The role of laryngology evaluation in Parkinson dysphonia therapy: Process evaluation and outcomes

Abstract

Objectives:

To describe the clinical evaluation process for patients with Parkinson disease (PD)-related dysphonia at an academic voice center, including potential delays in voice therapy initiation, the incidence of lesions at the time of screening laryngoscopy, and the frequency of procedural voice interventions.

Methods:

Retrospective single-institution chart review of patients (n = 113) with PD-related dysphonia seen in consultation by a fellowship-trained laryngologist and speech language pathologist (SLP) between 2014 – 2024 who underwent either SPEAK OUT!^® or Lee Silverman Voice Treatment^®. Sociodemographic information, referral/clinical encounter dates, medical history, laryngoscopy findings, and details regarding procedural interventions were collected.

Results:

Thirty-three patients satisfied inclusion criteria, including those who underwent appropriate laryngologist clearance prior to voice therapy (n = 29) and those who saw a laryngologist following treatment initiation in delayed fashion (n = 4). Time from SLP intake to therapy initiation occurred approximately 2 months sooner when patients did not see a laryngologist prior to treatment (mean 46 days versus 117 days, p < 0.05.) Screening laryngoscopy findings and voice-directed interventions ultimately performed did not significantly differ between cohorts.

Conclusion:

Requiring laryngology consultation for screening laryngoscopy prior to PD-directed voice therapy can delay care without affecting the procedural interventions ultimately performed. While all patients with PD-related dysphonia should undergo laryngoscopy at some time during treatment, these patients may benefit from timelier therapy if preceding medical clearance from laryngology is not required. Regardless of the timing of laryngologist consultation, select patients with PD may benefit from voice-directed interventions, and this warrants further study.

Introduction

Parkinson disease (PD) is a progressive neurodegenerative condition impacting nearly one million individuals in the United States.¹ While commonly reported symptoms include muscle rigidity, gait instability, and tremor, approximately 75 to 90 percent of PD patients also experience voice-related symptoms prodromally or throughout the course of their disease.¹ Voice alterations in PD include vocal tremor, hoarseness, breathiness/decreased intensity, change in pitch, and hypokinetic dysarthria.²

Commonly implemented PD-directed voice therapies include SPEAK OUT!^® and Lee Silverman Voice Treatment^® (LSVT), which seek to improve vocal loudness and functional communication over several sessions delivered by trained speech language pathologists (SLPs).^3,4 Existing literature supports the use of these PD-directed voice therapies. For instance, Behrman et al. found that with just 8 hours of SPEAK OUT!^® (12 sessions of 40-minute duration), patients have a significant improvement in acoustic measures of intensity, prosody, voice quality, and self-perception of voice.⁵ Rafferty et al. summarized international clinical practice guidelines for PD rehabilitation from Europe, United Kingdom, and Canada, all of which recommend early evaluation by SLP for evaluation of speech, swallowing, and communication-cognition.⁶ Evidence also shows that self-perceived speech impairment is often present in newly diagnosed PD and progresses significantly over time, highlighting the rationale for early voice therapy to improve or sustain speech function as the disease progresses.⁷

Prior to the initiation of PD-directed voice therapy, the American Academy of Otolaryngology dysphonia clinical practice guidelines recommend performing diagnostic laryngoscopy before initiating treatment and communicating these results to SLP colleagues.⁸ As such, the authors’ institution requires laryngology medical clearance with endoscopic evaluation prior to voice therapy initiation to exclude secondary causes of dysphonia such as vocal fold paralysis or a vocal fold lesion. It is presently unknown whether there is a greater incidence of actionable vocal fold lesions affecting voice quality in PD patients compared to the general dysphonia population, and relatedly, whether laryngology clearance is a necessary prerequisite in this select population with known PD status if no alarm features are present. In a recent study, Castro and colleagues queried laryngostroboscopy examinations from adult patients with PD and observed abnormal findings in 94% of patients, including characteristic findings of glottic insufficiency and vocal fold hypomobility/bowing, as well as nonspecific but generally nonactionable findings (e.g. arytenoid edema) in a smaller proportion (i.e., 22%) of patients.⁹ Previous literature has similarly examined laryngoscopic factors in PD-related dysphonia through an epidemiological lens, including a higher proportion of vocal fold motion abnormalities and lower proportion of bilateral vocal fold atrophy in women compared to men.¹⁰ Of note, these studies neither commented on the prevalence of any actionable incidental lesions noted on screening laryngoscopy (perhaps due to rarity), nor explored the process of voice therapy initiation or procedural interventions that were subsequently rendered.

The requirement to see a laryngologist prior to the initiation of voice therapy could potentially create a delay in care. For patients with PD-related dysphonia, a delay in voice therapy could potentially adversely impact quality of life by worsening communicative abilities and hindering social interaction, particularly among patients living longer with a diagnosis of PD. For instance, Boutsen et al. (2018) found longer PD duration fared negatively on various indices of SPEAK OUT!^® efficacy, including Voice Handicap Index (VHI) and Voice-Related Quality of Life (V-RQOL) measures; the authors therefore advocated for its initiation as soon as possible after the development of PD-related dysphonia.¹¹ Moreover, communication difficulties may contribute to feelings of isolation, being misunderstood, and depression, the latter which is rather prevalent (40%) among patients with PD and adversely affects quality of life.¹² Moreover, patients with PD tend to leave the workforce 4 – 7 years earlier than their cohorts, and mental health issues are a significant predictor for this early career loss.^13,14 Presently, to our knowledge, there is no existing literature surrounding the evaluative process and provision of PD-directed voice therapies, including timelines for care and whether the timing of voice therapy affects treatment outcomes or quality of life. With clear evidence that SPEAK OUT!^® can improve hypokinetic dysarthria and Voice-Related Quality of Life,⁶ the driving force behind the present study is to question whether delaying the initiation of speech care for known PD patients is supported by evidence suggesting a high preponderance of co-occurring actionable vocal fold lesions. Thus, the current study seeks to characterize the clinical evaluative process and timeline for voice therapy for patients with PD-related dysphonia at an academic voice center, as well as describe the prevalence of laryngoscopic findings and voice-directed interventions performed.

Methods

Study design

A retrospective single-institution electronic medical chart review was performed using data from medical billing records at an academic voice center based in a tertiary hospital in upstate New York. Inclusion criteria for patients included age 18 or older, a diagnosis of Parkinson disease (idiopathic or drug-induced), and a history of clinical encounters with both a fellowship-trained laryngologist (GTS, ELF, and two other faculty members) and an SLP trained in either SPEAK OUT!^® or LSVT^® techniques. Records satisfying these criteria with clinical encounter dates spanning January 2014 – September 2024 were queried and compiled for further electronic medical record review. Patients with a diagnosis of a Parkinson-plus syndrome (e.g. multiple system atrophy), history of head and neck cancer or radiotherapy, and patients seen predominantly for dysphagia or airway-related complaints were excluded from the study. The standard-of-care workflow for patients undergoing evaluation of PD-related dysphonia at the authors’ academic institution is depicted in Figure 1. In summary, patients diagnosed with PD by a neurologist or primary care provider were referred to the institution’s Division of Speech Language Pathology within the Department of Otolaryngology for consideration of PD-directed voice therapy. A team of SLPs trained in SPEAK OUT!^® or LSVT^® techniques then performed a baseline clinical and voice evaluation, including determination of patient candidacy for therapy. Patients who were deemed candidates for voice therapy were subsequently seen by a fellowship-trained laryngologist for medical clearance (including screening laryngoscopy) and then referred back to SLP for therapy initiation if interested in proceeding. Of note, a small subset of patients who met study inclusion criteria inadvertently began voice therapy prior to screening laryngoscopy by a laryngologist despite the institution’s standard-of-care workflow. Patients who circumvented the traditional clinical evaluation processes usually had a pre-existing relationship with Laryngology for an alternative clinical complaint (e.g. seen for dysphagia) and had previously underwent screening laryngoscopy prior to the development of PD-related dysphonia; these patients were included in the analysis as a sub-group to compare timelines for provision of care and procedural outcomes (see below).

Figure 1.

Institutional standardized workflow for patients referred to Otolaryngology/Speech Language Pathology for evaluation of PD-related dysphonia.

The present study was reviewed by the University of Rochester Institutional Review Board and given exempt status (study number 00009957).

Data of interest

Data were compiled by multiple authors via manual chart review after watching a web-based user training video aimed at standardizing data collection techniques. Variables of interest included sociodemographic information, referral and encounter-related dates of service, medical history (including relevant medications), type of voice therapy performed, flexible laryngoscopy findings (vocal fold atrophy, glottic insufficiency, vocal fold paresis/paralysis, and vocal fold lesions), and details pertaining to any voice-directed procedures (e.g. injection laryngoplasty) or surgeries (e.g. medialization thyroplasty) performed. Data were compiled in a university-sponsored REDCap electronic data repository and were only accessible to study personnel.¹⁵ The primary author secondarily audited each patient chart to ensure completeness and accuracy of collected data. Patient records were de-identified prior to export for analysis and kept secure on an encrypted device.

Statistical analysis

Descriptive statistics were tabulated for the aforementioned sociodemographic, historical, laryngoscopic, and procedural variables. Variables capturing the elapsed time between referral, SLP intake, laryngology clearance, and voice therapy initiation were also created to construct timelines for provision of care. Comparative analyses were performed among patients who met inclusion criteria on the basis of screening laryngoscopy timing (appropriately prior to voice therapy initiation versus inadvertently delayed until after voice therapy initiation), using independent t-tests and chi-square tests for continuous and categorical variables, respectively. The Fisher’s exact test correction and Mann-Whitney U test were applied when appropriate for statistical comparisons with small subgroup membership or nonparametric distributions for categorical and continuous variables, respectively. Statistical significance was set at an alpha level of 0.05. Statistical tests were performed using R software version 4.3.2.¹⁶

Results

Study sample

A total of 113 patients were identified through medical billing records. Upon further review during dedicated chart review, 22 patients were excluded due to a lack of qualifying PD diagnosis (e.g. parkinsonism or multiple system atrophy), 31 were excluded due to a lack of completion of voice therapy after the initial intake SLP session, 9 were excluded for receiving primarily dysphagia-directed SLP therapy as opposed to SPEAK OUT!^® or LSVT^®, and 18 were excluded for a lack of preceding referral to SLP. For these latter 18 cases, patients were seen first by the laryngologist, then subsequently an SLP in contrast to the traditional patient workflow.

The final study sample thus included 33 patients, including 29 patients who underwent laryngology screening laryngoscopy and medical clearance prior to voice therapy, as well as 4 patients who had inadvertently circumvented institutional processes in place to ensure physician clearance prior to therapy initiation and therefore saw a laryngologist at some point during the course of their voice therapy. Sociodemographic data are listed in Table 1. The mean age was 68 years (range 36 – 87); most patients were male, white, and non-Hispanic. The mean duration of PD diagnosis at the time of SLP intake evaluation was about 5 years (range 0 – 16). Nineteen (57.6%) patients underwent SPEAK OUT!^® therapy, while 14 (42.4%) participated in LSVT^®. Most (91%) patients had used dopaminergic medications following their PD diagnosis, while less than half had a smoking history or previously underwent deep brain stimulation.

Table 1.

Sociodemographic data.

Variable	Mean (SD) or n (%)
Age	68.2 (10.2)
Gender
Male	23 (69.7)
Female	10 (30.3)
Race/ethnicity
White/Non-Hispanic	32 (97.0)
Asian	1 (3.0)
Duration of PD diagnosis (years)	4.9 (3.9)
Smoking history	13 (39.4)
History of deep brain stimulation	10 (30.3)
Use of dopaminergic medications	30 (90.9)

SD = standard deviation; PD = Parkinson disease

Timeline of provision of care

Data highlighting the time elapsed between various dates of service are detailed in Table 2. The mean number of days between SLP intake and initiation of voice therapy was significantly longer for patients who underwent laryngologist clearance prior to beginning, versus those who inadvertently saw a laryngologist during the course of their voice therapy as opposed to prior (117 days versus 46 days; ranges 14 – 468 and 4 – 140, respectively; U = 93; z = 1.904, p = 0.028). For patients who underwent clearance prior to therapy, the mean times between SLP intake, laryngologist clearance, and initiation of voice therapy were 60 and 57 days, respectively.

Table 2.

Timelines for provision of care, based on timing of laryngology clearance.

	Clearance Prior to Therapy (n = 29) mean (SD)	Clearance During Therapy (n = 4) mean (SD)	t-value / z-score	p
Days from referral to SLP intake	34 (19.9)	17 (5.7)	1.162	0.25
Days from SLP intake to laryngologist clearance	60 (46.1)	-	-	-
Days from laryngologist clearance to voice therapy initiation	57 (64.9)	-	-	-
Days from referral to voice therapy initiation	153 (106.8)	36 (15.6)	1.524	0.07
Days from SLP intake to voice therapy initiation	117 (99)	46 (64.6)	1.904	0.028

SD = standard deviation; SLP = speech language pathology

Laryngoscopy findings and related interventions

Vocal fold atrophy was highly prevalent among both cohorts of patients (90 – 100%), and glottic insufficiency to a lesser degree (31 – 75%, see Table 3.) However, there were no statistically significant differences in laryngoscopy findings between those who underwent initial versus delayed laryngology clearance (Fisher’s exact values 0.125 – 1.000, p > 0.05). Six (21%) of patients undergoing appropriate clearance prior to therapy had a diagnosed vocal fold mobility impairment or lesion; of these, the sole patient with a vocal fold lesion was found to have a benign-appearing vocal process granuloma that resolved without intervention. Two (50%) patients who underwent delayed clearance after voice therapy initiation were also noted to have vocal fold paresis. Ten (34%) of patients in the traditional clearance pathway underwent injection laryngoplasty (IL), and 3 of those patients subsequently underwent bilateral medialization thyroplasty (MT). Among those who were cleared following voice therapy initiation in delayed fashion, 3 underwent IL, and similarly, 2 of these patients underwent MT. The frequency or type of interventions performed did not differ between the two cohorts (Fisher’s exact value 0.276, p > 0.05).

Table 3.

Screening laryngoscopy findings and procedures performed, based on timing of laryngology clearance.

	Clearance Prior to Therapy (n = 29) n (%)	Clearance During Therapy (n = 4) n (%)	Chi-square/Fisher’s exact value	p
Laryngoscopy findings			0.125 – 1.000	ns
Vocal fold atrophy	26 (90)	4 (100)
Glottic insufficiency	9 (31)	3 (75)
Vocal fold hypomobility/lesion	6 (21)	2 (50)
Procedures	10 (34)	3 (75)	0.276	ns
Injection laryngoplasty	10	3
Medialization thyroplasty	3	2

ns = not significant

Discussion

This study aimed to investigate the clinical evaluation and referral patterns of people with PD to PD-specific speech therapy at a single institution. We specifically aimed to evaluate whether there was a delay in treatment by requiring medical clearance with laryngoscopy prior to initiating voice therapy and the incidence of pre-existing laryngeal lesions. Secondarily, we evaluated the frequency of any procedural interventions resulting from the required medical clearance.

At our institution, patients with PD have an initial consultation (intake) with speech therapy but SPEAK OUT!^® (the standard PD-directed voice therapy provided at our institution since 2017; we formerly used the LSVT^® treatment program from 2014 – 2017) is not initiated until after clearance with screening laryngoscopy. We found that the average time from SLP intake to laryngoscopy was 60 days and voice therapy was initiated on average 57 days thereafter, resulting in a statistically significant average total delay in care of over 2 months compared to those who had inadvertently underwent laryngoscopy clearance after therapy had already been initiated. Moreover, when accounting for one of the 4 patients who underwent delayed clearance and had other mitigating personal factors precluding their initiation of voice therapy, the mean number of days between SLP evaluation and voice therapy initiation decreased from 46 to 14 days, reflecting nearly a 90% decrease in therapy initiation wait time. Our analysis found the incidence of pre-existing vocal fold lesions was 3% (1 in 33 patients), and this single lesion (benign vocal process granuloma) did not warrant interfering procedural or operative management.

This study is among the first to evaluate the incidence of pre-existing vocal fold lesions in patients with PD-related dysphonia. While we found the prevalence of pre-existing lesions to be extremely low, our results are congruent with other studies in that the presence of glottic insufficiency or atrophy was common (36% and 91%, respectively). Previous studies have evaluated the presence of coexisting abnormalities of vocal fold atrophy and glottic insufficiency but not the presence of lesions such as malignancy, polyps, or sulcus that may require procedural intervention prior to voice therapy initiation. Patients with PD-related dysphonia tend to have more severe vocal fold bowing than patients with age-related atrophy, correlating with worse Consensus Auditory-Perceptual Evaluation of Voice (CAPE-V) scores.¹⁷ Yiu et al. also observed that patients with PD and more severe vocal fold bowing had more difficulty with airway protection, which was predictive of swallowing safety. In their study, vocal fold atrophy was measured by using the bowing index (BI); scores indicative of severe BI (greater than 4.6) were predictive of Penetration-Aspiration Scale (PAS) scores greater than or equal to 4, with 87.0% sensitivity and 71.4% specificity.¹⁸ Although glottic insufficiency contributes to poor projection, asymmetric rigidity of intrinsic laryngeal muscles also plays a large role in PD-related voice disturbances.¹⁹ Thus, procedural treatment of glottic insufficiency and bowing alone is not sufficient to treat PD dysphonia, and it is vital that patients participate in PD-specific voice therapies to appropriately target PD-specific speech and communication deficits.

The authors do indeed acknowledge the concern for delayed diagnosis of dangerous lesions such as malignancy, worsening of pre-existing lesions such as nodules or polyps, or missing vocal fold paresis or paralysis by initiating therapy without laryngoscopic clearance. However, again, in our study, we found the incidence of pre-existing lesions in patients presenting with PD extremely low (3%). In our clinical practice, we also seek to expedite care of patients who have breathiness, roughness, or airway symptoms not characteristic of typical PD-related dysphonia. We are not suggesting that laryngoscopy be eliminated or delayed; rather, merely that therapy sessions be initiated while patients await an appropriately triaged consultation for laryngoscopy. Additionally, with similar sentiment conveyed in American Academy of Otolaryngology- Head and Neck Surgery Foundation (AAO-HNSF) dysphonia guidelines, we agree the presence of red flags or extenuating circumstances (recent surgical procedures involving the head, neck, or chest; recent endotracheal intubation; presence of concomitant neck mass; respiratory distress or stridor; history of tobacco abuse; and professional voice user) should warrant more timely referral for laryngoscopy.

The high prevalence of vocal fold atrophy and glottic insufficiency in this study cohort underscore the potential opportunity for procedural or operative interventions should voice therapy alone not prove adequate for symptomatic improvement. In the case of glottic insufficiency, reduced expiratory flow can create breathiness and poor projection, leading to fatigue and compensatory supraglottic hyperfunction. Up to nearly 90 percent of advanced PD-related dysphonia patients have persistent insufficiency often persisting despite pharmacologic treatment and PD-specific speech therapy.²⁰ Injection laryngoplasty (IL) and medialization thyroplasty (MT) are well-established for the treatment of insufficiency related to a variety of causes such as atrophy, paralysis, and scarring. They have also been used to treat PD-specific glottic insufficiency with the goal of improving speech intelligibility, as well as cough and swallow function.²⁰ A scoping review by Garabedian et al. reported a total sample of 131 patients with PD-related dysphonia and laryngoscopy-confirmed insufficiency who underwent augmentation procedures; 122 patients underwent bilateral or unilateral IL, 9 underwent MT, and 6 underwent both. Among individual studies, there were conflicting results, with one study reporting those with more severe symptoms had less vocal improvement from IL, while another study reported the same patients had the most improvement. The authors concluded that IL rather than permanent MT is generally performed, as temporary measures are preferred in patients with a progressive neurodegenerative disease.²⁰ The criteria for IL or MT in PD patients has not been well established. In the present study, 8 patients received IL only, none received MT only, and 5 received IL followed by MT. At our institution, it is our practice to trial IL on patients with PD-related dysphonia who fail to have adequate improvement with voice therapy alone. If they derive significant benefit from IL, we then offer MT. The literature on procedural intervention indications and outcomes for PD-related dysphonia is sparse, with only retrospective reviews and case series reported at present.

Although this study is a single academic institution’s experience, we believe it is generalizable to most institutions offering SPEAK OUT!^® or LSVT^®. In accordance with AAO-HNSF dysphonia clinical practice, we suspect that most SPEAK OUT!^®/LSVT^® programs require dysphonic PD patients to have laryngoscopy before voice therapy initiation. However, in the case of patients with known PD affecting speech and other motor patterns and a high probability of no other co-occurring vocal fold lesions, this has the potential to create a significant delay in voice therapy. Additionally, our study found that 31% patients who presented for laryngoscopy clearance were found to have glottic insufficiency, and 39% went on to undergo IL or MT. Although the aim of this study was not to evaluate outcomes of these procedures, we agree with previous studies that these interventions can have a positive impact on quality of life and future research is needed to ascertain which patients benefit the most, and in turn, to inform patient selection.

There are some important limitations to note. Weaknesses of our study included the retrospective nature and small cohort sizes with limited sociodemographic diversity, which might have potentially under-powered some of the statistical comparisons. Moreover, the two subgroups were imbalanced due to the relatively infrequent instances of inadvertent voice therapy initiation prior to physician clearance. While procedural outcomes were generally comparable irrespective of sub-group, small differences between the two sub-group sizes might not have been detected with the present statistical analysis. Additionally, physician or SLP time away from the clinic, or other patient or payer factors not captured in the dataset, might have influenced the timing of laryngologist or SLP encounters. Lastly, patients with a history of head and neck cancer, radiotherapy, or a Parkinson-plus syndrome were excluded from the study, and this may limit generalizability. For instance, previous work has suggested differences in phase closure, phase asymmetry, and amplitude among patients with idiopathic PD versus Parkinson-plus syndromes, which may inform decision-making regarding which interventions may be of greatest benefit.²¹ Nevertheless, existing studies on laryngologic interventions for dysphonia specifically related to PD are limited and this work advances the literature at a time when patients with PD are achieving longer life expectancies compared to years prior.²² Future efforts should involve larger cohorts, including additional analyses of patient outcomes following voice-directed interventions for PD-related dysphonia, as well as whether timing of voice therapy initiation or significant delay in care impacts therapy progression, effectiveness, and quality of life upon completion of treatment.

Conclusion

Vocal fold atrophy and glottic insufficiency are prevalent among patients with PD-related dysphonia. Requiring laryngology evaluation prior to initiation PD-directed voice therapy can delay care without affecting the procedural interventions ultimately performed. While all patients with PD-related dysphonia should undergo screening laryngoscopy during the course of voice therapy, patients may benefit from timelier care if medical clearance from laryngology is not a prerequisite to voice therapy initiation. Regardless of the timing of laryngologist consultation, select patients with PD may benefit from injection augmentation and medialization thyroplasty, but further study is needed.