Oral Oncology 115 (2021) 105189

Contents lists available at ScienceDirect

Oral Oncology
journal homepage: www.elsevier.com/locate/oraloncology

M. D. Anderson symptom inventory head neck (MDASI-HN) questionnaire:

Italian language psychometric validation in head and neck cancer patients
treated with radiotherapy ± systemic therapy – A study of the Italian
Association of Radiotherapy and Clinical Oncology (AIRO)
Anna Viganò a, Francesca De Felice b, *, Nicola Alessandro Iacovelli c, Daniela Alterio d,
Nadia Facchinetti c, Olga Oneta d, e, Almalina Bacigalupo f, Elena Tornari f, Stefano Ursino g,
Fabiola Paiar g, Orietta Caspiani h, Alessia Di Rito i, Daniela Musio b, Paolo Bossi l,
Patrizia Steca m, Barbara Alicja Jereczek-Fossa d, e, Andrea Greco a, 1, Ester Orlandi c, 1
a
Department of Human and Social Sciences, University of Bergamo, Italy
b
Department of Radiotherapy, Policlinico Umberto I “Sapienza” University of Rome, Rome, Italy
c
Radiotherapy 2 Unit, Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Milan, Italy
d
Division of Radiation Oncology, IEO, European Institute of Oncology IRCCS, Milan, Italy
e
Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
f
Radiation Oncology Policlinico San Martino IRCCS, Genova, Italy
g
Department of Radiation Oncology, S. Chiara University Hospital, Pisa, Italy
h
Radiation Oncology Department – Ospedale “S. Giovanni Calibita” Fatebenefratelli, Rome, Italy
i
Radiotherapy Unit, Ospedale Mons A.R. Dimiccoli, Barletta, Italy
l
Medical Oncology, Department of Medical and Surgical Specialties, Radiological Sciences and Public Health University of Brescia, ASST-Spedali Civili, Brescia, Italy
m
Department of Psychology, University of Milan “Bicocca”, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: Purpose: Head and neck cancer (HNC) patients are likely to develop severe side effects, which may persist long
Head and neck cancer after the end of treatment and may be responsible for decrease patient’s quality of life. The M.D. Anderson
Quality of life Symptom Inventory- Head and Neck Module (MDASI-HN) is a questionnaire developed to detect patient’s
MDASI-HN
symptom burden. To conduct an Italian language psychometric validation of MDASI-HN among Italian HNC
Psychometric validation
patients on behalf of the Italian Association of Radiotherapy and Clinical Oncology (AIRO) Head and Neck
Italian language
Working Group.
Method and materials: To assess construct validity, it was performed a confirmatory factor analysis (CFA) with
both a five-factor solution and three-factor solution, which were compared by a chi-square difference test. The
concurrent validity was evaluated by the correlation with EORTC QLQ-C30 and HN35, and it was also assessed
known-group validity. The internal consistency was tested using Cronbach’s alpha coefficient.
Results: In total 166 patients (71.7% male) were included in the study, most of patients (56.2%) had an
oropharynx cancer and received definitive chemoradiotherapy (51.2%). The chi-square difference test was sig­
nificant and indicated that the five-factor solution fits the data better than the other one. Regarding CFA, all
items had a significant saturation with their respective factors; besides, significant and strong correlations were
found among factors. Most of the correlations between MDASI-HN factors and EORTC QLQ-C30 and HN35 were
significant. It was found a good internal consistency.
Conclusion: The MDASI-HN is a valid, short, and easy patient-reported outcome questionnaire which would be
useful and efficient in clinical setting.

* Corresponding author at: Department of Radiotherapy, Policlinico Umberto I “Sapienza” University of Rome, Viale Regina Elena 326, 00161 Rome, Italy.
E-mail address: [email protected] (F. De Felice).
1
These authors contributed equally to the work.

https://doi.org/10.1016/j.oraloncology.2021.105189
Received 23 September 2020; Received in revised form 24 December 2020; Accepted 8 January 2021
Available online 4 February 2021
1368-8375/© 2021 Elsevier Ltd. All rights reserved.
A. Viganò et al. Oral Oncology 115 (2021) 105189

Introduction patients’ QoL. Here, we focused on the Italian language psychometric

validation of MDASI-HN questionnaire. Baseline data was relevant to
Head and neck squamous cell carcinoma (HNSCC) represents a het­ perform this analysis.
erogeneous and relative rare disease accounting for about 5% of all Consecutive HNSCC patients treated at seven Italian Oncologic
cancer diagnosis, worldwide [1]. In the last decades, improved diag­ Radiotherapy Departments, from January 2016 to December 2019, were
nostic and staging accuracy as well as more efficacious therapeutic ap­ screened to be included in the analysis. To note, patients consecutiveness
proaches established within a multidisciplinary context, including was guaranteed within each center. The study was approved by the Ethical
radiation oncologist, surgeons and medical oncologist, have improved Committee of Fondazione IRCCS Istituto Nazionale dei Tumori in Milan
clinical results for HNSCC patients [2]. Despite this, cancer itself and its (prot. INT 29/15). All patients signed study-specific informed consent
treatment may result in several complications and distressing symptoms and answered to the questionnaires after the physicians visit.
that negatively affect all aspects of patient’s quality of life (QoL) [3]. The inclusion criteria were: (a) patients with a biopsy-proven squa­
HNSCC patients face unique physical, emotional, and psychosocial mous cell carcinoma of the head and neck (including oral cavity,
challenges and life disruptions [4]. Furthermore, decreased level of oropharynx, larynx and hypopharynx) without evidence of distant me­
health and well-being may lead to poorer tolerance of therapy, un­ tastases, treated with curative intent; (b) age ≥ 18 years old; c) perfor­
planned treatment interruptions, or modification of the overall treat­ mance status, according to ECOG < 2; and (d) good knowledge of Italian
ment schedule [3,5]. language. Patients were excluded in case of history of cognitive or
Since QoL domains have been shown to predict survival in HNSCC psychiatric disorders, synchronous tumors, or previous RT to the head
patients [6,7], there has been a paradigm shift in the measurement of and neck region.
clinical outcomes from physician ratings to patient-reported outcome All patients received RT, with or without concomitant chemo­
(PRO) measures, in particular for radiotherapy (RT)-based treatment therapy, with definitive or adjuvant intent, based on stage of disease and
[8–10]. primary location. RT was delivered with intensity modulated techniques
The M. D. Anderson Symptom Inventory (MDASI) is a short, self- (IMRT or Volumetric Modulated arc therapy) to a total dose of 66–72 Gy
administered questionnaire which was developed to measure the (1.8–2.2 Gy per fraction) in the radical setting and 50–66 Gy (1.8–2.2 Gy
severity of cancer-related symptoms and their influence on daily activ­ per fraction) in the adjuvant one. Type of concomitant systemic therapy
ities [11]. A specific head and neck module (MDASI-HN) has been (platinum-based or cetuximab) was individualized based on patient’s
proposed [12] to directly assess relevant head and neck cancer (HNC) comorbidities. If needed, type of surgery was left to the surgeon’s
symptoms, such as radiation-induced mucositis, and an Italian version of discretion, and the type of induction chemotherapy was chosen by the
the MDASI-HN has been linguistically validated [13]; however, it has medical oncologist.
not been found a psychometric validation. A literature review showed
some differences among MDASI-HN factor analysis, except for the items Questionnaires and data collection
within the interference with daily life factor, which has been considered
as unique. For instance, Rosenthal and colleagues [12] considered the MDASI-HN is a 28 symptom items questionnaire (see Table 3 for the
13 core items as a unique factor, whereas several studies found a two- items): 13 general cancer-related symptoms, such as pain, fatigue and
factor solution [11,14–19]. Moreover, the 9 HNC items were consid­ nausea; 9 HNC-related symptoms, such as problems with mucus in the
ered as a dimension by Sanchez and colleagues [17], while as two mouth and difficulty in swallowing or chewing; 6 items to evaluate the
constructs by Rosenthal and colleagues [12]. Therefore, taking into effects of symptoms on daily life, including mood and enjoyment of life.
consideration these inconsistencies it has been decided to consider both Each item is rated on a 11-point scale from 0 (not at all) to 10 (as bad as
a five-factor solution (general, gastrointestinal, HNC and treatment- you can imagine), while the items that assess the interference of
related symptoms, and symptoms interference with daily activities fac­ symptoms on daily activities are rated from 0 (does not interfere) to 10
tors) and a three-factor solution (general cancer-related symptoms, (interfered completely). The coefficients of Cronbach are 0.88, 0.83, and
HNC-related symptoms, and symptoms interference with daily activities 0.92 respectively for the first, second, and third factors [12].
factors), whose comparison would allow to understand which is the EORTC QLQ-C30 and EORTC QLQ-HN35 questionnaires were
most suitable one. developed to evaluate cancer patients’ health related QoL; the latter one
In the context of patient reported outcomes (PRO) measures, MDASI- is specific for HNC patients and is often used in conjunction with the
HN could be useful for oncologic research purposes, adding prognostic former one [20–22].
value in HNC disease. However, MDASI-HN needs to be psychometri­ Data on patients’ clinical and demographic characteristics were also
cally validated in the Italian language to be considered a robust QoL collected. Toxicity was assessed and scored according to the common
measure; therefore, it was conducted a confirmatory factor analysis terminology criteria for adverse effects (CTCAE) version 4.0 and the
(CFA) to assess construct validity and the two above-mentioned factor radiation therapy oncology group/European organization for research
solutions were compared to each other. Besides, it was performed a and treatment of cancer (RTOG/EORTC) criteria. It was also evaluated
reliability test involving Cronbach alpha to evaluate internal consis­ the global stage, the ECOG performance status, the HPV status and
tency. The correlations with the EORTC QLQ-C30 and EORTC QLQ- whether patients underwent surgery.
HN35 were used to evaluate the concurrent validity of the MDASI-HN
and global stage of disease, patient’ ECOG performance status, HPV Statistical analysis
status and surgery were used for the known-group validity.
Data were analysed using IBM SPSS Statistics version 25 (IBM,
Methods Armonk, NY, USA) and Mplus software version 7 [23]. Confirmatory
factor analysis (CFA) was performed to assess factorial validity. The a
Study design priori sample size calculation was based on the common rules of thumb
proposing the ratio of the number of participants (N) to the number of
This is a multi-center longitudinal, prospective, observational study. measured variables (p); based on these assumptions, sample size should
Primary end-point was to psychometrically validate the MDASI core and be greater than the number of variables (N > p) and N:p ratios have to be
MDASI-HN questionnaires on behalf of the Italian Association of at least 5 with a minimum N > 100 [24,25]. Considering the number of
Radiotherapy and Clinical Oncology (AIRO) Head and Neck Working items, the minimum sample size required was of 140 participants. To
Group. Secondary endpoints were i) to analyze QoL changes over time; prevent multivariate non-normal distribution, robust maximum likeli­
ii) to assess the interference of psychological and physical variables on hood estimation was employed. Hu and Bentler’s [26] guidelines for

2
A. Viganò et al. Oral Oncology 115 (2021) 105189

various fit indices were used to determine whether the expected model fatigue, being distressed, and difficulty with voice/speech were normal
fits the data. The chi-square test statistic was used but considering its distributed.
sensitivity to sample size the root mean square error of approximation Among the items which are part of the general cancer symptoms
(RMSEA ≤ 0.08 is considered an acceptable fit) and the standardised core, nausea (item3) and vomiting (item12) had the lowest rate,
root mean square residual (SRMR ≤ 0.08 indicates an acceptable fit) whereas fatigue (item2), feeling distressed (item5), and feeling sad
were employed. (item11) had the highest rates. Regarding HNC-related symptoms items,
The 5-factor model and the 3-factor solution of MDASI-HN were difficulty with voice/speech (item17) was rated as the most painful
compared by a chi-square difference test using the Satorra-Bentler symptom. Cancer symptoms interfered more with patients’ mood
scaled chi-square test. (item24) and work (item25); furthermore, it had a negative impact on
The concurrent validity of the MDASI-HN was assessed by the cor­ enjoyment of life (item28).
relation with the EORTC QLQ-C30 and HN35 with the Pearson’s r co­
efficient of correlation. Correlations were evaluated weak (|0.10| < r < | Construct validity
0.29|), moderate (|0.30| < r < |0.49|), or strong (r > |0.50|). For the
known-group validity, in which Mann-Whitney U test and a t test for The Chi-square was significant, therefore considering the sensitivity
independent samples were employed, it was used the global stage (ac­ of it to the sample size, other fits statistics were considered, which
cording to TNM 7th edition), the ECOG performance status, HPV status, shown a good fit in both the five-factor solution (χ2 343 = 629.541, p ≤
and surgery. Questionnaires internal consistency was tested using 0.001; RMSEA 0.072; SRMR 0.069) and the three-factor solution (χ2
Cronbach’s alpha coefficient, which should be 0.70 or higher to meet the 347 = 676.595, p ≤ 0.001; RMSEA 0.077; SRMR 0.077). A chi-square
minimal standard for reliability. difference test using the Satorra-Bentler scaled chi-square test (27)
was used to compare the five-factor solution and three-factor solution.
Results The test result was significant (Δ χ2 4 = 29.204, p ≤ 0.001), indicating
that the five-factor solution fits the data better than the first one. The
Patient characteristics standardised factor loadings for the factors are shown in Fig. 1. Almost
all items had a strong, significant saturation with their respective five
In total 166 patients were enrolled. Socio-demographic and clinical factors (ranging from 0.57 to 0.93). Both correlations between core
characteristics are listed in Tables 1 and 2, respectively. symptoms factors (F1-F2) and HNC symptoms factors (F3-F4) whit their
main factors, general cancer-related symptom (FF1) and HNC-related
Description of symptoms symptom (FF2), were found strong, ranging from 0.76 to 1.02. More­
over, significant and strong correlations were found among the three
The average scores of participants responses ranged from 0.32 to second-order factors (FF1, FF2, F5), ranging between 0.69 and 0.86.
2.39, with standard deviation between 1.16 and 3.39 (Table 3). In most
of their responses, participants used all the rating scale values; a few Table 2
exceptions were related to item 7 and item 12, whose maximum score Patients’ clinical characteristics.
was 7 and 8 respectively. The results showed that only three items, - Demographic Frequencies (%)

Table 1 Tumour site

Patients’ socio-demographic characteristics. Hypopharynx 8 4.9
Larynx 29 17.9
Demographic Frequencies (%) Oral cavity 34 21
Sex Oropharynx 91 56.8
Male 119 71.7 Global stage (according to TNM 7th edition)
Female 47 28.3 I 5 3.3
Age (years), mean (SD) = 61.69 (11.01); range = 24–93 II 14 9.3
Marital status III 22 14.7
Single 10 6.0 IV 109 72.7
Married 102 61.4 ECOG status
Divorced/separated 22 13.3 0 105 64.4
Widowed 11 6.6 1 58 35.6
Missing 21 12.7 Tabacco smoker
Living situation Current 37 22.3
Alone 15 9.0 Former 74 44.6
Only with spouse/partner 63 38.0 Never 47 28.3
With spouse/partner and children 52 31.3 Unknown 8 4.8
Only with children 4 2.4 Alcohol
Only with other relatives 7 4.2 Current 39 23.5
Other 3 1.8 Former 18 10.9
Missing 22 13.3 Never 55 33.1
Educational level Unknown 54 32.5
None 1 0.6 HPV status
Lower school 17 10.2 Negative 32 19.3
Middle school 34 20.5 Positive 66 39.7
High school 70 42.2 Unknown 68 41.0
Graduate school 17 10.2 Surgery
Postgraduate school 5 3.0 Yes 52 31.7
Missing 22 13.3 No 112 68.3
Employment status Setting Radiation Therapy
Employed 58 35.0 Postoperative (45–66 Gy) with SyT 20 12.2
Unemployed 13 7.8 Postoperative (45–66 Gy) without SyT 32 19.5
Housewife 9 5.4 Definitive (66–72 Gy) with SyT 84 51.2
Retired 50 30.1 Definitive (66–72 Gy) without SyT 28 17.1
Retired, but with some work 13 7.8
Legend: ECOG = Eastern Cooperative Oncology Group; HPV = Human Papil­
Missing 23 13.9
loma Virus; Gy = Gray; SyT = systemic therapy.

3
A. Viganò et al. Oral Oncology 115 (2021) 105189

Table 3
Descriptive statistics of the MDASI-HN at baseline.
Missing Mean SD Min Max Skewness Kurtosis

General cancer-related symptoms

1-Your Pain at its worst 8 1.88 2.36 0 9 1.24 0.42
2-Your fatigue (tiredness) at its worst 8 2.39 2.51 0 10 0.96 0.11
3-Your nausea at its worst 7 0.65 1.68 0 10 3.44 12.86
4-Your disturbed sleep at its worst 9 1.86 2.36 0 9 1.28 0.75
5-Your feeling of being distressed (upset) at its worst 8 2.40 2.48 0 9 0.87 − 0.22
6-Your shortness of breath at its worst 8 1.38 2.21 0 10 1.84 2.78
7-Your problem with remembering things at its worst 9 1.10 1.71 0 7 1.78 2.57
8-Your problem with lack of appetite at its worst 9 1.19 1.90 0 9 2.05 4.13
9-Your feeling drowsy (sleepy) at its worst 9 1.82 2.38 0 10 1.48 1.42
10-Your having a dry mouth at its worst 9 2.08 2.77 0 10 1.31 0.59
11-Your feeling sad at its worst 10 2.24 2.58 0 10 1.15 0.47
12-Your vomiting at its worst 8 0.32 1.16 0 8 4.28 19.21
13-Your numbness or tingling at its worst 10 0.97 1.87 0 10 2.55 6.70

Head and neck cancer-related symptoms

14-Your problem with mucus in your mouth and throat at its worst 11 2.21 2.81 0 10 1.12 0.03
15-Your difficulty swallowing/chewing at its worst 14 2.18 2.86 0 10 1.25 0.37
16-Your choking/coughing (food/liquids going down the wrong pipe) 13 1.47 2.36 0 10 1.63 1.65
17-Your difficulty with voice/speech at its worst 11 2.83 3.39 0 10 0.94 − 0.56
18-Your skin pain/burning/rash at its worst 13 0.86 1.78 0 8 2.49 6.02
19-Your constipation at its worst 11 1.69 2.69 0 10 1.66 1.67
20-Your problem with tasting food at its worst 14 1.82 2.82 0 10 1.47 0.92
21-Your mouth/throat sores at their worst 11 1.03 2.01 0 10 2.19 4.30
22-Your problem with your teeth or gums at its worst 12 1.49 2.74 0 10 1.91 2.50

Symptoms interference on daily activities

23-General activity 14 2.09 2.96 0 10 1.46 0.96
24-Mood 13 2.41 2.72 0 10 1.14 0.40
25-Work (including work around the house) 12 2.52 3.29 0 10 1.10 − 0.15
26-Relations with other people 11 1.67 2.69 0 10 1.60 1.51
27-Walking 11 1.44 2.59 0 10 1.91 2.60
28-Enjoyment of life 12 2.77 3.29 0 10 1.01 − 0.26

Concurrent validity whereas ECOG status was significant for FF2 factor, namely t(108) =
− 2.52, p < 0.05. Consistent with the five-factor solution, HNC related
The concurrent validity was evaluated through the correlation be­ symptoms factor (FF2) resulted to be significant in relation to HPV, t
tween MDASI-HN and EORTC QLQ-C30 and HN35. All correlations (46.57) = 2.90, p < 0.01. Furthermore, surgery was significant for FF2
between MDASI-HN five factors and EORTC-QLQ-C30 dimensions were and F5, t(71.27) = − 3.90, p < 0.001 and t(84.78) = − 2.02, p < 0.05,
significant (Table 4), ranging from r = 0.20 to r = 0.69, with the respectively.
exception for the one between nausea/vomiting scale and HNC symp­
toms (F3). Similarly, MDASI-HN five factors and EORTC-QLQ-HN35
Internal consistency reliability
dimensions had significant correlations (Table 5), ranging from r =
0.17 to r = 0.61, except for the one between the speech scale and
The internal consistency of the MDASI factors, for both 5-factor so­
gastrointestinal symptoms factor (F2). Considering the three-factor so­
lutions and 3-factor, was very good for most of the factors, except for F2
lution of MDASI-HN, all the correlations were significant, ranging from
and F4, which had moderate values, specifically, Cronbach’s alpha was
r = 0.20 to r = 0.69 for EORTC QLQ-C30 (Table 4), and from r = 0.25 to
0.90 for F1, FF1, and FF2, 0.68 for F2, 0.89 for F3, 0.78 for F4, and 0.93
r = 0.52 for EORTC QLQ-HN35 (Table 5).
for F5.
For the known-group validity it was used Mann-Whitney U test and a
t test for independent samples. Since the two classes of global stage,
Discussion
ECOG status, HPV, and surgery were out of proportion, for each of these
variables there were randomly selected groups among the largest one in
The aim of this study was the psychometric validation of MDASI-HN
order to have similar sizes, namely, 25 patients for global stage, 60
questionnaire in order to better identify HNC patients baseline clinical
subjects for the ECOG status, 32 among the HPV positive patients group,
conditions to ameliorate their daily management. Several studies were
and 52 patients among those who did not undergo surgery. Considering
taken into consideration to define the CFA structure. The validation of
the five-factor solution, the global stage resulted to be significant for the
the MDASI-HN module conducted by Rosenthal and colleagues [12]
gastrointestinal symptoms factor (F2) (U F1 = 212, p > 0.005; U F2 =
found two-factor solution for the 9 items representing HNC symptoms.
135, p ≤ 0.05; U F3 = 192.5, p > 0.05; U F4 = 173.5, p > 0.05; U F5 =
Differently, Sanchez and colleagues [17] found that HN symptoms items
178, p > 0.05). A t test for independent samples was performed for
resulted in a one-factor solution instead of two-factor one. Nevertheless,
ECOG status, HPV, and surgery. The t test for the ECOG status resulted
it was found a two-factor solution for the core symptoms, which is
significant for F3 t(108) = − 2.06, p < 0.05 and F4 t(108) = − 2.57, p <
consistent with most of the other studies [11,14–16,18–19]. Taking into
0.05. Regarding HPV, the significance was found for F1, F3, and F4,
account these findings, it has been decided to consider both a 5-factor
respectively, t(47) = 2.23, p < 0.05, t(48.05) = 2.67, p < 0.01, and t
solution (general, gastrointestinal and HNC symptoms and treatment-
(46.56) = 2.64, p < 0.01. Significant values in the surgery variable were
related symptoms, and symptoms interference with daily activities fac­
found in F3, F4, and F5, specifically, t(69.71) = − 4.30, p < 0.001, t
tors) and a 3-factor solution (general cancer-related symptoms, HNC-
(78.84) = − 2.96, p < 0.01, and t(84.78) = − 2.02, p < 0.05.
related symptoms, and symptoms interference with daily activities
In the three-factor solution, global stage resulted not significant (U
factors).
FF1 = 1153, p > 0.05; U FF2 = 1085, p > 0.05; U F5 = 966.5, p > 0.05);
Factorial validity was assessed by CFA in which robust maximum

4
A. Viganò et al. Oral Oncology 115 (2021) 105189

Fig. 1. Confirmatory Factor Analysis for M.D. Anderson Symptom Inventory- Head and Neck.

likelihood estimation was applied; the indices used, RMSEA and SRMR, evaluated through the comparison with EORTC QLC-C30 and EORTC
shown a good fit. Regarding the standardised factor loadings for the five QLQ-HN35; most correlations were found significant, with a few
factors, all items had a strong, significant saturation with their respec­ exceptions.
tive factors. Similarly, the correlation between the main three factors, For the know-group validity, regarding the five-factor solution,
specifically, general cancer-related symptoms (FF1), HNC-related Mann-Whitney U test for the global stage resulted significant for the
symptoms (FF2), and symptoms interference with daily activities (F5) gastrointestinal symptoms factor (F2). Concerning ECOG, a t test for
had strong coefficients. The concurrent validity of MDASI-HN was independent samples was performed resulting significant for F3 and F4,

5
A. Viganò et al. Oral Oncology 115 (2021) 105189

Table 4
Concurrent validity of the MDASI-HN: correlation coefficients between EORTC QLQ-C30 dimensions and MDASI-HN factors.
EORCT QLQ- EORCT EORCT QLQ- EORCT QLQ- EORCT EORCT EORCT EORCT QLQ- EORCT QLQ-
C30 physical QLQ-C30 C30 cognitive C30 emotional QLQ-C30 QLQ-C30 QLQ-C30 C30 nausea/ C30 health
funct. role funct. funct. funct. social funct. fatigue scale pain scale vomiting scale status

General cancer- 0.534*** 0.564*** 0.379*** 0.522*** 0.331*** 0.694*** 0.616*** 0.409*** − 0.626**
related symptoms
(FF1)
HNC-related 0.409*** 0.436*** 0.211** 0.315*** 0.233** 0.507*** 0.546*** 0.202* − 0.456***
symptoms (FF2)
Symptoms 0.512*** 0.653*** 0.354*** 0.473*** 0.453*** 0.606*** 0.499*** 0.270** − 0.583***
interference on
daily activities (F5)
General symptoms 0.545*** 0.554*** 0.387*** 0.547*** 0.335*** 0.697*** 0.626*** 0.326*** − 0.610**
factor (F1)
Gastrointestinal 0.310*** 0.413*** 0.222** 0.223** 0.202* 0.445*** 0.377*** 0.639*** − 0.480***
symptoms factor
(F2)
HNC symptoms (F3) 0.373*** 0.397*** 0.167* 0.317*** 0.228** 0.423*** 0.449*** 0.149 − 0.434***
HNC treatment- 0.395*** 0.425*** 0.234** 0.269*** 0.209** 0.525*** 0.574*** 0.236** − 0.417***
related symptoms
(F4)
*
The correlation is significant at 0.05 (two-tailed).
**
The correlation is significant at 0.01 (two-tailed).
***
The correlation is significant at 0.001 (two-tailed).

Table 5
Concurrent validity of the MDASI-HN: correlation coefficients between EORTC QLQ-HN35 dimensions and MDASI-HN factors.
EORCT QLQ- EORCT QLQ-HN35 EORCT QLQ- EORCT QLQ- EORCT QLQ- EORCT QLQ-HN35 EORCT QLQ-
HN35 pain scale swallowing scale HN35 sense HN35 speech HN35 social eating social contact scale HN35 sexuality
scale scale scale scale

General cancer-related 0.407*** 0.337*** 0.278*** 0.348*** 0.329*** 0.353*** 0.273**

symptoms (FF1)
HNC-related symptoms 0.427*** 0.519*** 0.293*** 0.526*** 0.471*** 0.291*** 0.250**
(FF2)
Symptoms interference on 0.274*** 0.295*** 0.308*** 0.375*** 0.310*** 0.421*** 0.332***
daily activities (F5)
General symptoms factor 0.407*** 0.324*** 0.235** 0.367*** 0.326*** 0.364*** 0.273***
(F1)
Gastrointestinal symptoms 0.286*** 0.271*** 0.361*** 0.135 0.221** 0.172* 0.168*
factor (F2)
HNC symptoms (F3) 0.348*** 0.561*** 0.217** 0.609*** 0.440*** 0.266*** 0.269***
HNC treatment-related 0.449*** 0.394*** 0.330*** 0.368*** 0.437*** 0.282*** 0.194*
symptoms (F4)
*
The correlation is significant at 0.05 (two-tailed).
**
The correlation is significant at 0.01 (two-tailed).
***
The correlation is significant at 0.001 (two-tailed).

showing that patients with poorer ECOG score had higher HNC symp­ representativeness of the sample is mainly limited by the inclusion of
toms scores than those patients with better status. The HNC related male patients treated with definitive curative intent and with a preva­
symptoms factors (F3, F4) and general symptoms one (F1) resulted lence of oropharyngeal and stage IV cancer at diagnosis. In addition, the
significant when considering HPV status, highlighting that those pa­ most significant missing data is in relation to patients’ socio-
tients with HPV-negative disease had more problems concerning general demographic and clinical characteristics. But this does not affect the
cancer as well as cancer-specific aspects. Those patients who undergone psychometric validation of MDASI-HN. Indeed, all participants
surgery had worst symptoms scores about HNC symptoms and inter­ completed the MDASI-HN and it was expected to have some missing
ference with daily life seeing the statistical significance of F3, F4, F5. answers in relation to some items. Though PRO questionnaires may not
When considering the three-factor solution, the Mann-Whitney U test be fully comprehensive of subtle aspects of cancer patients’ experience,
showed no significance for the global stage, whereas, consistently with the MDASI-HN is a statistical valid instrument whose items concern
the above-mentioned results, ECOG status was significant for the HNC general cancer-related symptoms, HN specific ones, and the interference
related symptoms factor (FF2), which resulted to be significant also of physical symptoms with daily activities and, thus, subject’s QoL.
when considering HPV status. Regarding surgery, HNC specific symp­ Although MDASI-HN is sometimes considered as composed of three
toms (FF2) and symptoms interference with daily activities (F5) were factors, this study demonstrated that a five-factor solution allows to gain
significant. subtle insights of patients’ symptoms and QoL, highlighting the need to
Most of participants assessed fatigue (item2), feeling distressed consider this solution in order to have a deep understanding of patients’
(item5), feeling sad (item11), and difficulty with voice/speech (item17) condition.
as the most painful symptoms, highlighting that cancer symptoms The comparison with other QoL instruments showed that MDASI-HN
negatively interfered with enjoyment of life (item28). may be considered as the most comprehensive tool to investigate QoL of
We acknowledge some possible limitations that may have an influ­ HNC patients, being also a valid, short, and easy PRO questionnaire
ence on results generalization to HNC population. Firstly, the which would be useful as well as efficient in daily clinical setting.

6
A. Viganò et al. Oral Oncology 115 (2021) 105189

Funding sources [10] Cleeland CS, Mendoza TR, Wang XS, et al. Assessing symptom distress in cancer
patients: the M.D. Anderson Symptom Inventory. Cancer 2000;89(7):1634–46.
[11] Rosenthal DI, Mendoza TR, Chambers MS, Asper JA, Gning I, Kies MS, et al.
This research did not receive any specific grant from funding Measuring head and neck cancer symptom burden: the development and validation
agencies in the public, commercial, or not-for-profit sectors. of the M. D. Anderson symptom inventory, head and neck module. Head Neck.
2007;29(10):923–31.
[12] Greco A, Orlandi E, Mirabile A, Takanen S, Fallai C, Iacovelli NA, et al. Italian
Declaration of Competing Interest version of the M.D. Anderson Symptom Inventory-Head and Neck Module:
linguistic validation. Support Care Cancer 2015;23(12):3465–72.
The authors declare the following financial interests/personal re­ [13] Lin CC, Chang AP, Cleeland CS, Mendoza TR, Wang XS. Taiwanese version of the
M. D. Anderson symptom inventory: symptom assessment in cancer patients. J Pain
lationships which may be considered as potential competing interests: Symptom Manage 2007;33(2):180–8.
‘This research did not receive any specific grant from funding agencies in [14] Nejmi M, Wang XS, Mendoza TR, Gning I, Cleeland CS. Validation and application
the public, commercial, or not-for-profit sectors. The authors declare no of the Arabic version of the M. D. Anderson symptom inventory in Moroccan
patients with cancer. J Pain Symptom Manage 2010;40(1):75–86.
conflicts of interest and thank the Scientific Committee and Board of the [15] Okuyama T, Wang XS, Akechi T, Mendoza TR, Hosaka T, Cleeland CS, et al.
AIRO for the revision and approval of this paper’. Japanese version of the MD Anderson Symptom Inventory: a validation study.
J Pain Symptom Manage 2003;26(6):1093–104.
[16] Sánchez D, Chala A, Alvarez A, Payan C, Mendoza T, Cleeland C, et al.
References Psychometric validation of the M. D. Anderson Symptom Inventory-Head and Neck
Module in the Spanish Language. J Pain Symptom Manage 2016;51(6):1055–61.
[1] Orlandi E, Alfieri S, Simon C, Trama A, Licitra L, RARECAREnet Working Group. [17] Wang XS, Wang Y, Guo H, Mendoza TR, Hao XS, Cleeland CS. Chinese version of
Treatment challenges in and outside a network setting: Head and neck cancers. Eur the M. D. Anderson Symptom Inventory: validation and application of symptom
J Surg Oncol. 2019;45(1):40–5. measurement in cancer patients. Cancer 2004;101(8):1890–901.
[2] Elad S, Meyerowitz C. Late oral adverse effects of cancer treatments. In: ALERT [18] Wang XS, Laudico AV, Guo H, Mendoza TR, Matsuda ML, Yosuico VD, et al.
adverse late effects of cancer treatment medical radiology; 2013. p. 141–66. Filipino version of the M. D. Anderson Symptom Inventory: validation and
https://doi.org/10.1007/978-3-540-75863-1_6. multisymptom measurement in cancer patients. J Pain Symptom Manage 2006;31
[3] Rieke K, Schmid KK, Lydiatt W, Houfek J, Boilesen E, Watanabe-Galloway S. (6):542–52.
Depression and survival in head and neck cancer patients. Oral Oncol 2017;65: [19] Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, et al. The
76–82. European Organization for Research and Treatment of Cancer QLQ-C30: a quality-
[4] Wiederholt PA. Organized head and neck cancer care. In: Functional preservation of-life instrument for use in international clinical trials in oncology. J Natl Cancer
and quality of life in head and neck radiotherapy medical radiology; 2009. Inst 1993;85(5):365–76.
p. 307–16. https://doi.org/10.1007/978-3-540-73232-7_28. [20] Fayers P, Bottomley A, EORTC Quality of Life Group; Quality of Life Unit. Quality
[5] Osthus AA, Aarstad AK, Olofsson J, Aarstad HJ. Head and neck specific Health of life research within the EORTC-the EORTC QLQ-C30. European Organisation for
Related Quality of Life scores predict subsequent survival in successfully treated Research and Treatment of Cancer. Eur J Cancer 2002;38(Suppl 4):S125–33.
head and neck cancer patients: a prospective cohort study. Oral Oncol. 2011;47 [21] Bjordal K, Ahlner-Elmqvist M, Tollesson E, Jensen AB, Razavi D, Maher EJ, et al.
(10):974–9. Development of a European Organization for Research and Treatment of Cancer
[6] Trask PC, Hsu MA, McQuellon R. Other paradigms: health-related quality of life as (EORTC) questionnaire module to be used in quality of life assessments in head and
a measure in cancer treatment: its importance and relevance. Cancer J 2009;15(5): neck cancer patients. EORTC Quality of Life Study Group. Acta Oncol 1994;33(8):
435–40. 879–85.
[7] Cleeland CS, Sloan JA, ASCPRO Organizing Group. Assessing the Symptoms of [22] Muthen LK, Muthen B. Mplus users guide: statistical analysis with latent variables.
Cancer Using Patient-Reported Outcomes (ASCPRO): searching for standards. Los Angeles (CA): Muthén & Muthén; 2017.
J Pain Symptom Manage 2010;39(6):1077–85. [23] Li-tze Hu, Bentler PM. Cutoff criteria for fit indexes in covariance structure
[8] Holländer-Mieritz C, Johansen J, Taarnhøj GA, Johansen C, Vogelius IR, analysis: Conventional criteria versus new alternatives. Struct Eq Model: Multidisc
Kristensen CA, et al. Systematic use of patient reported outcome during J 1999;6(1):1–55.
radiotherapy for head and neck cancer: study protocol for the national DAHANCA [24] Satorra A, Bentler PM. A scaled difference chi-square test statistic for moment
38 trial. Acta Oncol 2020;59(5):603–7. structure analysis. Psychometrika 2001;66:507–14.
[9] Stephens RJ, Hopwood P, Girling DJ, Machin D. Randomized trials with quality of [25] Nunnally JC, Bernstein IJ. Teoríapsicométrica. Madrid: McGraw-Hill; 1995.
life endpoints: are doctors’ ratings of patients’ physical symptoms interchangeable
with patients’ self-ratings? Qual Life Res 1997;6(3):225–36.