Introduction
The voice is vital for effective communication. A sharp rise in advances in technology and instrumentation has aided our understanding of the physiology of voice, airway and swallowing. Around a third of the total population depend on their voice as a primary work tool, and 1 per cent of the world's population suffer from hoarseness or dysphonia at least once in their lifetime.Reference Behlau, Zambon and Madazio1,Reference Reiter, Hoffmann, Pickhard and Brosch2 Data on hoarseness as a symptom are available in the literature, with excellent guidelines being laid down regarding its patterns, triaging and management in the developed countries.Reference Stachler, Francis, Schwartz, Damask, Digoy and Krouse3 Patterns of dysphonia in the developing and underdeveloped world are varied, as optimal voice drops down in the priority list of health concerns.
Our study aimed to determine the epidemiology and the general management methods undertaken for common laryngology problems at our institute in India. The large sample size of this study, covering a population from a wide geographical region, gives us an idea about the pattern of hoarseness as a symptom and the scope of laryngology in the developing countries of the Indian subcontinent.
Materials and methods
Patients
Between January 2017 and December 2017, all patients with hoarseness who presented to the laryngology out-patient clinic and who underwent rigid or flexible laryngoscopy to diagnose the underlying cause were included in the study. There were no exclusion criteria. This retrospective observational study was conducted at a tertiary care teaching hospital in South India after approval by the institutional review board and the ethics committee (institutional review board number: 12587).
Methodology
Data for this retrospective observational study were retrieved from the computerised hospital information processing system. A detailed information profile was prepared for each individual, which included demographic and clinical information. Voice users were classified as level 1 users – elite voice performers or professionals (singers, theatre artists and performers), level 2 users – vocal professionals (e.g. clergy, radio jockeys, presenters, teachers), level 3 users – non-vocal professionals (e.g. doctors, lawyers, nurses, pharmacists) and level 4 users – non-vocal non-professionals.
Statistical analysis
The data were entered using EpiData software, version 3.1. All analyses were performed using SPSS statistical software, version 21.0 (IBM, Armonk, New York, USA). For continuous data, descriptive statistics of mean (standard deviation) and range were presented. For categorical data, numbers of patients and percentages were presented. The chi-square and Fisher's exact test (less cell count) were applied to find associations between categorical variables. All tests were two-sided at the α = 0.05 level of significance.
Results
Epidemiology
Of a total of 8320 new patients who attended the laryngology unit, 1033 presented with a change in voice and underwent laryngoscopy. The incidence of hoarseness in the patients seen at our unit was 12.4 per cent.
Among these patients with hoarseness, 70 per cent were males and 30 per cent females. The patients’ mean age was 46 ± 14.3 years (range, 10–91 years). The mean duration of hoarseness was 23 months, ranging from immediate onset of hoarseness (1 day) to 40 years.
The geographical distribution of the patients was wide, although the majority were from East India and Bangladesh. Other countries included neighbouring Myanmar, Sri Lanka, Bhutan and Nepal (Figure 1). Over a third of these patients (36.4 per cent) had a documented history of treatment or opinion from elsewhere before coming to our centre.
Risk factors and co-morbidities
Use of tobacco was recorded in 35.24 per cent, whereas only 7.6 per cent of patients admitted to consuming alcohol. The systemic illnesses documented were hypertension (16.2 per cent), diabetes (15.4 per cent), pulmonary pathology (12.9 per cent) and concomitant thyroid abnormalities (11.4 per cent).
Voice user levels
The majority of patients (50.73 per cent) were non-vocal non-professional voice users; only 3.2 per cent were level 1 voice users (elite voice performers) (Table 1).
*Total n = 1033. †Polyps, nodules, cysts, scars and sulci, arytenoid granuloma, Reinke's oedema, ectasia. ‡Puberphonia, muscle tension dysphonia, psychogenic. **Bacterial, viral, fungal laryngitis. §Haemangioma, papillomas, nerve sheath tumours. SD = standard deviation; RT = radiotherapy
Diagnosis
On histopathology, the majority of cases (82 per cent) were non-malignant (Table 1). Of the malignancies, squamous cell carcinoma was the most common (n = 177); two were variants of squamous cell carcinoma and the other eight were non-squamous.
The non-malignant conditions included benign mucosal lesions (e.g. polyps, nodules, cysts, scars or sulci, granulomas) (Figure 2), which accounted for the majority of cases (50.7 per cent, n = 473). The precise percentage distribution for each of these conditions was difficult to obtain, as there were many with more than one condition. For example, a large number of patients with laryngopharyngeal reflux, chronic laryngitis, polyps and scars had co-existing keratosis too.
A significant number of patients were diagnosed with vocal fold palsy (n = 105), with iatrogenic causes (32 per cent) being most common (Figure 3). Sixty-nine patients had pre-malignant leucoplakia or keratotic lesions; along with malignancies, such lesions contributed around 25 per cent of the total case load with hoarseness. The benign tumours (n = 11) seen were laryngeal papilloma, haemangiomas and nerve sheath tumours.
Of the 33 patients with neurogenic dysphonia, 10 had spasmodic dysphonia; the remaining cases were caused by Parkinson's disease, myasthenia gravis, amyotrophic lateral sclerosis, motor neurone disease, tremor and stroke. Muscle tension dysphonia (n = 20) and puberphonia (n = 12) were the most common functional voice disorders (n = 40). Regarding infective causes, 10 patients had laryngeal tuberculosis (TB) and 11 had fungal laryngitis.
Management
A total of 337 patients (33 per cent) were managed surgically, while 696 (67 per cent) were managed conservatively, with 565 patients requiring voice therapy. Only 39 per cent of patients with keratosis (n = 27) underwent a surgical procedure. A large proportion of patients with laryngeal squamous cell carcinoma (71 per cent, n = 125) opted for non-surgical management in the form of radiation or chemotherapy (organ preservation protocol). The majority of those with laryngeal trauma (75 per cent, n = 9), and 91 per cent of those with stenosis (n = 31), needed surgical intervention; only 18 per cent (n = 19) with nerve palsy opted for any form of surgical intervention.
Associations
In this study, we attempted to look for associations between risks and co-morbidities, voice user levels and diagnoses. We found significant associations between: level 3 and 4 voice users and polyps (p < 0.001) as well as keratotic pre-malignant lesion formation (p < 0.025); voice user level 1 and 2 and nodules (p < 0.004); voice user level 2 and 4 and scars and sulci (p < 0.009); and voice user level 2 and cysts (p < 0.034) (Table 2).
Voice user categories: level 1 – elite voice performers or professionals, level 2 – vocal professionals, level 3 users – non-vocal professionals, and level 4 users – non-vocal non-professionals. *Indicates statistical significance (p < 0.05)
We also observed a significant association between pulmonary co-morbidities and the presence of keratotic laryngeal lesions (p = 0.008). In most cases (80 per cent), laryngeal TB was secondary to pulmonary TB. As expected, there was a significant association between tobacco use (92.5 per cent were smokers) and laryngeal malignancies (p < 0.001). Smoking was also significantly associated with the formation of pre-malignant lesions (p = 0.02) and scars (p < 0.005).
In our population, we found significant associations between hypertension and the presence of nodules (p = 0.002) and polyps (p = 0.03), and between autoimmune disease and the presence of airway stenosis (p = 0.01). The presence of thyroid disorders and diabetes did not show an association with any voice pathology.
After the initial consultation, 42 per cent of patients (n = 434) required cross-referrals for optimal management, and 565 underwent speech therapy. Only 53 per cent (n = 547) of the study patients had a minimum follow up of three months.
Discussion
Hoarseness may result from a wide variety of causes. Laryngologists have to work with other medical sub-specialties as well as speech and language pathologists, whose contribution is indispensable for optimal management in most cases of dysphonia, making treatment a multidisciplinary team effort.
Epidemiological differences
Data show that dysphonia has a point prevalence of around 1 per cent among patients seeking treatment, with a female predominance (63.4 per cent).Reference Reiter, Hoffmann, Pickhard and Brosch2,Reference Cohen, Kim, Roy, Asche and Courey4 Indian reports of smaller study groups (100–110 patients) showed lower incidence rates of hoarseness (less than 0.5 per cent) and a male predominance (51–89 per cent).Reference Baitha, Raizada, Singh, Puttewar and Chaturvedi5–Reference Pal, Kaushal, Nagpure and Agarwal9 The majority of our patient population was male (70 per cent), and the incidence of hoarseness in our study was 12.4 per cent. While the high incidence is probably a result of having a specialised laryngology sub-specialty unit, it also highlights the scope for laryngology in this country. The male predominance, as against the literature in Western, more developed countries, may point towards the perceived lack of importance and limited attention given to non-emergent medical issues in women, regardless of whether the conditions affect quality of life.
Among 80 per cent of Indian patients, the average duration of hoarseness as a presenting complaint has been reported to be around 12 months,Reference Baitha, Raizada, Singh, Puttewar and Chaturvedi5 as compared with 1–3 months in more than 60 per cent of patients in Western literature.Reference Stachler, Francis, Schwartz, Damask, Digoy and Krouse3,Reference Brouha, Tromp, de Leeuw, Hordijk and Winnubst10 In our study, we found the mean duration of presenting complaints to be 24 months, clearly indicating a delay in seeking medical attention for hoarseness in the Indian subcontinent and a lack of awareness about the need to seek medical attention.
Diagnostic differences
Indian studies report a much lower rate of acute laryngitis (11–23 per cent vs around 40 per cent) and a much higher rate of malignant conditions (8–74 per cent vs around 2–3 per cent),Reference Baitha, Raizada, Singh, Puttewar and Chaturvedi5–Reference Pal, Kaushal, Nagpure and Agarwal9 with similar percentages of vocal fold palsy (3–10 per cent) and chronic laryngitis (9–10 per cent), compared with international data.Reference Reiter, Hoffmann, Pickhard and Brosch2,Reference Cohen, Kim, Roy, Asche and Courey4 Benign vocal fold lesions reported in Indian literature range from 8 to 60 per cent,Reference Baitha, Raizada, Singh, Puttewar and Chaturvedi5–Reference Pal, Kaushal, Nagpure and Agarwal9 compared with 10–31 per cent in Western literature.Reference Reiter, Hoffmann, Pickhard and Brosch2,Reference Cohen, Kim, Roy, Asche and Courey4 Our study showed the incidence of malignancy, vocal fold palsy and benign vocal fold lesions to be around 18 per cent, 10.5 per cent and 45 per cent, respectively.
Indian data indicate a 1–5 per cent incidence of laryngeal TB.Reference Baitha, Raizada, Singh, Puttewar and Chaturvedi5–Reference Batra, Motwani and Sagar7 Up to 14.5 per cent of head and neck TB cases have been reported to have laryngeal involvement.Reference Prasad, Sreedharan, Chakravarthy and Prasad11 While all 32 cases of laryngeal TB reported by Bhat et al.Reference Bhat, Latha, Upadhya and Hegde12 had pulmonary involvement, only 20 per cent of laryngeal TB patients reported by Gandhi et al.Reference Gandhi, Kulkarni, Mishra and Thekedar13 showed systemic involvement. With 8 out of 10 cases of laryngeal TB (80 per cent) being secondary to pulmonary TB, our findings were similar to those of Lim et al., where the majority had secondary laryngeal TB.Reference Lim, Kim, Choi, Kim, Kim and Choi14
Voice user level data
Voice professionals, such as teachers, broadcasters and musicians, whose professions are reliant on their voice, appear to be most at risk of developing a voice disorder.Reference Koufman and Isaacson15–Reference Timmermans, De Bodt, Wuyts, Boudewijns, Clement and Peeters17 Although level 1 voice users are typically well-qualified voice users, others often have insufficient or no voice use and vocal hygiene training, making them more likely to adopt inappropriate voice production techniques. Indian studies have reported that most dysphonic patients can be categorised as a level 4 voice user.Reference Batra, Motwani and Sagar7–Reference Pal, Kaushal, Nagpure and Agarwal9 That was also seen in this study, clearly illustrating the need for voice and vocal hygiene education in the general population. In a Belgian study, functional voice disorders (41 per cent), vocal fold nodules or hypertrophy (15 per cent), and reflux, laryngitis or inflammation (11 per cent) were found to be the most common pathologies among professional voice users.Reference Van Houtte, Van Lierde, D'Haeseleer and Claeys18
No studies correlating voice user levels or co-morbidities with pathology were found in the Indian literature search. We observed some interesting, statistically significant associations. First, vocal nodules among both level 1 and 2 voice users suggest that overuse with a lack of adequate rest could be a contributing factor among trained individuals. Second, among level 3 and 4 voice users, vocal polyps were more common, suggesting that sudden phonotrauma causes hoarseness in untrained voice users. Third, vocal cysts were seen more frequently in level 2 voice users.
Co-morbidities, habit-forming substances and dysphonia
Apart from well-proven correlations between tobacco use and laryngeal malignancies,Reference Sapkota, Gajalakshmi, Jetly, Roychowdhury, Dikshit and Brennan19–Reference Bobdey, Jain and Balasubramanium22 there are few significant associations between systemic co-morbidities and laryngeal pathologies in South Asian literature. In our patients, tobacco use was significantly associated not only with the presence of laryngeal malignancies, but also with pre-malignant or keratotic lesions, scars and nodules.
Chronic cough has been postulated to be related to an increased prevalence of benign vocal fold lesions.Reference Adessa, Xiao, Hull, Bowen, Milstein and Benninger23 The phonotrauma caused by the classical smoker's cough because of the chemical and thermal irritation, or because of underlying pulmonary co-morbidities, leads to increased keratinisation of delicate vocal fold epithelium, which could be the reason for these lesions.Reference Adessa, Xiao, Hull, Bowen, Milstein and Benninger23 This could explain why we found not only smoking but also pulmonary co-morbidities to be significantly associated with the presence of keratotic lesions on the vocal folds.
• This study, with a large sample and wide geographical distribution, provides a good overview of hoarseness in the Indian subcontinent
• Compared to Western literature, the data show large delays and male predominance in seeking medical attention
• This indicates that optimal voice is low on the list of health priorities in the subcontinent
• Significant associations were found between vocal professionalism level and co-morbidities and dysphonia aetiopathologies
• These associations might not imply clinical causation, but could be important for clinical management
The association of personality types with the formation of nodules and polyps has been extensively discussed in literature.Reference El Uali Abeida, Fernández Liesa, Vallés Varela, García Campayo, Rueda Gormedino and Ortiz García24–Reference Yano, Ichimura, Hoshino and Nozue29 In our cohort, the association of hypertension with vocal nodules as well as polyps could be the result of voice abuse and misuse in these individuals who are more subject to stress and tend to become anxious or angry.
Systemic TB was seen in 80 per cent of our patients with laryngeal TB, supporting the data indicating that isolated or primary laryngeal TB is uncommon.Reference Lim, Kim, Choi, Kim, Kim and Choi14,Reference Reis, Reis, da Costa, Lucena, Schubach and Oliveira30
An association between airway stenosis and autoimmune conditions has also been described in the literature (e.g. granulomatosis with polyangiitis, Churg–Strauss syndrome, immunoglobulin G4 disease, and idiopathic subglottic stenosis).Reference Aarnæs, Sandvik and Brøndbo31–Reference Kobraei, Song, Mathisen, Deshpande and Mark36 Patients present with hoarseness, cough or breathing difficulty either to the otolaryngologist or to the rheumatologist. After the initial management, the patients often need to be monitored every three to six months, as they tend to relapse.
Conclusion
This study, with its large sample size and the inclusion of patients from a wide geographical area, provides a good overview of hoarseness as a symptom in the developing Indian subcontinent. The hoarseness seen in our tertiary care referral centre was predominantly chronic, with acute-onset dysphonia being largely iatrogenic or traumatic in nature. An optimal voice seems to be lower on the list of health concerns in the subcontinent, especially among women. Polyps, nodules, malignancies and vocal fold palsies remain the major diagnoses, with scars and keratosis contributing a considerable number. A multidisciplinary approach to voice problems is often needed.
In addition to the commonly known associations between professional voice users and vocal fold nodules, and between tobacco use and malignancy and pre-malignant lesions, there appear to be interesting associations between hypertension and phonotraumatic lesions, between pulmonary co-morbidities and keratotic lesions, and between non-vocal professionals and non-professionals (voice user levels 3 and 4) and polyp formation. Even though these associations do not necessarily imply causation, exploring them further may provide useful information for management of these conditions. We hope that larger, multi-centre studies of hoarseness will be conducted in the future to understand this largely under-reported and under-emphasised symptom in our subcontinent.
Acknowledgements
The authors gratefully acknowledge the Departments of Biostatistics and the Department of Computerised Hospital Information Services at Christian Medical College, Vellore, India, for their expertise with analysis and data collection respectively.
Competing interests
None declared