Patient Satisfaction with Virtual vs In-Person Voice Therapy

Abstract

Objective:

To determine whether there is a difference in patient satisfaction between in-person and virtual voice therapy.

Methods:

Patient satisfaction answers to the National Research Corporation (NRC) Health patient survey were retrieved for two separate 11-month periods. The first was for an in-person cohort, from April 2019 to February 2020. The second was for a virtual cohort between April 2020 and February 2021. Two group t-tests or Wilcoxon rank sum tests were used to compare responses between the in-person and virtual cohorts. The effect of modality of therapy by gender, age, and race was examined by testing interactions with separate ANOVA models.

Results:

Responses were compared between 224 patient satisfaction surveys for the virtual cohort and 309 patient satisfaction surveys for the in-person cohort. Overall, responses were highly favorable in all categories. There were no differences between the in-person and virtual cohorts’ responses with respect to three main categories: likelihood of future referral of clinic or provider; communication with provider; and comprehension of the treatment plan. The interaction between modality of therapy delivery and age was significant for the question, “Did you know what to do after your visit,” with 18–44 year olds in the in-person group reporting a better understanding of the treatment plan compared to the 18–44 year olds in the virtual therapy cohort (p=0.004). There were no interactions between modality of therapy and gender, or race.

Conclusion:

Virtual delivery of voice therapy was associated with comparable visit satisfaction scores to in-person delivery, with both delivery modalities demonstrating very high satisfaction. Future studies are needed to identify which patients and conditions are most suited for virtual vs in-person delivery of speech-language pathology services in voice clinics.

Introduction

Telemedicine is the provision of healthcare services at a distance via telecommunication tools. It can increase access to health services in remote or rural locations and during emergency situations.^1,2 Following the onset of the coronavirus disease 2019 (COVID-19) pandemic, a shift to telemedicine quickly occurred to reduce in-person contact during health care visits.^3,4 Multiple reports of telemedicine implementation were published worldwide from Hong Kong⁵ to Australia⁶, France⁷, Israel⁸, Italy⁹, Norway³, and the US^10–12. Voice therapy was part of this shift as many clinics shifted speech language pathology (SLP) services from in-person to virtual delivery^10–12 in part due to rationale that tasks involved in voice therapy, especially singing or shouting, could be aerosolizing, and therefore could increase the transmission of SARS-CoV-2.¹³

Telerehabilitation services, or virtual therapy is a subgroup of telemedicine services provided by rehabilitation providers. Rehabilitation services are uniquely positioned to benefit from a virtual platform given that these services need multiple repeat visits to achieve therapy goals. Seeing the patient in the home allows the therapist to provide care within the patient’s “natural habitat,” and has led some to refer to telemedicine as the new house call.¹⁴

SLP services are a subset of rehabilitation services focusing on communication disorders. Effectiveness of virtual SLP services as compared to in-person services has been demonstrated for swallow therapy, head and neck care, and speech therapy for aphasia and Parkinson’s disease.^15–23 Effectiveness of virtual therapy for dysphonia, as reflected by auditory-perceptual, acoustic, aerodynamic, stroboscopic and patient satisfaction measures, has been reported in 4 separate publications, with cohorts ranging from 7 to 47 participants (median of 12) evaluated for various diagnoses including vocal fold nodules, vocal fold edema, unilateral vocal fold paralysis, muscle tension dysphonia and dyspnea due to vocal fold dysfunction.^22–25 Benefits of a virtual platform have been reported to include lower attrition rates²⁵ and cost-efficiency.¹⁷

Patient reported satisfaction is another important element of patient-centered care. Patient satisfaction is a measure of quality which can be tied to reimbursements from the Center for Medicare and Medicaid through results of Hospital Consumer Assessment of Healthcare Providers and Systems surveys.²⁶ Patient satisfaction with telemedicine overall has been associated with improved outcomes, preferred modality, ease of use, low cost, improved communication and decreased travel time.²⁶ Good patient satisfaction has been reported for virtual therapy for stutter²⁷, autism spectrum disorders²⁸, speech services in rural settings²⁹ and swallowing and communication services for the head and neck patient population.¹⁷ Since the onset of the COVID-19 pandemic, multiple reports of telemedicine implementation have been published worldwide. Renewed efforts to evaluate patient satisfaction reported comparable results in some studies ^7,8,30 while others showed a preference for in-person over virtual delivery of health care services.³¹

There is little evidence specifically about patient satisfaction with virtual delivery of voice therapy. Voice therapy services have unique features compared to other forms of SLP therapies. A stable connection with good audio and video feed is mandatory for an effective therapy session due to the need for excellent audio fidelity to perceive subtle sound differences and to follow specific orofacial instructions. Furthermore, a virtual platform does not allow certain therapeutic modalities requiring contact with the patient, such as circumlaryngeal massage, to be performed. Patient satisfaction with voice therapy was examined by Mashima et al., in a small 2003 study which included 47 patients who were randomly assigned to an in-person or a virtual platform approach. Patient satisfaction ratings on the process and outcome of voice therapy were rated on a 5-point scale.²⁵ The overall mean ratings indicated a generally positive response to services provided and outcome of therapy, with no difference in ratings between the in-person cohort vs the virtual cohort. Of note, the virtual delivery was performed within the clinic, with the patient being on site but in a separate room from the therapist. The authors noted that this was a proof-of-concept study in a highly controlled environment rather than a real life comparison of virtual vs in-person applicability. Subsequently, in a 2015 study of 10 women undergoing voice therapy for vocal fold nodules, positive patient experience was reported with virtual voice therapy, but there was no comparison group of patients receiving in-person therapy.²³ Instead, patients completed a 16-item questionnaire before and after therapy.²³ The questionnaire pre-therapy highlighted the hesitancy of proceeding with teletherapy. However, after treatment this theme significantly improved, and ultimately participant preference for virtual therapy over in person therapy increased.

In addition to the focus on delivery modality of therapy, other factors may also influence patient satisfaction with virtual rehabilitation services, like gender³⁰ and age^32,33. Tenforde et al., found that women rated their level of overall satisfaction significantly higher than male participants in a study of rehabilitation services of which 30% were SLP services.³⁰A systematic review evaluating barriers to adoption of telemedicine reported age as a barrier given the gap in technology acceptance in older patients.³² In contrast, a study focused on US Veterans (mean age=72) satisfaction with home virtual programs for medical management showed high patient satisfaction.³³

The necessity for a virtual platform was understood at the start of the COVID-19 pandemic given limitations in personal protective equipment availability and profound uncertainty about the course of this new disease. It is important to carefully consider under what conditions virtual therapy delivery should be continued in the future. As healthcare providers strive to deliver patient-centered care, at least two key questions arise: (1) is patient satisfaction different for voice therapy delivered virtually vs. in person, and (2) is patient satisfaction different across certain sociodemographic characteristics?

The primary aim of the present study is to compare patient satisfaction scores from virtual SLP therapy visits over an 11-month period during the COVID-19 pandemic, with scores from traditional in-person therapy visits immediately prior to the COVID-19 pandemic, at a voice clinic in an academic medical center. We hypothesized that patient satisfaction would be equivalent between the two cohorts. Secondarily, we aimed to determine whether patient sociodemographic factors influenced patient satisfaction.

Methods

Since January 2019, the Lions Voice Clinic at the University of Minnesota in Minneapolis, Minnesota has invited all new patients to complete patient satisfaction questionnaires. The National Research Corporation (NRC) Health Patient Survey is a Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) approved survey. It has been previously used to assess patient satisfaction in other studies.^31,34 Patients received the survey through email, phone or text. Patient satisfaction is assessed in the outpatient otolaryngology clinic by NRC Health with a 14-item questionnaire. The answers were then de-identified and aggregated through the NRC to preserve anonymity. Starting in March of 2020 all voice therapy at the Lions Voice Clinic was converted to virtual delivery. Appointment duration was unchanged.

Institutional Review Board (IRB) exemption was granted for this study since no patient records or personal health information was accessed. NRC patient satisfaction records were retrieved between April 2019 to February 2020 (for in-person cohort); and between April 2020 and February 2021 (for virtual cohort) for all voice therapy providers at the clinic. March of 2020 was not retrieved because this was a transition period for the clinic due to the onset of the COVID-19 pandemic.

Six questions of the survey were identified as applicable to patient satisfaction with voice therapy. These were divided into three themes: (1) future referral, (2) communication, and (3) comprehension of the treatment plan (Table 1). Other questions focused on the patient’s interaction and communication with the nurses and were not included as they were not directly related to a typical voice therapy experience. Subgroup analyses were performed using the following available categories: gender (male, female); age (18–44, 45–65, 65+) and race (white, Asian, Black, Hispanic, Native American). Given the small sample sizes in the race category, Asian, Black, Hispanic, Native American were combined in one subgroup.

Table 1 -.

Patient questionnaire

Theme	Patient Question	Abbreviation for Analysis	Patient Answer Options
Future referral	How likely would you be to recommend this provider to your family and friends?	Would recommend provider	0 to 10 0-least likely 10- most likely
	How likely would you be to recommend this clinic to your family and friends?	Would recommend clinic
Communication	Did you have enough input or say in your care?	Enough input/say in care	1-no 2- yes somewhat 3-yes, mostly 4-yes, definitely
	Did the care providers listen carefully to you?	Care providers listened
Comprehension of treatment plan	Did the care providers give you enough information about your health and treatment?	Enough info about treatment
	Did you know what to do after your visit?	Know what to do

Descriptive statistics were used to summarize the data collected. Demographics were compared between cohorts (t-test for age and Fisher’s exact tests for categorical variables).Two group t-tests or Wilcoxon rank sum tests were used to compare question responses between the in-person and virtual cohorts. The effect of gender, age, and race on in-person and virtual cohort differences was examined by testing interactions with separate ANOVA models (or ANOVA on ranks). If an interaction was significant, post-hoc analyses examined cohort differences in subgroups. P-values less than 0.05 were considered statistically significant. SAS V9.4 (SAS Institute Inc., Cary, NC) was used for the analysis.

Results

For the in-person cohort, 1083 patients were contacted and 354 patients responded (32.7% response rate). Of these, 309 patients had patient satisfaction survey responses for the questions of interest and were included in our analysis. For the virtual cohort, 927 patients were contacted and 284 patients responded (30.6% response rate). Of these, 224 patients were included in our analysis. Demographics are listed in Table 2.

Table 2 -.

Patient Demographics

	In-person n=309	Virtual n=224	P-value†
Gender, n (%)			0.7153
Female	199 (64)	140 (63)
Male	110 (36)	84 (38)
Race, n (%)			0.0605
Asian	11 (4)	8 (4)
Black	4 (1)	2 (1)
Hispanic	0	1 (0.4)
Native American	2 (0.6)	0
White	246 (80)	193 (86)
Other	12 (4)	1 (0.4)
Declined	34 (11)	19 (8)
Age, mean (SD)	53.9 (20.9)	53.6 (22.0)	0.8550
0–17	36 (12)	29 (13)
18–44	52 (17)	35 (16)
45–64	91 (29)	67 (30)
65+	130 (42)	93 (42)

^†P-values are from a t-test for age (continuous) and Fisher's exact tests for categorical demographics. No statistically significant difference in demographics between the in-person and virtual cohorts.

No significant differences were detected between the virtual cohort vs the in-person cohort on any of the six questions and the three themes evaluated. (Table 3)

Table 3 -.

Patient Satisfaction Scores

Theme	Patient Question (score range)	In-person n=309	Virtual n=224	P-value†
Future referral	Would recommend provider (0–10)	9.3 (1.4) [10] n=287	9.3 (1.5) [10] n=205	0.9313
	Would recommend clinic (0–10)	9.1 (1.5) [10] n=290	9.2 (1.5) [10] n=205	0.6741
Communication	Enough input/say in care (1–4)	3.6 (0.7) [4] n=118	3.6 (0.7) [4] n=59	0.6893
	Care providers listened (1–4)	3.7 (0.6) [4] n=304	3.8 (0.6) [4] n=221	0.4578
Comprehension of treatment plan	Enough info about treatment (1–4)	3.6 (0.7) [4] n=305	3.5 (0.8) [4] n=221	0.4192
	Know what to do (1–4)	3.6 (0.7) [4] n=303	3.6 (0.7) [4] n=218	0.2600

Mean (SD) [Median] are presented in the table.

^†P-values are from t-tests (for 0–10 scale questions) or Wilcoxon rank sum tests (ordinal scale questions)

The first theme reviewed was the likelihood of referring the provider or the clinic to family or friends. Both questions resulted in a very high likelihood of recommendation (>9 out of 10) for all groups, without a significant difference between the in-person and virtual cohorts. Regarding the likelihood of referring the provider, for the in-person cohort, 287 patients (93%) answered this question, with a mean score of 9.3 (SD 1.4), while for the virtual cohort, 205 (92%) patients answered this question, with a mean score of 9.3 (SD 1.5) (p-value 0.9313). Regarding the likelihood of referring the clinic, for the in-person cohort, 290 patients (94%) answered this question, with a mean score of 9.1 (SD 1.5), while for the virtual cohort, 205 (92%) patients answered this question, with a mean score of 9.2 (SD 1.5) (p-value 0.6741).

The second theme evaluated was communication between the provider and the patient, with resulting responses between “3- yes, mostly” and “4-yes, definitely” for both groups. Regarding the question “Did you (patient) have enough input or say in your care?”, for the in-person cohort, 118 patients (38%) answered this question, with a mean score of 3.6 (SD 0.7), while for the virtual cohort, 59 (26%) patients answered this question, with a mean score of 3.6 (SD 0.7) (p-value 0.6893). Regarding the question “Did the care providers listen carefully to you?” for the in-person cohort, 304 patients (98%) answered this question, with a mean score of 3.7 (SD 0.6), while for the virtual cohort, 221 (99%) patients answered this question, with a mean score of 3.8 (SD 0.6) (p-value 0.4578).

The last theme studied was the patient’s comprehension of the treatment plan, also with resulting responses between “3- yes, mostly” and “4-yes, definitely” for both groups. Regarding the questions “Did the care provider give you (the patient) enough information about your health and treatment?”, for the in-person cohort, 305 patients (99%) answered this question, with a mean score of 3.6 (SD 0.7), while for the virtual cohort, 221 (99%) patients answered this question, with a mean score of 3.5 (SD 0.8) (p-value 0.4192). Regarding the question, “Did you (the patient) know what to do after your visit”, for the in-person cohort, 303 patients (98%) answered this question, with a mean score of 3.6 (SD 0.7), while for the virtual cohort, 218 (97%) patients answered this question, with a mean score of 3.6 (SD 0.7) (p-value 0.2600).

The interaction of modality of therapy and gender, modality of therapy and age, and modality of therapy and race was examined and found to be mainly not significant. Age did appear to make a difference in one instance. For the question “Did you know what to do after your visit,” a significant group by age interaction was identified(p=0.0297). In those aged 18–44 years old, higher mean (SD) understanding was reported in the in-person group (3.7 (0.6)) vs the virtual group (3.3 (0.8)), p=0.004. This was significant at a Bonferroni adjusted level of p<0.0167. Cohort differences were not found in the other age categories, ages 45–64 (p=0.6907) and 65+ (p=0.8168) (Figure 1).

Figure 1. Patient question: “Did you know what to do after your visit”.

There was a significant interaction between modality of therapy and age (p=0.0297). 18–44 year olds in the virtual group reported a lower understanding of the treatment plan compared to those in the in-person cohort (p=0.004). Color indicates modality of therapy. Error bars indicate +/− SE. **=p<0.01.

There were no significant interactions between modality of therapy and gender, or modality of therapy and race. For the question, “How likely would you be to recommend this clinic to your family and friends?”, there was a significant main effect for age regardless of modality of therapy (p=0.0074). Those 65 and above (n=223, 42%) reported higher mean likelihood (SD) of recommending the clinic when compared to those 18–44 (n=87, 16%) [9.3 (1.3) vs 8.8 (1.9), p=0.004]. This was significant at a Bonferroni adjusted level of (p<0.0167).

Discussion

Expansion of telemedicine has been reported before the COVID-19 pandemic¹ and has greatly increased since the pandemic, given the implementation of emergency protocols which required social distancing and personal protective equipment use. Voice therapy can particularly benefit from virtual delivery since therapy visits are frequent and may be aerosolizing. Therefore, since the COVID-19 pandemic, the adoption of virtual voice therapy has greatly increased¹¹ and in the first several months of the pandemic was the sole platform for voice therapy services in our voice clinic. In this study, a large cohort of patients undergoing virtual voice therapy responded to patient satisfaction questions in a similar manner to a cohort of patients seen in-person during an immediately pre-pandemic period of 11 months. The patients’ likelihood of recommending the clinic or the SLP provider remained high and comparable to the prior in-person time period. Similarly, questions regarding communication, and comprehension of the treatment plan also had favorable and similar results. Patient satisfaction did not differ by gender and race. It appeared that age did play a factor, though this finding is of unknown clinical significance.

Multiple reports in the literature are concordant with our finding of comparable patient satisfaction with virtual vs in-person care delivery, including telemedicine overall^8,26 and rehab services more specifically.³⁰ A study in Israel (n= 540) reported similar very high satisfaction levels for the same themes as in this study (future referral, communication and treatment plan comprehension) from both a 2019 in-person cohort and a 2020 virtual cohort of patients receiving a range of services by physicians, psychologists, nurses, social workers, dietitians, speech therapists, genetic consultants and others.⁸ Since the needs of telerehabilitation may be different compared to other medical services, Tenforde et al., examined patient satisfaction in 205 participants receiving rehabilitation services, of which 30% were SLP services.³⁰ Our findings were similar to the excellent and very good responses reported by at least 87% of respondents with respect to addressing concern and questions during the virtual visit, communication with the therapist, developing a treatment plan, convenience, overall visit satisfaction and value in having a future telehealth visit. Echoing what was seen in smaller studies by Fu et al., and Mashima et al.,^23,25 patients were satisfied with the three themes evaluated in this study: future referral, communication and comprehension of treatment plan. Other interesting themes mentioned in the prior studies such as level of comfort, accessibility and preference for the virtual platform were not assessed in the current study as they were not part of the NRC survey.

In contrast to our observation of equivalent patient satisfaction between virtual and traditional voice therapy, Itamura et al., reported more favorable responses on NRC scores from patients seen in-person than those seen virtually in an otolaryngology clinic,³¹ while Burns et al., found a higher preference for virtual therapy over in-person therapy in a randomized controlled study in a cohort of head and neck patients.¹⁷ There are multiple possible explanations for these differences in findings compared to our study. In the Itamura et al., study, patients were undergoing otolaryngology evaluation for a wide variety of concerns and not virtual voice therapy alone, which likely necessitated a subsequent in-person physical examination as well as endoscopy and/or microscopy as appropriate.³⁵ Secondly, the time period examined was over the 2 months at the start of the pandemic, and therefore included the transition phase of virtual platform implementation for both providers and patients. Finally, in the Burns et al. study, the patient cohort was limited to only head and neck patients in Australia who had to travel a long way for in-person therapy appointments which could explain their preference for virtual services.¹⁷ Complementing these studies, our study adds the perspective of patients receiving a specific type of care (voice therapy) over a longer time frame pre- and post- virtual platform implementation.

The variable impact of various sociodemographic factors in our study is in accordance with prior mixed observations in the literature. Tenforde et al.,³⁰reported increased patient satisfaction for women with rehab services, while Fieux et al.,⁷ did not identify gender as a predictor of patient satisfaction with otolaryngology services. Although in our study, gender did not play a role, most patients were women which is typical of studies in voice. With respect to age, Crotty et al.,³⁶ and Young et al.,³³ showed favorable patient satisfaction with virtual rehab services after hospital discharge, while Fieux et al.,⁷and Tenforde et al.,³⁰ did not identify age as a predictor of patient satisfaction with otolaryngology and telerehabilitation services, respectively. We did observe a difference in patient satisfaction based on age in our study for the themes of future referral and cmprehension the treatment plan. However, other questions did not have an apparent age effect, and overall, regardless of therapy delivery modality, the responses were highly favorable in all categories. Thus the clinical significance of a potential effect of age finding is not clear. With respect to race, our study, similar to existing literature,³⁷ included limited racial diversity, so the role of race as a factor in patient satisfaction could not be evaluated.

Strengths of this study include its large sample size, broad timeline, and the inclusion of a comparison group. Our location, in the Midwest region of the US, had a more gradual onset of COVID than other regions of the country and other countries around the world, adding a complementary perspective to studies from other geographic locations. The population that we serve also includes a mixture of urban and rural respondents, reinforcing the potential value of offering virtual services to patients outside of urban areas. In addition, data collection is done by a third party, the NRC, reducing potential bias.

Limitations

Limitations of the study include its observational nature as well as a limited response rate which is typical for this type of methodology.³¹ Both participation in virtual therapy and ability to respond to the NRC survey imply access to internet services, which are not universally available. In addition, the anonymity of the survey, which facilitates honest responses, also limits our ability to examine other patient or visit characteristics that may affect responses, such as socioeconomic status, specific diagnosis (e.g., dysphonia, chronic cough, or paradoxical vocal fold motion) or subtype of virtual therapy (e.g., phone vs video platform). Finally, the limited racial diversity of our sample, which reflects the state in which the institution is based, as well as the specialized nature of this clinic, may limit generalizability to other populations and settings.

Future directions

Despite these limitations, our findings indicate that patient satisfaction with virtual visits is comparable to in-person visits for SLP therapy appointments in a voice clinic. Because satisfaction with the visit cannot be assumed to be equivalent to treatment outcomes or overall treatment satisfaction, future studies will need to investigate these topics. In addition, because responses from patients receiving virtual care in the first period of the pandemic may have been influenced by gratitude to receive any care at all in that era, it will be important to determine whether these findings are replicated in patients who could receive either in-person or virtual therapy. Future studies are needed to further characterize the impact of care modality on patient satisfaction and outcomes, as well as, the influence of patient and clinical factors.

Conclusions

Virtual delivery of voice therapy was associated with comparable visit satisfaction scores to in-person delivery, with both delivery modalities demonstrating very high satisfaction. These findings support continuing to offer virtual delivery of voice therapy beyond the immediate needs of the COVID-19 pandemic. Future studies are needed to identify which patients and conditions are particularly suited for virtual and in-person delivery of SLP services in voice clinics.