- Email: [email protected]
Hispanic Ethnicity as a Predictor of Voice Therapy Adherence
ARTICLE IN PRESS
Hispanic Ethnicity as a Predictor of Voice Therapy Adherence David E. Rosow, Jennylee Diaz, Debbie R. Pan, and Adam T. Lloyd, Miami, Florida Summary: Background. Voice therapy is a well-studied, evidence-based treatment in the management of voice disorders, yet it is known that adherence rates are generally decreased due to a variety of identified factors. In light of this fact, a high rate of nonadherence to voice therapy has been anecdotally observed in the Hispanic community comprising a sizable portion of the patient population in South Florida. Objective. We sought to analyze the rates of voice therapy attendance for patients who underwent treatment for benign vocal fold nodules at a single tertiary-care academic medical center. Based on our anecdotal observations, we hypothesized that Hispanic patients would have a significantly lower rate of voice therapy attendance compared to non-Hispanic patients. Study Design. Retrospective cohort study. Methods. A retrospective chart review was performed for Hispanic and non-Hispanic patients aged 18 years and older who were diagnosed in a single hospital-based otolaryngology department with benign vocal fold nodules between 2013 and 2018. Patients with other glottic pathology or those who were not recommended voice therapy as initial treatment were excluded. Demographic data, including ethnicity, home address, and preferred language by self-report (English vs. Spanish), were obtained and analyzed. Median income levels for patients were determined by postal codes. “Adherent” status was given to patients who attended at least one voice therapy session. Statistical comparisons of continuous quantitative variables were made using Student’s t test, ordinal quantitative variables using Mann-Whitney U test, and categorical variables using Fischer’s exact test. Statistical significance was determined as P < 0.05. Results. One hundred eleven patients met inclusion criteria. The population was 85% female, with an average age of 41 years. Overall voice therapy adherence rate was 68%. Forty-eight percent of patients self-identified as Hispanic, and of this cohort, 42% spoke Spanish as a preferred language. Differences in annual income levels were noted between non-Hispanic and Hispanic patients ($61,799 vs. $51,697, P = 0.017), as well between English-preferring and Spanishpreferring patients ($60,276 vs. $43,504, P = 0.0014). Thirty of 53 (57%) of Hispanic patients were adherent to voice therapy, compared to 45 of 58 (78%) non-Hispanic patients (P = 0.025). No significant differences were found in age, Voice Handicap Index-10 score, or number of sessions attended between the therapy-adherent patients in the Hispanic and non-Hispanic groups. Further differences in adherence rates were noted when the Hispanic group was subclassified into English and Spanish language preferences. Fifteen of 31 (48%) English-preferring Hispanic patients attended voice therapy compared to 45 of 58 (78%) non-Hispanic patients (P = 0.0085), while Spanish-preferring Hispanic patients had a 68% therapy adherence rate (15 of 22, P = 0.4). English-preferring Hispanic patients had higher average Voice Handicap Index-10 (22.0 vs. 14.9, P = 0.018) and lower total attended sessions (2 vs. 3.6, P = 0.024) than their non-Hispanic counterparts. Conclusion. We believe this is the first study demonstrating a significantly lower rate of voice therapy adherence in Hispanic versus non-Hispanic patients. Decreased utilization of a proven treatment strategy for vocal fold nodules puts these patients at increased risk of treatment failure and decreased voice-related quality of life. Clinicians must be aware of ethnicity-based healthcare disparities and encourage proven treatment adherence to ensure highest quality of life. Key Words: Voice−Voice disorders−Hispanic population−Voice therapy adherence−Voice therapy attendance.
INTRODUCTION Voice therapy is well understood to be a standard, evidencebased treatment option in the management of dysphonia due to a variety of voice-related diagnoses. Therapy involves direct and indirect rehabilitative treatment approaches focused on Accepted for publication September 20, 2019. The authors have no funding, financial relationships, or conflicts of interest to disclose. Presented at the 2018 Fall Voice Conference, Seattle, WA, USA, October 25-27, 2018. From the University of Miami Miller School of Medicine, Department of Otolaryngology, Miami, Florida. Address correspondence and reprint requests to David E. Rosow, University of Miami Miller School of Medicine, 1120 NW 14th Street, 5th Floor, Miami, FL 33136-1015. E-mail: [email protected] Journal of Voice, Vol. &&, No. &&, pp. &&−&& 0892-1997 © 2019 The Voice Foundation. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jvoice.2019.09.011
optimization of vocal function and it has been shown to be effective in reducing dysphonia regardless of the therapy approach used.1−3 Despite the known clinical effectiveness of voice therapy, decreased patient adherence is an ongoing problem.4−7 Failure of voice therapy initiation following clinical recommendation from a laryngologist and speech-language pathologist has been reported to be between 38% and 44%.5−8 Similarly, nonattendance to a voice evaluation with a speechlanguage pathologist following recommendation from a laryngologist has been reported to be 47% and voice therapy dropout rates have been found to be as high as 68%.6 The reasons underlying decreased therapy adherence are still being determined. Demographic factors such as age, gender, and race/ethnicity have not previously been associated with a higher rate of nonadherence. However, individuals with complex laryngeal diagnoses, more severe vocal
ARTICLE IN PRESS 2 impairment, medical comorbidities, occupational issues, and higher Voice Handicap Index (VHI) scores at initial presentation are less likely to attend voice therapy.8 Other factors that affect therapy adherence include self-efficacy and motivation for behavior change, coverage of services by insurance companies, distance from the clinic, the clinician−patient relationship (also termed therapeutic alliance), the level of understanding of the purpose of therapy, and a perceived concern that therapy is too difficult or “silly.”6−8,26 In accordance with suboptimal adherence rates to voice therapy, we have anecdotally observed poor attendance to voice therapy among the English- and Spanish-preferring Hispanic population that comprises a significant portion of patients seen at our facility. Healthcare disparities among minority populations exist and have been well-documented in a wide range of fields. Overall, Hispanics are less likely to seek care when necessary compared to non-Hispanics, and this is influenced by various socioeconomic cultural, linguistic factors that are discussed later. The topic of healthcare disparities among Hispanics is a relevant and significant area of study given the projected rise of this population over the next few decades. The term “Hispanic” refers people “of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin regardless of race.”9 As of July 2017, the Hispanic population of the United States was 58.9 million or 18.1% of the nation’s total population, making people of Hispanic origin the nation’s largest ethnic or racial minority.10 Also, it is projected that the Hispanic population will increase to 119 million by 2060.11 As such, it is becoming increasingly important to understand the medical needs of this population, but no known studies currently exist on adherence to voice therapy among Hispanics. This study seeks to fill this knowledge gap by comparing the rates of adherence to voice therapy between Hispanic and non-Hispanic groups. METHODS Study design and participants A retrospective medical record review was conducted for all patients seen for voice care services at a tertiary academic medical center. Institutional Review Board approval was obtained for this study. Inclusion criteria included patients aged 18 years and older who were diagnosed with benign vocal fold nodules between 2013 and 2018. Patients with other glottic pathology or those for whom voice therapy was not recommended as initial treatment were excluded. Preferred language (English or Spanish) and ethnicity (nonHispanic or Hispanic) were based on self-identified information as reported in the medical records. In this manuscript, the labels of either English- or Spanish-speaking are intended to represent the patient’s language of preference and are not meant to convey language exclusivity. All participants were seen by a laryngologist and one of five speech-language pathologists who are specialized in voice disorders, two of whom are Hispanic and fluent bilingual Spanish speakers. With rare exceptions, all Spanish-
Journal of Voice, Vol. &&, No. &&, 2019
preferring patients are seen for initial evaluation and subsequent therapy by one of the Spanish-fluent speech-language pathologists. Demographic and clinical data, including age, home postal code, gender, ethnicity, and Voice Handicap Index10 scores, were collected and analyzed. Voice therapy adherence was defined as attendance to at least one therapy session, and patients who did not attend at least one therapy session were deemed nonadherent. While annual income for each individual patient was not available, US Census data were used to determine median income level for each patient’s ZIP code. These figures were used as projected indicators of socioeconomic status and compared between the different groups. Statistical analyses Statistical comparisons of continuous quantitative variables were made using Student’s t test, ordinal quantitative variables were compared with the Mann-Whitney U test, and categorical variables were compared with Fischer’s exact test. Statistical significance was defined as P < 0.05. RESULTS Demographic data A total of 111 patients met inclusion criteria. Based on selfreporting, 53 (48%) patients were Hispanic and 58 (52%) were non-Hispanic. The average age of the study population was 41 years, and the majority were female (n = 94; 85%). Specifically, the average age of Hispanic patients with a preferred language of Spanish was 44.5 years, and for those Hispanics with a preferred language of English, it was 38.3 years. This difference was not significant (P = 0.12). The average age of the non-Hispanic group was 41 years. Among the Hispanic group, 22 (42%) of 53 patients self-reported Spanish as their preferred language, whereas the remaining 31 (58%) preferred English. All patients in the non-Hispanic group reported English as their preferred language and no other languages of preference were self-reported within our study population. Voice therapy adherence Overall adherence to voice therapy among all patients (Hispanic and non-Hispanic) was 68%. Non-Hispanic patients had significantly higher adherence to voice therapy compared to all Hispanic patients. Forty-five of 58 (78%) nonHispanic patients were adherent to voice therapy, compared to 30 of 53 (57%) non-Hispanic patients (P = 0.025). No significant differences were seen in age or VHI scores when comparing the therapy-adherent patients in either the Hispanic or non-Hispanic groups. There was also no significant difference in the number of therapy sessions attended between these two groups. When the Hispanic group was further classified into English- and Spanish-speaking, the differences became more pronounced: only 15 of 31 (48%) of English-speaking Hispanic patients attended voice therapy compared to 45 of
ARTICLE IN PRESS David E. Rosow, et al
Hispanic Voice Therapy Adherence
TABLE 1. Number of Patients Stratified by Ethnicity, Language Preference, and Voice Therapy Adherence Status (n = 111) Ethnicity Hispanic (48%)
Language preference Spanish (42%) English (58%)
Non-Hispanic (52%)
Spanish (0%)
English (100%)
Adherence status
Number
Adherent
15
Nonadherent Adherent Nonadherent Adherent
7 15 16 0
Nonadherent Adherent
0 45
Nonadherent
13
58 78% non-Hispanic patients (P = 0.0085), while Spanishspeaking Hispanic patients had a 68% therapy adherence rate (15 of 22, P = 0.4). English-speaking Hispanic patients also had significantly higher average VHI among therapyadherent patients compared to nonadherent patients (22.0 vs. 14.9, P = 0.018), as well as a significantly lower number of therapy sessions attended than their non-Hispanic counterparts (2 vs. 3.6, P = 0.024). Data stratification of number of patients by ethnicity, preferred language, and voice therapy adherence status is summarized in Table 1. Projected annual income The average projected annual income (PAI) for all patients was calculated as $56,984. Significant differences were seen in income level when patients were stratified according to ethnicity, as well as preferred language (Table 2). Hispanic patients had an average PAI of $51,697, while this figure was $61,799 for non-Hispanic patients (P = 0.017). Average PAI for Spanish-preferring patients was $43,505 vs. $60,276 for all English-preferring patients (P = 0.0014). This difference between English- and Spanish-preferring patients was conserved within the Hispanic cohort as well ($57,432 vs. $43,505, P = 0.0021). No difference was seen between Hispanic and non-Hispanic patients who preferred English ($57,432 vs. $61,799, P = 0.40). TABLE 2. Breakdown of Projected Annual Income (PAI) by Ethnicity and Language Preference Ethnicity
Language preference
Hispanic (48%)
Spanish (42%) English (58%) Spanish (0%) English (100%)
Non-Hispanic (52%)
PAI $43,505 $57,432 $61,799
3 DISCUSSION To our knowledge, there are no known studies that examine adherence to voice therapy in the US Hispanic population. Based on evidence from other fields, there may be multiple reasons for decreased adherence to voice therapy among this population. Overall, it is understood that racial and ethnic healthcare disparities negatively impact the Hispanic population in the United States, even when income, age, insurance status, and condition severity are similar to other groups.12 These disparities have been examined in many areas of healthcare, including but not limited to general mental health,13 bipolar disorder,14 depression traumatic brain injury,15 type II diabetes,16 stroke,17 and inpatient rehabilitative services including speech-language pathology.18 Common healthcare barriers among Hispanics include lack of insurance or reduced access to healthcare due to lower rates of full-time employment compared to other groups, fear of seeking services due to immigration status, lack of cultural competence by healthcare providers, and language barriers (limited English proficiency).19 Hispanic patients are more likely to refuse treatment compared to other groups, and they are less likely to receive treatment for common medical conditions compared to non-Hispanics.12 Overall, they are less likely than other groups to have health insurance.13 Additionally, the role of fatalismo, a cultural construct that purports that destiny or a higher power have more control over one’s life than oneself, may also play a role in decreased healthcare service utilization among some Hispanics.13 Anecdotally, a number of Hispanic patients who speak limited English have informally discussed with us that their understanding of their voice diagnoses and the rationale for voice therapy were made easier with a clinician that spoke Spanish. This brings into question whether adherence is affected by the clinician’s ability to communicate in the patient’s preferred language. Inadequate interpretation has been shown to impact quality of care among patients who are limited in English proficiency (LEP), and this may also affect attendance to appointments.20 Patients who require interpretation services but do not receive it are less likely to understand their diagnoses and have reduced satisfaction with their care.21 The use of ad hoc or untrained interpreters (including family members, caregivers, medical and nonmedical staff, and strangers) has also been shown to result in inferior quality of care. However, the use of trained medical interpreters or bilingual providers is associated with increased patient satisfaction and overall quality of care.22 There is limited evidence on attendance to healthcare appointments based on the primary language of communication between LEP patients, healthcare providers, and interpreters. One study conducted with LEP asthmatic patients found that they were more likely to miss appointments with monolingual English-speaking physicians than those with bilingual providers.23 A prospective study conducted with 714 patients in an emergency department found that LEP patients who required but did not receive
ARTICLE IN PRESS 4 interpretation services were twice as likely to be discharged from the ED without a follow-up appointment compared to patients who received interpretation services.24 Further studies are necessary to examine the role that language plays (either with bilingual providers or with interpreters) in attendance to healthcare appointments among LEP patients. For our purposes, this should also be studied in relation to voice therapy adherence. The aim of this study was to assess the rates of adherence to recommended voice therapy between Hispanics and nonHispanics. The results indicate that Hispanics (both English and Spanish speakers) are significantly less likely to attend voice therapy compared to the non-Hispanic population. There was no difference in adherence based on demographic information such as age and gender, though projected annual income differed between Hispanic and non-Hispanic patients. When this difference was broken down further into language preference, the Spanish-preferring group was noted to have a significantly lower income than English-speaking patients, while English-preferring Hispanic patients had no difference in income level. Interestingly, these English-preferring Hispanic patients were even less likely to attend therapy compared to both Spanish-preferring Hispanic patients and non-Hispanics, and this English-preferring group also presented with a significantly higher VHI score and fewer number of therapy sessions attended compared to non-Hispanics. These findings suggest that the nature of the observed disparity in therapy adherence may be rooted more in differences of culture, rather than language. Based on our findings, these differences also do not appear to be rooted in socioeconomic causes, as the lower-income group of Hispanic patients had better therapy attendance than the higher-income group. Further study is warranted to examine whether there are any differences in the healthcare needs of this subpopulation. There are limitations to this study as well as other areas to consider for further investigation. We did not obtain exact income data for each patient but relied instead on averaged US Census data for patients’ ZIP codes. While this method may provide a general assessment of socioeconomic status, it lacks specificity and obscures variations in income within each ZIP code.27 Ideally, individual income levels would be prospectively collected to better understand the impact of socioeconomic status on therapy adherence. Additionally, the Hispanic population is not monolithic, but rather composed of multiple races and many different cultural groups, each with their own challenges to healthcare access.25 Future studies should therefore consider investigating the racial make-up of the Hispanic population in relation to therapy adherence. Certain Hispanic groups, particularly Mexicans and Mexican-Americans, do not make up a large percentage of the South Florida community and are therefore poorly represented in this study. A small but nontrivial portion of the Hispanic patients in this study traveled from Central America, South America, and the Caribbean to receive voice therapy, especially in cases where they previously did not receive sufficient benefit from therapy services in their home countries. Therefore, challenges arising from
Journal of Voice, Vol. &&, No. &&, 2019
travel and finances may have had further implications on Hispanic patients’ adherence to therapy. Additionally, due to the retrospective design of this study, preferred languages may not be accurately reflected per the electronic medical record. It is also important to stress that while our study showed an association between Hispanic ethnicity and difference in therapy adherence, it did not show causation. Therefore, we cannot necessarily attribute the observed difference to the overall underutilization of health care resources by Hispanic patients, as discussed above. Future prospective studies with regression analysis would be better able to answer the question of why we noted this cultural difference. The retrospective design of this study carries inherent selection bias and limits our ability to communicate with patients regarding specific barriers or reasons for nonadherence to therapy. Prospective investigation would reduce bias and allow for further identification of both internal and external challenges that arise in this specific population. The size of the Spanish-preferring population was relatively small compared to the English-speaking groups, and future studies will need to study populations of more similar size. We also chose to examine therapy adherence in patients with benign vocal fold nodules because at our institution, they are consistently prescribed voice therapy as a first treatment prior to any other medical intervention. Thus, this study analyzes adherence to therapy in a specific group that may be considered relatively less severe than other laryngeal diagnoses, such as unilateral vocal fold paralysis or glottic cancer. It is conceivable that in diagnoses considered more “serious,” this disparity in therapy adherence might not be present. CONCLUSION This is the first known study that analyzes adherence rates to voice therapy in the Hispanic population. Hispanics as a group are overall less likely to begin voice therapy compared with non-Hispanics, with nonadherence being more pronounced in among English-speaking Hispanics. Multiple healthcare barriers may influence these poorer adherence rates, and further prospective study is warranted, in light of the increasing Hispanic population in the United States, to examine specific challenges or barriers to voice therapy attendance. REFERENCES 1. Roy N, Merrill RM, Gray SD, et al. Voice disorders in the general population: Prevalence, risk factors, and occupational impact. Laryngoscope. 2005;115:1988–1995. https://doi.org/10.1097/01.mlg.0000179174.32345.41. 2. Speyer R. Effects of voice therapy: a systematic review. J Voice. 2008;22:565–580. https://doi.org/10.1016/j.jvoice.2006.10.005. 3. Yiu EM-L, Lo MCM, Barrett EA. A systematic review of resonant voice therapy. Int J Speech Lang Pathol. 2017;19:17–29. https://doi. org/10.1080/17549507.2016.1226953. 4. Behrman A. Facilitating behavioral change in voice therapy: the relevance of motivational interviewing. Am J Speech Lang Pathol. 2006;15:215–225. https://doi.org/10.1044/1058-0360(2006/020). 5. Portone C, Johns MM, Hapner ER. A review of patient adherence to the recommendation for voice therapy. J Voice. 2008;22:192–196. https://doi.org/10.1016/j.jvoice.2006.09.009.
ARTICLE IN PRESS David E. Rosow, et al
Hispanic Voice Therapy Adherence
6. Hapner E, Portone-Maira C, Johns MM. A study of voice therapy dropout. J Voice. 2009;23:337–340. https://doi.org/10.1016/j.jvoice.2007.10.009. 7. van Leer E, Connor NP. Predicting and influencing voice therapy adherence using social−cognitive factors and mobile video. Am J Speech Lang Pathol. 2015;24:164–176. https://doi.org/10.1044/2015_AJSLP-12-0123. 8. Smith BE, Kempster GB, Sims HS. Patient factors related to voice therapy attendance and outcomes. J Voice. 2010;24:694–701. https:// doi.org/10.1016/j.jvoice.2009.03.004. 9. US Census Bureau. Hispanic Origin. 2018. Available at; https://www. census.gov/topics/population/hispanic-origin/about.html. 10. US Census Bureau. Hispanic Heritage Month. 2018. Available at; https://www.census.gov/newsroom/facts-for-features/2018/hispanicheritage-month.html. 11. US Census Bureau. Projections of the Size and Composition of the U.S. Population: 2014 to 2060. 2015. Available at; https://census.gov/content/dam/Census/library/publications/2015/demo/p25-1143.pdf. 12. Nelson A. Unequal treatment: confronting racial and ethnic disparities in health care. J Natl Med Assoc. 2002;94:666–668. https://doi.org/ 10.1001/jama.290.18.2487. 13. Anastasia EA, Bridges AJ. Understanding service utilization disparities and depression in Latinos: the role of fatalismo. J Immigr Minor Health. 2015;17:1758–1764. https://doi.org/10.1007/s10903-015-0196-y. 14. Salcedo S, McMaster KJ, Johnson SL. Disparities in treatment and service utilization among Hispanics and non-Hispanic whites with bipolar disorder. J Racial Ethn Health Disparities. 2017;4:354–363. https://doi.org/10.1007/s40615-016-0236-x. 15. Budnick HC, Tyroch AH, Milan SA. ScienceDirect Association for Academic Surgery Ethnic disparities in traumatic brain injury care referral in a Hispanic-majority population. J Surg Res. 2019;215:231– 238. https://doi.org/10.1016/j.jss.2017.03.062. 16. Cercosimo E, Musi N. Improving treatment in Hispanic/Latino patients. Am J Med. 2011;124(10 SUPPL.):S16–S21. https://doi.org/ 10.1016/j.amjmed.2011.07.019.
5 17. Romano JG, Sacco RL. Quantifying and addressing persistent stroke disparities in Hispanics. Ann Neurol. 2013;74:759–761. https://doi.org/ 10.1002/ana.24025. 18. Burnett DM, Kolakowsky-Hayner SA, Slater D, et al. Ethnographic analysis of traumatic brain injury patients in the national Model Systems database. Arch Phys Med Rehabil. 2003;84:263–267. https://doi. org/10.1053/apmr.2003.50091. 19. Figueroa JF, Reimold KE, Zheng J, et al. Differences in patient experience between Hispanic and non-Hispanic white patients across U.S. hospitals. J Healthc Qual. 2018;40:292–300. https://doi.org/10.1097/ JHQ.0000000000000113. 20. Flores G. The impact of medical interpreter services on the quality of health care: a systematic review. Med Care Res Rev. 2005;62:255–299. https://doi.org/10.1177/1077558705275416. 21. Baker D, Parker RV, Williams M, et al. Use and effectiveness of interpreters in an emergency department. J Am Med Assoc. 1996;275:783– 788. https://doi.org/10.1001/jama.1996.03530340047028. 22. Baker D, Hayes R, Fortier J. Interpreter use and satisfaction with interpersonal aspects of care for Spanish-speaking patients. Med Care. 1998;36:1461–1470. 23. Manson A. Language concordance as a determinant of patient compliance and emergency room use in patients with asthma. Med Care. 1988;26:1119–1128. 24. Sarver J, Baker DW. Effect of language barriers on follow-up appointments after an emergency department visit. J Gen Int Med. 2000;15:256–264. https://doi.org/10.1046/j.1525-1497.2000.015004256.x. 25. Weinick RM, Jacobs EA, Stone LC, et al. Challenging the myth of a monolithic Hispanic population. 2004; 42:313−320. https://doi.org/ 10.1097/01.mlr.0000118705.27241.7c. 26. van Leer E, Connor NP. Patient perceptions of voice therapy adherence. J Voice. 2010;24:458–469. https://doi.org/10.1016/j.jvoice.2008.12.009. 27. US Census Bureau. 2017 American Community Survey 1-year estimates. 2017. Available at; http://factfinder.census.gov.
Hispanic Ethnicity as a Predictor of Voice Therapy Adherence David E. Rosow, Jennylee Diaz, Debbie R. Pan, and Adam T. Lloyd, Miami, Florida Summary: Background. Voice therapy is a well-studied, evidence-based treatment in the management of voice disorders, yet it is known that adherence rates are generally decreased due to a variety of identified factors. In light of this fact, a high rate of nonadherence to voice therapy has been anecdotally observed in the Hispanic community comprising a sizable portion of the patient population in South Florida. Objective. We sought to analyze the rates of voice therapy attendance for patients who underwent treatment for benign vocal fold nodules at a single tertiary-care academic medical center. Based on our anecdotal observations, we hypothesized that Hispanic patients would have a significantly lower rate of voice therapy attendance compared to non-Hispanic patients. Study Design. Retrospective cohort study. Methods. A retrospective chart review was performed for Hispanic and non-Hispanic patients aged 18 years and older who were diagnosed in a single hospital-based otolaryngology department with benign vocal fold nodules between 2013 and 2018. Patients with other glottic pathology or those who were not recommended voice therapy as initial treatment were excluded. Demographic data, including ethnicity, home address, and preferred language by self-report (English vs. Spanish), were obtained and analyzed. Median income levels for patients were determined by postal codes. “Adherent” status was given to patients who attended at least one voice therapy session. Statistical comparisons of continuous quantitative variables were made using Student’s t test, ordinal quantitative variables using Mann-Whitney U test, and categorical variables using Fischer’s exact test. Statistical significance was determined as P < 0.05. Results. One hundred eleven patients met inclusion criteria. The population was 85% female, with an average age of 41 years. Overall voice therapy adherence rate was 68%. Forty-eight percent of patients self-identified as Hispanic, and of this cohort, 42% spoke Spanish as a preferred language. Differences in annual income levels were noted between non-Hispanic and Hispanic patients ($61,799 vs. $51,697, P = 0.017), as well between English-preferring and Spanishpreferring patients ($60,276 vs. $43,504, P = 0.0014). Thirty of 53 (57%) of Hispanic patients were adherent to voice therapy, compared to 45 of 58 (78%) non-Hispanic patients (P = 0.025). No significant differences were found in age, Voice Handicap Index-10 score, or number of sessions attended between the therapy-adherent patients in the Hispanic and non-Hispanic groups. Further differences in adherence rates were noted when the Hispanic group was subclassified into English and Spanish language preferences. Fifteen of 31 (48%) English-preferring Hispanic patients attended voice therapy compared to 45 of 58 (78%) non-Hispanic patients (P = 0.0085), while Spanish-preferring Hispanic patients had a 68% therapy adherence rate (15 of 22, P = 0.4). English-preferring Hispanic patients had higher average Voice Handicap Index-10 (22.0 vs. 14.9, P = 0.018) and lower total attended sessions (2 vs. 3.6, P = 0.024) than their non-Hispanic counterparts. Conclusion. We believe this is the first study demonstrating a significantly lower rate of voice therapy adherence in Hispanic versus non-Hispanic patients. Decreased utilization of a proven treatment strategy for vocal fold nodules puts these patients at increased risk of treatment failure and decreased voice-related quality of life. Clinicians must be aware of ethnicity-based healthcare disparities and encourage proven treatment adherence to ensure highest quality of life. Key Words: Voice−Voice disorders−Hispanic population−Voice therapy adherence−Voice therapy attendance.
INTRODUCTION Voice therapy is well understood to be a standard, evidencebased treatment option in the management of dysphonia due to a variety of voice-related diagnoses. Therapy involves direct and indirect rehabilitative treatment approaches focused on Accepted for publication September 20, 2019. The authors have no funding, financial relationships, or conflicts of interest to disclose. Presented at the 2018 Fall Voice Conference, Seattle, WA, USA, October 25-27, 2018. From the University of Miami Miller School of Medicine, Department of Otolaryngology, Miami, Florida. Address correspondence and reprint requests to David E. Rosow, University of Miami Miller School of Medicine, 1120 NW 14th Street, 5th Floor, Miami, FL 33136-1015. E-mail: [email protected] Journal of Voice, Vol. &&, No. &&, pp. &&−&& 0892-1997 © 2019 The Voice Foundation. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jvoice.2019.09.011
optimization of vocal function and it has been shown to be effective in reducing dysphonia regardless of the therapy approach used.1−3 Despite the known clinical effectiveness of voice therapy, decreased patient adherence is an ongoing problem.4−7 Failure of voice therapy initiation following clinical recommendation from a laryngologist and speech-language pathologist has been reported to be between 38% and 44%.5−8 Similarly, nonattendance to a voice evaluation with a speechlanguage pathologist following recommendation from a laryngologist has been reported to be 47% and voice therapy dropout rates have been found to be as high as 68%.6 The reasons underlying decreased therapy adherence are still being determined. Demographic factors such as age, gender, and race/ethnicity have not previously been associated with a higher rate of nonadherence. However, individuals with complex laryngeal diagnoses, more severe vocal
ARTICLE IN PRESS 2 impairment, medical comorbidities, occupational issues, and higher Voice Handicap Index (VHI) scores at initial presentation are less likely to attend voice therapy.8 Other factors that affect therapy adherence include self-efficacy and motivation for behavior change, coverage of services by insurance companies, distance from the clinic, the clinician−patient relationship (also termed therapeutic alliance), the level of understanding of the purpose of therapy, and a perceived concern that therapy is too difficult or “silly.”6−8,26 In accordance with suboptimal adherence rates to voice therapy, we have anecdotally observed poor attendance to voice therapy among the English- and Spanish-preferring Hispanic population that comprises a significant portion of patients seen at our facility. Healthcare disparities among minority populations exist and have been well-documented in a wide range of fields. Overall, Hispanics are less likely to seek care when necessary compared to non-Hispanics, and this is influenced by various socioeconomic cultural, linguistic factors that are discussed later. The topic of healthcare disparities among Hispanics is a relevant and significant area of study given the projected rise of this population over the next few decades. The term “Hispanic” refers people “of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin regardless of race.”9 As of July 2017, the Hispanic population of the United States was 58.9 million or 18.1% of the nation’s total population, making people of Hispanic origin the nation’s largest ethnic or racial minority.10 Also, it is projected that the Hispanic population will increase to 119 million by 2060.11 As such, it is becoming increasingly important to understand the medical needs of this population, but no known studies currently exist on adherence to voice therapy among Hispanics. This study seeks to fill this knowledge gap by comparing the rates of adherence to voice therapy between Hispanic and non-Hispanic groups. METHODS Study design and participants A retrospective medical record review was conducted for all patients seen for voice care services at a tertiary academic medical center. Institutional Review Board approval was obtained for this study. Inclusion criteria included patients aged 18 years and older who were diagnosed with benign vocal fold nodules between 2013 and 2018. Patients with other glottic pathology or those for whom voice therapy was not recommended as initial treatment were excluded. Preferred language (English or Spanish) and ethnicity (nonHispanic or Hispanic) were based on self-identified information as reported in the medical records. In this manuscript, the labels of either English- or Spanish-speaking are intended to represent the patient’s language of preference and are not meant to convey language exclusivity. All participants were seen by a laryngologist and one of five speech-language pathologists who are specialized in voice disorders, two of whom are Hispanic and fluent bilingual Spanish speakers. With rare exceptions, all Spanish-
Journal of Voice, Vol. &&, No. &&, 2019
preferring patients are seen for initial evaluation and subsequent therapy by one of the Spanish-fluent speech-language pathologists. Demographic and clinical data, including age, home postal code, gender, ethnicity, and Voice Handicap Index10 scores, were collected and analyzed. Voice therapy adherence was defined as attendance to at least one therapy session, and patients who did not attend at least one therapy session were deemed nonadherent. While annual income for each individual patient was not available, US Census data were used to determine median income level for each patient’s ZIP code. These figures were used as projected indicators of socioeconomic status and compared between the different groups. Statistical analyses Statistical comparisons of continuous quantitative variables were made using Student’s t test, ordinal quantitative variables were compared with the Mann-Whitney U test, and categorical variables were compared with Fischer’s exact test. Statistical significance was defined as P < 0.05. RESULTS Demographic data A total of 111 patients met inclusion criteria. Based on selfreporting, 53 (48%) patients were Hispanic and 58 (52%) were non-Hispanic. The average age of the study population was 41 years, and the majority were female (n = 94; 85%). Specifically, the average age of Hispanic patients with a preferred language of Spanish was 44.5 years, and for those Hispanics with a preferred language of English, it was 38.3 years. This difference was not significant (P = 0.12). The average age of the non-Hispanic group was 41 years. Among the Hispanic group, 22 (42%) of 53 patients self-reported Spanish as their preferred language, whereas the remaining 31 (58%) preferred English. All patients in the non-Hispanic group reported English as their preferred language and no other languages of preference were self-reported within our study population. Voice therapy adherence Overall adherence to voice therapy among all patients (Hispanic and non-Hispanic) was 68%. Non-Hispanic patients had significantly higher adherence to voice therapy compared to all Hispanic patients. Forty-five of 58 (78%) nonHispanic patients were adherent to voice therapy, compared to 30 of 53 (57%) non-Hispanic patients (P = 0.025). No significant differences were seen in age or VHI scores when comparing the therapy-adherent patients in either the Hispanic or non-Hispanic groups. There was also no significant difference in the number of therapy sessions attended between these two groups. When the Hispanic group was further classified into English- and Spanish-speaking, the differences became more pronounced: only 15 of 31 (48%) of English-speaking Hispanic patients attended voice therapy compared to 45 of
ARTICLE IN PRESS David E. Rosow, et al
Hispanic Voice Therapy Adherence
TABLE 1. Number of Patients Stratified by Ethnicity, Language Preference, and Voice Therapy Adherence Status (n = 111) Ethnicity Hispanic (48%)
Language preference Spanish (42%) English (58%)
Non-Hispanic (52%)
Spanish (0%)
English (100%)
Adherence status
Number
Adherent
15
Nonadherent Adherent Nonadherent Adherent
7 15 16 0
Nonadherent Adherent
0 45
Nonadherent
13
58 78% non-Hispanic patients (P = 0.0085), while Spanishspeaking Hispanic patients had a 68% therapy adherence rate (15 of 22, P = 0.4). English-speaking Hispanic patients also had significantly higher average VHI among therapyadherent patients compared to nonadherent patients (22.0 vs. 14.9, P = 0.018), as well as a significantly lower number of therapy sessions attended than their non-Hispanic counterparts (2 vs. 3.6, P = 0.024). Data stratification of number of patients by ethnicity, preferred language, and voice therapy adherence status is summarized in Table 1. Projected annual income The average projected annual income (PAI) for all patients was calculated as $56,984. Significant differences were seen in income level when patients were stratified according to ethnicity, as well as preferred language (Table 2). Hispanic patients had an average PAI of $51,697, while this figure was $61,799 for non-Hispanic patients (P = 0.017). Average PAI for Spanish-preferring patients was $43,505 vs. $60,276 for all English-preferring patients (P = 0.0014). This difference between English- and Spanish-preferring patients was conserved within the Hispanic cohort as well ($57,432 vs. $43,505, P = 0.0021). No difference was seen between Hispanic and non-Hispanic patients who preferred English ($57,432 vs. $61,799, P = 0.40). TABLE 2. Breakdown of Projected Annual Income (PAI) by Ethnicity and Language Preference Ethnicity
Language preference
Hispanic (48%)
Spanish (42%) English (58%) Spanish (0%) English (100%)
Non-Hispanic (52%)
PAI $43,505 $57,432 $61,799
3 DISCUSSION To our knowledge, there are no known studies that examine adherence to voice therapy in the US Hispanic population. Based on evidence from other fields, there may be multiple reasons for decreased adherence to voice therapy among this population. Overall, it is understood that racial and ethnic healthcare disparities negatively impact the Hispanic population in the United States, even when income, age, insurance status, and condition severity are similar to other groups.12 These disparities have been examined in many areas of healthcare, including but not limited to general mental health,13 bipolar disorder,14 depression traumatic brain injury,15 type II diabetes,16 stroke,17 and inpatient rehabilitative services including speech-language pathology.18 Common healthcare barriers among Hispanics include lack of insurance or reduced access to healthcare due to lower rates of full-time employment compared to other groups, fear of seeking services due to immigration status, lack of cultural competence by healthcare providers, and language barriers (limited English proficiency).19 Hispanic patients are more likely to refuse treatment compared to other groups, and they are less likely to receive treatment for common medical conditions compared to non-Hispanics.12 Overall, they are less likely than other groups to have health insurance.13 Additionally, the role of fatalismo, a cultural construct that purports that destiny or a higher power have more control over one’s life than oneself, may also play a role in decreased healthcare service utilization among some Hispanics.13 Anecdotally, a number of Hispanic patients who speak limited English have informally discussed with us that their understanding of their voice diagnoses and the rationale for voice therapy were made easier with a clinician that spoke Spanish. This brings into question whether adherence is affected by the clinician’s ability to communicate in the patient’s preferred language. Inadequate interpretation has been shown to impact quality of care among patients who are limited in English proficiency (LEP), and this may also affect attendance to appointments.20 Patients who require interpretation services but do not receive it are less likely to understand their diagnoses and have reduced satisfaction with their care.21 The use of ad hoc or untrained interpreters (including family members, caregivers, medical and nonmedical staff, and strangers) has also been shown to result in inferior quality of care. However, the use of trained medical interpreters or bilingual providers is associated with increased patient satisfaction and overall quality of care.22 There is limited evidence on attendance to healthcare appointments based on the primary language of communication between LEP patients, healthcare providers, and interpreters. One study conducted with LEP asthmatic patients found that they were more likely to miss appointments with monolingual English-speaking physicians than those with bilingual providers.23 A prospective study conducted with 714 patients in an emergency department found that LEP patients who required but did not receive
ARTICLE IN PRESS 4 interpretation services were twice as likely to be discharged from the ED without a follow-up appointment compared to patients who received interpretation services.24 Further studies are necessary to examine the role that language plays (either with bilingual providers or with interpreters) in attendance to healthcare appointments among LEP patients. For our purposes, this should also be studied in relation to voice therapy adherence. The aim of this study was to assess the rates of adherence to recommended voice therapy between Hispanics and nonHispanics. The results indicate that Hispanics (both English and Spanish speakers) are significantly less likely to attend voice therapy compared to the non-Hispanic population. There was no difference in adherence based on demographic information such as age and gender, though projected annual income differed between Hispanic and non-Hispanic patients. When this difference was broken down further into language preference, the Spanish-preferring group was noted to have a significantly lower income than English-speaking patients, while English-preferring Hispanic patients had no difference in income level. Interestingly, these English-preferring Hispanic patients were even less likely to attend therapy compared to both Spanish-preferring Hispanic patients and non-Hispanics, and this English-preferring group also presented with a significantly higher VHI score and fewer number of therapy sessions attended compared to non-Hispanics. These findings suggest that the nature of the observed disparity in therapy adherence may be rooted more in differences of culture, rather than language. Based on our findings, these differences also do not appear to be rooted in socioeconomic causes, as the lower-income group of Hispanic patients had better therapy attendance than the higher-income group. Further study is warranted to examine whether there are any differences in the healthcare needs of this subpopulation. There are limitations to this study as well as other areas to consider for further investigation. We did not obtain exact income data for each patient but relied instead on averaged US Census data for patients’ ZIP codes. While this method may provide a general assessment of socioeconomic status, it lacks specificity and obscures variations in income within each ZIP code.27 Ideally, individual income levels would be prospectively collected to better understand the impact of socioeconomic status on therapy adherence. Additionally, the Hispanic population is not monolithic, but rather composed of multiple races and many different cultural groups, each with their own challenges to healthcare access.25 Future studies should therefore consider investigating the racial make-up of the Hispanic population in relation to therapy adherence. Certain Hispanic groups, particularly Mexicans and Mexican-Americans, do not make up a large percentage of the South Florida community and are therefore poorly represented in this study. A small but nontrivial portion of the Hispanic patients in this study traveled from Central America, South America, and the Caribbean to receive voice therapy, especially in cases where they previously did not receive sufficient benefit from therapy services in their home countries. Therefore, challenges arising from
Journal of Voice, Vol. &&, No. &&, 2019
travel and finances may have had further implications on Hispanic patients’ adherence to therapy. Additionally, due to the retrospective design of this study, preferred languages may not be accurately reflected per the electronic medical record. It is also important to stress that while our study showed an association between Hispanic ethnicity and difference in therapy adherence, it did not show causation. Therefore, we cannot necessarily attribute the observed difference to the overall underutilization of health care resources by Hispanic patients, as discussed above. Future prospective studies with regression analysis would be better able to answer the question of why we noted this cultural difference. The retrospective design of this study carries inherent selection bias and limits our ability to communicate with patients regarding specific barriers or reasons for nonadherence to therapy. Prospective investigation would reduce bias and allow for further identification of both internal and external challenges that arise in this specific population. The size of the Spanish-preferring population was relatively small compared to the English-speaking groups, and future studies will need to study populations of more similar size. We also chose to examine therapy adherence in patients with benign vocal fold nodules because at our institution, they are consistently prescribed voice therapy as a first treatment prior to any other medical intervention. Thus, this study analyzes adherence to therapy in a specific group that may be considered relatively less severe than other laryngeal diagnoses, such as unilateral vocal fold paralysis or glottic cancer. It is conceivable that in diagnoses considered more “serious,” this disparity in therapy adherence might not be present. CONCLUSION This is the first known study that analyzes adherence rates to voice therapy in the Hispanic population. Hispanics as a group are overall less likely to begin voice therapy compared with non-Hispanics, with nonadherence being more pronounced in among English-speaking Hispanics. Multiple healthcare barriers may influence these poorer adherence rates, and further prospective study is warranted, in light of the increasing Hispanic population in the United States, to examine specific challenges or barriers to voice therapy attendance. REFERENCES 1. Roy N, Merrill RM, Gray SD, et al. Voice disorders in the general population: Prevalence, risk factors, and occupational impact. Laryngoscope. 2005;115:1988–1995. https://doi.org/10.1097/01.mlg.0000179174.32345.41. 2. Speyer R. Effects of voice therapy: a systematic review. J Voice. 2008;22:565–580. https://doi.org/10.1016/j.jvoice.2006.10.005. 3. Yiu EM-L, Lo MCM, Barrett EA. A systematic review of resonant voice therapy. Int J Speech Lang Pathol. 2017;19:17–29. https://doi. org/10.1080/17549507.2016.1226953. 4. Behrman A. Facilitating behavioral change in voice therapy: the relevance of motivational interviewing. Am J Speech Lang Pathol. 2006;15:215–225. https://doi.org/10.1044/1058-0360(2006/020). 5. Portone C, Johns MM, Hapner ER. A review of patient adherence to the recommendation for voice therapy. J Voice. 2008;22:192–196. https://doi.org/10.1016/j.jvoice.2006.09.009.
ARTICLE IN PRESS David E. Rosow, et al
Hispanic Voice Therapy Adherence
6. Hapner E, Portone-Maira C, Johns MM. A study of voice therapy dropout. J Voice. 2009;23:337–340. https://doi.org/10.1016/j.jvoice.2007.10.009. 7. van Leer E, Connor NP. Predicting and influencing voice therapy adherence using social−cognitive factors and mobile video. Am J Speech Lang Pathol. 2015;24:164–176. https://doi.org/10.1044/2015_AJSLP-12-0123. 8. Smith BE, Kempster GB, Sims HS. Patient factors related to voice therapy attendance and outcomes. J Voice. 2010;24:694–701. https:// doi.org/10.1016/j.jvoice.2009.03.004. 9. US Census Bureau. Hispanic Origin. 2018. Available at; https://www. census.gov/topics/population/hispanic-origin/about.html. 10. US Census Bureau. Hispanic Heritage Month. 2018. Available at; https://www.census.gov/newsroom/facts-for-features/2018/hispanicheritage-month.html. 11. US Census Bureau. Projections of the Size and Composition of the U.S. Population: 2014 to 2060. 2015. Available at; https://census.gov/content/dam/Census/library/publications/2015/demo/p25-1143.pdf. 12. Nelson A. Unequal treatment: confronting racial and ethnic disparities in health care. J Natl Med Assoc. 2002;94:666–668. https://doi.org/ 10.1001/jama.290.18.2487. 13. Anastasia EA, Bridges AJ. Understanding service utilization disparities and depression in Latinos: the role of fatalismo. J Immigr Minor Health. 2015;17:1758–1764. https://doi.org/10.1007/s10903-015-0196-y. 14. Salcedo S, McMaster KJ, Johnson SL. Disparities in treatment and service utilization among Hispanics and non-Hispanic whites with bipolar disorder. J Racial Ethn Health Disparities. 2017;4:354–363. https://doi.org/10.1007/s40615-016-0236-x. 15. Budnick HC, Tyroch AH, Milan SA. ScienceDirect Association for Academic Surgery Ethnic disparities in traumatic brain injury care referral in a Hispanic-majority population. J Surg Res. 2019;215:231– 238. https://doi.org/10.1016/j.jss.2017.03.062. 16. Cercosimo E, Musi N. Improving treatment in Hispanic/Latino patients. Am J Med. 2011;124(10 SUPPL.):S16–S21. https://doi.org/ 10.1016/j.amjmed.2011.07.019.
5 17. Romano JG, Sacco RL. Quantifying and addressing persistent stroke disparities in Hispanics. Ann Neurol. 2013;74:759–761. https://doi.org/ 10.1002/ana.24025. 18. Burnett DM, Kolakowsky-Hayner SA, Slater D, et al. Ethnographic analysis of traumatic brain injury patients in the national Model Systems database. Arch Phys Med Rehabil. 2003;84:263–267. https://doi. org/10.1053/apmr.2003.50091. 19. Figueroa JF, Reimold KE, Zheng J, et al. Differences in patient experience between Hispanic and non-Hispanic white patients across U.S. hospitals. J Healthc Qual. 2018;40:292–300. https://doi.org/10.1097/ JHQ.0000000000000113. 20. Flores G. The impact of medical interpreter services on the quality of health care: a systematic review. Med Care Res Rev. 2005;62:255–299. https://doi.org/10.1177/1077558705275416. 21. Baker D, Parker RV, Williams M, et al. Use and effectiveness of interpreters in an emergency department. J Am Med Assoc. 1996;275:783– 788. https://doi.org/10.1001/jama.1996.03530340047028. 22. Baker D, Hayes R, Fortier J. Interpreter use and satisfaction with interpersonal aspects of care for Spanish-speaking patients. Med Care. 1998;36:1461–1470. 23. Manson A. Language concordance as a determinant of patient compliance and emergency room use in patients with asthma. Med Care. 1988;26:1119–1128. 24. Sarver J, Baker DW. Effect of language barriers on follow-up appointments after an emergency department visit. J Gen Int Med. 2000;15:256–264. https://doi.org/10.1046/j.1525-1497.2000.015004256.x. 25. Weinick RM, Jacobs EA, Stone LC, et al. Challenging the myth of a monolithic Hispanic population. 2004; 42:313−320. https://doi.org/ 10.1097/01.mlr.0000118705.27241.7c. 26. van Leer E, Connor NP. Patient perceptions of voice therapy adherence. J Voice. 2010;24:458–469. https://doi.org/10.1016/j.jvoice.2008.12.009. 27. US Census Bureau. 2017 American Community Survey 1-year estimates. 2017. Available at; http://factfinder.census.gov.