ORIGINAL ARTICLE
Voice outcomes following endolaryngeal surgery; are we achieving our aims? Ansari, S. & MacKenzie, K. Departmet of Otolaryngology, Head & Neck Surgery, Glasgow Royal Infirmary, Glasgow, UK Accepted for publication 17 March 2015 Clin. Otolaryngol. 2015, 40, 580–586
Objective: To determine voice outcomes in endolaryngeal surgery. Methods: Prospective assessment of patients undergoing endolaryngeal surgery in one head and neck team in North Glasgow from 2007 to 2014. Each patient completed a VoiSS (the Voice Symptom Scale) questionnaire at pre-assessment and post-endolaryngeal surgery. Per-operatively, the aim of the procedure was recorded from a choice of seven options in addition to the clinical conditions and the procedure. Results: One thousand and sixty-six patient episodes were recorded. Thirty-two had incomplete data at pre-assessment
and were excluded. Of the remainder, 703 (68%) had complete paired data sets. ‘To improve voice’ (n = 169, 24%) and ‘to achieve a biopsy with no deterioration in voice’ (n = 185, 26%) categories had significant outcomes in keeping with the operative aims. Conclusion: This study demonstrated that it is possible to assess patient reported voice outcomes in universal heterogeneous endolaryngeal surgery cohort over a prolonged period. The voice outcomes were in keeping with the stated aims of the endolaryngeal surgical procedures.
Endolaryngeal surgery is an integral part of both the general and specialist ENT surgeon’s workload. The principal aim of such surgery is to diagnose and potentially treat pathologies of the upper aerodigestive tract with due consideration for the patient’s voice. The indications for endolaryngeal surgery are many and varied. These range from detailed examination of the larynx and hypopharynx under general anaesthesia, without the need for further instrumentation, to those that require intricate microscopic/endoscopic procedures on the laryngeal structures, principally the vocal cords.1–8 Although obtaining the histological diagnosis is one of the key outcomes of endolaryngeal surgery, the other principal outcome should be the effect on the quality of voice. Obtaining the histological diagnosis should not be at the expense of deterioration of voice quality and in a proportion of cases improvement in the voice quality will be the key outcome.9 To date, most published assessments of voice outcomes relate to interventions for specific pathologies and surgical techniques. This includes pathologies such as vocal cord paralysis, vocal cord polyps, Reinke’s oedema, laryngeal
papillomatosis and techniques such as specific microlaryngeal cold-steel instrumentation, carbon dioxide laser and the microdebrider. Not only are these reports frequently in relation to very specific pathologies, but also often come from institutions with a highly specialised skills set and specific referral patterns, for example those that would be expected in a tertiary referral centre.10–13 The question arises about how relevant such studies are to a general and specialised ENT practice, and in turn applicable to the general population. Frequently, assessments of these surgical interventions are made in relation to voice outcome measures which may be in the form of acoustic analysis or perceptual assessments.14 Although of relevance, these more research orientated assessment tools are often not readily available within routine otorhinolaryngological practice. With the development of patient reported outcome measures for voice such as the VoiSS (the Voice Symptom Scale), the VHI (the Vocal Handicap Inventory) and the VPQ (the Voice Performance Questionnaire), there is now the possibility of utilising these more readily accessible voice assessment tools in the population of patients undergoing endolaryngeal surgery.15–18 To allow comprehensive informed consent, it is important to assess voice outcomes in relation to the stated aims of all endolaryngeal surgical procedures. Ideally, this should take place prospectively in the general and specialist ENT laryngeal practice, be universally applied, and involve all
Correspondence: S. Ansari, ENT Department Glasgow Royal Infirmary, Alexandra Parade, Glasgow G4 0SF, UK. Tel.: 07789080568; Fax: 0141 211 4999; e-mail: [email protected]
580
© 2015 John Wiley & Sons Ltd Clinical Otolaryngology 40, 580–586
Voice outcomes of endolaryngeal surgery
surgeons within the team. Only by carrying out such generic evaluation will it be possible for the role of endolaryngeal surgery to be assessed in routine clinical practice.19 The aim of this study was to assess the voice outcomes with respect to the stated aims of surgery in a prospective universal audit of endolaryngeal surgery. Materials and methods Clinical study
All endolaryngeal surgery performed by one head and neck firm over a period of seven years, from May 2007 to May 2014, was audited prospectively. This included procedures carried out by the consultant incharge and trainees under direct supervision. The firm has a specialist profile of laryngology/head and neck surgery and manages all forms of benign and malignant disease.
581
assessment tool. It is a 30-item questionnaire addressing the physical, emotional and functional effects of voice on the quality of life. It results in an overall score between 0 and 120 with greater voice problems resulting in a higher score and with normal voice generally having a score less than 12.17,18,20,21 All patients undergoing endoscopic laryngeal surgery in the audit period completed the VoiSS questionnaire at preoperative assessment. At the time of the surgical procedure, the aim of the surgery was recorded from a choice of seven options (Table 1), in addition to the clinical condition and the endolaryngeal procedure. Three months postoperatively, the VoiSS questionnaire was posted to all the patients in a stamped addressed envelope and they were asked to complete and return the questionnaire. The data were entered onto a Microsoft Access database specifically created for recording endolaryngeal surgery. Analysis
Outcome measure
The primary outcome measure was voice. This was assessed using Voice Symptom Scale (VoiSS). The VoiSS questionnaire is a widely accepted and validated quality of life voice
Data were extracted from the MS Access database, and statistical analysis was carried out using Mini Tab 16. Wilcoxon’s rank sum test was used for analysis of nonparametric data.
Table 1. Table showing comparative analysis of demographs, aims of procedure and diagnoses in both complete and incomplete data sets
n Sex M F Mean age Mean preop VoiSS Aim of the procedure To exclude pathology To achieve biopsy with no deterioration in voice To improve voice quality To take biopsy of gross tumours To improve airway To resect tumours Others Total Diagnosis Malignant laryngeal pathologies Globus pharyngeus Carcinoma in situ/dysplasia Vocal cord nodules/cysts/polyps Papillomatosis Vocal cord paralysis Laryngitis Reinke’s oedema Functional dysphonia Others (webs, stenosis, stricture) Total
© 2015 John Wiley & Sons Ltd Clinical Otolaryngology 40, 580–586
Complete data set
Incomplete data set
703
331
351 352 57 40 n (%) 216(31) 185(26) 169(24) 77(11) 27(4) 21(3) 8(1) 703(100) n (%) 207(29) 99(14) 86(12) 83(12) 58(8) 41(6) 36(5) 33(5) 13(2) 47(7) 703(100)
165 166 58 43 n (%) 119(36) 60(18) 77(23) 51(15) 12(4) 7(2) 5(2) 331(100) n (%) 92(28) 59(18) 38(11.5) 38(11.5) 23(7) 11(3) 21(6) 23(7) 5(2) 21(6) 331(100)
582 S. Ansari & K. MacKenzie
Ethical consideration
As this study is an audit of current surgical clinical practice, specific ethical approval from the regional ethics committee was not required. The study was however discussed with the NHS Greater Glasgow and Clyde Clinical Effectiveness Group. Results and analysis
One thousand and sixty-six endolaryngeal surgical procedures were carried out in the seven-year study period (Fig. 1). Thirty-two were excluded. This was due to either patient demise in the postoperative period due to nonrelated cause or preoperative questionnaires not being adequately completed. As all patients attending preoperative assessment were asked to complete the VoiSS questionnaire, this was taken as an implied consent. If patients did not complete the form at preoperative assessment, they were excluded on the basis that they did not wish to participate. Seven hundred and three (68%) patients had complete clinical and surgical data with paired preoperative and postoperative VoiSS questionnaires. Comparative analysis of complete and incomplete data sets was carried out in Table 1. As both groups were not only similar in demographics but also in diagnosis and aim of the procedure categories, the groups were concluded to be comparable. It was decided not to proceed with an intention to treat analysis but to analyse the ‘complete data set’ cohort. The majority of endoscopic laryngeal surgery was performed to exclude pathology without taking laryngeal biopsies (31%) and consisted of those with previously treated head and neck cancer or those where there was a concerning group of symptoms. The next most frequent aim was to achieve a diagnosis by sampling the larynx but without having any impact on voice (26%). This group included patients with dysplasia, hyperkeratosis and low volume squamous disease. Approximately 24% of the endoscopic Total procedures 1066 May 2007 – May 2014
Excluded 32
Included 1034
- Deceased - No Pre-op data
(Implied consent)
Complete dataset 703 (68%)
Fig. 1. Total numbers of patients.
laryngeal workload was in the ‘to improve voice quality’ group which included benign pathologies such as vocal cord polyps, cysts, nodules and papillomatosis (Table 1). The preoperative and postoperative VoiSS scores in each of the seven groups are shown in Table 2. There was a statistically significant improvement in the overall median VoiSS scores for all seven hundred and three episodes (P < 0.005) (Table 2). In the ‘to improve voice quality’ group, there was a statistically significant improvement in subjective self-rated voice outcome. However, in patients where the aim was ‘to achieve biopsy without deterioration in their voice’, there was no statistically significant difference in their preoperative and postoperative scores, consistent with the aim of surgery (Table 2). The heterogeneity of our practice is evident from Table 3. The majority of the workload involved malignant pathologies, two hundred and seven (29%) with a further eighty-six (12%) have dysplasia or carcinoma in situ. This group with dysplasia or carcinoma in situ required a biopsy without any deterioration in voice. Other notable groups were benign laryngeal pathologies, with the aim to improve voice, such as vocal cord nodules/cysts/polyps, Reinke’s oedema and laryngeal papillomatosis. The preoperative and postoperative scores in each of these diagnostic groups are shown in Table 3. There was significant improvement in the VoiSS scores in patients with vocal cord nodules/cysts/polyps, papillomatosis, vocal cord palsy and Reinke’s oedema groups as anticipated. The scatter plot in Fig. 2 shows the individual differences in Voice Symptom Scale (VoiSS) scores pre- and postoperatively across all seven cohorts. Analysis of surgical aims
To exclude pathology (n = 216). This group consists of patients with globus pharyngeus and dysphonia undergoing endolaryngeal surgery to exclude a malignant process and also groups undergoing surveillance following treatment for head and neck cancer. Overall, there was no statistical improvement in VoiSS scores across the cohort. To achieve biopsy without deterioration in voice (n = 185). This group included patients with non-invasive malignant pathologies such as keratosis, dysplasia, carcinoma in situ and low volume squamous cell cancer (T1/T2). There was no significant change in the VoiSS scores in this subgroup. The aim of achieving a biopsy without a deterioration in voice was achieved. To improve voice group (n = 169). This cohort included benign laryngeal pathologies such as vocal cord cysts and © 2015 John Wiley & Sons Ltd Clinical Otolaryngology 40, 580–586
Voice outcomes of endolaryngeal surgery
583
Table 2. Table showing median and interquartile range of preoperative and postoperative VoiSS scores according to the aims of the procedure with their statistical analysis (P) using Wilcoxon’s signed rank test Preop VoiSS
Postop VoiSS
Aim
n
Median
IQR
Median
IQR
P
To exclude pathology To achieve biopsy with no deterioration in voice To improve voice quality To take biopsy of gross tumours To improve airway To resect tumours Others Total
216 185 169 77 27 21 8 703
35 38 51 26 42 21 18 39
30.75 38 30 36.5 38 27 35 35
32 34 35 30 39 19 25 33
37 46.5 36.5 45.5 38 32 32 40
0.584 0.943
Voice outcomes following endolaryngeal surgery; are we achieving our aims? Ansari, S. & MacKenzie, K. Departmet of Otolaryngology, Head & Neck Surgery, Glasgow Royal Infirmary, Glasgow, UK Accepted for publication 17 March 2015 Clin. Otolaryngol. 2015, 40, 580–586
Objective: To determine voice outcomes in endolaryngeal surgery. Methods: Prospective assessment of patients undergoing endolaryngeal surgery in one head and neck team in North Glasgow from 2007 to 2014. Each patient completed a VoiSS (the Voice Symptom Scale) questionnaire at pre-assessment and post-endolaryngeal surgery. Per-operatively, the aim of the procedure was recorded from a choice of seven options in addition to the clinical conditions and the procedure. Results: One thousand and sixty-six patient episodes were recorded. Thirty-two had incomplete data at pre-assessment
and were excluded. Of the remainder, 703 (68%) had complete paired data sets. ‘To improve voice’ (n = 169, 24%) and ‘to achieve a biopsy with no deterioration in voice’ (n = 185, 26%) categories had significant outcomes in keeping with the operative aims. Conclusion: This study demonstrated that it is possible to assess patient reported voice outcomes in universal heterogeneous endolaryngeal surgery cohort over a prolonged period. The voice outcomes were in keeping with the stated aims of the endolaryngeal surgical procedures.
Endolaryngeal surgery is an integral part of both the general and specialist ENT surgeon’s workload. The principal aim of such surgery is to diagnose and potentially treat pathologies of the upper aerodigestive tract with due consideration for the patient’s voice. The indications for endolaryngeal surgery are many and varied. These range from detailed examination of the larynx and hypopharynx under general anaesthesia, without the need for further instrumentation, to those that require intricate microscopic/endoscopic procedures on the laryngeal structures, principally the vocal cords.1–8 Although obtaining the histological diagnosis is one of the key outcomes of endolaryngeal surgery, the other principal outcome should be the effect on the quality of voice. Obtaining the histological diagnosis should not be at the expense of deterioration of voice quality and in a proportion of cases improvement in the voice quality will be the key outcome.9 To date, most published assessments of voice outcomes relate to interventions for specific pathologies and surgical techniques. This includes pathologies such as vocal cord paralysis, vocal cord polyps, Reinke’s oedema, laryngeal
papillomatosis and techniques such as specific microlaryngeal cold-steel instrumentation, carbon dioxide laser and the microdebrider. Not only are these reports frequently in relation to very specific pathologies, but also often come from institutions with a highly specialised skills set and specific referral patterns, for example those that would be expected in a tertiary referral centre.10–13 The question arises about how relevant such studies are to a general and specialised ENT practice, and in turn applicable to the general population. Frequently, assessments of these surgical interventions are made in relation to voice outcome measures which may be in the form of acoustic analysis or perceptual assessments.14 Although of relevance, these more research orientated assessment tools are often not readily available within routine otorhinolaryngological practice. With the development of patient reported outcome measures for voice such as the VoiSS (the Voice Symptom Scale), the VHI (the Vocal Handicap Inventory) and the VPQ (the Voice Performance Questionnaire), there is now the possibility of utilising these more readily accessible voice assessment tools in the population of patients undergoing endolaryngeal surgery.15–18 To allow comprehensive informed consent, it is important to assess voice outcomes in relation to the stated aims of all endolaryngeal surgical procedures. Ideally, this should take place prospectively in the general and specialist ENT laryngeal practice, be universally applied, and involve all
Correspondence: S. Ansari, ENT Department Glasgow Royal Infirmary, Alexandra Parade, Glasgow G4 0SF, UK. Tel.: 07789080568; Fax: 0141 211 4999; e-mail: [email protected]
580
© 2015 John Wiley & Sons Ltd Clinical Otolaryngology 40, 580–586
Voice outcomes of endolaryngeal surgery
surgeons within the team. Only by carrying out such generic evaluation will it be possible for the role of endolaryngeal surgery to be assessed in routine clinical practice.19 The aim of this study was to assess the voice outcomes with respect to the stated aims of surgery in a prospective universal audit of endolaryngeal surgery. Materials and methods Clinical study
All endolaryngeal surgery performed by one head and neck firm over a period of seven years, from May 2007 to May 2014, was audited prospectively. This included procedures carried out by the consultant incharge and trainees under direct supervision. The firm has a specialist profile of laryngology/head and neck surgery and manages all forms of benign and malignant disease.
581
assessment tool. It is a 30-item questionnaire addressing the physical, emotional and functional effects of voice on the quality of life. It results in an overall score between 0 and 120 with greater voice problems resulting in a higher score and with normal voice generally having a score less than 12.17,18,20,21 All patients undergoing endoscopic laryngeal surgery in the audit period completed the VoiSS questionnaire at preoperative assessment. At the time of the surgical procedure, the aim of the surgery was recorded from a choice of seven options (Table 1), in addition to the clinical condition and the endolaryngeal procedure. Three months postoperatively, the VoiSS questionnaire was posted to all the patients in a stamped addressed envelope and they were asked to complete and return the questionnaire. The data were entered onto a Microsoft Access database specifically created for recording endolaryngeal surgery. Analysis
Outcome measure
The primary outcome measure was voice. This was assessed using Voice Symptom Scale (VoiSS). The VoiSS questionnaire is a widely accepted and validated quality of life voice
Data were extracted from the MS Access database, and statistical analysis was carried out using Mini Tab 16. Wilcoxon’s rank sum test was used for analysis of nonparametric data.
Table 1. Table showing comparative analysis of demographs, aims of procedure and diagnoses in both complete and incomplete data sets
n Sex M F Mean age Mean preop VoiSS Aim of the procedure To exclude pathology To achieve biopsy with no deterioration in voice To improve voice quality To take biopsy of gross tumours To improve airway To resect tumours Others Total Diagnosis Malignant laryngeal pathologies Globus pharyngeus Carcinoma in situ/dysplasia Vocal cord nodules/cysts/polyps Papillomatosis Vocal cord paralysis Laryngitis Reinke’s oedema Functional dysphonia Others (webs, stenosis, stricture) Total
© 2015 John Wiley & Sons Ltd Clinical Otolaryngology 40, 580–586
Complete data set
Incomplete data set
703
331
351 352 57 40 n (%) 216(31) 185(26) 169(24) 77(11) 27(4) 21(3) 8(1) 703(100) n (%) 207(29) 99(14) 86(12) 83(12) 58(8) 41(6) 36(5) 33(5) 13(2) 47(7) 703(100)
165 166 58 43 n (%) 119(36) 60(18) 77(23) 51(15) 12(4) 7(2) 5(2) 331(100) n (%) 92(28) 59(18) 38(11.5) 38(11.5) 23(7) 11(3) 21(6) 23(7) 5(2) 21(6) 331(100)
582 S. Ansari & K. MacKenzie
Ethical consideration
As this study is an audit of current surgical clinical practice, specific ethical approval from the regional ethics committee was not required. The study was however discussed with the NHS Greater Glasgow and Clyde Clinical Effectiveness Group. Results and analysis
One thousand and sixty-six endolaryngeal surgical procedures were carried out in the seven-year study period (Fig. 1). Thirty-two were excluded. This was due to either patient demise in the postoperative period due to nonrelated cause or preoperative questionnaires not being adequately completed. As all patients attending preoperative assessment were asked to complete the VoiSS questionnaire, this was taken as an implied consent. If patients did not complete the form at preoperative assessment, they were excluded on the basis that they did not wish to participate. Seven hundred and three (68%) patients had complete clinical and surgical data with paired preoperative and postoperative VoiSS questionnaires. Comparative analysis of complete and incomplete data sets was carried out in Table 1. As both groups were not only similar in demographics but also in diagnosis and aim of the procedure categories, the groups were concluded to be comparable. It was decided not to proceed with an intention to treat analysis but to analyse the ‘complete data set’ cohort. The majority of endoscopic laryngeal surgery was performed to exclude pathology without taking laryngeal biopsies (31%) and consisted of those with previously treated head and neck cancer or those where there was a concerning group of symptoms. The next most frequent aim was to achieve a diagnosis by sampling the larynx but without having any impact on voice (26%). This group included patients with dysplasia, hyperkeratosis and low volume squamous disease. Approximately 24% of the endoscopic Total procedures 1066 May 2007 – May 2014
Excluded 32
Included 1034
- Deceased - No Pre-op data
(Implied consent)
Complete dataset 703 (68%)
Fig. 1. Total numbers of patients.
laryngeal workload was in the ‘to improve voice quality’ group which included benign pathologies such as vocal cord polyps, cysts, nodules and papillomatosis (Table 1). The preoperative and postoperative VoiSS scores in each of the seven groups are shown in Table 2. There was a statistically significant improvement in the overall median VoiSS scores for all seven hundred and three episodes (P < 0.005) (Table 2). In the ‘to improve voice quality’ group, there was a statistically significant improvement in subjective self-rated voice outcome. However, in patients where the aim was ‘to achieve biopsy without deterioration in their voice’, there was no statistically significant difference in their preoperative and postoperative scores, consistent with the aim of surgery (Table 2). The heterogeneity of our practice is evident from Table 3. The majority of the workload involved malignant pathologies, two hundred and seven (29%) with a further eighty-six (12%) have dysplasia or carcinoma in situ. This group with dysplasia or carcinoma in situ required a biopsy without any deterioration in voice. Other notable groups were benign laryngeal pathologies, with the aim to improve voice, such as vocal cord nodules/cysts/polyps, Reinke’s oedema and laryngeal papillomatosis. The preoperative and postoperative scores in each of these diagnostic groups are shown in Table 3. There was significant improvement in the VoiSS scores in patients with vocal cord nodules/cysts/polyps, papillomatosis, vocal cord palsy and Reinke’s oedema groups as anticipated. The scatter plot in Fig. 2 shows the individual differences in Voice Symptom Scale (VoiSS) scores pre- and postoperatively across all seven cohorts. Analysis of surgical aims
To exclude pathology (n = 216). This group consists of patients with globus pharyngeus and dysphonia undergoing endolaryngeal surgery to exclude a malignant process and also groups undergoing surveillance following treatment for head and neck cancer. Overall, there was no statistical improvement in VoiSS scores across the cohort. To achieve biopsy without deterioration in voice (n = 185). This group included patients with non-invasive malignant pathologies such as keratosis, dysplasia, carcinoma in situ and low volume squamous cell cancer (T1/T2). There was no significant change in the VoiSS scores in this subgroup. The aim of achieving a biopsy without a deterioration in voice was achieved. To improve voice group (n = 169). This cohort included benign laryngeal pathologies such as vocal cord cysts and © 2015 John Wiley & Sons Ltd Clinical Otolaryngology 40, 580–586
Voice outcomes of endolaryngeal surgery
583
Table 2. Table showing median and interquartile range of preoperative and postoperative VoiSS scores according to the aims of the procedure with their statistical analysis (P) using Wilcoxon’s signed rank test Preop VoiSS
Postop VoiSS
Aim
n
Median
IQR
Median
IQR
P
To exclude pathology To achieve biopsy with no deterioration in voice To improve voice quality To take biopsy of gross tumours To improve airway To resect tumours Others Total
216 185 169 77 27 21 8 703
35 38 51 26 42 21 18 39
30.75 38 30 36.5 38 27 35 35
32 34 35 30 39 19 25 33
37 46.5 36.5 45.5 38 32 32 40
0.584 0.943
