Health-related quality of life and voice following radiotherapy for laryngeal cancer – a comparison between glottic and supraglottic tumours

Abstract

Background. To provide further information about short-term effects on voice quality and health-related quality of life (HRQL) following radiotherapy for laryngeal cancer, comparing glottic and supraglottic tumours.

Material and methods. Sixty-seven male patients who received radiotherapy for glottic and supraglottic tumours were included. Voice recordings were performed and patient reported outcome questionnaires filled in before radiotherapy and one-month post-completion. Acoustic and temporal measures were produced from the recordings and compared to an age-matched, vocally healthy control group (n = 23).

Results. Deterioration of HRQL post-radiotherapy was noted for both the glottic and supraglottic cohort. Pre- radiotherapy the glottic cohort demonstrated acoustic and temporal measures that were statistically different from healthy controls, with improvements post-radiotherapy where several measures showed no statistically significant differences compared to healthy controls. The supraglottic cohort revealed voice measures comparable to the healthy controls but demonstrated statistically significant impairments in HRQL compared to the glottic cohort following radiotherapy.

Conclusion. Patients with supraglottic tumours may experience more eating and swallowing related problems prior to radiotherapy compared to glottic tumours and demonstrate a more significant reduction in HRQL following treatment. However, in terms of voice measurements, subjective measures correlate poorly with objective acoustic parameters, thereby emphasising the importance of patient-reported outcomes as part of treatment evaluation.

Radiotherapy is in a majority of cases the primary treatment for early laryngeal cancer, and good survival outcome has been reported [1]. Focus is therefore shifting to other effect measures of treatment, which include voice quality and health-related quality of life (HRQL) [2]. Patient-reported outcome (PRO) is now one of several important factors when considering choice of treatment and is becoming increasingly recognised with both cancer-specific and site-specific HRQL instruments in widespread use [3–7].

Several studies have investigated effects on voice following radiotherapy for laryngeal cancer. However, the outcomes are inconclusive as studies are small with varying results. For instance, Bibby et al. showed in their study including 30 patients with early glottic cancer that the voice improved significantly post- radiotherapy in terms of jitter, shimmer, harmonics-to-noise ratio (HNR) and maximum phonation time (MPT) [8]. Additionally, Niedzielska et al. investigated 45 patients with early laryngeal cancer who demonstrated improvement post-radiotherapy regarding jitter and shimmer but the values remained outwith normal [9]. In general, the voice deteriorates directly following radiotherapy to later improve over the course of the first year [10].

Despite the results in previously mentioned studies, only a few studies with large enough sample sizes exist from which conclusions regarding voice outcome and HRQL can be drawn. Results are mainly reported for glottic cancers only and rarely subdivide to include supraglottic tumours as a separate entity. Therefore, larger studies are needed addressing the mentioned issues. This study aims to investigate the short-term effects on voice quality and HRQL following radiotherapy in laryngeal cancer, comparing glottic and supraglottic tumours.

Material

All patients with suspected laryngeal cancer in the western part of Sweden are discussed at a weekly tumour conference at the Otorhinolaryngology Department at Sahlgrenska University Hospital, where diagnosis is established and treatment decided upon. All male patients diagnosed with glottic and supraglottic tumours from March 2000 to December 2011 (including a two-year interruption) that were to receive curatively intended radiotherapy ± chemotherapy were asked to participate in the study. The reason for only including male patients in the analyses in this study was that acoustic values vary greatly between male and female voices as well as women with laryngeal cancer being few. Inclusion criteria were sufficient knowledge of the Swedish language to independently answer the questionnaires as well as absence of severe cognitive impairment (e.g. dementia). A total of 139 male patients were eligible and asked to participate, of which 75 agreed. During the course of the study seven patients discontinued their participation and one received a tracheostomy leading to exclusion. In total, 67 patients were included in the analyses. Comorbidity was measured using the Adult Comorbidity Index Evaluation-27 (ACE-27) and recorded for all patients with laryngeal cancer during the study period, i.e. for both the study cohort as well as the non-participants [11,12].

A healthy control group comprising of 23 age-matched male volunteers were recruited from relatives of patients or visitors at Sahlgrenska University Hospital. The controls did not perceive any voice problems and showed normal vocal fold status when examined by an otolaryngology specialist. Voice and HRQL data were recorded in the same manners as for the study group. However, only 20 of the 23 healthy controls completed the questionnaires.

Method

Tumour classification and treatment

Tumours were classified according to global standard, the TNM-staging system by the International Union Against Cancer (UICC) involving tumour size (T), regional lymph nodes involvement (N) and distant meastasis (M) [13]. The tumours were also classified according to anatomical localisation; the glottis involves the vocal cords, and anterior and posterior commisures. The supraglottic larynx consists of the epiglottis, false vocal cords, ventricles, arytenoids and aryepiglottic folds.

All study patients received radiotherapy as definitive treatment. Nineteen patients received radiotherapy twice daily and 48 patients received radiotherapy once daily. All patients with glottic Tis-T1 disease received the latter regimen. Patients with supraglottic tumours or T2-T4 glottic tumours received radiotherapy either once or twice daily with additional lymph node irradiation. The total radiation dose for all patients ranged from 62.4 Gy to 68 Gy. Three patients with supraglottic tumours received induction chemotherapy.

Voice recording

Voice recordings were done twice: prior to radiotherapy and approximately one month following its completion. Recordings included reading of a standard passage and the maximum sustained vowel /a/ repeated three times. A headset microphone (Sennheiser MKE 2-p) was set at a distance of 12 cm from the corner of the mouth. Recordings were made at a sampling frequency of 44, 1 kHz with a Panasonic Professional Digital Audio Tape (DAT) Recorder SV-3800. Prior to analysis, all recordings were transferred from a DAT to a computer hard drive as an audio file (.wav) using the program Swell Soundfile Editor, version 4.5 (Saven Hitech).

Acoustic analysis

Voices were analysed using Voxalys, a plugin programme to Praat (Praat: Doing phonetics by computer). Acoustic analyses were made partly in accordance with recommendations made by the European Laryngological Society (ELS) [14]. Jitter, shimmer and harmonics-to-noise ratio (HNR) values were analysed from two seconds of the middle of the second sustained vowel /a/. Mean speaking fundamental frequency (F0) indicates the pitch of the voice, i.e. high-low, and was measured from the reading of the standard passage. Maximum phonation time (MPT) was measured as the longest time in seconds that the patients could sustain the vowel /a/ in one exhalation. In healthy older men MPT should be approximately 14.6 seconds [15]. The perturbation measures jitter and shimmer reflects irregularity in the waveform in frequency and amplitude, high values are considered pathological. HNR reflects harmonicity related to noise in the voice signal, a low value represents a pathological voice.

Patient-reported outcome

During each recording session, patients filled in questions about smoking habits and sociodemographic data as well as the following PRO instruments: Swedish Self-Evaluation of Communication Experiences after Laryngeal Cancer questionnaire (S-SECEL) [4,6,16,17], the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30) [18] and the EORTC QLQ specific for Head and Neck Cancer (EORTC QLQ-H&N35) [19].

Swedish Self-Evaluation of Communication Experiences after Laryngeal Cancer

The original SECEL, a self-administered questionnaire developed to assess communication dysfunction in patients with laryngectomies, has demonstrated satisfactory psychometric properties [16]. The Swedish version (S-SECEL) was adapted for use in patients receiving different treatments for laryngeal cancer. S-SECEL is an instrument consisting of 35 items covering communication experiences and dysfunction in laryngeal cancer patients. It has proved reliable and shown both convergent and discriminant validity and satisfactory internal consistency [4,17]. The first 34 items are divided into three domains. The General domain (5 items) focuses on the patient's general attitudes regarding acknowledgement of the illness as well as being relaxed and calm. The Environmental domain (14 items) describes how the patients experience their voice in different environments. The Attitudinal domain (15 items) focuses on speech attitudes. Each item is rated on a four-point categorical scale ranging from 0 (never) to 3 (always) and recalls the last 30 days. Scoring of domains and a total score is carried out by simple addition. Thus, the summary domain scores range from 0–15 for General, 0–42 for Environmental, 0–45 for Attitudinal and 0–102 for Total. A higher score indicates greater perceived communicative dysfunction [17].

EORTC QLQ C30 and EORTC QLQ-H&N35

The EORTC QLQ-C30 is a cancer-specific questionnaire that evaluates HRQL in cancer patients [18]. It consists of five function scales, a global quality of life scale, three symptom scales and six single items, totalling 30 questions that describe the patients’ symptoms and functional level during the last week. To address additional symptoms associated specifically with head and neck (H&N) cancer and its treatment, a complementary 35-item module can be used, the EORTC QLQ-H&N35 [18]. Calculated scale scores range from 0 to 100. On the functioning scales and global quality of life scales, a score of 100 represents maximum functioning, whereas on the symptom scales and single items a score of 100 equates to worst possible symptoms. For the EORTC QLQ-C30, a change of 10 points is considered a clinically significant change [20]. This value is frequently used when interpreting results in the H&N35 as well. Prior to starting the study, it was hypothesised which scales would be affected following radiotherapy as well as possibly indicating differences between the glottic and supraglottic cohort. These selected scales are presented in Table III.

Statistical analysis

The SPSS version 20.0 for Mac was used in the analysis. Non-parametric tests were used throughout due to the nature of the data. Significance level was set to < 0.05. Paired intra-group comparisons of values before and after radiotherapy were carried out with Wilcoxon signed rank test. Inter-group comparisons between the glottic or the supraglottic and the healthy controls were performed with the Mann Whitney U-test. Differences in baseline characteristics between the patients with glottic and supraglottic tumours were calculated using χ² and Mantel-Haenzel test for ordered categorical data.

Ethical aspects

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Regional Ethical Review Board in Gothenburg, Sweden. Before inclusion, all participants gave their informed consent.

Results

There were no statistically significant differences between the patients included in the study and the patients who declined participation with regard to comorbidity according to the ACE-27. Patient characteristics of the included patients divided into two groups, glottic and supraglottic tumours, are presented in Table I. The supraglottic tumours were statistically significantly larger in size and required a feeding tube to a greater extent during radiotherapy than the glottic tumours.

Table I. Patient characteristics subdivided for glottic and supraglottic tumours.

n	Glottic 54	Supraglottic 13	p-value
Mean age (SD)	64.4 (10.9) n (%)	65.6 (13.9) n (%)	0.515
Tumour size
Tis	2 (4)	0	< 0.001
T1	38 (70)	3 (23)
T2	13 (24)	4 (31)
T3	1 (2)	5 (38)
T4	0	1 (8)
Comorbidity (ACE-27)
None	24 (44.5)	4 (31)	0.491
Mild	19 (35)	6 (46)
Moderate	11 (20.5)	3 (23)
Severe	0	0
Smoking habits
Smoking	21 (39)	4 (31)	0.334
Quit smoking > 12 months ago	27 (50)	8 (61)
Non-smoker	6 (11)	1 (8)
Respiratory disease
COPD	4 (7.5)	2 (15)	0.660
Other respiratory disease	4 (7.5)	1 (8)
None	46 (85)	10 (77)
Feeding tube needed during and/or post RT
Yes	7 (13)	8 (62)	0.001
No	47 (87)	5 (38)

ACE-27, Adult Comorbidity Evaluation; COPD, chronic obstructive pulmonary disease; RT, radiotherapy.

Acoustic and temporal measures

Acoustic and temporal measures are presented in Table II. There were no statistically significant differences between the supraglottic and glottic tumour cohorts between the two occasions. The glottic cohort showed statistically significantly inferior baseline values in all acoustic and temporal parameters pre-radiotherapy compared to the healthy controls, yet post-radiotherapy only for MPT and shimmer.

Table II. Voice analysis and S-SECEL results pre- and post-radiotherapy for the patients with glottic and supraglottic tumours. The results are presented as mean (SD) p-values for comparisons with healthy controls.

	Glottic n = 54				Supraglottic n = 13				p-value ∆ between groups pre-RT	p-value ∆ between groups post-RT	Healthy controls n = 23
	Pre-RT	Post-RT	∆ within group pre-post RT	p-value within group	Pre-RT	Post-RT	∆ within group pre-post RT	p-value within group
Temporal
MPT (s)	11.1 (6.8) p = 0.001	12.0 (8.7) p = 0.005	1.0 (7.8)	0.668	16.1 (8.5) p = 0.771	16.8 (9.0) p = 0.913	0.6 (8.8)	0.966	0.058	0.111	16.9 (5.5)
Acoustic
Jitter	1.08 (1.04) p = 0.020	1.07 (1.53) p = 0.067	= 0.02 (1.83)	0.446	0.58 (0.25) p = 0.526	0.74 (0.36) p = 0.071	0.15 (0.31)	0.175	0.189	0.662	0.67 (0.78)
Shimmer	0.64 (0.40) p < 0.001	0.53 (0.33) p = 0.001	= 0.10 (0.48)	0.278	0.49 (0.26) p = 0.057	0.49 (0.25) p = 0.012	0.03 (0.26)	0.683	0.303	0.993	0.32 (0.18)
HNR	15.6 (6.4) p = 0.004	17.3 (5.4) p = 0.060	1.7 (7.6)	0.110	17.3 (4.2) p = 0.115	18.0 (4.4) p = 0.308	0.6 (3.9)	0.790	0.528	0.726	20.2 (5.0)
F0 mean	134.2 (26.2) p = 0.011	123.6 (27.2) p = 0.552	10.7 (26.0)	0.011	128.1 (25.0) p = 0.327	121.2 (23.6) p = 0.744	5.8 (32.7)	0.520	0.415	0.875	118.0 (19.3)
S-SECEL											n = 20
General	5.3 (2.8) p = 0.018	4.8 (2.6) p = 0.069	= 0.4 (2.2)	0.117	6.2 (1.9) p = 0.001	5.0 (2.1) p = 0.084	= 1.2 (3.3)	0.231	0.197	0.757	3.5 (2.2)
Environmental	14.3 (8.6) p < 0.001	13.8 (8.3) p < 0.001	= 0.4 (7.4)	0.510	15.4 (9.9) p < 0.001	12.0 (8.1) p = 0.003	= 3.4 (11.9)	0.216	0.729	0.567	3.5 (4.2)
Attitudinal	6.2 (5.9) p < 0.001	6.5 (5.9) p < 0.001	0.3 (5.7)	0.940	8.0 (6.5) p < 0.001	8.1 (7.8) p = 0.010	0.1 (10.0)	0.826	0.340	0.751	1.2 (1.9)
Total	25.8 (15.4) p < 0.001	25.1 (15.0) p < 0.001	= 0.5 (13.3)	0.566	29.5 (15.7) p < 0.001	25.1 (15.9) p = 0.003	= 4.5 (23.0)	0.380	0.439	0.981	7.4 (5.6)

F0, fundamental frequency; HNR, harmonics-to-noise ratio; MPT, maximum phonation time; RT, radiotherapy; SD, standard deviation; S-SECEL, Swedish Self-Evaluation of Communication Experiences after Laryngeal Cancer questionnaire. Maximum scores for the S-SECEL subscales: General: 15. Environmental: 42. Attitudinal: 45. Total: 102. A higher value indicates greater perceived communicative dysfunction.

The supraglottic tumours had values within normal both prior to and following radiotherapy for all acoustic and temporal measures. Fundamental frequency was also within normal range for the two groups both before and after radiotherapy. There was however, a significant decrease in mean fundamental frequency for the glottic subgroup.

Patient-reported outcome

Data from S-SECEL are presented in Table II. Both the glottic and the supraglottic cohort presented with statistically significant higher values (less communicative ability) than the healthy controls pre- radiotherapy.

Data from the EORTC QLQ-C30 and QLQ-H&N35 are presented in Table III. Pre-radiotherapy the supraglottic cohort showed inferior results regarding the physical, cognitive and social eating scales compared to the glottic cohort. Post-radiotherapy, statistically significant differences were observed in several EORTC QLQ-C30 scales, where patients with supraglottic tumours reported greater dysfunction than the glottic tumours. Patients with supraglottic tumours reported statistically and clinically significant deteriorations in social functioning scales and the symptom scales appetite loss and nausea/vomiting. The glottic cohort showed statistically, but not clinically, significant deteriorations regarding the physical and social functioning scales as well as the symptom scales nausea/vomiting and appetite loss.

Table III. EORTC QLQ-C30 and H&N35 pre- and post-radiotherapy for the patients with glottic and supraglottic tumours. The results are presented as mean 95% CI.

	Glottic n = 54				Supraglottic n = 13				p-value ∆ between groups pre-RT	p-value ∆ between groups post-RT	p-value difference between ∆
	Pre-RT	Post-RT	∆ within group	p-value ∆ within group	Pre-RT	Post-RT	∆ within group	p-value ∆ within group
EORTC QLQ C-30
Function scales
Physical	89.9 86.0; 93.8	82.5 77.8; 87.3	= 7.4 = 11.7; = 3.1	< 0.001	75.9 63.6; 88.2	71.8 64.2; 79.4	= 4.1 = 14.2; 6.0	0.421	0.009	0.017	0.920
Role	83.0 77.7; 88.4	75.8 69.6; 82.0	= 6.9 = 13.5; = 0.3	0.084	79.5 61.5; 97.5	69.2 55.7; 82.8	= 10.3 = 23.0; 2.4	0.072*	0.867	0.374	0.152
Social	87.0 82.4; 91.7	77.4 71.0; 83.7	= 9.4 = 15.6; = 3.2	0.001	83.3 71.7; 95.0	62.5 48.9; 76.1	= 23.6 = 41.4; 5.9	0.016*	0.450	0.029	0.097
Emotional	76.1 70.5; 81.6	79.6 74.5; 84.6	3.8 = 1.2; 8.8	0.069	76.9 65.1; 88.8	76.4 62.0; 90.8	= 1.4 = 5.8; 3.0	0.813	0.815	0.745	0.346
Cognitive	86.4 81.8; 91.0	82.4 77.1; 87.7	= 4.4 = 10.2; 1.4	0.248	67.9 54.6; 81.2	62.5 44.4; 80.6	= 5.6 = 21.4; 10.3	0.555	0.004	0.017	0.942
Global QOL	71.0 66.6; 75.4	66.4 61.8; 70.9	= 4.4 = 9.7; 0.9	0.167	68.6 52.8; 84.4	63.5 52.5; 74.4	= 5.1 = 20.3; 10.1	0.357	0.860	0.863	0.609
Symptom scales
Nausea/vomiting	2.8 0.7; 4.9	8.5 4.5; 12.5	5.7 1.7; 9.5	0.004	5.1 = 3.5; 13.7	32.1 18.8; 45.4	26.9 10.0; 43.6	0.006*	0.880	< 0.0001	< 0.001
Single item
Appetite loss	8.0 3.1; 13.0	18.2 9.9; 26.6	10.1 1.5; 18.6	0.01*	15.4 = 2.3; 33.1	41.0 16.1; 65.9	25.6 = 3.0; 54.3	0.141*	0.304	0.041	0.142
EORTC QLQ H&N-35
Speech	36.0 30.1; 41.9	32.5 26.4; 38.6	= 3.4 = 10.8; 4.1	0.193	39.3 28.1; 50.5	45.3 31.2; 59.4	6.0 = 8.2; 20.1	0.391	0.576	0.070	0.121
Swallowing	8.1 2.7; 13.5	12.3 6.8; 17.8	4.1 = 2.4; 10.7	0.110	21.8 1.2; 42.4	37.2 21.3; 53.1	15.4 = 12.9; 43.7	0.457	0.137	< 0.001	0.319
Social eating	8.2 2.9; 13.4	10.4 5.0; 15.8	2.0 = 4.3; 8.4	0.493	28.8 10.3; 47.4	32.7 12.5; 52.9	3.9 = 24.5; 32.2	0.734	0.002	0.031	0.890

*Clinically relevant change. CI, confidence interval; EORTC QLQ-C30, the European Organisation for Research and Treatment Quality of Life Questionnaire Core-30; EORTC QLQ-H&N35, the Head and Neck Cancer module; RT, radiotherapy. Higher scores on the function scales indicate a better function. A higher score on the symptom scale, single item, as well as the H&N35 indicates worse symptoms.

In the EORTC QLQ-H&N35, the inter-group differences were statistically significant for the social eating and swallowing scales post-radiotherapy where patients with supraglottic tumours presented with inferior function.

Discussion

Early stage laryngeal cancer is frequently treated using radiotherapy. This study highlights voice quality, communicative ability and HRQL in a glottic and supraglottic cohort. All patients reported a declining trend in HRQL measures following radiotherapy, although not all statistically significant. Statistically and clinically significant differences were found between the supraglottic and the glottic cohort post-radiotherapy, where the supraglottic cohort presented with impaired HRQL for the majority of scales measured compared to the glottic cohort. Regarding most of the acoustic measures, post-radiotherapy both cohorts showed values comparable to the voices of age-matched healthy controls.

S-SECEL was used to measure communicative dysfunction. Although both groups displayed improving trends, there were no significant inter-group differences pre-radiotherapy or following treatment nor were any significant differences within the cohorts recorded. However, a comparison between the healthy control group and the study patients revealed that both cancer cohorts presented with inferior scores compared to the healthy control group, where the difference was statistically significant. The environmental subscale in particular seems to highlight the vocal challenges that laryngeal cancer patients experience; mean scores range from 12 to 15.4 compared to 3.0 for the healthy control group. It implies that the S-SECEL, adapted for patients treated for laryngeal cancer, covers relevant aspects important for communication to this cohort, which is in accordance with the study by Johansson et al. [6]. Moreover, the results are similar to other studies using the Voice Handicap Index (VHI) questionnaire, which generally imply impairment following radiotherapy for both early and advanced laryngeal cancer [2,3].

In the acoustic measurements, only the fundamental frequency in the glottic cohort yielded a statistically significant improvement. However, all patients improved slightly regarding acoustic and temporal measurements following treatment. Glottic tumours were found to have consistently inferior acoustic and temporal measures compared to healthy controls pre-radiotherapy and improved post-radiotherapy for several measures. The supraglottic cohort presented with values comparable to the healthy controls at both time-point measurements. This could be explained by the localisation of the tumour itself, where a lesion located on the vocal fold would affect the voice quality more than lesions above the vocal folds. A study by Roh et al. [21] reports similar findings, where patients with supraglottic tumours presented with near-normal values post-laser surgery, with values not changing significantly pre- and post-surgery.

The supraglottic cohort in this study showed both clinically and statistically significant deteriorations in the EORTC QLQ-C30 social and role domain compared to baseline. This may be explained by more thorough examination of the symptom scales, where the supraglottic tumours report more nausea and vomiting, increasing appetite loss and dysphagia than the glottic cohort. These factors could impact on participation in daily life, reflected in the role and social domains. Moreover, supraglottic tumours were larger and patients did to a greater extent require a feeding tube during radiotherapy. These findings are supported by Langius et al. who found that malnutrition during radiotherapy for T1-T2 laryngeal cancer was associated with supraglottic as well as T2 tumours [22]. Also, having received an enteral feeding tube is expected to negatively affect HRQL following radiotherapy.

However, to be able to draw any conclusions regarding whether the effect on HRQL is due to tumour localisation or tumour size, further studies are needed, involving larger cohorts of patients with tumours in each T-classification as well as each localisation.

An important aspect to discuss is the timing of the measurement points with the second measurement being made closely after the end of radiotherapy. In this phase patients are still suffering from acute side effects from the therapy, affecting among other things voice and swallowing, with expected influence on HRQL. Some of these effects are however temporary, why further conclusions require a long-term follow-up.

The strengths of this study lie in its relatively large sample size as well as its prospective design. It is, to our knowledge, the only prospective study that combines general HRQL, self-perceived communicative dysfunction and objective acoustic measures when comparing glottic and supraglottic laryngeal cancer following radiotherapy. Similar studies do exist but are either retrospective with baseline measurements missing [23–26] or, if present, do not separate data according to tumour localisation [2,5,7]. Moreover, the vast majority reports only on glottic tumours. Oridate et al. [26] did subdivide to include a supraglottic cohort but had no baseline data, making conclusions difficult to draw.

A limitation to this study could be that only male patients were included, due to the acoustic analysis and the small number of women in the laryngeal cancer population. This would mean that the results only apply to the male patients. Therefore, future studies are required to address application and analysis of results for female voices.

Conclusion

Patients with supraglottic tumours experience to a greater extent a significant impairment in HRQL following oncologic treatment compared to glottic tumours. However, this is not reflected in communication dysfunction or objective acoustic measures, where glottic tumours tend to present with inferior voice quality, probably due to the localisation of the tumour.

Acknowledgement

This study was supported by the Swedish Cancer Society, Lions Cancer Foundation West, Assar Gabrielsson Foundation, the Laryng Foundation, the Research and Development Council (FoU) and the Sahlgrenska University Hospital Foundation. We would like to thank the dedicated speech-language pathologists in the Västra Götaland County for helping in collecting the data and Carina Åberg for her help in including the patients.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.