 60.7 ± 8.8

duration of illness 5.7 +/− 3.9

Hoehn and Yahr 3.1 +/− 1.1 Sham group: Age 57.4 ± 10.0

duration of illness 6.5 +/− 3.7

Hoehn and Yahr 3.5 +/− 1.0

Gender distribution not given. NS difference between groups.







cortical excitability






standard swallow therapy



stimulation


• 30 min/day, 5 days/week, 2 weeks



between-blocks

 pause.


**Table 2.** *Cont.*


Mini-Mental State Exam; electrical stimulation; NR—not reported; NS—not significant;

electrical stimulation;

stimulation;

TOR-BSST—Toronto

 Bedside Swallowing Screening test;

MRI—magnetic

rTMS—repetitive

 transcranial magnetic stimulation; SLT—Speech

 resonance imaging; MS—multiple

 sclerosis; MT—Motor Threshold;

OD—oropharyngeal

 dysphagia; OST—oral sensorimotor

 and Language Therapist;

VFSS—videofluoroscopic

 swallowing study.

NIHSS—National

 treatment; TBI—traumatic

 brain injury;

 Institutes of Health Stroke Scale; PAS—penetration—aspiration

 scale; tDCS—transcranial

 direct current

PES—pharyngeal

NMES—neuromuscular

**Table 3.**Outcome of rTMS and tDCS for people with oropharyngeal dysphagia.






*J.* 











230



SLT—speech

current stimulation

 and language therapist;

UPDRS—unified

 Parkinson's disease rating scale;

SSA—standardised

 swallowing assessment;

SWAL-QOL—swallowing

VFSS—videofluoroscopic

 swallowing study; WST—water

 quality of life;

TBI—traumatic

 swallow test.

 brain injury;

tDCS—transcranial

 direct

#### *3.3. Risk of Bias Assessment and Methodological Quality*

The Begg and Mazumdar rank correlation procedure produced a tau of −0.036 (twotailed *p* = 0.902) and 0.178 (two-tailed *p* = 0.536) for rTMS and tDCS, respectively. The rTMS meta-analysis incorporates data from 8 studies, which yield a *z*-value of 2.348 (two-tailed *p*-value = 0.019). The fail-safe N is 4. This means that 4 'null' studies need to be located and included for the combined two-tailed *p*-value to exceed 0.050. That means there would be need to be 0.5 missing studies for every observed study for the effect to be nullified. The tDCS meta-analysis incorporates data from 8 studies yielding a *z*-value of 4.857 (two-tailed *p*-value < 0.001). The fail-safe N is 42 indicating 42 'null' studies need to be located and included for the combined two-tailed *p*-value to exceed 0.050; there would be need to be 5.3 missing studies for every observed study for the effect to be nullified. Both of these procedures (i.e., Begg and Mazumdar rank correlation and fail-safe N test) indicate the absence of publication bias.

Figures 2 and 3 present, respectively, the risk of bias summary per domain for all included studies combined and for individual studies, assessed using the Revised Cochrane Collaboration tool for assessing risk of bias (RoB 2) [21]. The majority of studies had low risk of bias with very few exceptions.

**Figure 2.** Risk of bias summary for all included studies (*n* = 24) in accordance with RoB 2 [21].

*3.4. Meta-Analysis: Effects of interventions*

3.4.1. rTMS Meta-Analysis

Eight studies using rTMS [32,33,35,42–44,50,51] were included in the meta-analysis. Of these, three studies provided data for two different interventions groups [32,35,36]. Six studies were excluded as OD was not confirmed by instrumental assessment and one study was excluded as rTMS was combined with NMES.

Overall within-group analysis. Pre-post intervention effect sizes ranged from 0.085 to 2.068 (Figure 4) with seven studies showing large effect sizes (Hedges' *g* > 0.8). Pre-post interventions produced a significant, large effect size (Hedges' *g* = 1.038).

**Figure 3.** Risk of bias summary for individual studies (*n* = 24) in accordance with RoB <sup>2</sup> [21,27–34,36–50]. *Note***.** If one or more yellow circles (domains) have been identified for a particular study, the Overall score (last column) shows an exclamation mark, indicating that the study shows some concerns (yellow circle with exclamation mark).


**Figure 4.** rTMS within intervention group pre-post meta-analysis [32,33,35,42–44,50,51]. *Notes*. Kim et al. (2011a): high frequency, Kim et al. (2011b): low frequency; Park et al. (2017a): unilateral stimulation, Park et al. (2017b): bilateral stimulation; Tarameshu et al. (2019a): rTMS, Tarameshu et al. (2019b): rTMS plus DT.

Overall between-group analysis. A significant, small post-intervention between-group total effect size was calculated in favour of rTMS (random-effects model: *z(7)* = 2.338, *p* = 0.019, Hedges' *g* = 0.355, and 95% CI = 0.057–0.652; Figure 5). Between-study heterogeneity was non-significant (*Q*(7) = 6.763, *p* = 0.454).

**Figure 5.** rTMS between group post meta-analysis [32,34,35,37,47,49]. *Notes.* Kim et al. (2011a): high frequency versus sham, Kim et al. (2011b): low frequency versus sham; Park et al. (2017a): unilateral stimulation versus sham, Park et al. (2017b): bilateral stimulation versus sham.

Between-subgroup analyses. Subgroup analyses were conducted to compare time between pre- and post-intervention measurement, stimulation sites (bilateral, contralesional and ipsi-lesional sites), pulse ranges (low: ≤600; medium; >600 and <10,000; high: ≥10,000 pulses), stimulation frequencies (1, 5 and 10 Hz), and optional behavioural training (rTMS versus rTMS + DT; Table 4). No subgroup comparisons for outcome measures were conducted as all but one study used PAS. Studies including a longer time span between pre- and post-interventions (indicating longer stimulation times) showed increased positive effect sizes compared to one-day interventions, which showed negligible effect sizes. When comparing stimulation sites, non-significant, positive effect sizes were obtained for all three stimulation groups with large ranges in effect sizes within groups. Pulse range comparisons indicated an increased significant, positive effect for higher pulse ranges. Effect sizes were only significant for large numbers of pulses delivered. Subanalyses comparing stimulation frequencies did not indicate obvious tendencies between groups. rTMS in combination with DT showed non-significant, small positive effect sizes in one study, whereas DT alone showed similar significant, small effects sizes.

**Table 4.** Between subgroup meta-analyses per type of neurostimulation comparing intervention groups of included studies.



**Table 4.** *Cont.*

*Note.* \* Significant. *Notes.* CI—confidence interval; DOSS—dysphagia outcome and severity scale; DSRS dysphagia severity rating scale; DT—dysphagia therapy; FDS—functional dysphagia scale; rTMS—repetitive transcranial magnetic stimulation.

#### 3.4.2. tDCS Meta-Analysis

A total of eight studies using tDCS in stroke patients were included in the metaanalysis [27,29,36–41]. One study was excluded as having too few data for meta-analysis [31]. Overall within-group analysis. The overall pre-post intervention effect size was 1.385, with effect sizes ranging from 0.432 (small effect) to 3.365 (high effect; Figure 6). Studies showed small (*n* = 2), moderate (*n* = 1), and high effect sizes (*n* = 5).

**Figure 6.** tDCS within intervention group pre-post meta-analysis [27,29,36–41].

Overall between-group analysis. A moderate but significant post-intervention betweengroup total effect size in favour of tDCS was found using a random-effects model (*z*(7) = 3.332, *p* = 0.001, Hedges' *g* = 0.655, and 95% CI = 0.270–1.040; Figure 7). Between-study heterogeneity was significant (*Q*(7) = 15.034, and *p* = 0.036), with *I <sup>2</sup>* showing that heterogeneity accounted for 53.4% of variation in effect sizes across studies.


**Figure 7.** tDCS between group post meta-analysis [27,29,36–41].

Between subgroup analyses. Subgroup analyses were conducted comparing time between pre- and post-intervention measurements, outcome measures, total stimulation times and stimulation current (Table 4). Increasing the number of days between pre- and post-intervention showed a strong tendency towards increased positive effect sizes, with significant effect sizes for two and four-week periods. Comparisons between measures resulted in significant, large positive effect sizes for visuoperceptual evaluation of instrumental assessment, but negligible effects when using an oral intake measure. Effect sizes for comparisons between total stimulation times indicated increased effects when using longer stimulation times. Significant, large effects were demonstrated for stimulation times of 300 min and longer. Additionally, higher stimulation currents resulted in increased significant, large positive effect sizes.

#### **4. Discussion**

This systematic review (Part II) aimed to determine the effects of rTMS and tDCS in people with OD. This systematic review and meta-analysis of RCT studies were completed in accordance with PRISMA procedures [19,20]. No populations were excluded based on medical diagnoses.

#### *4.1. Systematic Review Findings*

Like the systematic review on effects of NMES and PES in people with OD (Part I) [75], methodological problems were identified relating to unclear definitions of OD and differences in methods of confirming the presence of OD (i.e., using instrumental assessment, patient self-report or clinical assessment). Consequently, to reduce heterogeneity in participant characteristics between RCTs, only studies using instrumental assessment to confirm diagnosis of OD were included in meta-analyses. As most studies included stroke patients only, no meta-analysis could be performed to determine effects per medical diagnosis.

With the exception of one study [33], all rTMS studies included in the meta-analysis used the PAS to evaluate intervention effects. For the tDCS studies, as heterogeneity in outcome measures was larger, data on three different clinical outcome measures were used when conducting the meta-analysis. All rTMS studies used sham stimulation as a comparison group with the exception of one study which included a rTMS plus DT group [33]. For the tDCS studies, all but one study [31] combined neurostimulation with simultaneous DT. When comparing the degree of heterogeneity in study designs between brain neurostimulation (i.e., rTMS and tDCS) and peripheral neurostimulation (i.e., NMES and PES), those in the peripheral neurostimulation group were more diverse, creating greater challenges for conducting meta-analyses. Non-invasive brain stimulation studies tended to recruit smaller sample populations compared to peripheral studies [75].

#### 4.1.1. rTMS

This review prioritised reducing heterogeneity for purposes of meta-analysis. In contrast to previously published reviews that did not confirm OD by instrumental assessment, those studies were excluded from this meta-analysis. With the exception of Bath, Lee [13], earlier reviews identified significant beneficial effects of rTMS. Therefore, even though comparing the current meta-analysis with analyses from previous reviews may be challenging due to the inclusion of different outcome data, the findings from these studies seem in line with each other and this review.

#### 4.1.2. tDCS

Fewer RCTs were identified for tDCS compared with rTMS. Eight out of nine studies were eligible for meta-analysis, with one study excluded due to insufficient data; this was the only study to include non-stroke patients (presbydysphagia) [31]. Again, as previous reviews on tDCS [10,12,13,16–18] applied different criteria for inclusion and study methodology (e.g., differences in selection of electronic databases and publication years), final numbers of studies used for these meta-analyses ranged between two and seven publications, with reviews published before 2020 including four or fewer studies. When comparing the present results with the two most recent reviews [10,18] (both including seven studies), the beneficial effects of tDCS identified by this review were confirmed by significant, small-to-moderate effects in favour of tDCS.

#### 4.1.3. Moderators

Several factors may have had an impact on conducting meta-analyses and results. Comparing previous reviews, different decisions were made concerning criteria for metaanalyses. For example, Bath, Lee [13] excluded comparison groups with active treatment components and Chiang, Lin [12] excluded chronic stroke patients. Chronicity of stroke has shown to influence effect sizes [10,18], but selecting different primary outcomes may also result in deviating findings. For instance, Bath, Lee [13] did not find any positive effects for either rTMS or tDCS on primary outcome measures defined as death or dependency at the end of trials. Additionally, underlying medical diagnoses of OD are expected to affect meta-analyses. However, no conclusions could be drawn as very few studies of non-stroke patients were included in this review, thus no meta-analysis differentiating between diagnoses was conducted.

Similar reasons for hindering comparisons between RCTs are present in the current review, for example, spontaneous recovery and stroke severity, as were identified in the systematic review on effects of NMES and PES in people with OD (Part I) [75]. To account for the possibility of spontaneous recovery in participants, only between-subgroup metaanalyses were conducted using post-intervention data. However, the effects of stroke severity linked to OD severity remains unclear as RCTs usually did not report on the severity of stroke in sufficient detail.

Lastly, brain neurostimulation between RCTs may differ with respect to stimulation protocols (e.g., stimulation site, number and duration of treatment sessions and period) and technical parameters (e.g., frequency or number of pulses). The relatively low numbers of RCTs included in this review meant that meta-analysis could not incorporate all potential moderators. However, many of the included studies lacked sufficient details on technical parameters to allow further comparisons.

#### *4.2. Limitations*

Although this review followed PRISMA guidelines and aimed at reducing bias, some limitations may have had an impact on the results as presented. Only RCTs published in English were eligible in this review. Thus, some RCTs may have been excluded based on language criteria when their findings could have contributed to the current meta-analysis. Moreover, the high degree of heterogeneity between included studies hampered metaanalyses. Therefore, the results of meta-analyses and generalisations made should be interpreted with care.

#### **5. Conclusions**

The results of this systematic review suggest that both rTMS and tDCS show promising effects in people with OD. Meta-analysis for RCTs identified large pre-post intervention effect sizes for both types of brain neurostimulation. In addition, this analysis found significant, small and moderate post-intervention between-group effects in favour of rTMS and tCDS, respectively. However, comparisons between studies remain uncertain and challenging due to high heterogeneity in stimulation protocols and experimental parameters, potential moderators of stimulation effects, small samples sizes, and inconsistent methodological reporting.

These findings suggest that there is a need for RCTs including larger sample sizes to support future meta-analyses that will be able to adequately account for the presence of moderators. In addition, international consensus on standardised study protocols and reporting guidelines is required to support comparisons between studies.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm11040993/s1, Table S1: PRISMA 2020 for Abstracts Checklist, Table S2: PRISMA 2020 Checklist.

**Author Contributions:** Conceptualization: R.S., R.C., A.-L.S., L.B. and S.H., Formal analysis: R.S. and R.C., Methodology: R.S. and R.C., Project administration: R.S. and R.C.; Validation: R.S. and R.C.; Writing—review & editing: R.S., R.C., A.-L.S., L.B., S.H., T.P., M.B. and A.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **Behavioural Interventions in People with Oropharyngeal Dysphagia: A Systematic Review and Meta-Analysis of Randomised Clinical Trials**

**Renée Speyer 1,2,3,\*, Reinie Cordier 2,4, Anna-Liisa Sutt 5,6, Lianne Remijn 7, Bas Joris Heijnen 3, Mathieu Balaguer 8, Timothy Pommée 8, Michelle McInerney <sup>9</sup> and Liza Bergström 10,11**


**Abstract:** Objective: To determine the effects of behavioural interventions in people with oropharyngeal dysphagia. Methods: Systematic literature searches were conducted to retrieve randomized controlled trials in four different databases (CINAHL, Embase, PsycINFO, and PubMed). The methodological quality of eligible articles was assessed using the Revised Cochrane risk-of-bias tool for randomised trials (RoB 2), after which meta-analyses were performed using a random-effects model. Results: A total of 37 studies were included. Overall, a significant, large pre-post interventions effect size was found. To compare different types of interventions, all behavioural interventions and conventional dysphagia treatment comparison groups were categorised into compensatory, rehabilitative, and combined compensatory and rehabilitative interventions. Overall, significant treatment effects were identified favouring behavioural interventions. In particular, large effect sizes were found when comparing rehabilitative interventions with no dysphagia treatment, and combined interventions with compensatory conventional dysphagia treatment. When comparing selected interventions versus conventional dysphagia treatment, significant, large effect sizes were found in favour of Shaker exercise, chin tuck against resistance exercise, and expiratory muscle strength training. Conclusions: Behavioural interventions show promising effects in people with oropharyngeal dysphagia. However, due to high heterogeneity between studies, generalisations of meta-analyses need to be interpreted with care.

**Keywords:** deglutition; swallowing disorders; RCT; intervention; compensation; rehabilitation

#### **1. Introduction**

Swallowing disorders, or oropharyngeal dysphagia (OD), can be the result of many underlying conditions such as stroke, progressive neurological diseases, and acquired brain

**Citation:** Speyer, R.; Cordier, R.; Sutt, A.-L.; Remijn, L.; Heijnen, B.J.; Balaguer, M.; Pommée, T.; McInerney, M.; Bergström, L. Behavioural Interventions in People with Oropharyngeal Dysphagia: A Systematic Review and Meta-Analysis of Randomised Clinical Trials. *J. Clin. Med.* **2022**, *11*, 685. https://doi.org/10.3390/ jcm11030685

Academic Editor: Adrien Eshraghi

Received: 7 December 2021 Accepted: 23 January 2022 Published: 28 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

injury. They may also be the consequence of treatment side effects; for example, radiation or surgical interventions in patients with head and neck oncological disorders. Prevalence of OD in the general population ranges from 2.3 to 16% [1]. However, depending on underlying disease severity and outcome measures used (e.g., instrumental assessment, screening or patient self-report) [2], prevalence estimates can be as high as 80% in stroke and Parkinson's disease patients, up to 30% in traumatic brain injury patients, and over 90% in patients with community-acquired pneumonia [3]. Also, pooled prevalence estimates for swallowing problems in people with cerebral palsy determined by meta-analyses are as high as 50.4% [4].

OD may have severe effects on a person's health as dysphagia can lead to dehydration, malnutrition, and aspiration pneumonia. OD also has a high disease burden and poses a major societal challenge, which is associated with significant psychological and social burden, resulting in reduced quality-of-life for both patients and caregivers [5].

The treatment of OD may include surgical, pharmacological and behavioural interventions. Behavioural interventions include: bolus modification and management (e.g., adjusting the viscosity, volume, temperature and/or acidity of food and drinks), motor behavioural techniques or oromotor exercises, general body and head postural adjustments, swallowing manoeuvres (e.g., manoeuvres to improve food propulsion into the pharynx and airway protection), and sensory and neurophysiologic stimulation (e.g., neuromuscular electrical stimulation [NMES]) [6].

An increasing number of reviews have been published over the last two decades on the treatment effects of behavioural interventions in people with OD. However, only one systematic review [7] summarised the effects of swallowing therapy as applied by speech and language therapists without restrictions on subject populations or study designs. Furthermore, while most reviews have focussed on selected types of interventions and patient populations, very few reviews use criteria related to study designs (e.g., [8,9] solely including randomised controlled trials [RCTs], ranked as the highest level of evidence [10]).

This systematic review aimed to determine the effects of behavioural interventions in people with OD based on the highest level of evidence (RCTs) only. Behavioural interventions comprised any intervention by a dysphagia expert, excluding surgical and pharmacological interventions. Clinicians being referred to as dysphagia experts include speech therapists, occupational therapists, or physiotherapists, but may incorporate other disciplines depending on national healthcare and education systems. Finally, neurostimulation techniques were considered out of scope of this current review.

#### **2. Methods**

The methodology and reporting of this systematic review were based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement and checklist. The PRISMA 2020 statement and checklist (Supplementary Tables S1 and S2) aim to enhance the essential and transparent reporting of systematic reviews [11,12]. The protocol for this review was registered at PROSPERO, the international prospective register of systematic reviews (registration number: CRD42020179842).

#### *2.1. Information Sources*

To identify studies, literature searches were conducted on 6 March 2021, across these four databases: CINAHL, Embase, PsycINFO, and PubMed. Publications dates ranged from 1937–2021, 1902–2021, 1887–2021, and late 1700s–2021, respectively. Additional searches included checking the reference lists of eligible articles.

#### *2.2. Search Strategies*

Electronic search strategies were performed in all four databases using subheadings (e.g., MeSH and Thesaurus terms) and free text terms. Two strings of terms were combined: (1) dysphagia and (2) randomised controlled trial. The full electronic search strategies are reported in Table 1.



#### *2.3. Inclusion and Exclusion Criteria*

The following criteria for inclusion were applied: (1) participants had a diagnosis of OD; (2) behavioural interventions were aimed at reducing swallowing or feeding problems; (3) studies included a comparison group; (4) participants were randomly assigned to one of the study arms or groups; (5) studies were published in English.

Studies focussing on drooling, self-feeding, gastro-oesophageal reflux or oesophageal dysphagia (e.g., dysphagia resulting from oesophageal carcinoma or esophagitis) were excluded. Further excluded studies were those describing drug-induced swallowing problems, temporary swallowing problems caused by oedema post-surgery (e.g., anterior cervical discectomy), or swallowing problems associated with adverse effects of interventions such as inflammation and oedema resulting from recent radiotherapy (≤three months after intervention) or thyroidectomy. Studies reporting solely on feeding tube removal after intervention that did not provide data on swallowing or feeding problems, were also excluded. Studies on behavioural eating problems including bulimia, anorexia, and picky eaters, were out of scope of this review. Finally, only original research was included, thus excluding, for example, conference abstracts, doctoral theses and reviews.

#### *2.4. Systematic Review*

*Methodological Quality and Risk of Bias.* The Revised Cochrane risk-of-bias tool for randomised trials (RoB 2) [13] was used to assess the methodological quality of the included studies. The RoB 2 tool provides a framework for evaluating the risk of bias in the findings of any type of randomised trial. The tool is structured along five domains through which bias might be introduced into the study results: (1) the randomisation process; (2) deviations from intended interventions; (3) missing outcome data; (4) measurement of the outcome; (5) selection of the reported result.

*Data Collection Process.* A data extraction form was created to extract data from the included studies under the following categories: methodological quality, participant diagnosis, inclusion criteria, sample size, age, gender, intervention goal, intervention agent/delivery/dosage, intervention condition, outcome measures and treatment outcome.

*Data, Items and Synthesis of Results.* Two independent raters reviewed all titles and abstracts, then original articles, for eligibility. Inclusion of studies was based on consensus between raters. To ensure rating accuracy, two group sessions were held to discuss ratings of one hundred randomly selected records to achieve consensus before rating the remaining abstracts. Where consensus could not be reached between the first two raters, a third party was consulted for resolution. Methodological quality assessment was also rated by two independent researchers, after which consensus was reached with involvement of a third reviewer, when necessary. No evident bias in article selection or methodological study quality rating was present as none of the reviewers had formal or informal affiliations with any of the authors of the included studies. At this stage reviewers did not exclude studies based on type of intervention (e.g., behavioural intervention, neurostimulation).

During data collection, data points across all studies were extracted using comprehensive data extraction forms. Risk of bias was assessed per individual study using RoB 2 [13]. The main summary measures for assessing treatment outcome were effect sizes and significance of findings.

#### *2.5. Meta-Analysis*

Data was extracted from relevant studies to compare the effect sizes for the following: (1) pre-post outcome measures of OD and (2) mean difference in outcome measures from pre to post between different types of behavioural interventions. All interventions were categorised into compensatory (e.g., body and postural adjustments, or bolus modification), rehabilitative (e.g., oromotor exercises or Shaker exercise), combined compensatory and rehabilitative interventions, and no dysphagia intervention. Only studies using instrumental assessment (videofluoroscopic swallow study [VFSS] or fiberoptic endoscopic evaluation of swallowing [FEES]) to confirm OD were included. Outcome measures based on visuoperceptual evaluation of instrumental assessment and clinical non-instrumental assessments, were eligible for inclusion in meta-analyses. However, if both types of data were available, instrumental assessment was preferred over non-instrumental assessment outcome data. Oral intake measures, screening tools and patient self-report measures were excluded from meta-analyses. Measures other than the authors' primary outcomes may have been selected if these measures helped to reduce heterogeneity between studies.

To compare effect sizes, group means, standard deviations, and sample sizes for pre- and post-measurements were entered into Comprehensive Meta-Analysis Version 3.3.070 [14]. If only non-parametric data were available (i.e., medians, interquartile ranges), then data were converted into parametric data for meta-analyses. Participants in studies of multiple intervention groups were analysed separately. Where studies used the same participants, only one study was included in the meta-analysis. If studies provided insufficient data for meta-analyses, authors were contacted by e-mail for additional data.

Effect sizes were calculated in Comprehensive Meta-Analysis using a random-effects model. Due to variations in participant characteristics, intervention approaches, and outcome measurements, studies were unlikely to have similar true effects. Heterogeneity was estimated using the *Q* statistic to determine the spread of effect sizes about the mean and *I* <sup>2</sup> was used to estimate the ratio of true variance to total variance. *I* 2-values of less than 50%, 50% to 74%, and higher than 75% indicate low, moderate, and high heterogeneity, respectively [15]. Using the Hedges *g* formula for standardized mean difference with a confidence interval of 95%, effect sizes were calculated and interpreted using Cohen's *d* convention: *g* ≤ 0.2 as no or negligible effect; 0.2 < *g* ≤ 0.5 as minor effect; 0.5 < *g* ≤ 0.8 as moderate effect; and *g* > 0.8 as large effect [16].

Forest plots of effect sizes for OD outcome scores were generated for pre-post behavioural interventions. Due to blended configurations of intervention groupings across studies it was not possible to compare a homogenous behavioural intervention group against a comparison group that did not have a behavioural component. For this reason, only a subgroup between group analysis was conducted (and not an overall between group analysis) to explore effect sizes as a function of various moderators. Behavioural interventions (compensatory, rehabilitative, or combined compensatory and rehabilitative interventions) were compared with conventional dysphagia treatment (CDT), or no dysphagia therapy groups. Other subgroup analyses were conducted to compare effect sizes between selected interventions (i.e., Shaker exercise, Chin Tuck Against Resistance exercise [CTAR], and Expiratory Muscle Strength Training [EMST]), medical diagnoses, and outcome measures. Only between-subgroup meta-analyses were conducted using post-intervention data, to account for possible spontaneous recovery during the period of intervention.

Using Comprehensive Data Analysis software, publication bias was assessed following the Begg and Muzumdar's rank correlation test and the fail-safe N test. The Begg and Muzumdar's rank correlation test reports the rank correlation between the standardised effect size and the variances of these effects [17]. This statistical procedure produces tau as well as a two tailed *p* value; values of zero indicate no relationship, whereas deviations away from zero indicate a relationship. High standard error would be associated with larger effect sizes if asymmetry is caused by publication bias. Tau would be positive if larger effects are presented by low values, while tau would be negative if larger effects are represented by high values.

The fail-safe N test calculates how many studies with effect size zero could be added to the meta-analysis before the result lost statistical significance. That is, the number of missing studies that would be required to nullify the effect [18]. If this number is relatively small, then there is cause for concern. However, if this number is large, it can be stated with confidence that the treatment effect, while possibly inflated by the exclusion of some studies, is not nil.

#### **3. Results**

#### *3.1. Study Selection*

A total of 8059 studies were retrieved across four databases: CINAHL (*n* = 239), Embase (*n* = 4550), PsycINFO (*n* = 231), and PubMed (*n* = 3039). After removal of duplicate titles and abstracts (*n* = 1113), a total of 6946 records remained. After assessing titles and abstracts, 261 original articles were identified. Full-text records were accessed to verify all inclusion criteria. During full-text assessment, articles were divided into different types of interventions, as this systematic review reports on behavioural interventions only. Based on the inclusion criteria, 36 articles were included, after which one study was identified through reference checking of the included articles. Figure 1 presents the flow diagram of the article selection process according to PRISMA.

#### *3.2. Description of Studies*

All 37 included studies are described in detail in Tables 2 and 3. Table 2 reports on study characteristics, definitions and methods of diagnosing oropharyngeal dysphagia, and details on participant groups. Information such as medical diagnosis, sample size, age and gender, is provided on all study groups. Table 3 presents intervention goals, intervention components, outcome measures and treatment outcome of each included study.

*Participants* (Table 2). The 37 studies included a total of 2656 participants (mean = 72; SD = 124.5), with the sample sizes across studies ranging from 10 [30] to 742 participants [38]. All but two studies reported the mean age of participants [38,49], which was 65.6 years (SD = 8.8). Participant age range was only reported in five studies, ranging between 55 [36] and 95 [38] years. The mean percentage of male participants across all studies was 55.8% (SD = 13.7).

Most studies included stroke patients (*n* = 24). Other diagnoses included: patients with Parkinson's disease [19,39,52], acquired brain injury [30], multiple sclerosis [51] and nasopharyngeal cancer [50]. Two studies included a mixed patient population with Parkinson's disease or dementia [38], and stroke or head and neck cancer patients after chemoradiation [40]. Five studies did not provide further details on diagnoses [28,38,49,54,55]. The most frequent method for confirming OD was VFSS (*n* = 17), with only four studies using FEES (*n* = 4) [20,31,38,40]. Seven studies used non-instrumental clinical assessments, five studies used a screening tool [28,29,39,48,56], and four studies used patient self-reported dysphagia [49,52,54,55]. The included studies were conducted across fifteen countries, with studies most frequently conducted in Korea (*n* = 13), USA (*n* = 6), China (*n* = 3) and Japan (*n* = 3).

**Figure 1.** Flow diagram of the reviewing process according to PRISMA.

*Outcome measures* (Table 3). Many different outcome measures were used across the included studies targeting different domains within the area of OD. The most frequently used measures were the Penetration Aspiration Scale (PAS; 15 studies), the Functional Oral Intake Scale (FOIS; 8 studies), various water swallow tests (4 studies), and the Mann Assessment of Swallowing Ability (MASA; 3 studies). All other outcome measures were used in one or two studies only, confirming the substantial heterogeneity in outcome measures.

*Interventions* (Table 3). The included 37 studies comprised a range of behavioural interventions, delivered by various health professionals. The interventions were most frequently implemented by single allied health disciplines: occupational therapists in ten studies, speech pathologists in eight studies, physical therapists in two studies [36,48], and nursing staff in one study [55]. In five studies, more than one discipline was involved [23,27,28,33,48], and two studies reported caregivers as the intervention agent either as a single agent [24] or in addition to occupational therapists [22]. Nine studies did not specify disciplines involved in providing the interventions. The intervention dosage varied greatly, ranging from one training session [54] to exercise 3 times daily, 7 days per week for 42 days [25].

*Behavioural intervention groups* (Table 3). Of the 37 included studies, seven studies comprised three participant groups [19–21,23,25,26,38], whereas all other studies included two groups. Based on authors' description of therapy contents, all intervention groups were categorised into compensatory, rehabilitative, and combined compensatory and rehabilitative interventions. Ten studies included different types of intervention groups (i.e., compensatory, rehabilitative and/or combined compensatory and rehabilitative intervention groups). Five studies included only compensatory groups [20,24,38,39,55], ten studies included only rehabilitative groups, and thirteen studies included only groups combining compensatory and rehabilitative interventions.


*J.*

**Table2.**Studycharacteristicsofstudiesonbehavioural interventionsforpeoplewithoropharyngealdysphagia.











gastronomy tube, problems with the oesophageal phase of dysphagia

*J. Clin. Med.* **2022**, *11*, 685






disability score = Hoehn and Yahr disability score; K-MMSE or MMSE-K = Mini-mental examination Korean version; K- NIHSS = Korean version of National Institute of Health Stroke Scale; MASA = Mann Assessment of Swallowing Ability; MBS = Modified Barium Swallow; MIE = Minimally Invasive Oesophagectomy; MDTP = McNeill Dysphagia Therapy Program; MMSE = Mini-Mental State Examination; MOCA = Montreal Cognitive Assessment; NIHSS = National Institute of Health Stroke Scale; NMES = Neuromuscular Electrical stimulation; NR = Not reported; OD = Oropharyngeal dysphagia; PAS = Penetration-Aspiration Scale; PD = Parkinson's disease; P-DHI = Persian Dysphagia Handicap Index; PDQ-39: Parkinson's Disease Questionnaire-39; PNF = proprioceptive neuromuscular facilitation; RCT = Randomised Controlled Trial; SLP: Speech-Language Pathology; SRS = Swallowing Rating Scale; SSA = Standardized Swallowing Assessment; SIS-6 = Swallowing Impairment Score; SWAL-QOL = Swallow Quality-of-Life Questionnaire; tDCS = transcranial Direct Current Stimulation; UC = Usual Care; VDS = Video-fluoroscopic Dysphagia Scale; VFSS = Video-Fluoroscopic Swallowing Study; WHO = World Health Organisation; WST = Water Swallow Test; TWST = Timed Water-Swallow Test.


**Table 3.** Outcome of behavioural interventions for people with oropharyngeal dysphagia.




*J. Clin.Med.*685


*J.Clin.Med.*685


**Table3.***Cont.*

 **2022**, *11*,




Swallow-Oesophagectomy; DysphagiaTherapyProgram;Expiratoryreported;Oropharyngeal dysphagia; PAS = Penetration-Aspiration Scale; PD = Parkinson's disease; P-DHI = Persian Dysphagia Handicap Index; PNF= Proprioceptive Neuromuscular Facilitation; PRRS = Pharyngeal Residue Rating Scale; QoL = Quality of life; RCT = Randomised Controlled Trial; SIS-6 = Swallowing Impairment Score; SWAL-QOL= Swallow Quality-of-Life Questionnaire; VDS= Video-fluoroscopic dysphagia scale; VFSS = Video-Fluoroscopic Swallowing Study; TWST= Timed Water-Swallow Test; VDS = Videofluoroscopic Dysphagia Scale; VFSE = Videofluoroscopic Examination.

Most studies (*n* = 23) included a comparison group that received a type of dysphagia treatment often referred to as traditional therapy, standard swallow therapy, or conventional dysphagia treatment (CDT). Some studies also used the term usual care for CDT groups. CDT treatment could include counselling and the provision of information about swallowing and dysphagia, compensatory strategies (e.g., bolus modification and adjusted head positioning), rehabilitation, oromotor exercises and/or thermal stimulation. Three studies included a comparison group receiving medical standard care without dysphagia treatment [20,51,56]. In three studies, patients underwent sham dysphagia training [36,43,53]. Several studies compared two or three behavioural interventions without having a CDT or medical standard care group included [33,34,46,49,50,55].

#### *3.3. Risk of Bias Assessment*

The Begg and Mazumdar rank correlation procedure produced a tau of 0.305 (twotailed *p* = 0.113), indicating there is no evidence of publication bias. This meta-analysis incorporates data from 15 studies, which yield a *z*-value of 7.528 (two-tailed *p* < 0.001). The fail-safe N is 207. This means that 207 'null' studies need to be located and included for the combined two-tailed *p*-value to exceed 0.050. That is, there would need to be 13.8 missing studies for every observed study for the effect to be nullified. Both of these procedures (i.e., Begg and Mazumdar rank correlation and fail-safe N) indicate the absence of publication bias.

#### *3.4. Methodological Quality*

Risk of bias of the included RCTs was assessed using the RoB 2 tool. Figures 2 and 3 present the risk of bias summary per domain for individual studies and for all included studies. Most studies showed low risk of bias per domain, but more than half of the included studies (19/37) scored overall as having some concerns, with three studies identified as being at high risk.

#### *3.5. Meta—Analysis: Effect of Interventions*

Twenty-one studies were included in the meta-analyses [21,22,24,25,28–31,34,35,40– 46,49,51,52,54]. All study groups were categorised into compensatory interventions, rehabilitative interventions, combined compensatory and rehabilitative interventions, and no dysphagia intervention. Seventeen studies were excluded from meta-analyses: one study included patients with self-reported swallowing difficulties without confirmed OD diagnosis by instrumental assessment (VFSS or FEES) [48], four studies did not report on instrumental or clinical non-instrumental outcome data [20,28,37,40], ten studies provided

insufficient data for meta-analysis [21,24,27,34,38,39,48,51,56,57], and two studies were excluded to reduce heterogeneity between studies [32,53].

**Figure 3.** Risk of bias summary for individual studies (*n* = 37) in accordance with RoB2 [19–55]. *Note*. If one or more yellow or red circles (domains) have been identified for a particular study, the Overall score (last column) shows an exclamation mark, indicating that either the study shows some concerns (yellow circle with exclamation mark) or is at high risk (red circle with exclamation mark).

*Overall, within group analysis*. (Figure 4). A significant, large pre-post intervention effect size was calculated using a random-effects model (*z(35)* = 8.047, *p* < 0.001, Hedges' *g* = 1.139, and 95% CI = 0.862–1.416). Pre-post intervention effects varied greatly between studies, ranging from 0.058 to 5.732. Of the 36 intervention groups included in the metaanalysis, 19 groups showed large effect sizes (Hedges' *g* > 0.8), six groups showed moderate effects sizes (0.5 < Hedges' *g* ≤ 0.8), seven groups showed minor effect sizes (0.2 < Hedges' *g* ≤ 0.5), and four groups showed negligible effect sizes (Hedges' *g* ≤ 0.2). Between-study heterogeneity was significant (*Q*(35) = 152.938, and *p* < 0.001), with *I <sup>2</sup>* showing heterogeneity accounted for 77.115% of variation in effect sizes across studies.


**Figure 4.** Within intervention group pre-post meta-analysis [21,22,24,25,28–31,40–46,49,51,52,54,56]. *Note.* Refer to Table 2 for explanation of the subgroups.

*Between subgroup analyses*. Subgroup analyses (Table 4) were conducted comparing different types of interventions: behavioural interventions were compared with conventional dysphagia treatment (CDT), or no dysphagia therapy groups (Figure 5). Both behavioural interventions and CDT were categorised into mainly compensatory, rehabilitative, and combined compensatory and rehabilitative interventions. Overall, significant treatment effects were identified favouring behavioural interventions. In particular, large effect sizes were found when comparing rehabilitative interventions with no CDT, and combined interventions with compensatory CDT. When comparing selected interventions based on commonalities across studies against CDT, significant, large effect sizes were found in favour of Shaker exercise, chin tuck against resistance exercise (CTAR), and expiratory muscle strength training (EMST). Most studies were conducted in stroke populations and showed significant, moderate effect sizes. Comparisons between outcome measures indicated at significant effects for PAS only.


**Table 4.** Between subgroup meta-analyses comparing intervention groups of included studies.

*Notes*. \* Significant.

**Figure 5.** Between subgroup meta-analysis for different types of interventions: behavioural interventions compared with conventional dysphagia treatment (CDT) or no dysphagia therapy [21,22,25,29–31,34,35,41–44,46,52]. *Note.* Refer to Table 2 for explanation of the subgroups.

#### **4. Discussion**

This systematic review aimed to determine the effects of behavioural interventions in people with OD based on the highest level of evidence (RCTs) only. Findings from the literature were reported using PRISMA and meta-analysis procedures.

#### *4.1. Systematic Review Findings*

In total, 37 behavioural RCTs in OD were identified. Considering the high prevalence [3] and severe impact of OD on health [57], quality of life [5,58], and health-economics [59], the limited number of high-level evidence studies is concerning. RCTs are costly and usually require extensive funding [60]. Possibly, the general lack of awareness of OD [61] might place funding applications in this research area at a disadvantage when competing with well-known, life-threatening diseases such as cancer or stroke. Although OD is a symptom of these diseases, and many other underlying conditions, limited public knowledge persists, resulting in reduced understanding and recognition of the devastating consequences of OD, in both health-care and non-health-care practitioners [61].

Further, although RCTs are characterised by random allocation and allocation concealment, few of the included studies included sufficient reporting on the processes of randomization and blinding. These finding are in line with current literature on quality assessments of RCTs [62,63], confirming that the risk of selection bias [63] and the success of blinding methods in RCTs [62] can often not be ascertained due to frequent poor reporting.

When comparing behavioural RCTs in OD, several methodological challenges arise. Authors may use different definitions for OD or fail to provide sufficient details when reporting on the swallowing problems of the included patient populations. Also, several studies used non-instrumental assessments (i.e., patient self-report or a screening tool) to identify or confirm OD, making the comparison between studies precarious. The use of a screening tool is especially problematic in identifying OD and cannot act as confirmation of OD. A screening tool's purpose is merely to identify patients at *risk* of OD, after which further assessment may confirm or refute the diagnosis [2]. Additionally, although instrumental assessment is considered the optimal tool for confirming OD diagnosis, VFSS and FEES protocols may differ (e.g., using different numbers of swallow trials, viscosities, and volumes).

Studies used a wide range of outcome measures to evaluate treatment effects. Since OD is a multidimensional phenomenon [64], different dimensions of OD may result in different therapy outcomes. For example, changes in dysphagia-related quality of life or oral intake do not necessarily correlate with findings from instrumental assessment. As such, to reduce heterogeneity in meta-analyses, patient self-report and oral intake measures were excluded. Also, some studies included outcome measures with poor or unknown psychometric properties, which in turn undermines the interpretation of treatment effects as data may not be valid or reliable. In addition, measures with weak responsiveness characteristics are not sensitive to treatment changes and should therefore be avoided as outcome measures aiming to determine intervention effects [2].

Most studies included a combined rehabilitative and compensatory intervention group or a rehabilitative intervention group, with only a few studies including exclusively compensatory groups. As the interventions classified as CDT comparison groups showed large variation as well, CDT comparison groups were categorised into similar group types (compensatory and/or rehabilitative CDT). Overall, terminology in the literature referring to CDT comparison groups was varied and complex. This was especially pertinent when interventions were not described in sufficient detail and descriptive terms such as "usual care" or "traditional therapy" did not provide further clarity on the type or content of CDT provided. Despite using categories to group different types of interventions, some degree of heterogeneity was inevitable. Interventions used different types of exercises or care, in distinct dosages, and were applied by different health care professionals. Therefore, it is challenging to identify the "active" ingredients of individual interventions, especially as most studies combined the use of different treatment strategies.

#### *4.2. Meta-Analysis Findings*

When considering meta-analyses for behavioural interventions, overall significant treatment effects were identified as favouring behavioural interventions over CDT and withholding dysphagia therapy. Most promising intervention approaches were rehabilitative interventions, which were associated with large effect sizes. Additionally, rehabilitative interventions such as Shaker exercise, CTAR exercise, and EMST showed significant, large effect sizes. However, since most studies included in the meta-analysis provided data on stroke patients only, future research still needs to confirm these findings in other diagnostic populations such as Parkinson's disease, acquired brain injury or patients with head and neck oncology. As stated above, patient self-report and oral intake measures were excluded from meta-analyses to increase homogeneity between studies. Though self-report and oral intake data might be interesting for future meta-analyses, this would require additional RCTs to be published, as currently there is limited data available in the literature. Finally, future studies should report on treatment dosage and duration in more detail. Due to high heterogeneity between studies and incomplete reporting, no subgroup meta-analyses could be conducted for these variables.

#### *4.3. Limitations*

Although reporting of this review followed the PRISMA guidelines to reduce bias, some limitations are inherent to this study. As only RCTs published in English were included, some RCTs may have been excluded based on language criteria. In addition, meta-analyses were restricted because of heterogeneity of the included studies. As such, comparisons across studies are challenging and, generalisations and meta-analyses results should be interpreted with caution.

#### **5. Conclusions**

Meta-analyses for behavioural studies in oropharyngeal dysphagia identified an overall, significant, large pre-post interventions effect size. Significant treatment effects were identified favouring behavioural interventions over conventional dysphagia treatment. Notably, large effect sizes were found when comparing rehabilitative interventions with no dysphagia treatment and combined interventions with compensatory conventional dysphagia treatment. Selected interventions compared with conventional dysphagia treatment showed significant, large effect sizes in favour of Shaker exercise, CTAR, and EMST.

Behavioural interventions show promising effects in people with oropharyngeal dysphagia. Still, generalisations from this meta-analysis need to be interpreted with care due to high heterogeneity across studies.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm11030685/s1, Table S1: PRISMA 2020 for Abstracts Checklist; Table S2: PRISMA 2020 Checklist.

**Author Contributions:** Conceptualization: R.S., R.C., A.-L.S., L.B. Formal analysis: R.S., R.C. Methodology: R.S., R.C. Project administration: R.S., R.C. Validation: R.S., R.C. Writing–review & editing: R.S., R.C., A.-L.S., L.R., B.J.H., M.B., T.P., M.M., L.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Journal of Clinical Medicine* Editorial Office E-mail: jcm@mdpi.com www.mdpi.com/journal/jcm

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com

ISBN 978-3-0365-4084-9