Keywords
quality of life - outcomes - cancer - dysphagia - swallowing - head and neck cancer
Introduction
Treatment for head and neck cancer (HNC) can have a major impact on the swallowing
function, and it is frequently identified by patients as a priority after treatment.[1]
[2]
[3]
[4] Nonsurgical approaches are favored for certain subsites of HNC due to the functional
morbidity associated with surgery combined with radiotherapy. However, the use of
radiotherapy and chemoradiotherapy can result in very significant deficits and poor
health related quality of life (HRQoL), with the potential for late treatment effects.
Early detection of patients with swallowing problems allows the opportunity for Speech
and Language Therapy (SLT) interventions which can aid adaption and promote better
outcomes. Patient Reported Outcome Measures (PROMS) are a standard means of identifying
the perspective of the patient and can help to screen patients for dysfunctional symptoms
that may delay or inhibit rehabilitation.[5] Clinical assessment of swallowing should always be performed at all stages of the
treatment pathway,[6] as the subjective deficits do not necessarily match the objective findings.[7]
[8] In order for PROMS and objective measures to be routinely used in clinical care,
they need to be straightforward and have minimum patient-clinical completion burden.
Several PROMs have been used to evaluate the perception of the ability to swallow,
and the wider impact on everyday life and socioemotional function and ‘quality of
life’.[9]
[10]
[11]
[12]
[13] No one questionnaire is a gold standard, and each has its strengths and weaknesses.
Assessing a patient using a battery of questionnaires can be time consuming to complete
and involve duplication of concepts and swallowing-related issues. The University
of Washington Head and Neck Quality of Life (UW-QoLv4)[14] is a commonly used HNC specific PROM. It can be used for screening, to assist in
the identification of those who report poor swallow function who could benefit from
additional intervention and support,[15] and can be included into routine practice.[16] Strong correlations have been found in previous studies with the M D Anderson Dysphagia
Inventory (MDADI), the Sydney Swallow Questionnaire (SSQ) and the Swallowing Quality
of Life Questionnaire (SWAL-QOL).[7]
[16] The Patients Concerns Inventory (PCI) is a prompt list that has been developed to
identify the concerns that patients would like to discuss during their consultation.[17] It helps focus the consultation onto patient needs and can be used as part of an
enhanced consultation package in combination with the UWQoL and promotes multidisciplinary
care. There is a randomized control trial evaluation currently taking place, looking
at the regular use of the UW-QoL and PCI in clinic consultations during the first
year, when the primary outcomes are a clinically meaningful and significant difference
in overall QoL, emotional dysfunction, and distress.
There are several clinician-rated scales that capture information about oral intake
and complement the PROMs. These include the Performance Status Scale for Head and
Neck Cancer (PSS-HN) Normalcy of Diet scale[18] and the Functional Oral Intake Scale (FOIS).[19]
Previous studies have found correlations between the PSS-HN and the MDADI,[20] the MD Anderson Symptom Inventory – Head and Neck (MDASI-HN)[12] and the FOIS,[21] suggesting strong relationships between patient reported swallow function, symptom
severity and oral intake.
Taste changes and dry mouth are commonly reported symptoms following (chemo-) radiotherapy
and are reported by patients as important issues following treatment.[1]
[2]
[21] They are strong drivers of oral intake and have been found to be significantly correlated
with the FOIS in long-term HNC survivors.[21]
The UW-QoL taste and saliva domains are widely used in this group[1]
[2]
[22] and the UW-QoL saliva domain has also been shown to be suitable as a screening tool
for dry mouth.[23]
Recent studies have found that PROMs are poorly aligned with Fiberoptic Endoscopic
Evaluation of Swallowing (FEES),[7]
[8] suggesting PROMs may not detect clinically significant dysphagia. Objective assessments,
such as FEES, are not always accessible, and cannot be easily incorporated into a
routine clinical assessment. The Water Swallow Test (WST) has been shown as a means
of quickly assessing swallowing performance and has been shown to be a useful indicator
of outcome.[24]
[25] It can identify patients with oropharyngeal dysphagia postsurgery[26] and aspiration post (chemo-) radiotherapy,[25] who may require intervention and instrumental assessments. It is widely used as
a clinical swallowing assessment tool in combination with objective assessment such
as videofluoroscopy, trismus measures, and PROMS including the MDADI and PSS-HN.[22]
[27]
[28]
[29]
[30]
[31]
[32] It has been found to be a pretreatment predictor of function at12 months and has
been shown to detect significant changes in function in patients under long-term follow
up.[33]
[34] Studies that have looked at correlations between WST and PROMS have found some correlations
with both the MDADI and PSS-HN.[8]
As summarized in the previous section, there is a large number of outcome measures
available to clinicians for use with the HNC population. The consideration is to which
combination of assessments to use routinely in a busy multidisciplinary team (MDT)
clinical setting. They need to be simple and quick to use to present a realistic means
to help focus resources, by having sufficient sensitivity to detect changes in function
following treatment, and to identify those patients who might benefit from additional
support and clinical intervention. Several studies have found strong correlations
between different PROMS to guide their use in the clinical setting. However, to date,
there have been no prospective studies looking at direct correlation over time between
the FOIS, PSS-HN, UW-QoL and WST. These measures have the potential to be widely integrated
into the clinical practice. The objective of the present study was to compare four
post-treatment outcome measures of swallowing function, over time, in a cohort of
HNC patients treated by (chemo-)radiotherapy.
Ethical Considerations
The measures used were part of the regular care in the follow-up of our HNC patients.
Data for the present study were collected as part of the routine clinical practice.
The present study was granted local audit approval but did not require formal IRAS
ethics approval.
Method
This was a prospective cohort study. Subjects were consecutively assessed at 3 months
and 12 months as part of normal care in an outpatient hospital setting during attendance
at a HNC MDT clinic. Subjects had unknown primary or primary squamous cell cancer
of the oropharynx, nasopharynx or hypopharynx stage T1-4, N0- 3, M0 disease.[35] Treatment was with (chemo-)radiotherapy, including induction, with curative intent.
Treatment schedules included (1) chemoradiotherapy (cisplatin 40 mg/m2 in 6 cycles)
combined with 60/30 Gy (2.1 Gy per fraction) over a 6-week period, (2) chemoradiotherapy
(cisplatin 40 mg/m2 in 6 cycles) combined with 65/30 Gy over a 6-week period (3) Cetuximab
combined with 65 /30 Gy over a 6-week period or (4) radiotherapy 60/30 Gy over a 6-week
period. Induction was with 3 cycles of TPF; Cisplatin, Docataxel and 5 FU. Percutaneous
Endoscopic Gastrostomy (PEG) tube was used as the method of supplementary feeding.
Tubes were placed prophylactically, and subjects were encouraged to maximize oral
intake throughout the treatment, as per current practice in our unit.
UW-QoL v4
The UWQoL questionnaire has 12 domains. The current study used the swallow, saliva,
and taste domains ([Table 1]).
Table 1
Outcome measures
|
UW-QoL swallow
|
|
100
|
I can swallow as well as ever
|
|
70
|
I cannot swallow certain foods
|
|
30
|
I can only swallow liquid food
|
|
0
|
I cannot swallow because it “goes down the wrong way” and chokes me
|
|
UW-QoL saliva
|
|
100
|
My saliva is of normal consistency
|
|
70
|
I have less saliva than normal, but it is enough
|
|
30
|
I have too little saliva
|
|
0
|
I have no saliva
|
|
UW-QoL taste
|
|
|
100
|
I can taste food normally
|
|
70
|
I can taste most foods normally
|
|
30
|
I can taste some food
|
|
0
|
I cannot taste any foods
|
|
PSS-HN Normalcy of Diet
|
|
100
|
Full diet (no restrictions)
|
|
90
|
Peanuts
|
|
80
|
All meat
|
|
70
|
Carrots, celery
|
|
60
|
Dry bread and crackers
|
|
50
|
Soft, chewable foods (e.g., macaroni, canned/soft fruits, cooked vegetables, fish,
hamburger, small pieces of meat)
|
|
40
|
Soft foods requiring no chewing (e.g., mashed potatoes, apple sauce, pudding)
|
|
30
|
Pureed foods (in blender)
|
|
20
|
Warm liquids
|
|
10
|
Cold liquids
|
|
0
|
Non-oral feeding (tube fed)
|
|
FOIS- Functional Oral Intake Scale
|
|
1
|
Nil By Mouth (NBM)
|
|
1
|
Tube dependent with minimal attempts of food/liquid
|
|
2
|
Tube dependent with consistent oral intake of food or liquid
|
|
3
|
Total oral diet of a single consistency
|
|
4
|
Total oral diet with multiple consistencies, but requiring special preparation or
compensations
|
|
5
|
Total oral diet with multiple consistencies without special preparation, but with
specific food limitations
|
|
6
|
Total oral diet with no restrictions
|
PSS-HN Normalcy of Diet
The diet texture restrictions of the patients were rated using the PSS-HN Normalcy
of Diet scale ([Table 1]).
FOIS
The degree of oral intake and supplementary feeding via tube was recorded using the
FOIS. ([Table 1])
100 mL Water Swallow Test
The WST involved patients being instructed to swallow 100 mL of water ‘as quickly
as is comfortably possible’[24] The number of swallows taken was counted simultaneously by the researcher, by feeling
the thyroid cartilage for laryngeal elevation. Time for swallowing was measured from
when the water first touched a patient's lips to when the larynx came to rest. Three
swallowing performance parameters were calculated (1) swallow volume (millilitres
per swallow = mL swallowed divided by number of swallows taken) (2) swallow capacity
(milliliters per second = mL swallowed divided by time taken) (3) swallow speed (time
per swallow = time taken divided by number of swallows). Patients who choked during
swallowing were asked to stop immediately, regardless of whether they had finished
drinking the water.
Analysis
The data were analyzed using IBM SPSS Statistics for Windows, Version 23 (IBM Corp.,
Armonk, NY, USA) using a range of statistics appropriate to the data type. The Spearman
Correlation coefficient (rs
) was used to explore the relationships between variables. Where data were missing,
cases were excluded from the data analysis.
Results
Sample
[Table 2] summarizes the disease, treatment patient characteristics for the study sample.
The cohort consisted of 49 patients, comprising 41 males and 8 females. The age range
was between 44 and 80 years old and the mean age was 59.9 years old. The most common
site of disease was the oropharynx and T stage was predominantly T1/T2. The majority
of patients were treated with chemoradiotherapy.
Table 2
Demographics
|
n = 49
|
|
|
Age (years old):
|
|
Minimum / Maximum
|
44
|
80
|
|
Range
|
36
|
|
|
Mean / standard deviation
|
59.9
|
8.69
|
|
No. of patients
|
%
|
|
Gender:
|
|
Male
|
41
|
84
|
|
Female
|
8
|
16
|
|
Site:
|
|
oropharynx
|
40
|
82
|
|
hypopharynx
|
5
|
10
|
|
nasopharynx
|
2
|
4
|
|
Cancer of unknown primary
|
2
|
4
|
|
T-stage:
|
|
T1
|
9
|
18
|
|
T2
|
27
|
55
|
|
T3
|
4
|
8
|
|
T4
|
7
|
14
|
|
Tx
|
2
|
4
|
|
Nodal Stage:
|
|
N0
|
6
|
12
|
|
N1
|
9
|
18
|
|
N2
|
8
|
16
|
|
N2a
|
4
|
8
|
|
N2b
|
20
|
41
|
|
N2c
|
1
|
2
|
|
N3
|
1
|
2
|
|
Treatment:
|
|
Cisplatin 60 Gy /30
|
4
|
8
|
|
Cisplatin 65 Gy /30
|
30
|
61
|
|
Cetuximab 65 Gy /30
|
9
|
18
|
|
Radiotherapy only
|
6
|
12
|
|
Induction
|
4
|
8
|
Data were collected on a total of 45 patients at 3 months post-treatment, although
this had decreased to a total of 20 patients at 12 months post-treatment. Details
of data completion are shown in [Table 3].
Table 3
Data completion
|
Completed
|
Completed
|
Drop-out
|
Drop-out
|
|
n
|
%
|
n
|
%
|
|
At 3 months:
|
|
|
|
|
|
3-month UW swallow
|
45
|
91.8%
|
4
|
8.2%
|
|
3-month UWtaste
|
45
|
91.8%
|
4
|
8.2%
|
|
3-month PSSHN (normalcy of diet)
|
45
|
91.8%
|
4
|
8.2%
|
|
3-month FOIS
|
45
|
91.8%
|
4
|
8.2%
|
|
3-month WST
|
33
|
67.3%
|
16
|
32.7%
|
|
At 12 Months:
|
|
|
|
|
|
12-month UWQoL swallow
|
20
|
40.8%
|
29
|
59.2%
|
|
12-month UWQoL taste
|
20
|
40.8%
|
29
|
59.2%
|
|
12-month PSS-HN (normalcy of diet)
|
20
|
40.8%
|
29
|
59.2%
|
|
12-month FOIS
|
20
|
40.8%
|
29
|
59.2%
|
|
12-month WST
|
17
|
34.7%
|
32
|
65.3%
|
Abbreviations: FOIS, Functional Oral Intake Scale; PSS-HN, Performance Status Scale
for Head and Neck Cancer; UWQoL, University of Washington Head and Neck Quality of
Life; WST, Water Swallow Test.
At 3 months, 75% of the patients reported impairment in swallowing, scoring < 100
on the UW-QoL swallow item, which reduced to 25% at 12 months. This was reflected
in changes in the use of the PEG tube with 21 patients using the tube at 3 months
(n = 45); this number dropped to 1 at 12 months (n = 20).
Correlations analysis found several relationships between measures. Cohen[36] suggested guidelines for grouping the strength of relationship between outcomes,
which can be used to interpret the coefficients.
PSS-HN and FOIS ([Table 4])
Table 4
Correlations
|
|
|
3m PSSHN
|
3m FOIS
|
3m UWQoL Saliva
|
3m UWQoL taste
|
12m PSSHN
|
12m FOIS
|
12m UWQoL saliva
|
12m UWQoL taste
|
|
Spearman rho
|
3-month PSS-HN
|
Correlation Coefficient
|
1.000
|
.579**
|
.090
|
.285
|
.761**
|
.657**
|
.194
|
.651**
|
|
|
n
|
45
|
45
|
44
|
45
|
17
|
17
|
17
|
17
|
|
3-month FOIS
|
Correlation Coefficient
|
.579**
|
1.000
|
.146
|
.310*
|
.416
|
.464
|
.331
|
.519*
|
|
|
n
|
45
|
45
|
44
|
45
|
17
|
17
|
17
|
17
|
|
3-month UWQoL Saliva
|
Correlation Coefficient
|
.090
|
.146
|
1.000
|
.226
|
.159
|
.026
|
.633**
|
.340
|
|
|
n
|
44
|
44
|
44
|
44
|
17
|
17
|
17
|
17
|
|
3-month UWQoL taste
|
Correlation Coefficient
|
.285
|
.310*
|
.226
|
1.000
|
.123
|
.013
|
-.191
|
.401
|
|
|
n
|
45
|
45
|
44
|
45
|
17
|
17
|
17
|
17
|
|
12-month PSS-HN
|
Correlation
|
.761**
|
.416
|
.159
|
.123
|
1.000
|
.823**
|
.375
|
.588**
|
|
|
n
|
17
|
17
|
17
|
17
|
20
|
20
|
20
|
20
|
|
12-month FOIS
|
Correlation
|
.657**
|
.464
|
.026
|
.013
|
.823**
|
1.000
|
.325
|
.496*
|
|
|
n
|
17
|
17
|
17
|
17
|
20
|
20
|
20
|
20
|
|
12-month UWQoL saliva
|
Correlation
|
.194
|
.331
|
.633**
|
-.191
|
.375
|
.325
|
1.000
|
.423
|
|
|
n
|
17
|
17
|
17
|
17
|
20
|
20
|
20
|
20
|
|
12-month UWQoL taste
|
Correlation
|
.651**
|
.519*
|
.340
|
.401
|
.588**
|
.496*
|
.423
|
1.000
|
|
|
n
|
17
|
17
|
17
|
17
|
20
|
20
|
20
|
20
|
Abbreviations: FOIS, Functional Oral Intake Scale; PSS-HN, Performance Status Scale
for Head and Neck Cancer; UWQoL, University of Washington Head and Neck Quality of
Life.
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
This analysis showed that the 3-month PSS-HN was significantly correlated with both
the 12-month PSS-HN (rs
= 0.761, n = 17) and the 12-month FOIS (rs
= 0.657, n = 17). The 3-month FOIS was significantly correlated with the 3-month PSS-HN (rs
= 0.579, n = 45, [Fig. 1]). Finally, the 12-month PSS-HN was significantly correlated with the 12-month FOIS
(rs
= 0.823, n = 20, [Fig. 2]).
Fig. 1 Correlation between 3 month Functional Oral Intake Scale and 3 month Performance
Status Scale for Head and Neck Cancer.
Fig. 2 Correlation between 12 month Functional Oral Intake Scale and 12 month Performance
Status Scale for Head and Neck Cancer.
UW-QoL Taste and UW-QoL Saliva Domain with PSS-HN and FOIS ([Table 4])
This analysis showed that the 12-month UW-QoL taste score was significantly correlated
with both the 3-month PSS-HN (rs
= 0.651, n = 17) and the 3-month FOIS (rs
= 0.519, n = 17). The 12-month UW-QoL taste was also significantly correlated with the 12-month PSS-HN
(rs
= 0.588, n = 20) and with the 12-month FOIS (rs
= 0.496, n = 20).
The UW-QoL saliva scores were not significantly correlated with 3-month or 12-month
PSS-HN and FOIS. However, the 3-month saliva score was significantly correlated with
the 12-month saliva score (rs
= 0.633, n = 17). Eight patients reported a slight improvement in saliva scores at 12 months, with
the remainder reporting no change.
WST and UW-QoL Swallowing Domain ([Table 4])
Correlations analyses were used to identify any relationships between the WST clinical
swallowing measures and the UW-QoL swallowing domain, at 3 months and 12 months post-treatment.
The strongest relationship was found at 12 months post-treatment between the three
WST measures and UW-QoL swallowing, as follows: 12-month UW-QoL swallow with 12-month
WST Capacity (rs
= 0.759, n = 17), 12-month UW-QoL swallow with 12-month WST Volume (rs
= 0.591, n = 17) and 12-month UW-QoL swallow with 12-month WST Speed (rs
= 0.588, n = 17). The correlations analysis suggested a relationship between the WST and UW-QoL
reported swallowing outcomes at 12 months post-treatment, so this was explored further
using Multivariate General Linear Model analysis using UW-QoL swallow as the predictor
variable and the 3 clinical WST swallowing measures as dependent variables. The 3-month
model was, as expected, nonsignificant. The 12-month data produced a better model.
However, a Stepwise model retained only one variable, and that was WST Capacity (R2
= 0.40; F = 10.203; p < 0.01). The authors here add a caveat that the sample size is obviously too small
for any robust predictive analysis, so these results should be treated with caution.
Discussion
We found several strong relationships between our PSS-HN, FOIS, UW-QoL and WST data
in the 12 months following treatment. To the best of our knowledge, this is the only
study that has compared the four measures, in a prospective, longitudinal study, across
more than one time point.
Our results show, for the first time, that the PSS-HN and FOIS have significant correlations
across two time points. We showed that in 3-month PSS-HN scores relating to 12-month
PSS-HN scores, there was also a relationship between the two measures. This is similar
to other studies which have reported 3-month PSS-HN scores which remained low at 12
months,[20]
[33] and to other PROMS studies which have shown pre-treatment PSS-HN can predict function
at 12 months, and have also shown very little change in MDADI and UW-QoL between 3
and 12 months.[2]
[33] Improvements in PSS-HN scores between 3 and 12 months have been reported in the
literature; however, this was in a specific group of patients given parotid sparing
intensity modulated radiotherapy (IMRT), which differs from the cohort in the present
study.[22] Clinicians need to select which measure to use, based on which information they
want. The PSS-HN provides information about diet texture, whereas the FOIS records
the degree of oral intake and supplementary feeding via tube. We suggest that the
use of either provides valuable information and can give an idea of expected changes
in function over time.
Taste is a common side effect reported by patients following treatment and can affect
their diet choices.[1]
[21] Our results showed that 12-month UW-QoL taste was significantly correlated with
both the 3-month PSS-HN and the 3-month FOIS. The 12-month UW-QoL taste was also significantly
correlated with the 12-month PSS-HN and the 12-month FOIS. These results are similar
to long-term studies that have found correlations with the MDASI-HN and FOIS,[21] and suggests that perceived impairment of taste is an important measure. We recommend
the use of the UW-QoL taste domain as a valuable tool to guide intervention. It can
enhance the discussion between the clinician and patient about expected change, and
tailor rehabilitation to optimize oral intake post-treatment.
In addition, our data suggest, for the first time, a relationship between WST and
the UW-QoL swallow domain. The WST was found to have some predictive power in terms
of PROMs, with WST capacity being the best potential predictor of UW-QoL swallow score.
This is similar to other studies that showed that WST capacity is sensitive to change[24] and found that it had a moderate relationship with the MDADI.[8] Our results were insufficient to propose the UW-QoL swallow as a stand-alone screening
tool in this setting. However, this suggests that assessment of the perception of
the patients of their swallow function may have the potential to help to highlight
patients who may require intervention.
Dry mouth is frequently reported as an issue post-treatment.[1]
[2] A relationship was found between 3-month and 12-month UW-QoL saliva scores, with
some variability across the cohort. This may be explained by the size and heterogeneous
nature of the sample in relation to site and treatment schedule, and further studies
with a larger cohort would be valuable.
Limitations of the Study
The data have many limitations. Data collection was performed prospectively, thus
limiting recall bias; however, the relatively small sample restricted the scope for
the statistical analyses. A multivariate analysis was attempted using UW-QoL swallowing
as the dependent variable, but the sample size was an issue for the robustness of
this test. As part of the current study, data were collected without additional funding,
as part of routine clinical follow-up appointments. This was dependent on availability
of staff, and patient attendance. These factors, in addition to disease-specific and
all-cause mortality, account for the missing data.
The target for data collection was 40 patients with no a priori statistic power calculation.
This resulted in a small sample size, restricted range of responses, and/or poor distribution
of responses to some questions. As a consequence, the results should be interpreted
with caution.
The current study used the WST as a clinical assessment tool, as it was quick and
easy to use in a routine clinical setting. Further research incorporating the instrumental
assessment of swallowing would be valuable.
The sample in the present study was heterogeneous in terms of tumor size, location
and treatment schedule. While the sample was very relevant in terms of the context
of this study, it would be valuable to collect more data to support PROMs that can
be applied to specific patient groups.
Conclusion
Several measures were found to have clinical significance, and 3-month outcomes gave
an indication of performance at 1 year. This supports the use of relatively simple
assessments as a realistic way to collect outcome data in a clinic setting. The UW-QoL
swallow item and the WST are easy to incorporate into routine care for all HNC patients
and should be used as standard post-treatment outcome measures of the swallowing function,
as part of the routine clinical assessment. Clinicians can select additional questionnaires,
which assess taste and diet, to enhance the clinical interaction and tailor rehabilitation.
Patients assessed as having poor swallow performance on the WST can be referred for
further instrumental assessments such as FEES and video fluoroscopy to help inform
and guide the appropriate intervention. Both the UW-QoL swallow question and the WST
can be performed in the out-patient consultation setting, and the ease of assessment
means that they can be performed by different members of the head and neck team, and
not limited to SLT. These measures can serve to help screen patients for dysfunction
and focus allocation of resources for those who would benefit from more comprehensive
assessment and intervention by SLT. Further research is needed to help underpin the
robustness of basing routine screening on the UW-QoL swallow question and the WST
and to establish their role in repeated longitudinal assessment to monitor the impact
of intervention strategies.
