Abril - 2017

atient

oncerns

nventory

for

head

and

neck

cancer

: B

razilian

cultural

adaptation

ssoc

ras

2017; 63(4):311-319

313

Based on the literature available for this type of re-

search, there is no consensus and no general criteria to

establish the required sample size for a patient-reported

outcome instrument validation study. A sample size of

at least 50 to 100 participants is generally recommended.

These numbers of subjects ensure stability of the vari-

ance-covariance matrix.

A psychometrically valid survey instrument will assess

what it is meant to be measured, and the PCI evaluates

whether patients want to discuss items – in practice, they

either want to discuss an item or they don’t. Due to its

“checklist” format, PCI is not suited for psychometric

testing in the usual sense applied to the majority of

HRQOL survey instruments. Translation and cultural

adaptation are the most important steps to assure that

the properties of the original instrument are maintained

in another language. Content validity is the most relevant

of the PCI properties, reflecting a belief that the questions

adequately cover the content of the field of study. In the

original language format, content validity was thorough-

ly established, which ought to be maintained after prop-

er translation into Brazilian Portuguese and subsequent

verification through back-translation.

However, even considering that the most important

step of cultural validation of PCI into Brazilian Portuguese

is the rigorous process of translation and back-translation,

we decided to test the construct validity of the Brazilian

version by comparing its results with the UW-QOL scores.

Construct validity is present if the scale behaves according

to hypothesized relationships. We hypothesized that the

composite score of the UW-QOL should be associated

and correlated with the number of issues selected for

discussion by the patients on the PCI.

The validated Brazilian Portuguese format of the

UW-QOLv4 was used because it is a disease-specific, con-

cise and well-validated QOL questionnaire and because its

domains overlap with many of the concerns listed on the

PCI. Each domain item on the UW-QOL scale is scored

from 0 to 100, with the composite score being the mean of

the 12 domains. Higher scores are indicative of better QOL.

In our study, the UW-QOL results were analyzed consider-

ing two composite scores: “physical function” (simple aver-

age of the domain scores for swallowing, chewing, speech,

saliva, taste and appearance) and “socio-emotional function”

(simple average of the domain scores for activity, recreation,

pain, mood, anxiety and shoulders).

Statistical analysis

Nonparametric Mann-Whitney test (two categories) or

Kruskal-Wallis test (three categories) were used to evalu-

ate the association between number of items/specialists

selected and patient characteristics. The Mann-Whitney

test was also used to associate the composite scores of the

UW-QOL (physical, socio-emotional) with the specific

items/specialists selected by the patients. The associa-

tions between the scores of the UW-QOL and age, time

of diagnosis and the number of items/specialists se-

lected were assessed using Spearman’s correlation coef-

ficient. Associations between the specific items/special-

ists selected and patient characteristics were evaluated

using Chi-square test or Fisher’s exact test. The level of

significance was set to 1%. The software used was R ver-

sion 3.0.1. (R is a language and environment for statisti-

cal computing and graphics. R provides a wide variety

of statistical and graphical techniques, and is highly

extensible. It is available as free software under the terms

of the Free Software Foundation’s GNU General Public

License in source code form).

esults

Translation process and pretest survey

As part of the translation and adaptation process, we

conducted a pretest survey with 20 patients with UADT

cancers (eight oral cavity, six larynx, two oropharynx, three

nasopharynx and one maxillary sinus) who completed

the Brazilian Portuguese version of the PCI.

There were no significant discrepancies between the

translation and back-translation versions, despite the

grammatical and cultural differences between the Brazil-

ian and English populations.

However, in the pretest survey, some patients did not

understand the meaning of the following terms in the

first translated version: mucous, deglutition, smell, taste,

percutaneous gastrostomy (PEG), fear of adverse events

and coping strategies. The terms home care/Family

Health Program (originally described as home care/dis-

trict nurse) were erroneously interpreted as any family

member or person providing help/medical assistance.

The term “activity” was interpreted differently by patients,

being understood as physical exercises or as activities of

daily living. Patients were also confused by the terms

salivation and dry mouth and were unable to distinguish

and clearly define them.

Following the adequacy of equivalence and adapta-

tions for the Brazilian population, the consensus research

team then constructed a final version in Brazilian Portu-

guese. This final version was defined replacing or supple-

menting the terms that were misunderstood, misinter-

preted or indistinguishable, for phlegm/secretions,

swallowing (deglutition), olfaction (smell), taste (sense